A zwitterionic zirconium complex that catalyzes hydroamination of aminoalkenes at room temperature

Article abstract of DOI:10.1039/b918989k

The zwitterionic cyclopentadienyl-bis(2-oxazolinyl)borate diamidozirconium(iv) complex is a precatalyst for the cyclization of aminoalkenes to five- and six-membered rings under mild conditions.

Full text of DOI:10.1039/b918989k

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
A zwitterionic zirconium complex that catalyzes hydroamination
of aminoalkenes at room temperaturew
Kuntal Manna, Arkady Ellern and Aaron D. Sadow*
Received (in Berkeley, CA, USA) 14th September 2009, Accepted 3rd November 2009
First published as an Advance Article on the web 23rd November 2009
DOI: 10.1039/b918989k
The zwitterionic cyclopentadienyl-bis(2-oxazolinyl)borate
diamidozirconium(IV) complex is a precatalyst for the cyclization
of aminoalkenes to five- and six-membered rings under mild
conditions.
(ꢀ8.1 ppm, acetonitrile-d3) that was characterized as a
four-coordinate acetonitrile adduct.13 This material was not
isolated but instead was purified by column chromatography
on silica gel (hexane : EtOAc : Et3N = 12 : 7 : 1) to give
H[PhB(C5H5)(OxMe )2] (H[1]).
2
The isolated H[1] is also a mixture of three species according
to 11B NMR spectroscopy (ꢀ15.3, ꢀ15.6, and ꢀ16.0 ppm).
Bent-sandwich Group 4 compounds are among the most
versatile of catalyst classes, mediating a range of processes
involving insertion and s-bond metathesis steps.1 Much of
their chemistry depends on three mutually adjacent frontier
orbitals located in the wedge of the 14-electron Cp2M-fragment
(Cp = C5H5).2 Modified ligands, such as the ansa-bis(cyclo-
pentadienyl)3 or mixed cyclopentadienyl-amido ligands,4 or
the incorporation of a borate in the ligand periphery,5 have
enhanced reactivity. Combining these concepts, a borate-
containing, mixed Cp–Ln (Ln = donor group) ligand could
provide reactive compounds by reducing the effective electron
count while maintaining the critical mutually-cis configuration
of orbitals. Herein, we report zwitterionic zirconium and
hafnium diamido compounds containing such a dianionic
mixed cyclopentadienyl-bis(oxazolinyl)borato ligand that are
reactive precatalysts for aminoalkene hydroamination.
(1)
However, only one 15N NMR resonance was observed at
ꢀ172 ppm in a 1H–15N HBQC experiment, 45 ppm upfield
of free 4,4-dimethyl-2-oxazoline (ꢀ127 ppm). In addition, only
one nCN was observed in the infrared region at 1589 cmꢀ1
,
and free 4,4-dimethyl-2-oxazoline was not detected
(nCN = 1630 cmꢀ1).14 For comparison, in hydrogen tris-
(4,4-dimethyl-2-oxazolinyl)phenylborate,15 two nCN were
observed at 1627 and 1594 cmꢀ1 corresponding to non-
protonated and protonated oxazoline groups, respectively.
This spectroscopic evidence is consistent with both oxazoline
groups interacting with a proton in H[1].
Two general mechanisms have been proposed for intra-
molecular olefin hydroamination catalyzed by Group 4 metal
complexes. In a mechanism favored for cationic zirconocene
alkyls and constrained-geometry zirconium catalysts,6,7
turnover-limiting olefin insertion into a Zr–N bond gives the
new C–N bond. The alternative imido mechanism,8 proposed
for bis(amidate) and phenolate-oxazoline piano-stool-type
zirconium compounds,9,10 provides the C–N bond through a
[2p+2p] cycloaddition. Regardless of the pathway, Group
4-catalyzed hydroaminations typically require elevated
temperatures.11,12 In contrast, our zwitterionic zirconium
compound is active at room temperature.
A related ligand based on pyrazolyl (pz), rather than
oxazoline, is obtained from the redistribution of Sm(Cp)(Tp*)2
(Tp* = HB(3,5-Me2pz)3) as part of the samarium complex
{HB(C5H4)(Me2pz)2}Sm(k3–Tp*).16 Also, titanium(IV) diamides
coordinated by the dianionic bis(pyrrolyl)methane ligand
have Z5–Z1 ground state structures and are active as inter-
molecular alkyne hydroamination catalysts at 75 1C.17
Group 3 compounds containing the monoanionic ligand
[(C5H4)CPh2CH(Me2pz)2] (bpzcp)18 are isoelectronic with
the cyclopentadienyl-bis(oxazolinyl)borate Group 4 compounds
described here.
Reaction of Na[C5H5] and PhB(OxMe
)
(OxMe = 4,4-
2
2
2
dimethyl-2-oxazolinyl)13 for 12 h in THF at room temperature
The isomeric mixture of H[1] reacts with M(NMe2)4
(M = Zr, Hf) in benzene at room temperature to give
product formulated as Na[PhB(C5H5)(OxMe )2]
2
yields
a
(Na[1], eqn (1)). The 11B NMR spectrum of the crude
{PhB(C5H4)(OxMe )2}M(NMe2)2 in high isolated yield
2
material in acetonitrile-d3 contained three borate resonances
(Zr (2), 97.6%; Hf (3), 97.9%, eqn (2)). One set of diastereo-
topic oxazoline resonances was observed in the 1H NMR
spectrum (benzene-d6) consistent with a Cs-symmetric species.
(ꢀ15.4, ꢀ15.0, and ꢀ14.5 ppm). These resonances are
upfield relative to the starting material PhB(OxMe
)
2
2
Department of Chemistry, Iowa State University and AMES
Laboratory, 1605 Gilman Hall, Ames, IA, USA.
E-mail: sadow@iastate.edu; Fax: +1 515 294 0105;
Tel: +1 515 294 8069
(2)
w Electronic supplementary information (ESI) available: Procedures
and characterization data for compounds 1–4, representative kinetic
data, and X-ray crystallographic data for 4. CCDC 748394. For ESI
and crystallographic data in CIF or other electronic format see DOI:
10.1039/b918989k
These oxazoline resonances (in toluene-d8) were well resolved
and are equivalent from 300 K to 190 K. No fluxionality was
observed, suggesting that both oxazoline groups are
ꢁc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 339–341 | 339

View Article Online
Table 1 Hydroamination/cyclization of aminoalkenes
Conversion
Solvent Time/h (yield)a
Cat Substrate
Nt/hꢀ1b
1
2
3
4
5
2
2
2
3
2
2
2
2
2
2
C6D6
THF
THF
C6D6
C6D6
THF
C6D6
C6D6
THF
C6D6
11
11
23
24
11
11
17
33
11
11
90 (84)
68
87
29
85
0.82
0.62
0.38
0.12
0.77
0.65
0.35
0.18
0.26
0.84
6
7
72
59
8
9
10
62
29
92 (87)
Fig. 1 ORTEP diagram of {PhB(C5H4)(OxMe )2}Zr(NMe2)2THF (4).
2
11
12
2
2
C6D6
C6D6
11
48
87 (80)
0
0.79
0
The THF and C6H5 are drawn as stick bonds, and hydrogen atoms are
not shown. Zr1–C11, 2.518(2) A; Zr1–C12, 2.492(2) A; Zr1–C13,
2.528(2) A; Zr1–C14, 2.569(2) A; Zr1–C15, 2.547(2) A, Zr1–N3,
2.045(2) A; Zr1–N4, 2.062(1) A; Zr1–O3, 2.414(1) A. Cpcent–Zr1–N2,
99.41.
Conditions: 10% catalyst loading. Temperature, 23 1C.a Conversion
(%) was determined by 1H NMR spectroscopy. Isolated yield (%)
b
from reactions in C6H6. Since second-order kinetics were observed,
coordinated to the metal center. This configuration was
further supported by the infrared spectra, where only one
nCN band was detected (Zr: 1595; Hf: 1595 cmꢀ1, KBr).
A weak peak was observed for the hafnium derivative at
1549 cmꢀ1; note that only the symmetric mode is expected
Nt is time dependent and was calculated by {[substrate]0 ꢀ [substrate]t}/
{t ꢂ [catalyst]}.
time (i.e., Nt is time dependent). This rate dependence on time
contrasts typical lanthanide-mediated aminoalkene cyclization
kinetics, which show zero-order dependence on substrate
concentration. Attempts to increase conversion or improve
the rate by heating the reaction mixture were unsuccessful.
Finally, the formation of cyclohexyl-substituted piperidine
and pyrrolidene are catalyzed at similar rates (cf. Table 1,
entries 10 and 11). For comparison, the seemingly related
[(C5H4)SiMe2N-t-Bu]Zr(NMe2)2 catalyzes the cyclization of
to have significant intensity.
A peak corresponding to
uncoordinated oxazoline was not observed.
Compound 2 crystallizes as a THF adduct {PhB(C5H4)-
(OxMe )2}Zr(NMe2)2THF (4) with one oxazoline coordinated
2
to the zirconium center (Fig. 1).19 The geometry of the
complex is best described as a four-legged piano-stool. The
NMe2 ligands are transoid (N3–Zr–N4, 120.26(6)1), as are
the coordinated oxazoline and THF ligands (N2–Zr1–O3,
159.40(4)1). The cyclopentadienyl ring is bonded through all
five carbons, but in a twisted fashion such that the shortest
distance is Zr1–C12 (2.492(2) A) rather than Zr1–C11
(2.518(2) A, where C11 is the carbon bonded to boron).
The room temperature 1H NMR spectrum of 4 in toluene-d8
was identical to a combination of the independent spectra of 2
and THF. However, at 190 K, the diastereotopic methyl and
methylene signals were coalesced into broad resonances. This
fluxionality is the only NMR evidence for THF coordination.
Similar behavior was observed for the hafnium analog.
In contrast with 2, the IR spectrum of 4 (KBr) contained
two nCN bands: one at 1610 cmꢀ1 is assigned to the non-
coordinated oxazoline ring, and one at 1533 cmꢀ1 is due to the
2,2-dimethyl-4-penten-1-amine with
a turnover rate of
0.07 hꢀ1 at 100 1C (an insertion mechanism was proposed).7
The 4-legged piano-stool compound Cp*(L)Zr(NMe2)2 (L =
silicyloxazolinato) forms a more active catalyst (Nt = 2.35 hꢀ1
)
although temperatures of 90 1C are employed (imido mechanism
was proposed).10
The mechanism of 2-catalyzed hydroamination/cyclization
was investigated more closely to better understand these
unusual features. HNMe2 is observed by 1H NMR spectro-
scopy upon addition of the aminoalkene substrate to 2.
Although the active catalyst is not directly observed, it is
formed at room temperature within 5 min after addition of
aminoalkene 5 to 2, based on the appearance of 2-methyl-4,4-
1
bound group. Comparison of this spectrum with that of 2
further supports (Z5-C5H4)-k2-(OxMe
THF-free species.
diphenylpyrrolidine product in the H NMR spectrum of the
2
)
coordination in the
reaction mixture. For two half-lives, plots of ln[5] vs. time are
linear (see ESIw for kinetic data), and this is consistent with
first-order dependence on substrate concentration. After two
half-lives, however, the reaction rate significantly decreases,
and first-order plots are not linear over three half-lives.
Independent verification of the substrate order is provided
by the method of initial rates.20 In these experiments, the
initial substrate concentration was varied over several runs
while catalyst concentration was kept constant; the initial rate
(d[5]/dt) was then determined for each run. A plot of ln(initial
rate) vs. ln[5] is linear, and a slope of 0.94 verifies first-order
dependence on [5].20 Likely, the severe rate decrease after two
half-lives results from product inhibition, as the least hindered
2
Compound 2 is a precatalyst for the conversion of amino-
alkenes to corresponding pyrrolidenes and a piperidine (see
Table 1). As is typical, the cyclization efficiency increases as
the substrates’ substituents increase in size (Ph (5) 4 Me (6) 4
H (7)). However, there are several unusual features of this
catalysis. First, the reaction proceeds at room temperature,
which is uncommon for Group 4-catalyzed hydroaminations.
Second, THF only moderately inhibits the conversion and the
apparent turnover rate (compare entries 1 vs. 2 and 5 vs. 6).
Third, long reaction times only slightly increase the conversion,
and the apparent turnover rate (Nt) decreases with increasing
ꢁc
This journal is The Royal Society of Chemistry 2010
340 | Chem. Commun., 2010, 46, 339–341

View Article Online
versatility of bent-sandwich compounds in catalysis and
the enhanced reactivity demonstrated here, [PhB(C5H4)-
(OxMe )2]ZrX2-type compounds may offer new possibilities in
2
catalytic chemistry.
We thank the U.S. DOE Office of Basic Energy Science
(DE-AC02-07CH11358) and the ACS Green Chemistry
Institute-Petroleum Research Fund for financial support.
Dr Andreja Bakac is thanked for many helpful discussions.
Notes and references
1 A. Togni and R. L. Halterman, Metallocenes: synthesis, reactivity,
applications, Wiley-VCH, Weinheim, 1998.
2 J. W. Lauher and R. Hoffmann, J. Am. Chem. Soc., 1976, 98,
1729–1742.
Scheme 1
3 H. Lee, P. J. Desrosiers, I. Guzei, A. L. Rheingold and G. Parkin,
J. Am. Chem. Soc., 1998, 120, 3255–3256.
product shows the poorest conversion. Two experiments were
performed to support this idea: (1) the reaction rate is slowed
by addition of product, and (2) addition of more substrate
results in further conversion. A plot of kobs vs. [2] (ranging
from 6.8 to 40 mM) reveals a linear dependence on [2],
providing the empirical rate law: ꢀd[5]/dt = k0[5][2] (k0 =
2.0(3) ꢂ 10ꢀ3 Mꢀ1 sꢀ1). A large primary kinetic isotope effect
(kH/kD = 5.4) is determined by comparison of k0 obtained
from 5 and 5-d2 substrates. Interestingly, the rate constant for
cyclization of the less hindered substrate 6 is almost three
4 (a) P. J. Shapiro, E. Bunel, W. P. Schaefer and J. E. Bercaw,
Organometallics, 1990, 9, 867–869; (b) A. L. McKnight and
R. M. Waymouth, Chem. Rev., 1998, 98, 2587–2598.
5 (a) M. Bochmann, S. J. Lancaster and O. B. Robinson, J. Chem.
Soc., Chem. Commun., 1995, 2081–2082; (b) P. J. Shapiro, Eur. J.
Inorg. Chem., 2001, 321–326.
6 D. V. Gribkov and K. C. Hultzsch, Angew. Chem., Int. Ed., 2004,
43, 5542–5546.
7 B. D. Stubbert and T. J. Marks, J. Am. Chem. Soc., 2007, 129,
6149–6167.
8 P. J. Walsh, A. M. Baranger and R. G. Bergman, J. Am. Chem.
Soc., 1992, 114, 1708–1719.
times larger (k0 = 5.67(5) ꢂ 10ꢀ3 Mꢀ1
s
ꢀ1) than the rate
9 (a) M. C. Wood, D. C. Leitch, C. S. Yeung, J. A. Kozak and
L. L. Schafer, Angew. Chem., Int. Ed., 2007, 46, 354–358;
(b) D. A. Watson, M. Chiu and R. G. Bergman, Organometallics,
2006, 25, 4731–4733; (c) H. Kim, P. H. Lee and T. Livinghouse,
Chem. Commun., 2005, 5205–5207.
constant for 5. This difference is not reflected in the ‘turnover
rate’ since the product from 6 shows greater catalyst inhibition
than the product from 5.
Two mechanisms are consistent with the experimental rate
law. The first (Scheme 1a) involves rapid intramolecular
insertion followed by turnover-limiting protonolysis. This
overall sequence is similar to the well-established insertion
mechanism for hydroamination/cyclization,21 with the important
distinction that the relative rate constants for the insertion and
protonolysis steps in the classic mechanism are reversed in our
mechanism. This turnover-limiting protonolysis mechanism is
also consistent with a large kinetic isotope effect and rapid
catalyst initiation. However, facile alkene insertion into a
Zr–N bond is unusual, as is the requirement that protonolysis
be slower than alkene insertion.
10 A. L. Gott, A. J. Clarke, G. J. Clarkson and P. Scott, Chem.
Commun., 2008, 1422–1424.
11 T. E. Muller, K. C. Hultzsch, M. Yus, F. Foubelo and M. Tada,
¨
Chem. Rev., 2008, 108, 3795–3892.
12 L. Ackermann, R. G. Bergman and R. N. Loy, J. Am. Chem. Soc.,
2003, 125, 11956–11963.
13 J. F. Dunne, K. Manna, J. Wiench, M. Pruski and A. D. Sadow,
Dalton Trans., DOI: 10.1039/b911851a.
14 A. I. Meyers and E. W. Collington, J. Am. Chem. Soc., 1970, 92,
6676–6678.
15 J. F. Dunne, J. C. Su, A. Ellern and A. D. Sadow, Organometallics,
2008, 27, 2399–2401.
16 I. Lopes, G. Y. Lin, A. Domingos, R. McDonald, N. Marques and
J. Takats, J. Am. Chem. Soc., 1999, 121, 8110–8111.
17 Y. Shi, C. Hall, J. T. Ciszewski, C. Cao and A. L. Odom, Chem.
Commun., 2003, 586–587.
18 (a) A. Otero, J. Fernandez-Baeza, A. Antinolo, J. Tejeda, A.
Lara-Sanchez, L. Sanchez-Barba, A. M. Rodriguez and
M. A. Maestro, J. Am. Chem. Soc., 2004, 126, 1330–1331;
(b) A. Otero, J. Fernandez-Baeza, A. Lara-Sanchez, A. Antinolo,
J. Tejeda, E. Martinez-Caballero, I. Marquez-Segovia, I.
Lopez-Solera, L. F. Sanchez-Barba and C. Alonso-Moreno, Inorg.
Chem., 2008, 47, 4996–5005.
19 Crystal data for C29H45BN4O3Zr (4), M = 599.72, monoclinic,
P21/n, a = 12.7494(5), b = 17.2023(7), c = 13.5397(5) A,
b = 93.952(1)1, V = 2962.5(2) A3, T = 123(2) K, Z = 4, 31 553
reflections collected, 6824 independent reflections (Rint = 0.0382),
R = 0.0274, wR2 = 0.681 [I 4 2s(I)], R = 0.0357, wR2 = 0.0732
(all reflections).
A second mechanism involves reversible substrate coordination,
turnover-limiting a-abstraction to form a zirconium imido
intermediate, and subsequent rapid [2p+2p] cycloaddition
for C–N bond formation (Scheme 1b).9,10 The large isotope
effect is consistent with this mechanism, as a-NH abstraction
reactions have significant isotope effects.22 This mechanism is
also consistent with the observation that a secondary amine
substrate does not undergo cyclization.
Attempts to isolate or trap a zirconium imido in this system
have not yet been successful. However, the lack of reactivity of
the secondary amine 9 is unlikely to be due to steric hindrance.
In addition, related four-legged piano-stool zirconium compounds
are proposed to catalyze hydroamination/cyclization through
an imido pathway.10 For these reasons, we favor the imido
mechanism of Scheme 1b. These catalytic reactions show that
this zwitterionic zirconium compound is activated toward
hydroamination/cyclization of aminoalkenes. Given the general
20 (a) J. H. Espenson, Chemical kinetics and reaction mechanisms,
McGraw-Hill, New York, 1995; (b) K. J. Laidler, Chemical
Kinetics, McGraw-Hill, New York, 1965.
21 M. R. Gagne, C. L. Stern and T. J. Marks, J. Am. Chem. Soc.,
1992, 114, 275–294.
22 C. P. Schaller, C. C. Cummins and P. T. Wolczanski, J. Am. Chem.
Soc., 1996, 118, 591–611.
ꢁc
This journal is The Royal Society of Chemistry 2010
Chem. Commun., 2010, 46, 339–341 | 341