Crystal packing forces dictate η1- versus η2-coordination of benzyl groups in [guanidinate]zr(ch2ph)3

Article abstract of DOI:10.1021/om000123s

An X-ray structure determination of {CyNC-[N(SiMe₃)₂]NCy}Zr(CH₂Ph)₃, crystallized from pentane, differs from the solid-state structure previously reported for the same compound crystallized from toluene. In cont

Full text of DOI:10.1021/om000123s

Organometallics 2000, 19, 2809-2812
2809
Cr ysta l P a ck in g F or ces Dicta te η1- ver su s
η2-Coor d in a tion of Ben zyl Gr ou p s in
[Gu a n id in a te]Zr (CH2P h )3
Garth R. Giesbrecht, Glenn D. Whitener, and J ohn Arnold*
Department of Chemistry, University of California, Berkeley, and the Chemical Sciences
Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720-1460
Received February 9, 2000
Summary: An X-ray structure determination of {CyNC-
[N(SiMe3)2]NCy}Zr(CH2Ph)3, crystallized from pentane,
differs from the solid-state structure previously reported
for the same compound crystallized from toluene. In
contrast to the earlier report, we find no evidence for the
presence of an η2 interaction in the solid state by X-ray
electrophilicity of the metal center and, by the same
argument, the electron-donating ability of the ancillary
ligands. Nevertheless, in the absence of confirming
1
evidence by, for example, H or 13C NMR spectroscopy,
the difference in energy between η1 and η2 bonding
modes is uncertain.14,15
1
crystallography or in solution by means of H or 13C
Here we compare two independent crystallographic
determinations of the same zirconium tribenzyl com-
plex: in one, an η2 interaction is present; in the other
there is no evidence for such behavior. These data
confirm that simple crystal packing forces may be
sufficient to change hapticitites in metal benzyl com-
plexes; they further serve to highlight the difficulties
associated with using crystal structure data to reach
conclusions regarding chemical bonding, especially where
the energy differences involved are small.
NMR spectroscopy.
In tr od u ction
Benzyl ligands (Bz) are well-established in organo-
metallic chemistry and are distinct from other alkyls
by virtue of their ability to exhibit a range of hapticities
when binding to metals. In addition to the simple η1
case, bonding to the metal through the π-orbitals of the
aromatic ring may also occur, resulting in η2, η3, or even
η7 binding modes.1-4 In cases where two or more Bz
ligands are coordinated to a single metal, differing
modes have been seen2,4-13 and many studies of the
fluxional behavior of such species in solution have been
reported.
Resu lts a n d Discu ssion
Richeson and co-workers recently described the prepa-
ration of {CyNC[N(SiMe3)2]NCy}Zr(CH2Ph)3,16 accord-
ing to eq 1.
[{CyNC[N(SiMe3)2]NCy}ZrCl4][Li(thf)2](ether) +
toluene, RT
3PhCH2MgCl -LiCl, -3MgCl 8
2
{CyNC[N(SiMe3)2]NCy}Zr(η2-CH2Ph)(CH2Ph)2
(1)
A
It is generally believed that observation of η2 Bz
groups in the solid state is a good indicator of the
Orange crystals were obtained from concentrated tolu-
ene at -34 °C in good yield. An X-ray crystal structure
revealed an acute Zr-C-Cipso angle for one of the benzyl
groups, implying a single benzyl group bound to the
metal center in an η2 fashion and two η1 benzyl groups.
We had independently synthesized the same com-
pound via an alkane elimination reaction between the
guanidine {CyNC[N(SiMe3)2]NCy}H and Zr(CH2Ph)4
(1) Scholz, J .; Rehbaum, F.; Thiele, K. H.; Goddard, R.; Betz, P.;
Kruger, C. J . Organomet. Chem. 1993, 443, 93.
(2) Legzdins, P.; J ones, R. H.; Phillips, E. C.; Yee, V. C.; Trotter, J .;
Einstein, F. W. B. Organometallics 1991, 10, 986.
(3) Mintz, E. A.; Moloy, K. G.; Marks, T. J . J . Am. Chem. Soc. 1982,
104, 4692.
(4) Pellecchia, C.; Immirzi, A.; Pappalardo, D.; Peluso, A. Organo-
metallics 1994, 13, 3773.
(5) Ciruelo, G.; Cuenca, T.; Gomez, R.; Gomez-Sal, P.; Martin, A.;
Rodriguez, G.; Royo, P. J . Organomet. Chem. 1997, 547, 287.
(6) Pellecchia, C.; Grassi, A.; Immirzi, A. J . Am. Chem. Soc. 1993,
115, 1160.
(7) Girolami, G. S.; Wilkinson, G.; Thornton-Pett, M.; Hursthouse,
M. B. J . Chem. Soc., Dalton Trans. 1984, 2789.
(8) Bei, X.; Swenson, D. C.; J ordan, R. F. Organometallics 1997, 16,
3282.
(9) Tedesco, C.; Immirzi, A.; Proto, A. Acta Crystallogr. 1998, B54,
431.
(10) Tsukahara, T.; Swenson, D. C.; J ordan, R. F. Organometallics
1997, 16, 3303.
(11) Latesky, S. L.; McMullen, A. K.; Niccolai, G. P.; Rothwell, I.
P.; Huffman, J . C. Organometallics 1985, 4, 902.
(12) Edwards, P. G.; Andersen, R. A.; Zalkin, A. Organometallics
1984, 3, 293.
(14) J ordan et al. have estimated the η2-η1 isomerization barrier
to be 14.0 kcal/mol for a cationic zirconium species; a lower value could
be expected for a neutral complex (see ref 8). Legzdins et al. have
determined a barrier of ∼10 kcal/mol for a series of molybdenum and
tungsten bis(benzyl) nitrosyl complexes (see ref 2); similarly, the
tribenzyl species Cp*M(CH2Ph)3 (M ) Th, U) exhibit barriers of <10
kcal/mol (ref 3).
(15) A search of the Cambridge Structural Database reveals 44
crystallographically characterized zirconium benzyl species. Defining
an η2 interaction as one in which the Zr-C-Cipso angle is less than
100°, and the Zr-Cipso distance is less than 2.90 Å, then 11 compounds
fall into this category. Of these, seven are neutral species, five of which
have been shown by 1H and/or 13C NMR spectroscopy to maintain an
additional interaction with the metal center in solution.
(16) Wood, D.; Yap, G. P. A.; Richeson, D. S. Inorg. Chem. 1999, 38,
5788.
(13) Mena, M.; Pellinghelli, M. A.; Royo, P.; Serrano, R.; Tiripicchio,
A. J . Chem. Soc., Chem. Commun. 1986, 1118.
10.1021/om000123s CCC: $19.00 © 2000 American Chemical Society
Publication on Web 05/27/2000

2810 Organometallics, Vol. 19, No. 14, 2000
Notes
Gross structural features are the same in both
determinations: the zirconium centers are ligated by a
single bidentate guanidinate ligand and three benzyl
groups, and bond lengths and angles in the guanidinate
ligands are similar in both cases. Additionally, the
dihedral angle between the N(SiMe3)2 and NCN frag-
ments of the ligands are comparable (85.8(3)° for A,
82.0(1)° for B), suggesting that the contribution of a
zwitterionic resonance structure (II) is negligible is both
cases. As a result, disparities between the two struc-
tures cannot be attributed to differences in coordination
behavior of the ancillary guanidinate ligand.
In A, two of the benzyl groups are bound to the metal
in an η1 manner (Zr1-C27-C33 ) 116.0(5)°, Zr1-C34-
C40 ) 123.2(4)°), while the third displays an acute Zr-
C-Cipso angle consistent with η2 coordination (Zr1-
C20-C26 ) 88.7(4)°). Additionally, two of the benzyl
groups in A (C20-C26, C34-C40) are oriented away
from the guanidinate ligand, while the third (C27-C33)
is bent toward one of the cyclohexyl groups of the
ancillary ligand. In contrast, orientations of the benzyl
groups in B are more conventional: all three benzyl
groups are bound to the metal in an η1 fashion (Zr-C-
Cipso ) 104.6(2)-115.2(2)°) and are pointed away from
the ancillary ligand. The Zr-CH2 bond distances in B
(2.255(2)-2.289(3) Å) are close to those cited for A
(2.263(7)-2.288(6) Å), implying that, electronically, the
zirconium centers are similar. Distances between the
metal and the ipso carbons of the benzyl groups range
from 3.006(2) to 3.221(2) Å and are unremarkable;
nevertheless, in A, a short Zr-Cipso distance of 2.672(2)
Å is present, consistent with an η2-coordinated benzyl
group. (The remaining Zr-Cipso distances are 3.187 and
3.345 Å.)
F igu r e 1. ORTEP views of {CyNC[N(SiMe3)2]NCy}Zr-
(CH2Ph)3 crystallized from toluene (A)16 and pentane (B)
drawn with 30% probability ellipsoids.
13C and 1H NMR data of the compound prepared
according to eq 2 are identical with those previously
reported, indicating a single cyclohexyl substituent and
a single benzyl group in solution, and in neither case is
there any evidence for an η2-benzyl interaction. In cases
where significant π-bonding occurs between the aryl
group and the zirconium center in solution, Rothwell
has determined that the ortho protons of the benzyl
groups tend to shift upfield (>6.8 ppm).11 In addition,
the 1J CH of the -CH2 groups increases (>130 Hz) as the
Zr-C-Ph angle decreases, mirroring the increase in the
amount of sp2 character of the R-carbon. Solid-state
structures of related polybenzyl complexes frequently
contain benzyl ligands that are distorted to varying
degrees; still, these are equivalent in solution, where
time-averaged NMR properties are observed.17 For
example, the solid-state structure of Cp*Ti(CH2Ph)3
shows a single distorted -CH2Ph ligand13,18 and exhibits
a high-field ortho hydrogen resonance (δ 6.57) in solu-
(eq 2); in our case, cooling a concentrated pentane
solution to -30 °C also resulted in orange crystals that
were suitable for X-ray diffraction.
{CyNC[N(SiMe3)2]NCy}H +
Zr(CH2Ph)4 ether, -78 °C8
-toluene
{CyNC[N(SiMe3)2]NCy}Zr(CH2Ph)3
(2)
B
Physical and spectroscopic properties of the com-
pounds are identical. ORTEP views of the solid-state
structures A and B, shown in Figure 1, appear almost
identical apart from the disposition of the benzyl groups
(see below). Both crystallize in the space group P21/n,
with four molecules per unit cell and roughly equivalent
cell volumes. The axes, a, b, and c, and the angle â are
slightly different in the two polymorphs. Importantly,
contents of the unit cell are identical: there is no
evidence for the inclusion of solvent or other molecules
in either lattice.
(17) Crowther, D. J .; J ordan, R. F.; Baenziger, N. C.; Verma, A.
Organometallics 1990, 9, 2574.
(18) Mena, M.; Royo, P.; Serrano, R.; Pellinghelli, M. A.; Tiripicchio,
A. Organometallics 1989, 8, 476.

Notes
Ta ble 1. Cr ysta llogr a p h ic Da ta for A a n d B
Organometallics, Vol. 19, No. 14, 2000 2811
Ta ble 2. Bon d Len gth s (Å) for A a n d B
A
B
A
B
temp, K
λ, Å
space group
a, Å
b, Å
203(2)
0.71073
P21/n
10.483(1)
18.889(2)
20.459(3)
102.431(2)
3956.0(9)
4
1.228
0.369
0.0473
0.0918
4145
153(2)
0.71073
P21/n
10.810(1)
23.778(1)
15.755(1)
97.144(1)
4018.1(2)
4
1.19
0.45
0.026
0.031
Zr1-N1
Zr1-N2
Zr1-C20
Zr1-C27
Zr1-C34
N1-C6
2.241(4)
2.250(5)
2.268(5)
2.263(7)
2.288(6)
1.465(6)
1.461(6)
1.339(7)
1.344(7)
1.421(7)
1.756(5)
1.775(4)
1.466(8)
1.463(9)
1.491(8)
2.672(6)
3.187
2.249(2)
2.203(2)
2.289(3)
2.255(2)
2.266(3)
1.468(3)
1.471(3)
1.340(3)
1.337(3)
1.402(3)
1.776(2)
1.772(2)
1.491(4)
1.488(3)
1.483(4)
3.221(2)
3.198(2)
3.006(2)
c, Å
â, deg
N2-C12
N1-C13
N2-C13
N3-C13
N3-Si1
volume, Å3
Z
dcalcd, g/cm3
abs coeff, mm-1
R
N3-Si2
Rw
C20-C26
C27-C33
C34-C40
Zr1-C26
Zr1-C33
Zr1-C40
no. of reflns
no. of params
5213
659
415
tion, but a normal sp3 1J CH value of 122 Hz. In our case,
the chemical shift of the ortho protons in the H NMR
1
3.345
1
spectrum is unexceptional (δ 6.94), as is the J CH (123
Ta ble 3. Bon d An gles (d eg) for A a n d B
Hz). These data show that an η2 interaction is not
maintained in solution.
A
B
Since there is no spectroscopic evidence for an η2-
coordinated benzyl group in solution for the present
case, what is the source of this interaction in the solid-
state structure of A? A contribution from a zwitterionic
resonance structure (II), which would render the metal
center less electrophilic and less prone to engage in
additional interactions with benzyl groups, may be ruled
out due to the similar dihedral angles between N(SiMe3)2
and NCN planes in A and B. It appears that the
presence of an η2-coordinated benzyl group in the solid-
state structure of A is simply due to crystal packing
effects resulting from the slight difference in crystal-
lization conditions as compared to B. Thus, in the
absence of corroborating evidence from solution spec-
troscopy, conclusions based on X-ray crystallography
regarding electronic effects, and their influence on
chemical bonding, must be viewed with caution.
N1-Zr1-N2
N1-Zr1-C20
N1-Zr1-C27
N1-Zr1-C34
N2-Zr1-C20
N2-Zr1-C27
N1-Zr1-C34
C20-Zr1-C27
C20-Zr1-C34
C27-Zr1-C34
Zr1-C20-C26
Zr1-C27-C33
Zr1-C34-C40
N1-C13-N2
N1-C13-N3
N2-C13-N3
C13-N3-Si1
C13-N3-Si2
Si1-N3-Si2
59.77(17)
95.6(2)
59.38(6)
86.09(8)
108.46(9)
130.64(9)
139.54(8)
98.72(8)
94.25(8)
112.8(1)
93.4(1)
116.9(1)
115.2(2)
115.9(2)
104.6(2)
110.9(2)
125.0(2)
124.1(2)
117.9(1)
119.5(1)
118.1(1)
111.7(3)
138.9(2)
126.4(2)
113.8(2)
84.0(2)
119.6(3)
90.7(2)
99.9(3)
88.7(4)
116.0(5)
123.2(4)
113.1(5)
123.6(5)
123.3(6)
117.1(3)
119.2(4)
123.7(3)
was stirred at room temperature overnight. To this was added
water (0.50 g, 27.8 mmol) via syringe, causing the solution to
gradually turn cloudy. This was then stirred for 24 h. The
mixture was filtered through a frit lined with Celite to remove
LiCl to yield a pale yellow solution. Removal of the solvent
under vacuum gave a golden oil, which yielded colorless plates
upon mechanical agitation (4.92 g, 50%). Mp: 58-60 °C. 1H
NMR (C6D6): δ 3.84 (m, 1H, unique CyH), 3.45 (m, 1H, unique
Exp er im en ta l Section
Gen er al Con sider ation s. Standard Schlenk-line and glove-
box techniques were used unless stated otherwise. Diethyl
ether and pentane were purified by passage through a column
of activated alumina and degassed with argon.19 C6D6 was
vacuum transferred from sodium/benzophenone. 1,3-Dicyclo-
hexylcarbodiimide was purchased from Sigma and distilled
prior to use. LiN(SiMe3)2 and Zr(CH2Ph)4 were prepared
according to literature procedures.20,21 Melting points were
determined in sealed capillary tubes under nitrogen and are
uncorrected. 1H and 13C{1H} NMR spectra were recorded at
3
CyH), 3.09 (d, 1H, NH, J H-H ) 7.0 Hz), 2.11-0.90 (m, 20H,
C6H10), 0.21 (s, 18H, SiMe3). 13C{1H} NMR (C6D6): δ 147.22
(NCN), 56.95, 50.00, 36.21, 34.89, 33.92, 26.91, 26.76 and 25.73
(C6H11), 2.32 (SiMe3). IR (cm-1): 3441 (NH, m), 1636 (CdN,
s), 1346 (w), 1296 (m), 1251 (s), 1220 (s), 1154 (w), 1118 (m),
1009 (m), 956 (s), 938 (s), 889 (sh), 868 (sh), 840 (s), 757 (m),
684 (m), 639 (w). Anal. Calcd for C19H41N3Si2: C, 62.06; H,
11.24; N, 11.43. Found: C, 62.44; H, 11.13; N, 11.64.
1
ambient temperature on a Bruker AM-300 spectrometer. H
NMR chemical shifts are given relative to C6D5H (7.15 ppm).
13C{1H} NMR spectra are relative to C6D6 (128.39 ppm). IR
samples were prepared as Nujol mulls and taken between KBr
plates. Elemental analyses were determined at the College of
Chemistry, University of California, Berkeley. Single-crystal
X-ray structure determination was performed at CHEXRAY,
University of California, Berkeley.
{CyNC[N(SiMe3)2]NCy}H. 1,3-Dicyclohexylcarbodiimide
(5.52 g, 26.8 mmol) and LiN(SiMe3)2 (4.48 g, 26.8 mmol) were
combined in a 500 mL round-bottomed flask, and 300 mL of
diethyl ether was added. The resultant pale yellow solution
{CyNC[N(SiMe3)2]NCy}Zr (CH2P h )3. To an ethereal solu-
tion (30 mL) of Zr(CH2Ph)4 (2.01 g, 4.41 mmol) maintained at
-78 °C was added {[CyNC[N(SiMe3)2]NCy}H (1.62 g, 4.41
mmol) in diethyl ether (20 mL) via cannula. The reaction
mixture was left to stir and warm to room temperature
overnight. The solvent was then removed under vacuum and
the resultant yellow solid extracted with pentane. Cooling to
-30 °C yielded large yellow blocks (3.09 g, 96%). Mp: 141-
143 °C. 1H NMR (C6D6): δ 7.22-6.91 (m, 15H, C6H5), 3.50 (m,
2H, unique CyH), 2.50 (s, 6H, CH2), 1.70-0.90 (m, 20H, C6H10),
0.18 (s, 18H, SiMe3). 13C{1H} NMR (C6D6): δ 174.55 (NCN),
145.13, 129.80, 128.58 and 123.20 (C6H5), 77.77 (CH2), 56.75,
35.43, 26.81 and 26.04 (C6H11), 2.72 (SiMe3). IR (cm-1): 1589
(CdN, s), 1309 (m), 1253 (s), 1195 (s), 1139 (m), 1062 (w), 1025
(sh), 1014 (m), 971 (m), 929 (s), 860 (m), 840 (s), 825 (s), 798
(19) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J . Organometallics 1996, 15, 1518.
(20) Kruger, C. R.; Niederprum, H. Inorganic Syntheses; McGraw-
Hill: New York, 1966; Vol. 8.
(21) Zucchini, U.; Giannini, U.; Albizzati, E. J . Organomet. Chem.
1971, 26, 357.

2812 Organometallics, Vol. 19, No. 14, 2000
Notes
(sh), 759 (w), 742 (s), 694 (s), 671 (sh), 642 (s). Anal. Calcd. for
Su p p or tin g In for m a tion Ava ila ble: Full details of the
single-crystal X-ray analysis of {CyNC[N(SiMe3)2]NCy}Zr(CH2-
Ph)3. This material is available free of charge via the Internet
at http://pubs.acs.org.
C
40H61N3Si2Zr: C, 65.69; H, 8.41; N, 5.75. Found: C, 64.78;
H, 8.69; N, 5.87.
Ack n ow led gm en t. We thank the NSF for the
award of a predoctoral fellowship to G.D.W. and the
DOE for support of this work.
OM000123S