General preparation of (N3N)ZrX (N3N = N(CH 2CH2NSiMe3)33-)

Article abstract of DOI:10.1021/om8008684

A homoleptic triamidoamine zirconium complex featuring a metalated trimethylsilyl substituent, [κ⁵-(Me₃SiNCH ₂CH₂)₂NCH₂CH₂NSiMe ₂CH₂]Zr (1), was synthesized

Full text of DOI:10.1021/om8008684

Organometallics 2009, 28, 573–581
573
3-
General Preparation of (N3N)ZrX (N3N ) N(CH2CH2NSiMe3)3 )
Complexes from a Hydride Surrogate
Andrew J. Roering,† Annalese F. Maddox,† L. Taylor Elrod,† Stephanie M. Chan,†
Michael B. Ghebreab,† Kyle L. Donovan,† Jillian J. Davidson,† Russell P. Hughes,§
Tamila Shalumova,‡ Samantha N. MacMillan,‡, Joseph M. Tanski,‡ and Rory Waterman*,†
Department of Chemistry, UniVersity of Vermont, Burlington, Vermont 05405, Department of Chemistry,
Dartmouth College, HanoVer, New Hampshire 03755, and Department of Chemistry, Vassar College,
Poughkeepsie, New York 12604
ReceiVed September 5, 2008
A homoleptic triamidoamine zirconium complex featuring a metalated trimethylsilyl substituent, [κ5-
(Me3SiNCH2CH2)2NCH2CH2NSiMe2CH2]Zr (1), was synthesized by reaction of Zr(CH2Ph)4 with
N(CH2CH2NHSiMe3)3 followed by sublimation. Complex 1 is a general precursor to a family of complexes
with the formulation (N3N)ZrX (N3N ) N(CH2CH2NSiMe3)33-, X ) anionic ligand) by reactions that
parallel expected reactivity of a hydride derivative. Treatment of 1 with phosphines, amines, thiols, alkynes,
and phenol resulted in the formation of new, pseudo-C3V-symmetric (N3N)ZrX complexes (X ) phosphido,
amido, alkynyl, thiolate, or phenoxide) via element-H bond activation. Thus, the reactivity of complex
1 is that best described as a hydride surrogate. For example, complex 1 reacted with PhPH2 at ambient
temperature to provide (N3N)ZrPHPh (2) in 86% yield. Density functional theory studies and X-ray
crystal structures provide a general overview of the bonding in these complexes, which appears to be
highly ionic. In general, there is little evidence for ligand-to-metal π-bonding for the pseudoaxial X
ligand in these complexes except for strongly π-basic terminal amido ligands. The limited π-bonding
appears to be the result of competitive π-donation by the pseudoequatorial amido arms of the triamidoamine
ancillary ligand. Thus, the relative Zr-X bond energies are governed by the basicity of the anionic
ligand X. Solid-state structures of phosphido (3, 4, 5), amido (10), and thiolate (15) complexes support
the computational results.
Recently however, triamidoamine-supported zirconium com-
plexes have demonstrated reactivity in element-element bond-
forming reactions. These complexes affected the catalytic
dehydrocoupling of phosphines and, under appropriate condi-
tions, selectively forming diphosphine products.9 Similarly, these
complexes engaged in the catalytic dehydrocoupling of arsines,
and for sterically encumbered primary arsines, evidence for
R-arsinidene elimination was observed.10 Furthermore, these
zirconium species catalyzed the heterodehydrocoupling of
primary phosphines with silanes or germanes to provide P-E
(E ) Si, Ge) bonds.11 Recently, these complexes have facilitated
the synthesis of a phosphaalkene with perfect atom economy.12
The reported preparation of primary phosphido complexes
(N3N)ZrPHR (N3N ) N(CH2CH2NSiMe3)33-; R ) Ph, 2; Cy,
3)9 began with the synthesis of (N3N)ZrNMe2, discovered by
Verkade and co-workers,12 and utilized elements of the prepara-
tions of highly related (N′3N)ZrX complexes (N′3N )
N(CH2CH2NSitBuMe2)33-; X ) monoanionic ligand) reported
by Scott.13,14 This synthesis, though effective, was time-
consuming and provided an overall yield of less than 50% of
Introduction
Metal-catalyzed element-element bond formation has gar-
nered increased attention because of the main-group small
molecules, polymers, and materials that have already been
realized.1-5 Thus, simple, readily prepared metal catalysts have
become increasingly desirable for this chemistry. The com-
mercially available triamidoamine ligand has been used exten-
sively in early transition-metal chemistry. In particular, these
ligands have seen substantial application in supporting group 5
and 6 metals featuring multiply bonded ligands.6,7 Despite the
well-established value of triamidoamine ligands in synthetic
inorganic chemistry, these ligands have seen far less use in
catalysis.8
* Corresponding author. E-mail: rory.waterman@uvm.edu.
† University of Vermont.
§ Dartmouth College.
‡ Vassar College.
Current address: Department of Chemistry, Massachusetts Institute of
Technology.
(1) Herbert, D. E.; Mayer, U. F. J.; Manners, I. Angew. Chem., Int. Ed.
2007, 46, 5060–5081.
(2) Clark, T. J.; Lee, K.; Manners, I. Chem.-Eur. J. 2006, 12, 8634–
(9) Waterman, R. Organometallics 2007, 26, 2492–2494.
(10) Roering, A. J.; MacMillan, S. N.; Tanski, J. M.; Waterman, R.
Inorg. Chem. 2007, 46, 6855–6857.
(11) Roering, A. J.; Davidson, J. J.; MacMillan, S. N.; Tanski, J. M.;
Waterman, R. Dalton Trans. 2008, 4488–4498.
(12) MacMillan, S. N.; Tanski, J. M.; Waterman, R. Chem. Commun.
2007, 4172–4174.
8648.
(3) Masuda, J. D.; Hoskin, A. J.; Graham, T. W.; Beddie, C.; Fermin,
M. C.; Etkin, N.; Stephan, D. W. Chem.-Eur. J. 2006, 12, 8696–8707.
(4) Waterman, R. Curr. Org. Chem. 2008, 12, 1322–1339.
(5) Gauvin, F.; Harrod, J. F.; Woo, H. G. AdV. Organomet. Chem. 1998,
42, 363–405.
(6) Balazs, G.; Gregoriades, L. J.; Scheer, M. Organometallics 2007,
26, 3058–3075.
(7) Schrock, R. R. Acc. Chem. Res. 1997, 30, 9–16.
(8) Schrock, R. R. Acc. Chem. Res. 2005, 38, 955–962.
(13) Duan, Z.; Naiini, A. A.; Lee, J.-H.; Verkade, J. G. Inorg. Chem.
1995, 34, 5477–5482.
(14) Morton, C.; Munslow, I. J.; Sanders, C. J.; Alcock, N. W.; Scott,
P. Organometallics 1999, 18, 4608–4613.
10.1021/om8008684 CCC: $40.75
2009 American Chemical Society
Publication on Web 12/03/2008

574 Organometallics, Vol. 28, No. 2, 2009
Roering et al.
Scheme 1. Metalated Complex 1 Acting As a Hydride
Surrogate
the family of (N3N)ZrX complexes suggest that the bonding in
these complexes is highly ionic and that there is little π-bonding
character from most pseudoaxial X ligands with the exception
of strong π-bases such as amido ligands. The limited π-interac-
tion may explain the high reactivity of the Zr-X bond already
seen in stoichiometric and catalytic reactions.9-12
Results and Discussion
Preparation of Metalated Complex 1. Reaction of
Zr(CH2Ph)4 with 1 equiv of N(CH2CH2NHSiMe3)3 resulted in
liberation of toluene to give an oily brown solution. Monitoring
the reaction by 1H NMR spectroscopy (benzene-d6) revealed a
mixture of products including starting material as well as a new
pseudo-C3-symmetric complex. This complex was tentatively
assigned as (N3N)Zr(η1-CH2Ph) based partly on a diagnostic
methylene resonance at δ 2.44. Efforts to isolate or crystallize
this benzyl complex from a variety of solvents such as hexane,
toluene, and Et2O were ineffective, and only a mixture of
products including [κ5-N,N,N,N,C-(Me3SiNCH2CH2)2NCH2-
CH2NSiMe2CH2]Zr (1) was obtained. A highly related zirconium
benzyl complex, (N′3N)Zr(η1-CH2Ph), was isolated by the group
the desired phosphido derivatives 2 or 3. Given the value of
these zirconium complexes as catalysts,9-11 a simplified prepa-
ration of the metal complexes was sought. A zirconium hydride
complex, (N3N)ZrH, would be an ideal precursor. However,
(N3N)MH (M ) group 4 metal) complexes are exceedingly rare.
Schrock and co-workers have reported the observation of
(N3N)TiH, but this complex was not isolable due to a cyclo-
metalation process involving a trimethylsilyl substituent.15,16 In
related (N′3N)Zr complex chemistry, a hydride ligand was not
observed, and only a cyclometalated product was isolated. Facile
H/D exchange at the methyl substituents of the SitBuMe2 groups
with D2 provided strong evidence for the importance of such a
hydride complex.14 Later, Scott successfully thwarted this
metalation process by replacing the trialkylsilyl substituents with
aryl groups on the triamidoamine ligand framework. Unfortu-
nately, these new derivatives did not yield an isolable zirconium
hydride.17 Likewise, it was observed that any putative (N3N)ZrH
complex readily metalated to give [κ5-N,N,N,N,C-(Me3Si-
NCH2CH2)2NCH2CH2NSiMe2CH2]Zr (1) as the exclusive prod-
uct. The discovery that 1 is important in the dehydrocoupling
of phosphines suggested that this compound may provide a route
to (N3N)ZrX complexes.9
1
of Scott, and this complex displayed very similar H NMR
spectroscopic data.14
Heating the aforementioned reaction mixture under reduced
pressure followed by sublimation at temperatures greater than
100 °C resulted in the isolation of 1 as a colorless solid in 73%
yield (eq 2). Complex 1 can be similarly prepared by heating
isolated (N3N)ZrMe; however, the preparation of this complex
requires three synthetic steps from N(CH2CH2NHSiMe3)3.9
It was our hypothesis that the cyclometalated complex 1
would be a general synthetic precursor to (N3N)ZrX species by
acting as a hydride surrogate (Scheme 1). The inability to isolate
(N3N)ZrH makes complex 1 the closest accessible chemical
species, and it has been observed that (N3N)ZrH is unstable
with respect to cyclometalation to form 1 and H2.9 Furthermore,
complex 1 is important in the catalytic dehydrocoupling of
phosphines as the likely precursor to Zr-phosphido complexes.
A hydride complex would be expected to react with polar
element-hydrogen bonds (H-X) to afford (N3N)ZrX and
liberate H2 as the driving force (Scheme 1). Complex 1 may
also react with H-X bonds to form (N3N)ZrX and restore the
trimethylsilyl substituent. The latter reactivity would rely on
the strength of the resultant Zr-X bond and the methyl C-H
bond, which would presumably be a reasonable driving force.
Herein, we report a new preparation of the metalated complex
1 based on methodology described by Scott and co-workers7
and demonstrate that complex 1 is a general precursor to
(N3N)ZrX complexes, where X can be an alkynyl, amido,
phenoxide, phosphido, or thiolate ligand. This reactivity of
complex 1 is best described as a that of a hydride surrogate,
affording Zr-products that would be expected from reaction of
(N3N)ZrH with HX. Theoretical studies and solid-state data on
Interestingly, complex 1 prepared by this method exhibited
increased resolution in the H NMR spectrum. For example,
1
the zirconium methylene resonance at δ 0.314 could be
distinguished from the trimethylsilyl methyl resonances at δ
0.274. Otherwise, spectroscopic data for 1 was identical to
authentic samples prepared by the literature route.9 Sublimed
samples of 1 from the literature route gave the same spectro-
scopic data as this preparation. Attempts to hydrogenate either
the putative benzyl complex or 1 to observe or isolate a
zirconium hydride complex have thus far been unsuccessful.
Hydride Surrogate Reactivity of Complex 1. Complex 1
was first observed in the preparation of primary phosphido
derivatives (N3N)ZrPHR (R ) Ph, Cy) by reaction of RPH2
with (N3N)ZrMe. These studies suggested that 1 may be a
general precursor to (N3N)ZrX complexes with overall reactivity
that would parallel a hydride complex.9
Treatment of 1 with either phenyl- or cyclohexylphosphine
in Et2O at ambient temperature afforded of the corresponding
phosphido complex (N3N)ZrPHR (R ) Ph, 2; R ) Cy, 3) in 86
and 89% isolated yield, respectively (eq 2). Monitoring the
reaction by 1H and 31P NMR spectroscopy (benzene-d6) showed
quantitative conversion to the phosphido complex products.
Compounds 2 and 3 gave identical spectral data to authentic
samples prepared by the known method.9 Starting from meta-
lated complex 1, the new preparation of these phosphido
compounds was more efficient than previously reported. The
overall yield of phenylphosphido complex 2 was 64% from
commercially available Zr(CH2Ph)4. This is a significant
improvement over the 48% overall yield for 2 starting from
(15) Cummins, C. C.; Schrock, R. R.; Davis, W. M. Organometallics
1992, 11, 1452–1454.
(16) Schrock, R. R.; Cummins, C. C.; Wilhelm, T.; Lin, S.; Reid, S. M.;
Kol, M.; Davis, W. M. Organometallics 1996, 15, 1470–1476.
(17) Morton, C.; Gillespie, K. M.; Sanders, C. J.; Scott, P. J. Organomet.
Chem. 2000, 606, 141–146.

Zirconium Hydride Surrogate
Organometallics, Vol. 28, No. 2, 2009 575
Scheme 2. General Preparation of (N3N)ZrX Derivatives
from Compound 1
Figure 1. Molecular structure of (N3N)ZrPPh2 (4) with thermal
ellipsoids drawn at the 35% level. Hydrogen atoms are omitted for
clarity.
Table 1. Selected Bond Lengths (Å) and Angles (deg) for
(N3N)ZrPPh2 (4)
commercially available Zr(NMe2)4 by the published route.9,13
This increased efficiency is due to the synthetic steps saved by
use of complex 1 as a “(N3N)Zr” synthon. Thus, phosphido
complexes were prepared in a single step by the addition of
phosphine to complex 1.
Zr-P
P-C(16)
P-C(22)
2.7061(4)
1.835(1)
1.820(1)
Zr-N(1)
Zr-N(2)
Zr-N(3)
Zr-N(4)
2.071(1)
2.083(1)
2.045(1)
2.523(1)
Zr-P-C(16)
Zr-P-C(22)
C(16)-P-C(22)
119.06(5)
118.49(5)
103.74(6)
N(1)-Zr-P
N(2)-Zr-P
N(3)-Zr-P
N(4)-Zr-P
102.58(4)
110.06(3)
106.47(3)
175.87(3)
Treatment of 1 with 1 equiv of 1,2-diphosphinoethane in
ethereal solution at ambient temperature gave analytically pure,
light orange crystals of (N3N)ZrPHCH2CH2PH2 (5) in 76% yield
(Scheme 2). Spectroscopic data (1H, 13C NMR and IR) of
complex 5 were consistent with the formulation given. The 31
P
The use of complex 1 as a hydride surrogate in the preparation
of (N3N)ZrX complexes is general. Reaction of diphenylphos-
phine with 1 in benzene at ambient temperature afforded
analytically pure, bright red crystals of (N3N)ZrPPh2 (4) in 80%
yield (Scheme 2). Complex 4 displayed pseudo-C3-symmetry
with respect to the triamidoamine ligand in the 1H NMR
spectrum as well as a characteristic zirconium phosphido 31P
NMR resonance at δ 85.3.9,18,19 Free rotation of P-C and Zr-P
bonds was evident by the equivalent phenyl resonances in the
1H and 13C NMR spectra. Confirmation of the solution structure
of complex 4 came from a single-crystal X-ray diffraction study.
Single crystals of 4 were obtained by slow evaporation of Et2O
solvent over a period of ca. 2 h. The solid-state structure of 4
(Figure 1) features a trigonal-bipyramidal zirconium center (τ
) 1.0)20 with Zr-P ) 2.7061(4) Å and a long Zr-Namine bond,
Zr-N ) 2.523(1) Å (Table 1), and an approximately pyramidal
phosphorus atom (∑ at P ≈ 341°). Other terminal zirconium
phosphido complexes display similar structures and Zr-P bond
lengths in the solid state.10,19,21-23
NMR spectrum of 5 revealed two phosphorus environments,
one for the zirconium-bound phosphido ligand at δ -72.4 and
the other for the free phosphine at δ -133. Single crystals of 5
were obtained by cooling a concentrated Et2O solution of 5 at
-30 °C for extended periods. The molecular structure of 5,
shown in Figure 2, was found to be trigonal bipyramidal at the
zirconium center (τ ) 1.1) with a Zr-P ) 2.7185(7) Å (Table
2). The τ value greater than unity reflects the fact that
the zirconium center lies above the trigonal plane formed by
the amido nitrogen substituents of the N3N ligand. Though it is
greater than 1, the τ parameter indicates that complex 5 is much
closer to a trigonal-bipyramidal geometry than square-based
pyramid.
Zirconium-carbon bonds can be formed using hydride
surrogate 1 (Scheme 2). Terminal alkynes reacted rapidly with
1 to provide analytically pure, colorless microcrystals of alkynyl
derivatives (N3N)ZrCR’CꢀR (R ) Ph, 6; SiMe3, 7; Bu, 8) in
74-84% isolated yield. Complexes 6, 7, and 8 gave expected
spectroscopic data for terminal alkynyl complexes, and diag-
nostic features of these complexes are summarized in Table 3.
Reaction of 1 with phenol in Et2O at ambient temperature
afforded analytically pure, colorless microcrystals of
(N3N)ZrOPh (9) in 85% yield (Scheme 2). Complex 9 gave
expected and unremarkable spectroscopic data. Attempts to
synthesize alkoxide complexes have met with limited success.
In reactions of 1 with alcohols or water, a mixture of products
was obtained that appeared to contain the desired (N3N)ZrOR
(R ) alkyl, H), but pure complexes have not yet been isolated.
(18) Roddick, D. M.; Santarsiero, B. D.; Bercaw, J. E. J. Am. Chem.
Soc. 1985, 107, 4670–4678.
(19) Vaughan, G. A.; Hillhouse, G. L.; Rheingold, A. L. Organometallics
1989, 8, 1760–1765.
(20) Addison, A. W.; Rao, T. N.; Reedijk, J.; Rijn, J. v.; Verschoor,
G. C. J. Chem. Soc., Dalton Trans 1984, 134, 9–1356.
(21) Hey-Hawkins, E. Chem. ReV. 1994, 94, 1661–1717.
(22) Stephan, D. W. Angew. Chem., Int. Ed. 2000, 39, 315–329.
(23) Urnezius, E.; Klippenstein, S. J.; Protasiewicz, J. D. Inorg. Chim.
Acta 2000, 297, 181–190.

576 Organometallics, Vol. 28, No. 2, 2009
Roering et al.
Figure 3. Molecular structure of (N3N)ZrNHtBu (10) with thermal
ellipsoids drawn at the 35% level. Hydrogen atoms except for H(3),
located on N(3), are omitted for clarity. Selected bond lengths (Å):
Zr-N(3) ) 2.071(3), Zr-N(1) ) 2.088(1), Zr-N(2) ) 2.552(2),
N(3)-C(6) ) 1.492(3).
Figure 2. Molecular structure of (N3N)ZrPHCH2CH2PH2 (5) with
thermal ellipsoids drawn at the 35% level. Hydrogen atoms except
those located on P(1) and P(2) are omitted for clarity.
Table 2. Selected Bond Lengths (Å) and Angles (deg) for
(N3N)ZrPHCH2CH2PH2 (5)
(N3N)ZrNHNdCPh2 (12) in 25% yield (Scheme 2). Diagnostic
1
Zr-P(1)
P(1)-C(16)
Zr-N(4)
2.7185(7)
1.883(3)
2.540(1)
Zr-N(1)
Zr-N(2)
Zr-N(3)
2.055(1)
2.060(1)
2.061(1)
features of the N-H moiety include δ 8.068 in the H NMR
spectrum and νNH ) 3299 cm-1 in the infrared for 12. Complex
1 reacted rapidly with 1 equiv of benzophenone imine to give
(N3N)ZrNdCPh2 (13) as an analytically pure, orange powder
in 82% yield (eq 3). Spectroscopic features for the iminido
C(16)-P(1)-Zr
C(16)-P(1)-H(1)
Zr-P(1)-H(1)
107.21(8)
99.40(1)
88.40(1)
N(1)-Zr-P(1)
N(2)-Zr-P(1)
N(3)-Zr-P(1)
N(4)-Zr-P(1)
106.15(5)
107.11(5)
106.79(5)
179.47(4)
ligand included δ 174.8 in the 13C NMR spectrum and νCN
)
1641 cm-1 in the infrared.
Table 3. Selected 13C NMRa and IRb Spectroscopic Data for
Compounds 6, 7, and 8
6
7
8
CR
Cꢀ
νcc
131.77
147.36
2076
113.49
169.97
2029
107.74
138.43
2081
b
a In ppm. In cm-1
.
Thiols reacted readily with 1 to form Zr-S bonds (Scheme
2). Treatment of 1 with benzenethiol or tert-butylthiol in Et2O
at ambient temperature gave analytically pure, colorless crystals
of (N3N)ZrSPh (14) or (N3N)ZrStBu (15) in 44% and 64% yield,
respectively. Spectroscopic data (1H, 13C NMR and IR) for 14
and 15 supported the formulation given. Single crystals of 15
were grown from concentrated 20:1 toluene/hexane solvent
mixture cooled to -30 °C. The molecular structure of 15 (Figure
4) features trigonal-bipyramidal geometry at the zirconium
center (τ ) 0.94) with Zr-S ) 2.5154(4) Å (Table 4, 5). The
crystal structure also revealed sulfur displaying bent geometry
with a Zr-S-C ) 127.74(4)°. Other zirconium tert-butyl
thiolate complexes possess similar bond angles. For Cp2Zr-
(StBu)2, Zr-S-C ) 123° was observed, while arylthiolates
Cp2Zr(SPh)2 and Cp2Zr[S(p-C6H4Cl)]2 displayed Zr-S-C
angles of 119° and 109°, respectively.24-26 It appears that the
nature of the organothiol substituent has a great impact on the
Zr-S-C bond angle.
Complex 1 provides a facile route to terminal amido
complexes (Scheme 2). Treatment of 1 with 1 equiv of tert-
butyl amine in benzene at ambient temperature gave analytically
pure, colorless crystals of (N3N)ZrNHtBu (10) in 73% yield.
Key spectroscopic data for complex 10 include an NH resonance
1
at δ 3.856 in the H NMR spectrum and νNH ) 3210 cm-1 in
the infrared. Single crystals of 10 were grown from a concen-
trated Et2O solution kept at -30 °C for extended periods and
were subject to a single-crystal X-ray study. The molecular
structure of 10, shown in Figure 3, is truly C3V-symmetric with
respect to the triamidoamine ligand. The complex was found
to be trigonal bipyramidal at zirconium (τ ) 0.94) and planer
at the amido nitrogen, N(3). The planar amido nitrogen is
consistent with ligand-to-metal π-donation, a structural feature
that has been seen for (N3N)ZrNMe2.14
Similarly, reaction of 1 with aniline and in benzene at ambient
temperature afforded analytically pure, colorless crystals of
(N3N)ZrNHPh (11) in 92% yield (Scheme 2). Complex 11
showed pseudo-C3V-symmetry with respect to the triamidoamine
It is important to note that while the isolated yields of
(N3N)ZrX complexes varied, in all cases, monitoring the reaction
of 1 with any of these HX reagents by NMR spectroscopy
1
ligand and showed typical H and 13C NMR resonances with
1
diagnostic features NH ) δ 5.962 in the H NMR spectrum
(24) Ashby, M. T.; Alguindigue, S. S.; Khan, M. A. Inorg. Chim. Acta
1998, 270, 227–237.
(25) Howard, W. A.; Trnka, T. M.; Parkin, G. Inorg. Chem. 1995, 34,
5900–5909.
(26) Yam, V. W.-W.; Qi, G.-Z.; Cheung, K.-K. J. Organomet. Chem.
1997, 548, 289–294.
and νNH ) 3244 cm-1 in the infrared.
Other Zr-amido complexes can also be prepared by this route.
Treatment of benzophenone hydrazone with 1 in Et2O at ambient
temperature afforded analytically pure, yellow crystals of

Zirconium Hydride Surrogate
Organometallics, Vol. 28, No. 2, 2009 577
Figure 4. Molecular structure of (N3N)ZrStBu (15) with thermal
ellipsoids drawn at the 35% level. Hydrogen atoms omitted for
clarity.
Figure 5. Molecular structure of (N3N)ZrPHCy (3) with thermal
ellipsoids drawn at the 35% level. The 65% occupancy model for
the 2-fold cyclohexyl disorder is shown here with hydrogen atoms
except for H(1), located on P(1), omitted for clarity.
Table 4. Selected Bond Lengths (Å) and Angles (deg) for
(N3N)ZrStBu (15)
Table 6. Selected Bond Lengths (Å) and angles (deg) for the
Solid-State and Minimized Structures of 2 and 3
Zr-S
S-C(16)
Zr-N(4)
2.5154(4)
1.848(1)
2.530(1)
Zr-N(1)
Zr-N(2)
Zr-N(3)
2.066(1)
2.076(1)
2.079(1)
(N3N)ZrPHPh (2)
(N3N)ZrPHCy (3)
bond
calc
X-ray
calc
X-raya
Zr-S-C(16)
N(1)-Zr-S
N(2)-Zr--S
127.74(4)
101.67(3)
115.95(3)
N(3)-Zr-S
N(4)-Zr-S
102.02(3)
170.73(3)
Zr-P
2.739
1.839
2.633
2.104
118.0
175.6
107.7
316.4
2.7254(6)
1.830(2)
2.526(2)
2.059(5)
111.50(7)
175.44(4)
106.3
2.696
1.892
2.631
2.110
124.0
173.2
107.5
326.2
2.734(1)
1.890(4)
2.532(1)
2.066(5)
118.0(1)
171.06(6)
106.5
P-C
Zr-Namine
Zr-Namido
Zr-P-C
b
Table 5. Selected Bond Lengths (Å) and Angles (deg) for
(N3N)ZrPHCy (3)
Namine-Zr-P
Namido-Zr-Pb
Zr-P
P-C(21)
Zr-N(4)
2.734(1)
1.890(4)
2.532(1)
Zr-N(1)
Zr-N(2)
Zr-N(3)
2.071(1)
2.061(1)
2.066(1)
∑
at Pc
305.7
317
a Data from one of two orientations of the Cy moiety in the crystal.
b Simple average of the three amido substituents. c Sum of the angles at
phosphorus.
C(21)-P-Zr
C(21)-P-H(1)
Zr-P-H(1)
118.0(1)
99(2)
100(2)
N(1)-Zr-P
N(2)-Zr-P
N(3)-Zr-P
N(4)-Zr-P
98.02(6)
112.30(6)
109.24(5)
171.06(6)
A detailed picture of the bonding in these molecules was
obtained by natural bond order (NBO) and related natural
resonance theory (NRT) and natural population analyses (NPA)
of the molecules. From this set of calculations several general
trends regarding the bonding in these complexes have emerged.
First, there is a high degree of polarization resulting from very
ionic bonding between the amido arms of the triamidoamine
ligand and the zirconium center. The same is true for the
pseudoaxial X ligand, where X is amido, phosphido, hydride,
methide, or acetylide in these calculations. For example, in the
methide complex 16, the Zr-CH3 σ-bond is strongly polarized,
with ∼16% of the electron density on Zr and ∼84% on C.
Likewise, the Zr bonds to the amido arms have σ- and
π-components, with both components even more strongly
polarized compared to the Zr-C bond. The electron density in
the σ- component is only ∼10% on Zr and ∼90% on N,
compared to ∼6% on Zr and 94% on N in its π-counterpart.
Second, the bond between the axial amine nitrogen of the
triamidoamine ligand and zirconium is quite weak. This weak
bond is evident in the solid-state structures of these (N3N)ZrX
complexes by the Zr-Namine bond length of ca. 2.53 Å (Zr-N(4)
) 2.523(1)-2.540(1) Å, Vide supra). This weak bonding to the
fifth neutral ligand appears to promote a more tetrahedral-
like electronic structure at zirconium. This weak bond may also
be part of the reason that three amido arms of the N3N ligand
are not coplanar with Zr, as seen in the minimized structures
and in the solid-state data. NBO analysis reveals that the lone
(benzene-d6) showed quantitative conversion to the desired
(N3N)ZrX product. The variety of complexes that can be
prepared by using 1 as a hydride surrogate (Vida supra) suggests
that any moderate acidic HX reagent or reactive H-X bond
could yield a (N3N)ZrX complex.
Bonding. The bonding and electronic structure of these
complexes, key considerations for any catalytic reaction, were
investigated by DFT calculations performed at the B3LYP/
LACV3P**++ level of theory using the Jaguar program suite.
In these calculations, optimized gas-phase structures of com-
plexes 1-3 were obtained. Similarly structures of the hydride
(N3N)ZrH, a model alkynyl complex (N3N)ZrCt CH, and the
known, structurally characterized methyl derivative (N3N)ZrCH3
(16)27 were optimized. These optimized structures compared
very favorably with X-ray crystallographic structures for
complexes 2, 3, and 16. An X-ray diffraction study of 3 was
performed on single crystals grown from a concentrated ethereal
solution cooled to -30 °C, and a perspective view of the
molecular structure can be found in Figure 5. A comparison of
the bond lengths and angles for the solid-state structures of 2
and 3 with the gas-phase DFT calculation can be found in Table
6. The tabulated data show strong agreement between the
calculations and solid-state structures.
(27) MacMillan, S. N.; Tanski, J. M.; Waterman, R. Acta Crystallogr.
2008, E64, m477.

578 Organometallics, Vol. 28, No. 2, 2009
Roering et al.
Table 7. Bond Dissociation Energies (BDEs) for Gas-Phase
pair on the axial N in 16 is rather localized on the N (occupancy
) 1.85 electrons) with only slight stabilization (∼15 kcal/mol) by
delocalization into the Zr-C σ* orbital (giving a weak three-center/
four-electron bond). The use of second-order perturbative NBO
analysis for estimating such stabilization energies from the off-
diagonal Fock matrix element expressed in the NBO basis has been
thoroughly described and is now well established.28-30
Third, the amido substituents of the N3N ligand display strong
π-donation to the zirconium center, resulting in a significant,
though ionic, π-bonding component to the interaction. This
insight is also reflected in the solid-state data. The amido arms
are planar at nitrogen and have short Zr-N bonds of ca. 2.07
Å (2.045(1)-2.083(1) Å, Vide supra). A simple orbital analysis
on an idealized trigonal-bipyramidal complex suggests that three
π-bonds in the equatorial plane is symmetry forbidden. A more
tetrahedral-like geometry at zirconium, with the weakened bond
between zirconium and the amine nitrogen of the N3N ligand,
would relax the symmetry requirements of a trigonal-bipyra-
midal structure and allow for more ligand-to-metal π-donation
from the amido nitrogen atoms.
The degree of π-bonding from the N3N ligand likely limits
π-donation from the pseudoaxial X ligand to only the most
π-basic ligands. Therefore, all structurally characterized phos-
phido complexes have pyramidal phosphorus centers, and in
the two structurally characterized amido complexes, 10 and
(N3N)ZrNMe2,13 the nitrogen centers are planar. In the calcula-
tions, the phosphido ligands did not display any significant
π-bonding. It would be expected that the axial ligands that
display little π-donation may have weak M-X bonds and be
reactive. Indeed, the zirconium phosphido and arsenido deriva-
tives have exhibited substantial reactivity.9,11
Minimized Structures of Triamidoamine-Supported Zirconium
Complexesa
compound
(N3N)ZrH
(N3N)ZrCH3 (16)
(N3N)ZrPHPh (2)
(N3N)ZrPHCy (3)
(N3N)ZrCt CH
(N#3N)ZrH
(N#3N)ZrCH3
(N#3N)ZrNH2
(N#3N)ZrPH2
BDEb
81.7
72.4
55.4
56.2
126.0
78.1
71.8
102.8
65.3
56.5
58.5
(N#3N)ZrPHPh
(N#3N)ZrPHCy
a N# N ligand (N# N ) N(CH2CH2NSiH33-) was used in some cases
3
3
to simplify the computation. b BDEs (∆H298
)
in kcal mol-1 with
estimated errors of 1-2 kcal mol-1
.
( 0.5)], H2N-H [106.7 (107.57 ( 0.06)], (CH3)HP-H [80.4
(77.0 ( 3.0)], H3C-H [105.0 (105.0 ( 0.09)], and HCt C-H
[135.4 (133.32 ( 0.07)]. While it not clear whether such a
correction is required for transition-metal radicals, the same
correction was applied to the zirconium radicals to be consistent.
Because the zirconium radical is the same in a particular series
of BDE values, the trends in BDEs should be real, even if
absolute values may require further refinement. Though there
are no experimental BDE values with which to compare
computed numbers for this system, there is a decreasing order
of BDE for X ) -Ct CH > -H > -CH3, in agreement with
previously published work.32 Zirconium-phosphido bonds
(-PH2, -PHPh, and -PHCy) are significantly weaker than
corresponding bonds to -NH2, perhaps due to the increased
basicity of amido over phosphide and consistent with the greater
observed reactivity of the Zr-P linkage.4,10,12
In the original report of complexes 2 and 3, it was observed
that reaction of a zirconium precursor, such as 16, with an excess
of a 1-to-1 mixture of PhPH2 and CyPH2 gave exclusively 2.9
Now, with the solid-state structure of both complexes known,
it is clear that these complexes are isomorphous. The optimized
gas-phase structures of these two molecules are reported herein,
and the Zr-P BDEs for the two molecules are essentially
identical (Table 7). None of these new data allow a reasonable
hypothesis to be made as to the preferential formation of 2. It
seems likely that there are solution effects that are not being
accounted for in this analysis that may explain the observed
reactivity. Further study to understand the exact nature of the
bonding is clearly warranted for the potential impact on catalyst
design.
While cognizant of the caveats,31 careful application of DFT
does provide an opportunity to compute bond dissociation
enthalpies (BDEs), and the B3LYP functional coupled with the
LACV3P**++ basis set have been used to compute a large
number of such values for idealized transition metal com-
pounds.32 Consequently, the same techniques were applied to
evaluating BDE values for the Zr-X bonds in representative
(N3N)ZrX compounds. Values appear in Table 7, and correspond
to ∆H298 for the gas-phase homolytic process (eq 4),
(N3N)Zr-X f (N3N)Zr•+X•
(4)
for which zero-point energy corrected values of H298 were
calculated for all three components in their fully optimized and
unconstrained structures; radicals were treated using unrestricted
methods with the assumption of doublet ground states in all
cases. The “universal” B3LYP correction (addition of 2 kcal
Conclusion
mol-1 to the computed values of H298 33 was made for all radical
)
Complexes of the form (N3N)ZrX can be synthesized by
reaction of 1 with HX reagents. It appears that this reaction is
general, and it mimics the Zr products that would be expected
from reaction of (N3N)ZrH with HX. These zirconium com-
plexes have highly ionic bonding based on DFT calculations.
Solid-state and theoretical data on these complexes support little
ligand-to-metal π-donation except for amido derivatives where
significant π-bonding to the metal center can be observed. This
bonding is attributed to the greater π-basicity of the amido
ligand. The lack of strong π-bonding interactions between
zirconium and phosphido (and arsenido) ligands may give rise
to the observed reactivity.
species. Inclusion of this correction gave excellent agreement
between benchmark computational and experimental34 (in
parentheses) BDE values (kcal mol-1) for H2P-H [82.3 (83.9
(28) Leyssens, T.; Peeters, D.; Orpen, A. G.; Harvey, J. N. Organome-
tallics 2007, 26, 2637–2645.
(29) Leyssens, T.; Peeters, D. J. Org. Chem. 2008, 73, 2725–2730.
(30) Leyssens, T.; Peeters, D.; Orpen, A. G; Harvey, J. N. New J. Chem.
2005, 29, 1424–1430.
(31) Cundari, T. R., Crabtree, R., Mingos, M., Eds. The Application of
Modern Computational Chemistry Methods to Organometallic Systems. In
ComprehensiVe Organometallic Chemistry III; Elsevier Ltd., 2007; Vol. 1,
pp 639-669.
(32) Uddin, J.; Morales, C. M.; Maynard, J. H.; Landis, C. R.
Organometallics 2006, 25, 5566–5581.
Experimental Details
(33) Wright, J. S.; Rowley, C. N.; Chepelev, L. L. Mol. Phys. 2005,
103, 815–823.
(34) Luo, Y.-R. ComprehensiVe Handbook of Chemical Bond Energies;
CRC Press: Boca Raton, FL, 2007.
All manipulations were performed under a nitrogen atmosphere
with dry, oxygen-free solvents using an M. Braun glovebox or using

Zirconium Hydride Surrogate
Organometallics, Vol. 28, No. 2, 2009 579
standard Schlenk techniques. Benzene-d6 was purchased from
Cambridge Isotope Laboratory, then degassed and dried over NaK
alloy. Celite-454 was heated to a temperature greater than 180 °C
under dynamic vacuum for at least 8 h. Elemental analyses were
performed by Desert Analytics, Tucson, AZ. NMR spectra were
recorded with either a Bruker AXR or Varian 500 MHz spectrom-
eter in benzene-d6 and are reported with reference to residual solvent
resonances (δ 7.16 and 128.0) or external 85% H3PO4 (δ 0.0).
Infrared spectra were collected on a Perkin-Elmer System 2000
FT-IR spectrometer at a resolution of 1 cm-1. Starting materials
Zr(CH2Ph)4 and N(CH2CH2NHSiMe3)3 were prepared according
to literature procedures.35,36 All other chemicals were obtained from
commercial suppliers and dried by appropriate means.
Preparation of ([K5-N,N,N,N,C-(Me3SiNCH2CH2)2NCH2-
CH2NSiMe2CH2]Zr (1). A Schlenk tube was charged with Zr-
(PhCH2)4 (1.25 g, 2.75 mmol), N(CH2CH2NHSiMe3)3, (0.997 g,
2.75 mmol), and ca. 5 mL of toluene and fitted with a coldfinger.
The reaction was allowed to stir for ca. 3 h. Following removal of
the solvent under reduced pressure, the Schlenk tube was heated
in an oil bath under reduced pressure. After the residue reached
ca. 80 °C, the coldfinger was charged with a dry ice/acetone slurry.
Heating was continued to ca. 110 °C until sublimation deposited 1
as a colorless solid on the coldfinger (0.991 g, 2.20 mmol, 73%).
1H NMR (500.1 MHz): δ 3.600 (t, 2 H) 3.126 (m, 4 H), 2.344 (m,
6 H), 0.437 (s, 6 H, C H3), 0.314 (s, 2 H, ZrC H2), 0.274 (s, 18 H,
C H3). Complete data for complex 1 have been reported.9
Preparation of (N3N)ZrPHPh (2). A 50 mL round-bottom flask
was charged with 1 (100 mg, 0.222 mmol) and ca. 5 mL of benzene.
A 5 mL benzene solution of phenylphosphine (24.4 mg, 0.222
mmol) was added dropwise to the solution of 1, whereupon the
mixture immediately changed from colorless to red. After stirring
the red solution for 30 min, the solution was frozen and lyophilized
to give the title compound as an orange powder (1.068 g, 0.191
mmol, 86%). Spectroscopic data for 2 were identical to those
previously reported.9
change from colorless to pale yellow. The solution was stirred at
ambient temperature for 30 min followed by partial removal of
solvent under reduced pressure. The solution was filtered through
Celite and concentrated until incipient crystallization. The solids
were redissolved with gentle warming, and the homogeneous
solution was cooled to -30 °C to afford light orange crystals of 5
1
(0.182 g, 0.334 mmol, 76%). H NMR (500.1 MHz): δ 3.230 (t,
CH2, 6 H), 2.243 (t, CH2, 6 H), 2.730 (ddt, JPH ) 186 Hz, JHH
)
7.0 Hz, JHH ) 2.2 Hz, PH), 2.97 (dt, JPH ) 211 Hz, JHH ) 6.8 Hz,
PH2), 2.33 (m, CH2), 1.86 (m, CH2), 0.315 (s, CH3, 27 H). 13C{1H}
NMR (125.8 MHz): δ 63.03 (s, CH2), 47.31 (s, CH2), 20.63 (m,
CH2), 22.49 (pseudo d, CH2), 1.31 (s, CH3). 31P{1H} NMR (202.46
MHz): δ -72.4 (s, PH), -133 (s, PH2). IR (KBr, Nujol): 2349 s,
2340 s, 2283 m, 1411 m, 1376 s, 1335 s, 1261 s, 1245 s, 1144 s,
1052 w, 1024 m, 928 s, 897 m, 837 w, 782 s, 735 m, 678 s, 664 s,
624 s, 565 s, 453 m, 426 m cm-1. Anal. Calcd for C17H46N4P2Si3Zr:
C, 37.53; H, 8.52; N, 10.30. Found: C, 37.61; H, 8.45; N, 10.26.
Preparation of (N3N)ZrCt CPh (6). A 50 mL round-bottom
flask was charged with 1 (300 mg, 0.667 mmol), phenylacetylene
(68 mg, 0.667 mmol), and ca. 5 mL of benzene solvent. After
stirring the solution for 12 h the mixture was frozen, then
lyophilized, yielding 6 as a colorless powder (310 mg, 0.560 mmol,
84%). 1H NMR (500.1 MHz): δ 7.6159 (d, C6H5, 2 H), 6.920-6.994
(m, C6H5, 3 H), 3.275 (t, CH2, 6 H), 2.217 (t, CH2, 6 H), 0.421 (s,
CH3, 27 H). 13C{1H} NMR (125.8 MHz): δ 147.36 (s, Cꢀ) 131.77
(s, CR), 128.10 (s, CH), 127.09 (s, CH), 124.97 (s, CH) 106.54 (s
CH), 62.16 (s, CH2), 47.47 (s, CH2) 1.32 (s, CH3). IR (KBr, Nujol):
2076 s (νCC), 1594 w, 1485 s, 1377 w, 1260 s, 1244 s, 1204 m,
1143 m, 1061 s, 1022 m, 930 s, 903 m, 837 s, 782 s, 756m, 736 s,
679 w, 563 m, 538 m, 459 w cm-1. Anal. Calcd for C23H44N4Si3Zr:
C, 50.04; H, 8.03; N, 10.15. Found: C, 50.51; H, 8.14; N, 10.15.
Preparation of (N3N)ZrCt CSiMe3 (7). A 50 mL round-bottom
flask was charged with 1 (300 mg, 0.667 mmol), trimethylsily-
lacetylene (65.5 mg, 0.667 mmol), and ca. 5 mL of Et2O solvent.
After stirring for 2 h Et2O was removed under reduced pressure,
and the residue extracted into hexanes. The solution was filtered
through Celite, and the volume was reduced until incipient
crystallization. The solids were redissolved with gentle warming,
and the homogeneous solution was cooled to -30 °C, which yielded
colorless microcrystals of 7 (276 mg, 0.500 mmol, 75%). 1H NMR
(500.1 MHz): δ 3.225 (t, CH2, 6 H), 2.173 (t, CH2, 6 H), 0.396 (s,
CH3, 27H), 0.2623 (s, CH3, 9 H). 13C{1H} NMR (125.8 MHz): δ
169.97 (s, Cꢀ), 113.49 (s, CR), 62.52 (s, CH2), 47.70 (s, CH2), 1.62
(s, CH3), 0.37 (s, CH3). IR (KBr, Nujol): 2029 w (VCC), 1959 w,
1602 m, 1401 m, 1334 s, 1244 s, 1144 s, 1061 s, 1018 s, 930 s,
Preparation of (N3N)ZrPHCy (3). A 50 mL round-bottom flask
was charged with 1 (0.921 g, 2.00 mmol), and ca. 5 mL of benzene.
Cyclohexylphosphine (0.255 g, 2.20 mmol) was added dropwise.
The solution turned from colorless to yellow upon the addition of
PH2Cy. The solution was stirred for ca. 10 min. The solution was
then frozen and lyophilized, yielding yellow crystals of the title
compound (1.031 g, 1.780 mmol, 89%). Spectroscopic data for 3
were identical to those previously reported.9
Preparation of (N3N)ZrPPh2 (4). A 50 mL round-bottom flask
with charged with 1 (0.400 g, 0.889 mmol) and ca. 5 mL of
benzene. Ph2PH (0.166 g, 0.889 mmol) in ca. 5 mL of benzene
was added to the flask containing 1. The benzene solution turned
from colorless to red gradually over 10 min. The solution was
allowed to stir for an additional 2 h, then frozen and lyophilized to
give a red power of 4 (0.452 g, 0.771 mmol, 80%). 1H NMR (500.1
MHz): δ 7.732 (t, C6H5, 4 H), 7.173 (partially obscured by solvent),
6.932 (m, C6H5, 2 H), 3.312 (t, CH2, 6 H) 2.256 (t, CH2, 6 H),
0.209 (s, 27 H, C H3). 13C{1H} NMR (125.8 MHz): δ 142.03 (s,
Ph), 134.71 (d, JPC ) 13.8 Hz, Ph), 128.35 (d, JPC ) 8.8 Hz, Ph),
125.58 (s, Ph), 64.37 (s, CH2), 47.57 (s, CH2) 1.56 (s, CH3). 31P
NMR (202.4 MHz): δ 85.37 (s). IR (KBr, Nujol): 1578 s, 1560 m,
1432 m, 1333 m, 1299 w, 1264 s, 1243 s, 1143 m, 1047 s, 1026 s,
841 s cm-1
.
Preparation of (N3N)ZrCt CBu (8). A 50 mL round-bottom
flask was charged with 1 (0.200 g, 0.444 mmol), 1-hexyne (0.037
g, 0.444 mmol), and ca. 10 mL of Et2O. After stirring the solution
for 24 h, the solvent was removed under reduced pressure until
incipient crystallization. The solution was warmed to ambient
temperature, and then the homogeneous solution was cooled to -30
°C, which yielded colorless crystals of 8 (0.190 g, 0.355 mmol,
80%). 1H NMR (500.1 MHz): δ 3.245 (t, CH2, 6 H), 2.120 (t, CH2,
6 H), 2.135 (t, CH2, 2 H), 1.433 (m, CH2, 4 H), 0.837 (t, CH3, 3
H), 0.407 (s, CH3, 27 H). 13C NMR (125.8 MHz): δ 138.43 (s,
Cꢀ), 107.74 (s, CR), 62.41 (s, CH2), 47.68 (s, CH2), 31.49 (s, CH2),
22.31 (s, CH2), 19.88 (s, CH2), 13.78 (s, CH3), 1.69 (s, CH3). IR
(KBr, Nujol): 2081 (s, VCC), 1372 m, 1330 w, 1270 s, 1239 s,
1144 m, 1080 s, 928 m, 836 m, 779 w, 735 w cm-1. Anal. Calcd
for C21H48N4Si3Zr: C, 47.40; H, 9.09; N, 10.53. Found: C, 47.47;
H, 8.15; N, 10.63.
927 s, 898 s, 835 s, 780 s, 734 s, 696 s, 683 s, 565 m, 461 w cm-1
.
Anal. Calcd for C27H49N4PSi3Zr: C, 50.98; H, 7.76; N, 8.81. Found:
C, 50.75; H, 7.57; N, 8.95.
Preparation of (N3N)ZrPHC2H4PH2 (5). A scintillation vial
was charged with 1 (0.198 g, 0.440 mmol) and 2 mL of Et2O. To
the ethereal solution of 1, 1,2-diphosphinoethane (0.041 g, 0.440
mmol) in 2 mL of Et2O was added, resulting in an immediate color
Preparation of (N3N)ZrOPh (9). A 50 mL round-bottom flask
was charged with 1 (0.100 g, 0.222 mmol), phenol (0.021 g, 0.222
mmol), and ca. 5 mL of Et2O. The resulting solution was stirred
for 2.5 h, whereupon the solvent was removed under reduced
pressure until incipient crystallization. The solution was warmed
(35) Covert, K. J.; Mayol, A.-R.; Wolczanski, P. T. Inorg. Chim. Acta
1997, 263, 263–278.
(36) Gudat, D.; Verkade, J. G. Organometallics 1989, 8, 2772–2779.

580 Organometallics, Vol. 28, No. 2, 2009
Roering et al.
to ambient temperature and placed in a -30 °C freezer, yielding 9
as colorless microcrystals. The mother liquor was reduced further
to produce a second batch of crystals (103 mg, 0.189 mmol, 85%).
1H NMR (500.1 MHz): δ 7.191-7.210 (m, C6H5, 4 H), 6.810 (m,
C6H5, 1 H), 3.231 (t, CH2, 6 H), 2.3476 (t, CH2, 6 H), 0.236 (s,
CH3, 27 H). 13C NMR (125.8 MHz): δ 129.53 (s, CH), 128.28 (s,
CH), 121.11 (s, CH), 120.29 (s, C), 61.33 (s, CH2), 46.85 (s, CH2),
1.50 (s, CH3). IR (KBr, Nujol): 1590 m, 1493 m, 1480 s, 1281 s,
1244 s, 1162 w, 1061 m, 1022 w, 933 s, 906 m, 864 m, 837 s,
784 m, 757 m, 692 w, 625 w, 563 w cm-1. Anal. Calcd for
C21H44N4OSi3Zr: C, 46.36; H, 8.15; N, 10.30. Found: C, 46.06; H,
8.10; N, 10.15.
47.12 (s, CH2), 1.43 (s, CH3). IR (KBr, Nujol): 1641 (s, νCN),
1595 w, 1463 s, 1378 m, 1258 s, 1242 s, 1080 s, 1020 m, 931 s,
901 m, 837 s, 784 s, 741 m, 701 m, 631 m, 564 m, 478 w, 451 w
cm-1. Anal. Calcd for C28H49N5Si3Zr: C, 53.28; H, 7.82; N, 11.10.
Found: C, 52.98; H, 7.74; N, 11.18.
Preparation of (N3N)ZrSPh (14). A 50 mL round-bottom flask
was charged with 1 (0.106 g, 0.236 mmol) and 5 mL of Et2O, and
the solution was cooled to -30 °C. A cold 5 mL ethereal solution
of benzenethiol (0.026 g, 0.236 mmol) was added to the zirconium
solution. The resulting pale yellow solution was stirred for 24 h.
Volatile materials were removed under reduced pressure, and the
residue was redissolved in Et2O and filtered through Celite. The
solvent was removed again under reduced pressure, and the residue
was extracted into hexanes. The solution was concentrated, then
cooled to give colorless crystals of 14 (0.058 mg, 0.103 mmol,
Preparation of (N3N)NHtBu (10). A scintillation vial was
charged with 1 (0.049 g, 0.108 mmol) and ca. 2 mL of benzene.
To the solution of 1 was added a 2 mL benzene solution of tBuNH2
(0.009 g, 0.122 mmol), and the resulting solution stirred at ambient
temperature for 30 min. The solution was dried under reduced
pressure to give 10 as a colorless powder (0.021 g, 0.040 mmol,
1
44%). H NMR (500.1 MHz): δ 7.703 (d, CH, 2 H), 7.094 (m,
CH, 2 H), 7.079 (m, CH, 1 H), 3.256 (t, CH2, 6 H), 2.233 (t, CH2,
6 H), 0.2832 (s, CH3, 27 H). 13C{1H} NMR (125.8 MHz): δ 133.7
(s, Ph), 128.7 (s, Ph), 128.2 (s, Ph), 124.3 (s, Ph), 63.72 (s, CH2),
47.64 (s, CH2), 1.463 (s, CH3). IR (KBr, Nujol): 1578 m, 1377 s,
1245 s, 1143 w, 1085 w, 1056 m, 1024 m, 929 s, 900 sm, 837 s,
782 m, 737 m, 694 m, 563 s cm-1. Anal. Calcd for C21H44N4SSi3Zr:
C, 45.03; H, 7.92; N, 10.00. Found: C, 44.57; H, 7.90; N, 9.77.
Preparation of (N3N)ZrStBu (15). A 50 mL round-bottom flask
was charged with 1 (0.100 g, 0.222 mmol) and 5 mL of Et2O, and
the solution was cooled to -30 °C. A cold 5 mL Et2O solution of
tert-butylthiol (0.020 g, 0.222 mmol) was added to the zirconium
solution. The resulting pale yellow solution was stirred for 24 h.
Volatile materials were removed under reduced pressure, and the
residue was redissolved in Et2O and filtered through Celite. The
solvent was removed again under reduced pressure, and the resi-
due was extracted into a toluene/hexanes solvent mixture (20:1).
The solution was concentrated, then cooled to give colorless crystals
1
73%). H NMR (500.1 MHz): δ 3.856 (bs, N H, 1 H), 3.196 (t,
CH2, 6 H), 2.324 (t, CH2, 6 H), 1.451 (s, CH3, 9 H), 0.156 (s, CH3,
27 H). 13C{1H} NMR (125.8 MHz): 63.64 (s, CH2), 35.12 (s, CH2),
25.59 (s, C), 22.78 (s, CH3), 14.20 (s, CH3). IR (KBr, Nujol): 3210 s
(νNH), 2923 s, 2849 s, 1462 s, 1244 s cm-1. Anal. Calcd for
C19H49N5Si3Zr: C, 43.62; H, 9.44; N, 13.39. Found: C, 43.90; H,
9.88; N, 13.44.
Preparation of (N3N)ZrNHPh (11). A round-bottom flask was
charged with 1 (0.726 g, 1.61 mmol) and 2 mL of benzene. A
solution of aniline (0.151 g, 1.62 mmol) in 3 mL of benzene was
added to the Zr solution, and the resulting homogeneous solution
was stirred for 30 min before lyophilization. The resulting solid
was washed with pentane and decanted. The remaining colorless
solid was redissolved in benzene and filtered before being lyoph-
ilized to give 11 as a colorless powder (806 mg, 1.48 mmol, 92%).
1H NMR (500.1 MHz): δ 7.203 (t, C6H5, 2 H), 6.902 (d, C6H5, 2
H), 6.744 (d, C6H5, 1 H), 5.962 (NH, 1 H), 3.252 (t, CH2, 6 H),
2.312 (t, CH2, 6 H), 0.221 (s, CH3, 27 H). 13C{1H} NMR (125.8
MHz): δ 153.2 (s, CdN), 129.1 (s, Ph), 119.4 (s, Ph), 117.9 (s,
Ph), 61.79 (s, CH2), 47.07 (s, CH2), 1.48 (s, CH3). IR (KBr, Nujol):
3294 (s, νNH), 2921 s, 2341 s, 1597 s, 1489 s cm-1. Anal. Calcd
for C21H45N5Si3Zr: C, 46.44; H, 8.35; N, 12.90. Found: C, 47.06;
H, 8.49; N, 13.03.
Preparation of (N3N)ZrNHNdCPh2 (12). A round-bottom flask
was charged with 1 (0.052 g, 0.116 mmol), benzophenone hydra-
zone (0.023 g, 0.117 mmol), and 2 mL of Et2O. The resulting bright
yellow solution was stirred for 30 min at ambient temperature. The
volume of the solution was reduced until a solid appeared, and
then the solution was gently warmed to ambient temperature. The
homogeneous solution was cooled to -30 °C, yielding yellow
microcrystals of 12 (0.019 g, 0.029 mmol, 25%). 1H NMR (500.1
MHz): δ 8.068 (s, NH, 1 H), 7.820 (d, C6H5, 2 H), 7.604 (d, C6H5,
2 H), 7.290 (m, C6H5, 4 H), 7.146 (t, C6H5, 2 H), 3.508 (t, CH2, 6
H), 2.658 (t, CH2, 6 H), 0.2686 (s, CH3, 27 H). 13C{1H} NMR
(125.8 MHz): δ 148.10 (s, CdN), 137.72 (s, Ph), 132.20 (s, Ph),
127.26 (s, Ph), 125.60 (s, Ph), 58.86 (s, CH2), 40.16 (s, CH2), 0.22
(s, CH3). IR (KBr, Nujol): 3299 s (νNH), 2923 s, 2853 s, 1457 s,
1259 s cm-1. Anal. Calcd for C28H50N6Si3Zr: C, 52.04; H, 7.80; N,
13.00. Found: C, 52.20; H, 7.52; N, 12.45.
1
of 15 (0.077 mg, 0.142 mmol, 64%). H NMR (500.1 MHz): δ
3.194 (t, CH2, 6 H), 2.296 (t, CH2, 6 H), 1.724 (s, CH3, 9 H), 0.389
(s, CH3, 27 H). 13C{1H} NMR (125.8 MHz): δ 64.67 (s, CH2),
50.25 (s, C(CH3)3), 47.81 (s, CH2), 37.32 (s, CH3), 1.94 (s, CH3).
IR (KBr, Nujol): 1376 s, 1360 s, 1335 m, 1300 w, 1260 s 1244 s,
1147 m, 1056 s, 1023 m, 930 s, 838 s, 783 m, 736 m, 674 s, 623 s,
587 s, 564 s cm-1. Anal. Calcd for C19H48N4SSi3Zr: C, 42.25, H,
8.96, N, 10.37. Found: C, 42.07; H, 8.91, N, 10.75.
DFT Computations. Gas-phase structures were optimized using
the hybrid B3LYP functional37,38 and the triple-ꢁ LACV3P**++
basis set,39-42 which uses extended core potentials on heavy atoms
and a 6-311G**++ basis for other atoms, as implemented in the
Jaguar43 suite of programs. Natural bond orbital (NBO), natural
resonance theory (NRT), and natural population analyses (NPA)44-49
are also integrated with the Jaguar program and were run on
B3LYP/LACV3P**++ optimized structures. Radical species were
(37) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5652.
(38) Becke, A. D. J. Chem. Phys. 1993, 98, 1372–1377.
(39) Dunning, T. H.; Hay, P. J. Modern Theoretical Chemistry, Vol. 4:
Applications of Electronic Structure Theory; Plenum: New York, 1977; p
461.
(40) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 270–283.
(41) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299–310.
(42) Wadt, W. R.; Hay, P. J. J. Chem. Phys. 1985, 82, 284–298.
(43) Jaguar, version 7.0; Schro¨dinger, LLC: New York, 2007.
(44) Weinhold, F.; Landis, C. R. Valency and Bonding: A Natural Bond
Orbital Donor-Acceptor PerspectiVe, 1st ed.; Cambridge University Press:
Cambridge, 2005.
(45) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. ReV. 1988, 88,
899–926.
(46) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985,
83, 735–746.
(47) Foster, J. P.; Weinhold, F. J. Am. Chem. Soc. 1980, 102, 7211–
7218.
Preparation of (N3N)ZrNdCPh2 (13). A solution of benzophe-
none imine (0.009 g, 0.052 mmol) in 2 mL of benzene was added
to a 2 mL solution of 1 (26 mg, 0.058 mmol) in benzene at ambient
temperature, whereupon the colorless solution immediately changed
to orange. The clear orange solution was stirred for 2 h, then
1
lyophilized to an orange powder (0.027 g, 0.042 mmol, 82%). H
NMR (500.1 MHz): δ 7.297 (m, C6H5, 4 H), 7.114 (m, C6H5, 6
H), 3.488 (t, CH2, 6 H), 2.577 (t, CH2, 6 H), 0.256 (s, CH3, 27 H).
13C{1H} NMR (125.8 MHz): δ 174.81 (s, CdN), 142.08 (s, C6H5),
133.91 (s, C6H5), 129.30 (s, C6H5), 128.45 (s, C6H5), 61.42 (s, CH2),
(48) Reed, A. E.; Weinhold, F. J. Chem. Phys. 1983, 78, 4066–4073.
(49) Glendening, E. D.; Badenhoop, J. K.; Reed, A. K.; Carpenter, J. E.;
Bohmann, J. A.; Morales, C. M.; Weinhold, F. NBO 5.0; Theoretical
Chemistry Institute, University of Wisconsin: Madison, 2001.

Zirconium Hydride Surrogate
Organometallics, Vol. 28, No. 2, 2009 581
Table 8. Crystal Data and Structure Refinement Parameters for 3, 4, 6, 10, and 15
3
4
5
10
15
formula
M
C21H51N4PSi3Zr
566.12
C27H49N4PSi3Zr
636.16
C17H46N4P2Si3Zr
544.01
C19H49N5Si3Zr
523.12
C19H48N4SSi3Zr
540.16
cryst syst
color
a/Å
b/Å
c/Å
R/deg
ꢀ/deg
γ/deg
unit cell vol/Å3
space group
Z
orthorhombic
yellow
16.8829(9)
17.589(1)
20.347(1)
90
90
90
6042.1(6)
Pbca
8
monoclinic
orange
10.7744(7)
16.648(1)
19.357(1)
90
102.495(1)
90
3389.9(4)
P21/n
orthorhombic
orange
17.0290(9)
17.1007(9)
20.194(1)
90
90
90
5880.6(5)
Pbca
8
trigonal
colorless
11.9522(6)
11.9522(6)
11.6463(5)
90
90
120
1440.83(12)
P31/c
2
monoclinic
colorless
11.8438(5)
20.1540(8)
12.0195(5)
90
95.389(1)
90
2856.4(2)
P21/n
4
4
θ range/deg
µ/mm-1
N
1.95 to 30.46
0.551
81714
1.63 to 28.28
0.499
43983
1.97 to 26.62
0.615
64367
2.63 to 28.24
0.520
18203
1.98 to 28.70
0.596
38319
Nind
8889
8406
6160
2393
7374
Rint
0.0610
0.0320
0.0649
0.436; -0.378
1.007
0.0314
0.0250
0.0586
0.401; -0.280
1.045
0.0430
0.0283
0.0640
0.388; 0.543
1.054
0.0207
0.0144
0.0381
0.197; -0.188
1.068
0.0248
0.0212
0.0548
0.371; -0.282
1.045
a
R1 (I > 2σ(I))
b
wR2 (I > 2σ(I))
∆Fmax; ∆Fmin/e Å3
GoF on R1
b
a R1 ) |Fo| - | Fc|/∑|Fo|. wR2 ) {∑[w(Fo - Fc2)2]/∑[w(Fo )2]}1/2
.
2
2
t
optimized as doublets using unrestricted occupancy methods, and
the absence of significant spin contamination was confirmed by
checking the values of S2 against the expected doublet values of
0.750. All computed structures were confirmed as energy minima
by calculating the vibrational frequencies by second-derivative
analytic methods and confirming the absence of imaginary frequen-
cies. Thermodynamic quantities were calculated assuming an ideal
gas and are zero-point energy corrected. The “universal” B3LYP
correction33 was made for all radical species.34
X-ray Structure Determinations. X-ray diffraction data were
collected on a Bruker APEX 2 CCD platform diffractometer (Mo
ΚR, λ ) 0.71073 Å) at 125 K. Suitable crystals of each complex
were mounted in a nylon loop with Paratone-N cryoprotectant oil.
The structures were solved using direct methods and standard
difference map techniques and were refined by full-matrix least-
squares procedures on F2 with SHELXTL (version 6.14).50 All non-
hydrogen atoms were refined anisotropically. Hydrogen atoms on
carbon were included in calculated positions and were refined using
a riding model. The hydrogen atom on the phosphorus of 3, H(1),
and all hydrogen atoms on phosphorus atoms of 5, H(1), H(2), and
H(3), were located in the Fourier difference map and refined freely.
The hydrogen atom on the nitrogen of 10, H(1), was located in the
Fourier difference map and refined semifreely with the help of a
distance restraint. Crystal data and refinement details are presented
in Table 8. The cyclohexyl phosphido ligand of 3 exhibited a 2-fold
disorder that was modeled successfully, refining freely to 65/35
occupancy. The Bu amido of 10 exhibited disorder about the
crystallographic 3-fold axis that was modeled as a perfect 3-fold
disorder and refined with the help of distance restraints (SADI),
similarity restraints on displacement parameters (SIMU), and rigid
bond restraints on 1-2 and 1-3 distances and anisotropic displace-
ment parameters (DELU).
Acknowledgment. This work was supported by the U.S.
National Science Foundation (CHE-0747612 to R.W. and
CHE-0518170 to R.P.H.), the American Chemical Society
Petroleum Research Fund (466669-G3 to R.W.), and the
University of Vermont (UVM). Summer support for S.M.C.
was provided by the American Chemical Society (ACS)
Project SEED and the Green Mountain ACS Local Section.
Research support for J.J.D. was provided by a UVM
Undergraduate Research Endeavors Competitive Award.
X-ray crystallographic facilities were provided by the U.S.
National Science Foundation (CHE-0521237 to J.M.T.).
R.P.H. also thanks Professor Clark Landis (University of
Wisconsin) for valuable discussions.
Supporting Information Available: . Crystallographic data for
complexes 3, 4, 5, 10, and 15 (CIF) and tables of computational
results (pdf). These materials are available free of charge via the
Internet at http://pubs.acs.org.
(50) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112–122.
OM8008684