Mechanistic elucidation of intramolecular aminoalkene hydroamination catalyzed by a tethered bis(ureate) complex: Evidence for proton-assisted C-N bond formation at zirconium

Article abstract of DOI:10.1021/ja202448b

A broad mechanistic investigation regarding hydroamination reactions catalyzed by a tethered bis(ureate) zirconium species, [ureate(2-)]Zr(NMe ₂)₂(HNMe₂), is described. The cyclization of both primary and secondary aminoal

Full text of DOI:10.1021/ja202448b

ARTICLE
pubs.acs.org/JACS
Mechanistic Elucidation of Intramolecular Aminoalkene
Hydroamination Catalyzed by a Tethered Bis(ureate) Complex:
Evidence for Proton-Assisted CꢀN Bond Formation at Zirconium
David C. Leitch, Rachel H. Platel, and Laurel L. Schafer*
Department of Chemistry, University of British Columbia, 2036 Main Mall, Vancouver, British Columbia V6T 1Z1, Canada
S Supporting Information
b
ABSTRACT: A broad mechanistic investigation regarding hydro-
amination reactions catalyzed by a tethered bis(ureate) zirconium
species, [ureate(2ꢀ)]Zr(NMe2)2(HNMe2), is described. The cy-
clization of both primary and secondary aminoalkene substrates
gives similar kinetic profiles, with zero-order dependence on sub-
strate concentration up to ∼60ꢀ75% conversion, followed by first-
order dependence for the remainder of the reaction. The addition
of 2-methylpiperidine changes the observed substrate dependence
to first order throughout the reaction, but does not act as a com-
petitive inhibitor. The reactions are first order in precatalyst up to
loadings of ∼0.15 M, indicating that a well-defined, mononuclear catalytic species is operative. Several model complexes have been
structurally characterized, including dimeric imido and amido species, and evaluated for catalytic performance. These results
indicate that imido species need not be invoked as catalytically relevant intermediates, and that the mono(amido) complex
[ureate(2ꢀ)]Zr(NMe2)(Cl)(HNMe2) is much less active than its bis(amido) counterpart. Structural evidence suggests that this is
due to differences in coordination geometry between the mono- and bis(amido) complexes, and that an equatorial amido ligand is
required for efficient catalytic turnover. On the basis of the determination of kinetic isotope effects and stoichiometric reactivity, the
catalytic turnover-limiting step is proposed to be a concerted CꢀH, CꢀN bond-forming process with a highly ordered, unimolecular
transition state (ΔSq = ꢀ21 ( 1 eu). In addition to this key bond-forming step, the catalytic cycle involves an on-cycle pre-equilibrium
between six- and seven-coordinate intermediates, leading to the observed switch from zero- to first-order kinetics.
’ INTRODUCTION
support this proposed cycloaddition pathway with alkyne and
allene substrates.
Hydroamination reactions promoted by metal-based catalyst
systems are proving to be an increasingly successful and widely
applicable method for the synthesis of many important classes of
nitrogen containing compounds.1,2 Among the most promising
catalysts in terms of synthetic applicability, economic viability,
and low environmental impact are those incorporating group 4
metal ions. Since the first reports of group-4-mediated hydro-
amination in the early 1990s,3ꢀ5 a plethora of research has been
directed toward the discovery of efficient titanium and zirconium
catalysts.6ꢀ13 Elucidation of the mechanistic details of hydro-
amination reactions catalyzed by these systems is particularly
important for further catalyst development and optimization.
In a landmark contribution, Bergman and co-workers de-
scribed the mechanism of intermolecular alkyne hydroamination
catalyzed by Cp2Zr(NHR)2 complexes (Scheme 1A).3 The key
step of this proposed pathway is a [2 + 2] cycloaddition between
an imidozirconium intermediate and the alkyne substrate.6
Subsequent turnover-limiting protonolysis of the ZrꢀC bond
by an incoming amine substrate followed by an intramolecular
proton-transfer releases the hydroamination product and regen-
erates the imidozirconium species. Subsequent mechanistic
studies by Pohlki and Doye14 and Johnson and Bergman15
on the related titanium catalyst system Cp2TiMe2 strongly
As a result of substantial catalyst development efforts by many
research groups, the catalytic intramolecular hydroamination of
select aminoalkenes is now viable with a number of titanium and
zirconium complexes.16ꢀ47 Mechanistic work reported by Scott
and co-workers,28,44 and our own research group,20 strongly sup-
ports a Bergman-type cycloaddition mechanism3 for neutral catalysts
with sterically encumbering ligand systems. One common me-
chanistic probe for the intermediacy of a catalytically active imido
species is the demonstration that secondary aminoalkenes are
not cyclized. While the vast majority of group 4 catalysts exhibit
this substrate selectivity, a few reported examples do not obey
this trend. These include cationic species that only cyclize secondary
aminoalkenes16,17 and rare neutral complexes that cyclize both
primary and secondary substrates.25,29,31,35,43 A kinetic analysis
of catalytic reactions involving zirconium precatalysts supported
by constrained-geometry ligands led Stubbert and Marks to
propose a catalytic cycle wherein the key step involves an insertion
of the alkene into a zirconiumꢀamido bond (Scheme 1B),25
similar to the mechanism determined for rare-earth catalysts.48
Received: March 30, 2011
Published: August 18, 2011
r
2011 American Chemical Society
15453
dx.doi.org/10.1021/ja202448b J. Am. Chem. Soc. 2011, 133, 15453–15463
|

Journal of the American Chemical Society
ARTICLE
Scheme 1. (A) Simplified Cycloaddition Mechanism Proposed for Zirconocene Catalysts;3 (B) Simplified Insertion Mechanism
Proposed for CGCꢀZr Catalysts;25 and (C) Simplified Concerted CꢀN, CꢀH Bond Formation Mechanism Proposed for
OxazolinylborateꢀZr Catalysts55
In a previous communication, we reported the development of
a tethered bis(ureate) zirconium catalyst (1, Figure 1) for the
intermolecular hydroamination of alkynes and the intramolecu-
lar hydroamination of alkenes.35 This catalyst possesses unpre-
cedented reactivity, generality, and functional group tolerance for
a group 4 system. Precatalyst 1 performs effectively with both
primary and secondary amines, forms 5-, 6-, and 7-membered
rings via intramolecular hydroamination, operates in the pre-
sence of Lewis basic functional groups, and is chemoselective for
the hydroamination product over a potentially competing
α-alkylation reaction.49ꢀ53 Thus, this system represents a unique
example of a broadly applicable group 4 catalyst for alkene hydro-
amination, stemming directly from its ability to transform sub-
strates that are inert to many, if not all reported group 4 systems.
Understanding the source of this unique reactivity is therefore
vital to further hydroamination catalyst development.
Figure 1. Tethered bis(ureate) precatalyst for hydroamination reactions.
(Scheme 1A and B) are accurate descriptions of the catalytic
process. Instead, the evidence presented here points to a concerted
CꢀH, CꢀN bond-forming event as the turnover-limiting step, as
has been suggested within the field previously.48,54 Importantly, a
similar mechanistic model has been recently reported by Sadow
and co-workers for an oxazolinylborate-supported zirconium
catalyst.55 Their model involves proton-transfer from a protic
amido ligand to the alkene substrate in a six-membered transition
state (Scheme 1C). This previous work provides important
precedent for proton-assisted cyclohydroamination. In contrast
to this system, for which a zirconiumꢀimido complex is a key
In an effort to elucidate the mechanistic pathway by which 1
promotes hydroamination, we have undertaken a broad mechan-
istic investigation involving both a kinetic analysis of catalytic
reactions and an exploration of the stoichiometric chemistry of
the precatalyst. We have determined that neither the imido-based
cycloaddition mechanism nor the σ-bond insertion mechanism
15454
dx.doi.org/10.1021/ja202448b |J. Am. Chem. Soc. 2011, 133, 15453–15463

Journal of the American Chemical Society
ARTICLE
Figure 3. Plot of the consumption of 3 (c/c0, [3]0 = 0.75 M) versus time
(min) with added 2-methylpiperidine (0ꢀ6 equiv, up to 4.50 M); error
bars indicate (5%.
Figure 2. Plot of the consumption of aminoalkene substrates 2ꢀ4
(c/c0) versus time (min) through approximately four half-lives. Red data
points indicate exponential decay regime (not included in linear regres-
sion analysis). Error bars indicate (5%.
Table 1. Effect of Added 2-Methylpiperidine on the
Conversion of 3
intermediate, imido species do not appear to be catalytically
relevant in reactions involving precatalyst 1. This is an important
distinction that explains the ability of 1 to efficiently cyclize both
primary and secondary aminoalkenes.
’ RESULTS AND DISCUSSION
entry
[2-methylpiperidine] (M)a
kobs (10ꢀ2 M minꢀ1 a,b
)
Empirical Rate Laws. To determine an empirical rate law for
hydroamination reactions promoted by 1, a kinetic analysis of
aminoalkene cyclization has been undertaken. Figure 2 shows the
results of monitoring the conversion of typical aminoalkene
1
2
3
4
0.750
1.50
3.00
4.50
2.19 ( 0.02
2.25 ( 0.04
2.13 ( 0.04
2.05 ( 0.03
1
substrates (2ꢀ4) by H NMR spectroscopy relative to an in-
a First-order rate constants. b Errors estimated from regression analysis
of individual data points (single runs).
ternal standard (1,3,5-trimethoxybenzene) in deuterated benzene.
All of the reactions described herein have a high initial substrate
concentration of 0.750 M, in accord with the conditions de-
scribed previously for synthetic reactions.35 1H NMR signals
attributed to the protons of the ureate ancillary ligand can be
observed in the spectra obtained during kinetic monitoring;
unfortunately, these data are ambiguous as to the nature of the
catalyst resting state.
conversion, while that of 3 does not vary until ∼75% conversion.
Kinetic profiles of this type often result from competitive
inhibition by reaction products during the later stages of the
reaction. Such product inhibition kinetics have been observed
previously with rare-earth,62ꢀ66 actinide,25,57 and group 425
hydroamination precatalysts.
The plots in Figure 2 clearly show a linear decay of substrate
during the initial stage of the reaction (up to 65ꢀ75% conver-
sion, best-fit line), indicative of zero-order dependence on amino-
alkene concentration. Notably, cyclization to form 6-membered
products 6 and 7 proceeds faster than that for the conversion of 2
to 5. Finally, the conversion of the N-methylated secondary
aminoalkene 4 occurs at a 5-fold higher rate than the cyclization
of the analogous primary aminoalkene. This kind of catalytic
behavior appears to be in contrast to other neutral d0-metal-
based catalysts, which have been reported to show higher activity
for primary aminoalkenes over secondary substrates.29,43,54,56ꢀ61
Following the consumption of 2ꢀ4 reveals that at higher
substrate conversion the plot becomes nonlinear (red points),
exhibiting exponential decay consistent with first-order con-
sumption of substrate. Furthermore, the onset of the nonlinear
regime is dependent on the identity of the substrate: the rate of
consumption of substrates 2 and 4 changes after ∼65%
To investigate the possibility of competitive product inhibi-
tion during hydroamination catalysis, a series of reactions have
been carried out with varying amounts of added 2-methylpi-
peridine. This specific inhibitor was chosen to mimic the
structure of the hydroamination product 6. Figure 3 illustrates
the kinetic profiles of the conversion of 3 (105 °C) in the
presence of 0ꢀ6 equiv (up to 4.50 M) of 2-methylpiperidine.
Two striking features are observed for these reactions. First,
there is no significant change in rate between the experiments
with added inhibitor, despite the large excess of piperidine
added to the latter reactions. Indeed, the difference in the
pseudo-first-order kobs across this 6-fold concentration range is
less than 5% (Figure 3, Table 1). Thus, the reaction kinetics are
largely unaffected by increasing the concentration of 2-methyl-
piperidine, meaning that this product-mimic is not a compe-
titive inhibitor. More importantly, these results suggest that
15455
dx.doi.org/10.1021/ja202448b |J. Am. Chem. Soc. 2011, 133, 15453–15463

Journal of the American Chemical Society
ARTICLE
Table 2. Aminoalkene Cyclization at Varying Precatalyst
Concentrations
b
entry
[precatalyst] (M)a (mol %)
kobs (10ꢀ3 minꢀ1
)
1
2
3
4
5
6
7
8
0.0510 (6.8)
0.0578 (7.7)
0.0688 (9.2)
0.0750 (10.0)
0.104 (13.9)
0.125 (16.7)
0.127 (16.9)
0.142 (18.9)
6.72 ( 0.12
7.81 ( 0.15
9.14 ( 0.10
10.6 ( 0.2
16.4 ( 0.2
19.9 ( 0.4
20.0 ( 0.5
23.2 ( 0.4
Figure 4. Plot of zero-order kobs (minꢀ1) versus [precatalyst] (M),
showing first-order dependence up to ∼0.15 M. Error bars from esti-
mated errors tabulated in Table 2.
solution-phase characterization of the precatalyst35,68 also reveal
mononuclear complexes. Notably, hydroamination reactions
using Cp2TiMe2 as the precatalyst have a nonlinear catalyst
dependence with [Cp2TiMe2] > 0.02 M, due to deactivation by
reversible imido dimerization.14 The fact that there is first-order
behavior up to high concentrations of 1 (∼0.15 M) is note-
worthy, especially given the steric accessibility of the zirconium
center in this case, as this accessibility may be anticipated to
promote the formation of catalytically inactive dimers or higher
aggregates. To ascertain why catalyst deactivation by dimeriza-
tion is not kinetically relevant, model complexes derived from the
precatalyst (1) have been prepared and studied. Given that both
imido- and amido-based mechanisms are possible for the cycliza-
tion of primary aminoalkenes, studying the dynamic behavior of
imido and amido species may help to determine if one or both are
plausible catalytic intermediates.
Previously, we reported the synthesis of a bulky imido complex
(8) supported by the same tethered bis(ureate) ligand through
aminolysis of a dibenzyl complex.69 X-ray crystallography in-
dicates 8 is mononuclear in the solid state; however, this complex
is observed to dimerize in solution to form a 1:1 mixture of 8 and
9 (eq 3). It therefore follows that imido species present during
catalysis could be prone to this same dimerization. Notably, if the
mixture of 8 and 9 is treated with excess pyridine (∼10 equiv)
and heated (65 °C), only complex 8 is observed. While pyridine
is not present during catalytic reactions, this observation demon-
strates that neutral ligands, such as the amine groups of ami-
noalkene hydroamination substrates or heterocyclic products
formed, are capable of shifting the equilibrium in favor of the
mononuclear species.
a Precatalyst concentration determined at start of reaction by 1H NMR
spectroscopy through relative integration to internal standard, (5%.
b Errors estimated from regression analysis of individual data points
(single runs).
competitive inhibition by product is not the source of late-
reaction nonlinearity.
Second, the addition of 2-methylpiperidine does appear to
change the reaction order in substrate throughout the entire
course of the reaction. With no added piperidine, substrate
consumption is linear until ∼70% conversion (Figure 3, gray
line). In contrast, substrate consumption is exponential if the
piperidine is present. These data indicate that there is a change in
the turnover-limiting step induced by the addition of 2-methyl-
piperidine. It is therefore likely that a similar change in mechan-
ism, induced by high concentration of product (and concomitant
low concentration of substrate), is responsible for the observed
shift from zero- to first-order kinetics illustrated in Figure 2. Our
rationale for this behavior will be discussed in terms of our overall
mechanistic proposal in a subsequent section.
The above results indicate that there are two kinetic regimes
for the conversion of aminoalkenes 2ꢀ4, dependent upon the
extent of reaction, and that the hydroamination product is not a
competitive inhibitor of the reaction. To establish empirical rate
laws for these two stages of the reaction, the rate of conversion of
3 was measured for a series of precatalyst concentrations
(Table 2). A plot of the zero-order kobs values obtained for the
primary stage of the reaction versus [1] shows a linear correlation
in the range [1] = 0.05ꢀ0.15 M (Figure 4).67 Thus, there is a
first-order dependence on catalyst concentration, giving the
empirical rate law in eq 1 for the first stage of the hydroamination
reaction. At high conversion, the empirical rate law in eq 2 is
operative, with first-order dependence on both catalyst and
substrate.
primary stage : rate ¼ k½catalystꢁ
ð1Þ
ð2Þ
secondary stage : rate ¼ k0½substrateꢁ½catalystꢁ
The Nature of the Active Catalyst. The first-order depen-
dence on precatalyst concentration discussed in the previous section
suggests that a well-defined, mononuclear catalytic species is present
throughout the concentration range examined. Solid-state and
While the imido moiety of 8 is terminal, it is also very sterically
demanding, especially in comparison to the aminoalkene sub-
strates under investigation. To test if less bulky mononuclear
imido complexes are accessible, 2,6-dimethylaniline has been
15456
dx.doi.org/10.1021/ja202448b |J. Am. Chem. Soc. 2011, 133, 15453–15463

Journal of the American Chemical Society
ARTICLE
Scheme 2. Synthesis of Pyridine Adduct 10 and Dimeric Imido Complex 11
used to prepare an imido complex via the bis(amido) pyridine
adduct 10. A reaction between 10, the aniline derivative, and
excess pyridine at 65 °C yields a poorly soluble, pale yellow
compound in 64% yield (Scheme 2). NMR spectroscopy reveals
a ratio of ancillary ligand to 2,6-dimethylphenyl to pyridine of
1:1:1, consistent with an imido compound; however, the pre-
sence of only 1 equiv of pyridine suggests that the complex is a
dimer (11), as the more sterically protected mononuclear imido
complex 8 contains two stabilizing neutral ligands. Furthermore,
there is only one observed component in solutions of 11.
X-ray crystallography of 11 has confirmed the identity of the
complex as dimeric in the solid state (Figure 5). Each half of the
C2h-symmetric molecule is related by an inversion center. The
weakly bound pyridine ligands (ZrꢀN6: 2.557(1) Å) are in axial
positions at the seven-coordinate zirconium centers. The ureate
ligands deviate slightly from planarity, but to a much lesser
degree than observed previously for the 2,6-diisopropylphenyl
substituted dimer 9.69 In contrast to the monomer/dimer
equilibrium observed between 8 and 9, heating a mixture of 11
and excess pyridine to 65 °C does not result in any change to the
spectral features. This indicates that, even in the presence of excess
neutral ligands, 11 remains dimeric, or the monomer/dimer equi-
librium heavily favors the dimer. Presumably, less bulky imido
groups, such as those that would be formed with aminoalkene
substrate, would favor dimerization to an even greater degree.
Notably, treatment of 8, 9, or 11 with a variety of alkynes and
alkenes does not result in observable [2 + 2] cycloaddition
reactions, the key step in imido-based hydroamination mechan-
isms (Scheme 1A).3 The combination of facile dimerization
observed in the formation of 11, the fact that this dimer appears
to persist in solution even in the presence of an excess of neutral
ligand, the lack of evidence for dimerization or aggregation in
reactions with precatalyst 1 (Table 2, Figure 4), and the un-
reactive nature of the ZrꢀN imido bonds in complexes 8, 9, and
11 toward stoichiometric cycloaddition lead us to rule out imido
species as catalytically relevant intermediates for this system.
If imido species are not catalytically viable, then presumably
the catalyst derived from 1 operates exclusively via amido-based
intermediates that are structurally similar to precatalyst 1 and
pyridine adduct 10. Loss of the neutral ligand from either of these
complexes would result in a coordinatively unsaturated species
that could accommodate alkene coordination prior to insertion.
Six-coordinate alkyl complexes supported by the same tethered
bis(ureate) ligand have been isolated previously.69 Synthesis of
an analogous six-coordinate bis(amido) species has been at-
tempted by subjecting solid 10 to high vacuum for 48 h (eq 4).
Compound 10 was chosen for this purpose because the
Figure 5. ORTEP representation of the molecular structure of 11
(ellipsoids plotted at 50% probability, hydrogens omitted) with selected
bond lengths (Å), bond, and torsion angles (deg): ZrꢀN1, 2.2455(9);
ZrꢀO1, 2.2373(8); ZrꢀN5, 2.1653(9); ZrꢀN5*, 2.0338(9); ZrꢀN6,
2.557(1); C1ꢀN1, 1.321(1); C1ꢀO1, 1.291(1); C1ꢀN2, 1.361(1);
N1ꢀZrꢀO1, 57.91(3); N1ꢀZrꢀN3, 74.21(4); N5ꢀZrꢀN6, 87.17(4);
N5ꢀZrꢀN6*, 166.70(4); N5ꢀZrꢀN5*, 79.84(4); ZrꢀN6ꢀZr*, 100.16-
(4); N1ꢀC1ꢀN2ꢀC5, 15.2(2); N3ꢀC8ꢀN4ꢀC12, 42.6(2).
disappearance of pyridine can be easily monitored by NMR
spectroscopy; however, similar treatment of 1 gives the same
product 12. The solid 10 changes from yellow to colorless during
this period, and this new compound (12) has poor solubility in
common organic solvents. The NMR spectra for product 12 are
nearly identical to those of both 1 and 10.
Compound 12 can be recrystallized from toluene for an X-ray
crystallographic study (Figure 6), which confirms the structure as
an amido-bridged dimer. This complex is structurally similar to
the imido dimer 11, exhibiting analogous C2h symmetry. Each
zirconium is seven-coordinate, as in the mononuclear parent
complexes 1 and 10. Two amido ligands are bound exclusively to
one zirconium, with the remaining two bridging between the
metal centers. This results in much longer ZrꢀN bonds for these
15457
dx.doi.org/10.1021/ja202448b |J. Am. Chem. Soc. 2011, 133, 15453–15463

Journal of the American Chemical Society
ARTICLE
Figure 7. ORTEP representation of the molecular structure of 13
(ellipsoids plotted at 50% probability, hydrogens omitted) with selected
bond lengths (Å), bond, and torsion angles (deg): ZrꢀN1, 2.195(2);
ZrꢀO1, 2.180(2); ZrꢀN5, 2.171(2); ZrꢀN6, 2.397(2); ZrꢀCl,
2.596(1); C1ꢀN1, 1.353(3); C1ꢀO1, 1.309(2); C1ꢀN2, 1.322(3);
N1ꢀZrꢀO1, 59.73(6); N1ꢀZrꢀN3, 78.33(6); ClꢀZrꢀN5, 167.10(5);
sum of angles about N5, 359.6(6); sum of angles about N6, 341.0(4).
Figure 6. ORTEP representation of the molecular structure of 12
(ellipsoids plotted at 50% probability, hydrogens omitted) with selected
bond lengths (Å), bond, and torsion angles (deg): ZrꢀN1, 2.241(1);
ZrꢀO1, 2.226(1); ZrꢀN5, 2.108(1); ZrꢀN6, 2.323(1); ZrꢀN6*,
2.374(1); C1ꢀN1, 1.312(2); C1ꢀO2, 1.296(2); C1ꢀN3, 1.373(2);
N1ꢀZrꢀO2, 58.37(4); N1ꢀZrꢀN2, 74.43(4); N5ꢀZrꢀN6, 93.69(4);
N5ꢀZrꢀN6*, 170.01(4); N6ꢀZrꢀN6*, 76.34(4); ZrꢀN6ꢀZr*, 103.66-
(4); N1ꢀC1ꢀN3ꢀC8, 44.1(2).
equatorial plane of a pentagonal bipyramid. However, there is
one crucial difference: the anionic chloro and amido ligands are
axial, and therefore trans to one another, whereas in precatalyst 1,
the two amido ligands are oriented cis, forcing one dimethyla-
mido into an equatorial position. Previously reported results are
consistent with the formal insertion of internal alkynes into the
longer equatorial zirconiumꢀamido bond;68 we therefore antici-
pate that an intramolecular alkene insertion should also occur
preferentially into an equatorial ZrꢀNamido bond. Given that the
sole amido ligand of 13 adopts an axial configuration, an amido
ligand derived from an aminoalkene substrate will likewise have a
thermodynamic preference for an axial coordination site, which
may impact catalytic activity. Notably, the axial amido ligand in
13 has a remarkably long ZrꢀN5 distance of 2.171(2) Å, which is
even longer than the equatorial ZrꢀNamido bond of 1 (2.135(1)
Å). Although the ZrꢀN5 length may suggest diminished multi-
ple bond character, the geometry about N5 is planar, indicative
of sp2-hybridization and therefore some degree of Zrꢀamido
π-bonding.
In contrast to the aforementioned zirconium constrained
geometry systems,25 use of 13 as a precatalyst for the hydro-
amination of aminoalkenes 2, 3, and 4 leads to a dramatic
decrease in activity relative to the bis(amido) 1. Only minimal
conversion to product is observed under the reaction conditions,
even for the secondary aminoalkene 4 (Scheme 3). Clearly, a
weak Zrꢀamido bond is not a sufficient requirement for efficient
hydroamination with this class of catalyst. In addition, all
attempts to affect the stoichiometric insertion of internal alkynes
into the axial ZrꢀNamido bond of 13 using previously established
reaction conditions68 have been unsuccessful. These observa-
tions are consistent with the notion that an equatorial amido
ligand is required for high catalytic activity with this class of
catalyst.
bridging amido groups (2.323(1) Å) relative to the terminal
amido ligands (2.108(1) Å).
Attempts to observe a solution-phase equilibrium between 12
and a putative six-coordinate complex by variable-temperature
NMR spectroscopy have been unsuccessful; however, the treat-
ment of 12 with excess pyridine immediately regenerates 10 as
judged by the distinct color change from colorless to yellow, and
by the appearance of a broad pyridine signal at δ 8.61 ppm in the
1H NMR spectrum, which is shifted slightly downfield from free
pyridine (∼8.5 ppm). These results suggest that in the presence
of neutral donors, as is the case during catalysis, a neutral ligand
stabilized seven-coordinate species is favored. Importantly, 12
can be used in catalytic hydroamination reactions with no decrease
in reactivity relative to precatalyst 1, verifying that catalyst de-
activation via dimerization is not kinetically relevant. This is in
excellent agreement with the previously discussed kinetic data
that indicates first-order dependence on catalyst concentration
(Table 2, Figure 4). The fact that 1 can operate at very high
concentrations is important from a practical perspective as well,
because reactions involving precatalyst 1 can be carried out effici-
ently with minimal solvent, reducing overall waste generation.
If the active catalyst derived from 1 does not access an imido
intermediate, then presumably only one labile dimethylamido
ligand is required in the precatalyst. Previously, Stubbert and
Marks showed that a constrained-geometry-type zirconium hydro-
amination precatalyst with one dimethylamido ligand and one
chloride ligand performs better than its bis(dimethylamido)
analogue.25 Notably, this catalyst has been postulated to mediate
cyclohydroamination via a σ-bond insertionpathway(Scheme 1B).
Modifying the structure of precatalyst 1 in the same manner
would therefore provide an opportunity to test the viability of the
insertion mechanism, and also to potentially improve catalytic
activity. To investigate these questions, we have prepared an
analogue to precatalyst 1 that contains one chloride ligand and
one amido ligand (13). Complex 13 has been characterized by
X-ray crystallography, with the solid-state molecular structure
shown in Figure 7.
The Nature of the Turnover-Limiting Step. The mechanistic
evidence presented to this point definitively rules out an imido-
based catalytic cycle and is consistent with an insertion-type
catalytic cycle analogous to that proposed by Stubbert and Marks
for the CGCꢀZr systems (Scheme 1, B).25 Kinetic data support-
ing this interpretation include the turnover-limiting intra-
molecular insertion of the alkene into a zirconiumꢀamido bond,
and the zero-order substrate dependence observed using 1 as a
precatalyst. To further probe the nature of the turnover-limiting
At first glance, the complex is closely analogous to the
parent compound 1, with the bis(ureate) ligand occupying the
15458
dx.doi.org/10.1021/ja202448b |J. Am. Chem. Soc. 2011, 133, 15453–15463

Journal of the American Chemical Society
ARTICLE
Scheme 3. Low Activity of 13 When Used as a
Hydroamination Precatalyst
Notably, many lanthanide catalyst systems that are proposed
to operate via σ-bond insertion exhibit large primary kinetic
isotope effects (KIEs) when comparing N-proteo and N-deutero
substrates,48,65 meaning NꢀH bond cleavage is involved in the
turnover-limiting step. These results would seem to contradict
the notion of a simple intramolecular alkene insertion as being
turnover-limiting. Previously, Marks and co-workers have pro-
posed NꢀH bond involvement in the insertion transition state;48
however, there have been no further investigations reported on
this possibility with lanthanide catalysts. Recently, Sadow and co-
workers have presented an alternate mechanistic interpretation,
concerted CꢀH, CꢀN bond formation at magnesium- and
zirconium-based systems (Scheme 1C), including the observa-
tion of primary KIEs for NꢀH/D substrates.54,55
To gauge the effect of N-deuteration on reactions with 1 as the
precatalyst, the conversion of substrates d2-3 and d-4 has been
investigated. The cyclization of these aminoalkenes is much
slower than for their proteo counterparts 3 and 4, leading to
kH/kD values of 5.2 ( 0.4 for 3/d2-3 and 2.9 ( 0.2 for 4/d-4
(Figure 9). These primary KIEs are similar to the aforemen-
tioned values obtained with lanthanide catalysts (kH/kD
=
2.7ꢀ5.2),48,65 a magnesium catalyst (4.6),54 and a zirconium
catalyst (3.5).55 This implies that a protonolysis event is turn-
over-limiting, either as a discrete step or as part of a concerted
bond-forming process.
To investigate the possibility of stepwise insertion/protono-
lysis or concerted bond-formation, we endeavored to carry out
model insertion reactions into zirconiumꢀamido bonds sup-
ported by our tethered bis(ureate) ligand. As stated previously,
we have observed the insertion of internal alkynes into the
equatorial ZrꢀNMe2 ligand of 1.68 To model analogous alkene
insertion reactions, we have reacted 1 with an excess of a variety
of activated alkenes, including styrene derivatives and norbor-
nene. Unfortunately, after heating to 100ꢀ140 °C for several
days, these reactions do not result in appreciable alkene con-
sumption, and no discernible organometallic products have been
observed. This is in accord with the inability of 1 to catalyze all
attempted intermolecular alkene hydroamination reactions to
date. Instead, our attention moved to investigate intramolecular
insertion reactions, akin to those that would result in the
cyclization of aminoalkene substrates during successful catalytic
reactions.
Treatment of precatalyst 1 with either 1 or 2 equiv of the
N-methylated aminoalkene 4 at 65 °C for 4 h does not give an
insertion product, or even a substrate-derived bis(amido) com-
plex. Rather, complete cyclization of 4 to 7 is observed, with no
apparent change to the structure of the metal complex
(Scheme 4, top). The dimethylamine liberated through amino-
lysis to form the initial zirconiumꢀsubstrate complex presum-
ably reacts with any ZrꢀC bond that may be formed via insertion
to release the product 7. In a similar fashion, treating 1 with 2
equiv of the N-methylated-aminopentene substrate 14 results in
complete conversion to product 15 in 20 h without heating.
Remarkably, this result reveals that 1 is capable of mediating
intramolecular alkene hydroamination at room temperature.
Only two other zirconium systems are able to perform this
reaction under such mild conditions.37,55
Figure 8. Eyring plot in the temperature range 75ꢀ105 °C for the
cyclization of 3; activation parameters as indicated; kobs values represent
zero-order rate constants determined over ∼2 half-lives. Error on
activation parameters represents standard error determined from re-
gression analysis (see the Supporting Information for further details).
step, the temperature dependence of the cyclization rate of 3 has
been investigated by determining the zero-order kobs values at
10 °C intervals between 75 and 105 °C. The Eyring plot
generated from these data is shown in Figure 8, along with the
calculated activation parameters.
The ΔHq value of 20.1 ( 0.5 kcal/mol and ΔSq value of
ꢀ21 ( 1 eu are similar to those determined for analogous cyclohy-
droamination reactions involving rare-earth catalysts,48,62,63
actinide catalysts,25 and CGCꢀZr catalysts.25 All of these systems
are proposed to operate via σ-bond insertion, with the insertion
step being turnover-limiting. The large and negative ΔSq value
suggests a highly ordered transition state, as would be expected
from the proposed bicyclic transition state for alkene insertion
(Scheme 1B). It should be noted that many mechanistic studies
on imido-based catalytic cycles (Scheme 1A) have also yielded
similar activation parameters, so a definitive statement regarding
mechanism cannot be made solely from these data.
To prevent the generation of a proton source that results in
catalytic turnover to form the hydroamination products, the
dialkyl complex 16 can be used as an alternate starting material.
The absence of protic byproducts from aminolysis of 16 could
possibly enable observation of putative organometallic insertion
15459
dx.doi.org/10.1021/ja202448b |J. Am. Chem. Soc. 2011, 133, 15453–15463

Journal of the American Chemical Society
ARTICLE
Figure 9. Primary kinetic isotope effects observed in the cyclization of aminoalkenes 3 and 4.
Scheme 4. Attempted Observation of Intramolecular Insertion Reactions; Evidence for Proton-Assisted CꢀN Bond Formation
products; however, subjecting 16 to identical reaction conditions
to those described above leads exclusively to the substrate-derived
amido species (17,18) as judged by NMR spectroscopy, with the
alkene moieties intact (Scheme 4, bottom). As CꢀN bond
formation clearly occurs during stoichiometric reactions between
1 and 4, the failure to observe insertion products from the
reaction between 16 and 4 reveals that CꢀN bond formation in
the absence of a proton source is not favored. As has been
previously suggested,48,54,55 a proton source may be required for
CꢀN bond formation to occur. To further test this hypothesis,
intermediates 17,18 have been treated with 2 equiv of morpholine.
Upon heating to 65 °C for 4 h, complete conversion to product 4
is again observed. These experiments provide further evidence for
proton-assisted CꢀN bond formation during catalytic reactions.70
Mechanistic Proposal. On the basis of the accumulated data
regarding both thermodynamic and kinetic factors underlying
the reactivity of 1 and related compounds, a plausible catalytic
mechanism can be constructed. In a thermodynamic sense, it has
been established that in the presence of neutral donors, zirconium
dichloro,71 dialkyl,69 and bis(amido)35 species supported by the
tethered bis(ureate) ligand adopt a mononuclear, seven-coordi-
nate, base-stabilized bonding motif. In contrast, zirconium imido
fragments supported by this same ligand exhibit facile dimeriza-
tion, even in the presence of excess donor, such as pyridine
(Scheme 2, and eq 3); a mononuclear complex (8) can only be
accessed by using a very large imido N-substituent and a
significant excess of pyridine.69 Furthermore, these imido com-
pounds do not readily undergo [2 + 2] cycloaddition with
alkynes to give metallacyclic products. These facts suggest that
an imido-based, [2 + 2] cycloaddition mechanism3 for hydro-
amination involving 1 is unlikely.
formation of enamine side-products shows that hydroamination
results from these reactions. Aminolysis of these metallacycles
similarly forms enamines, thus establishing the feasibility of an
insertion/protonolysis sequence for an alkyne hydroamination
catalytic cycle with zirconium.68 However, the results of at-
tempted intramolecular insertions of alkenes into ZrꢀN bonds
suggest that an amine proton source is required for CꢀN bond
formation to occur with alkenes, in contrast to the previously
observed alkyne insertion chemistry (Scheme 4).
Kinetically, the cyclization of aminoalkenes 2ꢀ4 has a zero-
order dependence on substrate concentration, with a switch to
first-order dependence occurring at high conversion (Figure 2),
or with the addition of a product analogue (2-methylpiperidine,
Figure 3). The cyclization of 3 is first order in catalyst up to [1] ≈
0.15 M, revealing that the catalytically active species is well-
defined, and not susceptible to deactivation at higher loadings.
Given that both primary (2 and 3) and secondary (4) aminoalk-
enes exhibit similar kinetic profiles, it is reasonable that these
substrates engage in the same mechanistic pathway, with the re-
duced reactivity observed for the primary amine substrate being
attributed to some off-cycle imido complex formation. In addi-
tion, a primary KIE for the cyclization of N-deuterated substrates
d2-3, d-4 indicates that NꢀH/D bond cleavage and/or CꢀH/D
bond formation must be involved in the turnover-limiting step,
consistent with a concerted CꢀH, CꢀN bond-forming
process.54,55 Finally, the ΔSq for the cyclization of 3 is large
and negative, providing evidence for a highly ordered
transition state.
A proposed mechanism to account for all of these experi-
mental observations is pictured in Scheme 5. This catalytic cycle
is illustrated for the conversion of 3 to 6; however, it should be
equally accurate for the cyclization of other primary and secondary
aminoalkenes. Initial aminolysis of the precatalyst by excess
In contrast, σ-bond insertion of alkynes into the equatorial
amido ligand of 1 has been observed previously, and concomitant
15460
dx.doi.org/10.1021/ja202448b |J. Am. Chem. Soc. 2011, 133, 15453–15463

Journal of the American Chemical Society
ARTICLE
Scheme 5. Proposed Catalytic Cycle for Intramolecular Alkene Hydroamination Using 1, with Derived Differential Rate Law
(Steady-State Approximation for [A], [B], and [C])
substrate liberates dimethylamine to generate a structurally
analogous bis(amido) species (A), which has a third substrate
equivalent bound in an axial position. From this seven-coordinate
state, intramolecular hydroamination takes place via irreversible
CꢀH, CꢀN bond formation (k1) through a highly ordered
transition state (TS) involving a nucleophilic equatorial ZrꢀN
bond and an acidic NꢀH from the axially bound amine ligand.
One interpretation of this concerted process is analogous to a
transition state previously suggested by Marks and co-workers
for rare-earth catalysts.48 In this sense, a formal σ-bond insertion
occurs via a four-membered transition state, involving the equa-
torial amido ligand and the alkene. Simultaneous protonation can
also occur through a four-membered transition state between the
NꢀH bond of the axial neutral amine ligand and the partial
ZrꢀC bond. Alternatively, as suggested by Sadow and co-workers,
this process can be viewed not as two connected four membered
transition states, but as a single, chairlike six-membered transition
state with no direct interaction between zirconium and the alkene
(pictured in Scheme 1C).54,55 A combination of both of these
mechanistic extremes must also be considered. In either case, this
concerted process would account for the large negative ΔSq, the
primary KIEs, and the observed stoichiometric proton-assisted
CꢀN bond formation represented in Scheme 4.70
Once cyclization has occurred at the equatorial ZrꢀN bond,
the piperidine 6 would be neutrally bound in an equatorial
position, forcing an electronically disfavored trans arrangement
of two π-donating amido ligands (B). Rapid dissociation of 6
(k2) would then generate a six-coordinate intermediate (C). This
coordinatively unsaturated state can bind another equivalent of
substrate (k3) to regenerate a seven-coordinate species (A). This
substrate binding is reversible (kꢀ3), leading to an equilibrium
between catalytic intermediates A and C. On the basis of the
observed preference for the binding of neutral donors, it can be
deduced that this equilibrium favors the seven-coordinate species
(i.e., k3 > kꢀ3, or Keq > 1). Applying the steady-state approxima-
tion for the concentrations of all catalytic intermediates, the
differential rate law at the bottom of Scheme 5 can be derived. By
making several reasonable assumptions regarding the relative
magnitudes of the rate constants, this complex rate law can be
simplified into two regimes that account for all of the kinetic
behavior observed and are identical to the empirical rate laws
derived from kinetic data (eqs 1 and 2).
15461
dx.doi.org/10.1021/ja202448b |J. Am. Chem. Soc. 2011, 133, 15453–15463

Journal of the American Chemical Society
ARTICLE
In the first situation (primary stage of the reaction), neutral
ligand association/dissociation is facile, and the rate constants k2,
’ CONCLUSIONS
Through a combination of elucidating stoichiometric reactiv-
ity patterns, and analyzing the kinetic profiles of catalytic reac-
tions, a stereoelectronic proposal for the reactivity of zirconium
precatalyst 1 has been presented. The ability of the catalytic
species derived from 1 to efficiently cyclize both primary and
secondary aminoalkenes stems directly from the steric and
electronic environment imposed by the ancillary ligand. Use of
this tethered bis(ureate) ligand provides steric accessibility to the
metal center, enabling the favorable formation of seven-coordi-
nate bis(amido) complexes that would facilitate a concerted
CꢀH, CꢀN bond-forming event. Importantly, this renders
precatalyst 1 more active for the cyclization of secondary amino-
alkenes than primary substrates, behavior that is quite unique
among neutral d0-metal catalyst systems.29,43,54,56ꢀ61
The kinetic profile of aminoalkene cyclization indicates that
the reaction proceeds through a unimolecular, highly ordered
transition state. On the basis of the observation of proton-
assisted CꢀN bond formation in stoichiometric reactions, and
primary KIEs for cyclization of N-deuterated aminoalkenes, this
transition state is proposed to involve concerted CꢀH and CꢀN
bond formation. This can either be viewed as simultaneous σ-
bond insertion of the alkene into the ZrꢀN bond and proto-
nolysis of the partially formed ZrꢀC bond,48 or as simultaneous
nucleophilic attack by the amido ligand and protonation of the
alkene.54,55 These two interpretations differ in the role that
zirconium plays in the process; in the former, the metal is directly
involved in bond formation, while in the latter it is only indirectly
involved. Furthermore, competitive product inhibition does not
occur with this system, in contrast to many other d0-metal
catalyst systems.25,57,62ꢀ66 Instead, two different kinetic regimes
operate, depending on the relative concentrations of substrate
and product. This dual reaction manifold is due to an on-cycle
pre-equilibrium between a six-coordinate bis(amido) species and
a seven-coordinate bis(amido)amine intermediate that can un-
dergo concerted bond formation.
k3, and k are likely much larger than k1. In other words, the
ꢀ3
concerted bond formation (k1) is the turnover-limiting step in
the catalytic sequence. Because k1 , k2, k3, and kꢀ3, all of the
denominator terms containing k1 can be neglected. Furthermore,
as described in the previous paragraph, k3 > kꢀ3, meaning Keq > 1.
Finally, during the initial stage of the reaction, the concentration
of substrate is high; therefore, k2k3[3] . k2kꢀ3. Applying these
assumptions gives the simplified rate law in eq 5, which is zero-
order in [3] and first-order in [1]0. This is the same as the
empirical rate law determined from kinetic experiments (eq 1).
Furthermore, the only rate constant that contributes to the
overall rate is k1, which corresponds to the concerted bond-
forming step. This same rate law can be derived in an intuitive
fashion by assuming k1 , k2, k3, and kꢀ3, and [B] + [C] , [A] ≈
[1]0. In this sense, catalytic turnover is dependent only on the
concentration of the species A, which is much higher than that of
B or C, and the rate constant of the bond formation (k1).
The second stage of hydroamination catalysis, which exhibits
first-order dependence on [3], occurs at high substrate conver-
sion (eq 2), or with the addition of excess 2-methylpiperidine
(Figure 3). To simplify the rate law from Scheme 5 to account for
this kinetic behavior, the following assumptions can be made.
First, k1 is still much smaller than the other rate constants, so the
denominator terms containing k1 can be excluded in the same
fashion as for eq 5. Second, at high conversion of 3, the substrate
concentration will be lower; therefore, k2kꢀ3 . k2k3[3]. This has
the effect of changing the turnover-limiting step to include both
the equilibrium between A and C (Keq) and the bond-forming
step (k1) (eq 6).
The insight regarding hydroamination catalysis gleaned as a
result of this investigation sheds light on specific steric and
electronic properties that can enable group 4 catalysts to perform
hydroamination effectively with both primary and secondary
amine substrates. With a mechanistic model in hand, ongoing work
to prepare group 4 systems with further improved reactivity and
selectivity patterns is targeted. The design of such systems, as well
as further mechanistic and computational work exploring the pro-
posed concerted bond formation pathway, is currently underway.
Importantly, this second rate law accounts for both the
observed first-order dependence on [3] and the observed zero-
order dependence on added 2-methylpiperidine (Table 1,
Figure 3). Because of the structural similarity between 2-methyl-
piperidine and the hydroamination product 6, one can infer that
the reaction should also be zero-order in [6]. If 6 were acting as a
competitive inhibitor, the reaction should be inverse-order in
product (or added 2-methylpiperidine). Notably, the addition of
2-methylpiperidine is observed to change the reaction to first-
order in substrate throughout the course of the reaction, not just
when the concentration of 3 is low. This suggests that 2-methyl-
piperidine, and by analogy the hydroamination product, is
involved in the equilibrium between A and C. As the concentra-
tion of product increases, or if one or more equivalents of
2-methylpiperidine is added, the amido ligands on A and C will
increasingly be derived from bulky piperidines (product/in-
hibitor) rather than acyclic amines (substrate). The presence
of these bulky ligands would help to stabilize the six-coordinate
species C, thus decreasing the equilibrium constant Keq. If this
decrease is large enough, then the term k2k3[3] would be much
less than k2kꢀ3 (i.e., k2kꢀ3 . k2k3[3]), leading to the same rate
law given in eq 6. Work is currently underway to probe this
proposed equilibrium between complexes similar to the inter-
mediates A and C to assess this hypothesis.
’ ASSOCIATED CONTENT
S
Supporting Information. Complete experimental details,
b
kinetic plots, and CIFs for solid-state molecular structures. This
material is available free of charge via the Internet at http://pubs.
acs.org.
’ AUTHOR INFORMATION
Corresponding Author
schafer@chem.ubc.ca
’ ACKNOWLEDGMENT
We acknowledge funding from NSERC (CGSD to D.C.L.,
CRD Grant to L.L.S.), Imperial Oil Ltd., and Boehringer
15462
dx.doi.org/10.1021/ja202448b |J. Am. Chem. Soc. 2011, 133, 15453–15463

Journal of the American Chemical Society
ARTICLE
Ingelheim (Canada) Ltd. in support of this work. R.H.P. thanks
the Government of Canada (CBIE) for a Commonwealth post-
doctoral fellowship.
(40) Zi, G.; Zhang, F.; Liu, X.; Ai, L.; Song, H. J. Organomet. Chem.
2010, 695, 730.
(41) Xiang, L.; Zhang, F.; Zhang, J.; Song, H.; Zi, G. Inorg. Chem.
Commun. 2010, 13, 666.
(42) Wang, Q.; Song, H.; Zi, G. J. Organomet. Chem. 2010, 695,
1583.
(43) Hu, Y.-C.; Liang, C.-F.; Tsai, J.-H.; Yap, G. P. A.; Chang, Y.-T.;
Ong, T.-G. Organometallics 2010, 29, 3357.
(44) Allan, L. E. N.; Clarkson, G. J.; Fox, D. J.; Gott, A. L.; Scott, P.
J. Am. Chem. Soc. 2010, 132, 15308.
’ REFERENCES
(1) M€uller, T. E.; Beller, M. Chem. Rev. 1998, 98, 675.
(2) M€uller, T. E.; Hultzsch, K. C.; Yus, M.; Foubelo, F.; Tada, M.
Chem. Rev. 2008, 108, 3795.
(45) Janssen, T.; Severin, R.; Diekmann, M.; Friedemann, M.;
Haase, D.; Saak, W.; Doye, S.; Beckhaus, R. Organometallics 2010,
29, 1806.
(46) Antunes, M. A.; Munha, R. F.; Alves, L. G.; Schafer, L. L.;
Martins, A. M. J. Organomet. Chem. 2011, 696, 2.
(47) Ayinla, R. O.; Gibson, T.; Schafer, L. L. J. Organomet. Chem.
2011, 696, 50.
(3) Walsh, P. J.; Baranger, A. M.; Bergman, R. G. J. Am. Chem. Soc.
1992, 114, 1708.
(4) McGrane, P. L.; Jensen, M.; Livinghouse, T. J. Am. Chem. Soc.
1992, 114, 5459.
(5) McGrane, P. L.; Livinghouse, T. J. Am. Chem. Soc. 1993, 115,
11485.
(6) Duncan, A. P.; Bergman, R. G. Chem. Rec. 2002, 2, 431.
(7) Bytschkov, I.; Doye, S. Eur. J. Org. Chem. 2003, 935.
(8) Doye, S. Synlett 2004, 1653.
(48) Gagnꢀe, M. R.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1992,
114, 275.
(49) Bexrud, J. A.; Eisenberger, P.; Leitch, D. C.; Payne, P. R.;
Schafer, L. L. J. Am. Chem. Soc. 2009, 131, 2116.
(50) Roesky, P. W. Angew. Chem., Int. Ed. 2009, 48, 4892.
(51) Kubiak, R.; Prochnow, I.; Doye, S. Angew. Chem., Int. Ed. 2009,
48, 1153.
(52) Prochnow, I.; Kubiak, R.; Frey, O. N.; Beckhaus, R.; Doye, S.
ChemCatChem 2009, 1, 162.
(53) Kubiak, R.; Prochnow, I.; Doye, S. Angew. Chem., Int. Ed. 2010,
49, 2626.
(9) Odom, A. L. Dalton Trans. 2005, 225.
(10) Severin, R.; Doye, S. Chem. Soc. Rev. 2007, 36, 1407.
(11) Lee, A. V.; Schafer, L. L. Eur. J. Inorg. Chem. 2007, 2243.
(12) Eisenberger, P.; Schafer, L. L. Pure Appl. Chem. 2010, 82, 1503.
(13) Zi, G. J. Organomet. Chem. 2011, 696, 68.
(14) Pohlki, F.; Doye, S. Angew. Chem., Int. Ed. 2001, 40, 2305.
(15) Johnson, J. S.; Bergman, R. G. J. Am. Chem. Soc. 2001, 123,
2923.
(16) Knight, P. D.; Munslow, I.; O’Shaughnessy, P. N.; Scott, P.
Chem. Commun. 2004, 894.
(17) Gribkov, D. V.; Hultzsch, K. C. Angew. Chem., Int. Ed. 2004, 43,
5542.
(54) Dunne, J. F.; Fulton, D. B.; Ellern, A.; Sadow, A. D. J. Am. Chem.
Soc. 2010, 132, 17680.
(55) Manna, K.; Xu, S.; Sadow, A. D. Angew. Chem., Int. Ed. 2011,
50, 1865.
(18) Bexrud, J. A.; Beard, J. D.; Leitch, D. C.; Schafer, L. L. Org. Lett.
2005, 7, 1959.
(19) Kim, H.; Lee, P. H.; Livinghouse, T. Chem. Commun. 2005,
(56) Hong, S.; Marks, T. J. J. Am. Chem. Soc. 2002, 124, 7886.
(57) Stubbert, B. D.; Marks, T. J. J. Am. Chem. Soc. 2007, 129, 4253.
(58) Yuen, H. F.; Marks, T. J. Organometallics 2008, 27, 155.
(59) Yuen, H. F.; Marks, T. J. Organometallics 2009, 28, 2423.
(60) Stanlake, L. J. E.; Schafer, L. L. Organometallics 2009, 28, 3990.
(61) Broderick, E. M.; Gutzwiller, N. P.; Diaconescu, P. L. Organo-
metallics 2010, 29, 3242.
5205.
(20) Thomson, R. K.; Bexrud, J. A.; Schafer, L. L. Organometallics
2006, 25, 4069.
(21) Watson, D. A.; Chiu, M.; Bergman, R. G. Organometallics 2006,
25, 4731.
(62) Li, Y.; Marks, T. J. J. Am. Chem. Soc. 1996, 118, 9295.
(63) Hong, S.; Kawaoka, A. M.; Marks, T. J. J. Am. Chem. Soc. 2003,
125, 15878.
(64) Ryu, J. S.; Li, G. Y.; Marks, T. J. J. Am. Chem. Soc. 2003,
125, 12584.
(22) Lee, A. V.; Schafer, L. L. Organometallics 2006, 25, 5249.
(23) Kim, H.; Kim, Y. K.; Shim, J. H.; Kim, M.; Han, M.; Livinghouse,
T.; Lee, P. H. Adv. Synth. Catal. 2006, 348, 2609.
(24) Wood, M. C.; Leitch, D. C.; Yeung, C. S.; Kozak, J. A.; Schafer,
L. L. Angew. Chem., Int. Ed. 2007, 46, 354.
(65) Gribkov, D. V.; Hultzsch, K. C.; Hampel, F. J. Am. Chem. Soc.
2006, 128, 3748.
(25) Stubbert, B. D.; Marks, T. J. J. Am. Chem. Soc. 2007, 129, 6149.
(26) Marcsekova, K.; Loos, C.; Rominger, F.; Doye, S. Synlett
2007, 2564.
(27) Gott, A. L.; Clarke, A. J.; Clarkson, G. J.; Scott, P. Organo-
metallics 2007, 26, 1729.
(28) Gott, A. L.; Clarke, A. J.; Clarkson, G. J.; Scott, P. Chem.
Commun. 2008, 1422.
(29) Majumder, S.; Odom, A. L. Organometallics 2008, 27, 1174.
(30) Cho, J.; Hollis, K. T.; Helgert, T. R.; Valente, E. J. Chem.
Commun. 2008, 5001.
(31) M€uller, C.; Saak, W.; Doye, S. Eur. J. Org. Chem. 2008, 2731.
(32) Graebe, K.; Pohlki, F.; Doye, S. Eur. J. Org. Chem. 2008, 4815.
(33) M€uller, C.; Koch, R.; Doye, S. Chem.-Eur. J. 2008, 10430.
(34) Zi, G.; Liu, X.; Xiang, L.; Song, H. Organometallics 2009,
28, 1127.
(66) Giardello, M. A.; Conticello, V. P.; Brard, L.; Gagnꢀe, M. R.;
Marks, T. J. J. Am. Chem. Soc. 1994, 116, 10241.
(67) The plot of kobs versus [precatalyst] has a nonzero intercept due
to the moisture sensitivity of the catalyst and the impracticality of
completely removing trace water from the aminoalkene substrate.
(68) Leitch, D. C.; Turner, C. S.; Schafer, L. L. Angew. Chem., Int. Ed.
2010, 49, 6475.
(69) Leitch, D. C.; Schafer, L. L. Organometallics 2010, 29, 5162.
(70) It should be noted that another possible pathway is consistent
with the observed reactivity and kinetic profiles, fast reversible alkene
insertion followed by slow intramolecular proton-transfer between a
coordinated amine substrate and the ZrꢀC bond; however, this process
would require a β-amide elimination reaction to occur as the reverse of
alkene insertion. Observation of such a reaction has never been reported
for group 4 or rare earth complexes. This pathway has previously been
ruled out for lanthanide systems (see ref 48).
(35) Leitch, D. C.; Payne, P. R.; Dunbar, C. R.; Schafer, L. L. J. Am.
Chem. Soc. 2009, 131, 18246.
(36) Reznichenko, A. L.; Hultzsch, K. C. Organometallics 2010, 29,
24.
(37) Manna, K.; Ellern, A.; Sadow, A. D. Chem. Commun. 2010, 46,
339.
(71) Leitch, D. C.; Beard, J. D.; Thomson, R. K.; Wright, V. A.;
Patrick, B. O.; Schafer, L. L. Eur. J. Inorg. Chem. 2009, 2691.
(38) Bexrud, J. A.; Schafer, L. L. Dalton Trans. 2010, 39, 361.
(39) Zi, G.; Zhang, F.; Xiang, L.; Chen, Y.; Fang, W.; Song, H. Dalton
Trans. 2010, 39, 4048.
15463
dx.doi.org/10.1021/ja202448b |J. Am. Chem. Soc. 2011, 133, 15453–15463