Insertion of isocyanides into group 4 metal-carbon and metal-nitrogen bonds. Syntheses and DFT calculations

Article abstract of DOI:10.1021/om030211n

The reactions of [Ti(η⁵-Ind)(NMe₂)₂Me], [Ti(η⁵-Ind)(NMe₂)Cl₂], and [Zr(η⁵-Ind)(NMe₂)₂Cl] with 2,6-dimethylphenyl isocyanide (CN(2,6-Me₂Ph)) and t

Full text of DOI:10.1021/om030211n

4218
Organometallics 2003, 22, 4218-4228
In ser tion of Isocya n id es in to Gr ou p 4 Meta l-Ca r bon a n d
Meta l-Nitr ogen Bon d s. Syn th eses a n d DF T Ca lcu la tion s
Ana M. Martins,* J ose´ R. Ascenso, Cristina G. de Azevedo, Alberto R. Dias,
M. Teresa Duarte, J oa˜o F. da Silva, Lu´ıs F. Veiros, and Sandra S. Rodrigues
Centro de Qu´ımica Estrutural, Instituto Superior Te´cnico, Avenida Rovisco Pais no. 1,
1049-001 Lisboa, Portugal
Received March 19, 2003
The reactions of [Ti(η5-Ind)(NMe2)2Me], [Ti(η5-Ind)(NMe2)Cl2], and [Zr(η5-Ind)(NMe2)2Cl]
with 2,6-dimethylphenyl isocyanide (CN(2,6-Me2Ph)) and tert-butyl isocyanide (CN-t-Bu) have
been investigated. For the first complex, insertion occurs exclusively into the Ti-C bond to
give [Ti(η5-Ind)(NMe2)2{C(Me)dN-t-Bu}] (1), which, according to DFT results, is due to the
weakness of the Ti-CH3 bond when compared to Ti-NR2. [Ti(η5-Ind)(NMe2)Cl2] was found
to react with 1 equiv of CN(2,6-Me2Ph) to yield [Ti(η5-Ind){C(NMe2)dN(2,6-Me2Ph)}Cl2] (2).
Two equivalents of 2,6-dimethylphenyl isocyanide reacts with [Zr(η5-Ind)(NMe2)2Cl] to give
the double insertion compound [Zr(η5-Ind){C(NMe2)dN(2,6-Me2Ph)}2Cl] (3). The same
reaction performed with CN-t-Bu proceeds slowly and enabled the characterization, by NMR,
of [Zr(η5-Ind)(NMe2){C(NMe2)dN-t-Bu}Cl] (4) and [Zr(η5-Ind){C(NMe2)dN-t-Bu}2Cl] (5). The
molecular structures of 2 and 3 have been determined by X-ray diffraction. DFT calculations
of the insertion of CNH in the zirconium nitrogen bonds of [Zr(η5-Ind)(NR2)2Cl] (R ) H, Me)
have been performed and show that the progressive stability of the insertion products
accounts for the experimentally found double insertion product.
In tr od u ction
of transition metal-carbon bonds stimulated the syn-
thesis of transition metal-amido complexes aiming at
the comparison between the reactivity of metal-carbon
and metal-nitrogen bonds. This goal was largely frus-
trated by the lack of reactivity of many metal-amido
compounds toward insertion, and few examples of
metal-iminocarbamoyl derivatives, formed by insertion
of isocyanides into metal-amido bonds, have been
reported.6 These reactions are important in the general
context of the amination of organic substrates for which
titanium and zirconium imido complexes proved to be
useful catalysts.7
In this paper we report the insertion of CN(2,6-
Me2Ph) and CN-t-Bu into metal-carbon and metal-
nitrogen bonds of titanium and zirconium indenyl amido
complexes previously reported.8 The reaction mecha-
nism for the isocyanide insertion into Zr-N bonds was
studied by ab initio9 and DFT10 calculations, and the
Due to its implication in organic and organometallic
syntheses and also to its role in homogeneous catalysis,
the insertion of unsaturated reagents into M-C bonds
of transition metal complexes is one of the most impor-
tant chemical reactions. Among a great diversity of
substrates, the insertion of isocyanides was subject to
extensive studies.1 The participation of the resulting
iminoacyl ligands in several chemical transformations
is very wide, including reactions such as isomerization
into vinylamido and η3-azallyl,2 C-C coupling leading
to imines3 and enediamido or enamidolate ligands,4 and
intramolecular coupling with alkynes.5 The relevance
* Corresponding author. E-mail: ana.martins@ist.utl.pt.
(1) (a) Pohlki, F.; Doye S. Chem. Soc. Rev. 2003, 32, 104. (b) Galindo,
A.; Go´mez, M.; Go´mez-Sal, P.; Mart´ın, A.; del Rio, D.; Sa´nchez, F.
Organometallics 2002, 21, 293. (c) Thorn, M. G.; Fanwick, P. E;
Rothwell, I. P. Organometallics 1999, 18, 4442. (d) Alcalde, M. I.;
Go´mez-Sal, M. P.; Royo, P. Organometallics 1999, 18, 546. (e) Rietveld,
M. H. P.; Hagen, H.; van de Water, L.; Grove, D. M.; Kooijman, H.;
Veldman, N.; Spek, A. L.; van Koten, G. Organometallics 1997, 16,
168. (f) Scott, M. J .; Lippard, S. J . Organometallics 1997, 16, 5857. (g)
Durfee, L. D.; Rothwell, I. P. Chem. Rev. 1988, 88, 1059.
(2) (a) Amador, U.; Daff, P. J .; Poveda, M. L.; Ruiz, C.; Carmona, E.
J . Chem. Soc., Dalton Trans. 1997, 3145. (b) Filippou, A. C.; Gru¨nle-
itner, W.; Vo¨lk, C.; Kiprof, P. J . Organomet. Chem. 1991, 413, 181. (c)
Beshouri, S. M.; Chebi, D. E.; Fanwick, P. E.; Rothwell, I. P. Organo-
metallics 1990, 9, 2375.
(5) (a) Venne-Dunker, S.; Ahlers, W.; Erker, G.; Fro¨hlich, R. Eur.
J . Inorg. Chem. 2000, 1671. (b) Curtis, M. D.; Real, J .; Hirpo, W.;
Butler, W. M. Organometallics 1990, 9, 66.
(6) (a) Alcalde, M. I.; Go´mez-Sal, M. P.; Royo, P. Organometallics
2001, 20, 4623. (b) Sa´nchez-Nieves, J .; Royo, P.; Pellinghlli, M. A.;
Tiripicchio, A. Organometallics 2000, 19, 3161. (c) Thirupathi, N; Yap,
G. P. A.; Richeson, D. S. Organometallics 2000, 19, 2573. (d) Broder,
C. K.; Goeta, A. E.; Howard, J . A. K.; Hughes, A. K.; J ohnson, A. L.;
Malget, J . M.; Wade, K. J . Chem. Soc., Dalton Trans. 2000, 3526. (e)
Wu, Z.; Diminnie, J . B.; Xue, Z. Organometallics 1999, 18, 8, 1002. (f)
Galakhov, M.; Go´mez-Sal, M. P.; Mart´ın, A.; Mena, M.; Ye´lamos, C.
Eur. J . Inorg. Chem. 1998, 1319. (g) Zanella, P.; Brianese, N.; Casellato,
U.; Ossola, F.; Porchia, M.; Rossetto, G.; Graziani, R. J . Chem. Soc.,
Dalton Trans. 1987, 2039. (h) Chisholm, M. H.; Hammond, C. E.; Ho,
D.; Huffman, J . C. J . Am. Chem. Soc. 1986, 108, 7860.
(3) (a) Bashall, A.; Collier, P. E.; Gade, L. H.; McPartlin, M.;
Mountford, P.; Pugh, S. M.; Radogevic, S.; Schubart, M.; Scowen, I. J .;
Tro¨sch, D. J . M. Organometallics 2000, 19, 4784. (b) Cadierno, V.;
Zablocka, M.; Donnadieu, B.; Igau, A.; Majoral, J .-P.; Skowronska, A.
J . Am. Chem. Soc. 1999, 121, 11086.
(4) (a) Greidanus-Strom, G.; Carter, C. A. G.; Stryker, J . M.
Organometallics 2002, 21, 1011. (b) Ong, T.-G.; Wood, D.; Yap, G. P.
A.; Richeson, D. S. Organometallics 2002, 21, 1. (c) Chamberlain, L.
R.; Durfee, L. D.; Fanwick, P. E.; Kobriger, L. M.; Latesky, S. L.;
McMullen, A. K.; Steffey, B. D.; Rothwell, I. P.; Folting, K.; Huffman,
J . C. J . Am. Chem. Soc. 1987, 109, 6068.
(7) (a) Haak, E.; Bytschkov, I.; Doye, S. Angew. Chem., Int. Ed. 1999,
38, 3389. (b) Zuckerman, R. L.; Krska, S. W.; Bergman, R. G. J . Am.
Chem. Soc. 2000, 122, 753. (c) Fairfax, D.; Stein, M.; Livinghouse, T.;
J ensen, M. Organometallics 1997, 16, 1523. (d) Cao, C.; Ciszewski, J .
T.; Odom, A. L. Organometallics 2001, 20, 5011; Organometallics 2002,
21, 5148.
10.1021/om030211n CCC: $25.00 © 2003 American Chemical Society
Publication on Web 09/06/2003

Insertion of Isocyanides
Organometallics, Vol. 22, No. 21, 2003 4219
results were compared with the NMR experimental
findings of tert-butylisocyanide insertion in Zr-NMe2
bonds.
tions, is unstable. Successive insertion reactions, either
in the Ti-iminoacyl or the Ti-NMe2 bonds, or intramo-
lecular rearrangements involving the iminoacyl and
amido ligands may occur and be responsible for the
modifications observed in the proton spectra;13 however,
except for a set of signals attributable to indene, we were
not able to assign them. The elimination of indene from
titanium and zirconium indenyl bis(dimethylamido)
compounds has previously been reported and reflects
the anionic character of the indenyl ligand in this class
of complexes.8a,14
Resu lts a n d Discu ssion
Ch em ica l Stu d ies. In ser tion in to M-C Bon d s.
The reaction of equimolar amounts of CN-t-Bu and
[Ti(η5-Ind)(NMe2)2Me] led systematically to the forma-
tion of product mixtures. 1H NMR spectra of the
reaction’s crude showed broad bands, suggestive of the
formation of polymeric and/or paramagnetic aggregates,
which were not possible to identify. The reaction was
The reactions of CN-t-Bu with [Ti(η5-Ind)(NMe2)Me2]
and [Zr(η5-Ind)(NMe2)2Me] led to mixtures that were not
possible to characterize by NMR.
1
then performed in a NMR tube and followed by H and
13C{1H} NMR spectroscopy. The insertion process is
instantaneous. In fact, immediately after the addition
of the reagents, a new set of signals, compatible with
the formation of [Ti(η5-Ind)(NMe2)2{C(Me)dN-t-Bu}], 1,
appeared. The chemical shifts of the methyl group
presented in the carbon (δ 20.5 ppm) and proton (δ 2.32
ppm) spectra are indicative of the insertion of tert-butyl
isocyanide into the Ti-CH3 bond. Strong evidence of the
coordination of the iminoacyl ligand is the shift to lower
field of the RCdNR′ atom (δ 240.9 ppm), as observed
in other compounds such as [ZrCp(ArO)(η2-C(CH2Ph)N-
t-Bu)(CH2Ph)] (ArO ) 2,3,5,6-Ph4C6HO; 4,6-t-Bu2-2-
PhC6H2O) (δ 242-244 ppm),11 and [Zr{(iPrNH)C-
(NiPr)2}2(CH2Ph){C(CH2Ph)N(2,6-Me2Ph)}] (δ 252.86
ppm).4b This resonance is typical of dihapto ligands, very
commonly found in group 4 metal complexes. In addi-
tion, the NMe2 proton and carbon chemical shifts in 1
do not suffer any significant modification, denoting that
the insertion in the Ti-N bonds did not take place. The
indenyl ligand resonances consist of three multiplets
between δ 6 and 7.5 ppm. The four protons of the
benzene ring give rise to two multiplets (δ 7.21 and 6.75
ppm), and the three protons of the five-membered ring
appear together as a multiplet at δ 6.18 ppm. Spin
saturation transfer and NOEDIFF experiments are
suggestive of a fluxional process that equilibrates the
two halves of the indenyl ligand. Saturation of the t-Bu
singlet, at room temperature, showed the existence of
a similar NOE between these protons and both H2/H3
of the indenyl (see indenyl label). The rotation of the
η2-coordinated iminoacyl ligand is a fluxional process
often observed in solution2a,12 and accounts for the NMR
data.
In ser tion in to M-N Bon d s. The insertion of CN-
(2,6-Me2Ph) into the titanium-nitrogen bond of [Ti(η5-
Ind)(NMe2)Cl2] led to the synthesis of [Ti(η5-Ind)-
{C(NMe2)dN(2,6-Me2Ph)}Cl2], 2, which was isolated in
86% yield as dark red crystals. The NMR spectra of 2
show two inequivalent methyl resonances for the NMe2
moiety, as a result of hindered rotation around the
carbon-nitrogen bond. This suggests the conjugation
of the CdN(2,6-Me2Ph) double bond with the NMe2 lone
pair, resulting in delocalized partial multiple bonds
through the (2,6-Me2Ph)NdCdNMe2 frame. An NOE
difference spectrum enabled the assignment of the
methyl resonances since the interaction between the
endo NMe2 methyl (see iminocarbamoyl label) and the
xylyl methyl groups gave a positive NOE peak at δ 2.68
ppm when the resonance at δ 2.01 ppm (2,6-(CH3)2Ph)
was irradiated. The observation of a unique resonance
for the methyl xylyl groups is consistent with the free
rotation of the ring around the C-N bond. The quater-
nary carbon of the iminocarbamoyl ligand is, as ex-
pected, observed at low field (δ 199.5 ppm) and sugges-
tive of a dihapto coordination. The indenyl NMR
resonances show a pattern compatible with an average
Cs symmetry structure, implying the rotation of the
iminocarbamoyl ligand around its bond to the metal.
1
Variable-temperature H NMR experiments were per-
formed. In d8-toluene, the rotation of the iminocarbam-
oyl ligand is fast until 193 K since no changes have been
observed in the proton spectra. When the temperature
of a solution of 2 in C6D5Br was raised to 373 K, the
NMe2 signals did not suffer any modification. However,
above this temperature, a small peak, which gradually
increased in intensity until 413 K, began to appear
between the resonances of the two methyl groups, at δ
3.21 ppm. It was not possible to achieve the coalescence
temperature, which is higher than 413 K. This temper-
ature is indicative of a considerably high energy barrier
for the rotation around the CdNMe2 bond in comparison
to the values reported for organic compounds such as
amides, amidine, or enamines, for which those values
Although the reaction proceeds instantaneously, the
conversion of [Ti(η5-Ind)(NMe2)2Me] is not complete
when both reagents are mixed in stoichiometric quanti-
ties. The addition of an excess of isocyanide displaces
the reaction toward the formation of 1 and enables the
quantitative conversion of [Ti(η5-Ind)(NMe2)2Me] in [Ti-
(η5-Ind)(NMe2)2{C(Me)dN-t-Bu}], which, in these condi-
range from 50 to 90 kJ mol-1
.
(8) (a) Ascenso, J . R.; Azevedo, C. G.; Correia, M. J .; Dias, A. R.;
Duarte, M. T.; Silva, J . L. F.; Gomes, P. T.; Lourenc¸o, F.; Martins, A.
M.; Rodrigues, S. S. J . Organomet. Chem. 2001, 632, 58-66. (b)
Martins, A. M.; Ascenso, J . R.; Azevedo, C. G.; Calhorda, M. J .; Dias,
A. R.; Rodrigues, S. S.; Toupet, L.; de Leonardis, P.; Veiros, L. F. J .
Chem. Soc., Dalton Trans. 2000, 4332-4338.
The insertion of xylyl isocyanide in both metal-
dimethylamido bonds of [M(η5-Ind)(NMe2)2Cl] (M ) Ti,
Zr) was exclusively observed for the zirconium deriva-
(9) Hehre, W. J .; Radom, L.; Schleyer, P. v. R.; Pople, J . A. Ab Initio
Molecular Orbital Theory; J ohn Wiley & Sons: New York, 1986.
(10) Parr, R. G.; Yang, W. Density Functional Theory of Atoms and
Molecules; Oxford University Press: New York, 1989.
(11) Thorn, M. G.; Lee, J .; Fanwick, P. E.; Rothwell, I. P. J . Chem.
Soc., Dalton Trans. 2002, 3398.
(13) (a) Castro, A.; Galakhov, M. V.; Go´mez, M.; Go´mez-Sal, M. P.;
Mart´ın, A.; Sa´nchez, F.; Velasco, P. Eur. J . Inorg. Chem. 2000, 2047.
(b) Galakhov, M.; Go´mez, M.; J ime´nez, G.; Royo, P.; Pellinghelli, M.
A.; Tiripicchio, A. Organometallics 1995, 14, 2843.
(14) (a) Diamond, G. M.; J ordan, R. F.; Peterson, J . L. Organome-
tallics 1996, 15, 4030. (b) Diamond, G. M.; J ordan, R. F.; Peterson, J .
L. J . Am. Chem. Soc. 1996, 118, 8024.
(12) Castro, A.; Galakhov, M.V.; Go´mez, M.; Go´mez-Sal, M. P.;
Mart´ın, A.; Sa´nchez, F. J . Organomet. Chem. 2000, 595, 36.

4220 Organometallics, Vol. 22, No. 21, 2003
Martins et al.
Sch em e 1
Sch em e 2
Cl]. The isomerization to 5b took place very slowly in
the NMR tube. Raising the temperature led to the
appearance of indene resulting from partial decomposi-
tion of the products.
tive. Stereochemical constraints are most probably in
the origin of the lack of reactivity of the titanium
complex, and similar arguments may justify the absence
of insertion of CN-t-Bu into the Ti-NMe2 bonds of
[Ti(η5-Ind)(NMe2)Cl2] and [Ti(η5-Ind)(NMe2)2Cl]. Com-
plex [Zr(η5-Ind){C(NMe2)dN-2,6-Me2Ph}2Cl], 3, was
obtained in 77% yield as yellow crystals. Considering
that the rotation around the Zr-iminocarbamoyl bonds
is hindered, three possible isomers, with pseudo-Cs (3a
and 3c) and C1 (3b) symmetries, can be drawn for 3, as
depicted in Scheme 1.
The conformer 3c was the only product obtained when
the reaction was left for 24 h at room temperature, as
confirmed by NMR and single-crystal X-ray diffraction.
As for 2, the two methyl groups of each NMe2 fragment
in 3 give rise to two singlets, in agreement with the
extensive delocalization of electron density through the
N-C-N core of the iminocarbamoyl ligands. However,
in contrast to what is observed in 2, hindered rotation
around the N-xylyl bonds of 3 leads to the appearance
of different resonances for the xylyl methyl groups. As
pointed out by the X-ray and theoretical data (see
below), this is a stereochemical constraint rather than
an electronic effect, resulting from the close location of
the two xylyl moieties. The three protons of the indenyl
five-membered ring ligand appear at 6.53 ppm as a
complex peak. This pattern, previously encountered in
other titanium and zirconium indenyl complexes,8b is
attributed to local magnetic field anisotropy, created by
the unsaturated ligands.
1
The H NMR spectrum at room temperature of the
reaction mixture shows two sets of seven resonances
attributable to the indenyl ligands of 4 and 5a . The
observation of seven peaks for 4 attests to the existence
of a chiral metal center and is compatible with the free
rotation around the zirconium-iminocarbamoyl bond.
The indenyl pattern shown for 5a is, as for 4, consistent
with a C1 symmetry. However, in 5a , the data imply
hindered rotation of the iminocarbamoyl ligands, a
result that is mainly due to stereochemical constraints.
Surprisingly, for both compounds, the magnetic equiva-
lence of the two methyl groups in each CNMe2 imi-
nocarbamoyl fragment was observed. This is consistent
with free rotation around the C-NMe2 bonds and
reflects a reduced multiple-bond character comparing
to 2 and 3. The 13C{1H} spectrum of the reaction
mixture shows a great complexity of signals that, except
for those corresponding to the quaternary iminocarbam-
oyl carbons, were impossible to assign. The isomeriza-
tion to 5b is readily observed by NMR due to the
existence of an average symmetry plane. The indenyl
ligand leads to four proton and to five carbon reso-
nances. A gHMBC (1H-15N) heteronuclear correlation
performed for 5b allowed the distinction between the
two nitrogen atoms of the iminocarbamoyl ligands that
appeared at δ 448 (t-BuNC) and 304 ppm (CdNMe2).
These values present a dramatic deshielding relative
to the free tert-butylisocyanide chemical shift, observed
at δ -186.5 ppm.15
Cr ysta llogr a p h ic Stu d ies. The molecular struc-
tures of complexes 2 and 3 are depicted in Figure 1 and
Figure 2, respectively, and selected bond lengths and
angles are listed in Tables 1 and 2.
In 2, the titanium adopts a distorted piano stool
geometry, considering the medium point of the CdN(2,6-
Me2Ph) bond ((CN)*) and the C5 indenyl centroid (C5*)
as occupying the bonding positions. The coordination
angles C5*-Ti-X (X ) Cl, (CN)*) are larger than the
Cl-Ti-X angles (X ) Cl, (CN)*), and the indenyl plane
and that defined by Cl(1), Cl(2), and (CN)* are close to
parallel, with an angle of 5.2(7)° between them.
The failure to isolate a pure complex from the reaction
of [Zr(η5-Ind)(NMe2)2Cl] with 1 equiv of CN(2,6-Me2Ph)
led us to the study of the insertion reaction of isocya-
nides into zirconium-nitrogen bonds by NMR, using as
starting materials [Zr(η5-Ind)(NMe2)2Cl] and CN-t-Bu,
and by DFT, using [Zr(η5-Ind)(NR2)2Cl] (R ) Me or H)
and CNH as an isocyanide model.
The insertion of CN-t-Bu in the Zr-NMe2 bonds of
[Zr(η5-Ind)(NMe2)2Cl] is a slow reaction on the NMR
time scale and allowed the identification of all the
intermediates. The reaction sequence is depicted in
Scheme 2.
The formation of complexes [Zr(η5-Ind)(NMe2)-
{C(Me2N)dN-t-Bu}Cl], 4, and [Zr(η5-Ind){C(Me2N)d
N-t-Bu}2Cl], 5a , is almost simultaneous, either when 1
or 2 equiv of isocyanide is added to [Zr(η5-Ind)(NMe2)2-
(15) Pflug, J .; Bertulat, A.; Kehr, G.; Fro¨hlich, R.; Erker, G.
Organometallics 1999, 18, 3818.

Insertion of Isocyanides
Organometallics, Vol. 22, No. 21, 2003 4221
Ta ble 1. Selected Bon d Len gth s (Å) a n d An gles
(d eg) for 2
Distances, Å
Ti(1)-Cl(1)
Ti(1)-Cl(2)
Ti(1)-N(1)
Ti(1)-C(16)
N(2)-C(16)
N(2)-C(17)
N(2)-C(18)
N(1)-C(16)
N(1)-C(10)
2.293(2)
2.347(2)
1.982(4)
2.023(5)
1.308(6)
1.466(6)
1.461(6)
1.294(5)
1.427(5)
Angles, deg
C(16)-Ti(1)-N(1)
C(16)-Ti(1)-Cl(1)
C(16)-Ti(1)-Cl(2)
Cl(1)-Ti(1)-Cl(2)
N(1)-Ti(1)-Cl(1)
N(1)-Ti(1)-Cl(2)
Ti(1)-N(1)-C(10)
Ti(1)-N(1)-C(16)
C(16)-N(1)-C(10)
Ti(1)-C(16)-N(2)
Ti(1)-C(16)-N(1)
N(1)-C(16)-N(2)
C(16)-N(2)-C(18)
C(16)-N(2)-C(17)
C(18)-N(2)-C(17)
37.7(2)
90.9(1)
123.0(1)
98.06(6)
111.8(1)
88.2(1)
149.6(3)
72.9(3)
134.3(4)
155.9(4)
69.4(2)
134.6(4)
124.2(4)
120.0(4)
115.6(4)
101.8(6)
106.1(6)
115.1(6)
113.8(6)
119.2(6)
F igu r e 1. Molecular diagram of 2. H atoms were omitted
for clarity. Ellipsoids were drawn at the 40% probability
level with ORTEP3.
Cl(1)-Ti(1)-(C(16)-N(1))*
Cl(2)-Ti(1)-(C(16)-N(1))*
Cl(1)-Ti(1)-C5*
Cl(2)-Ti(1)-C5*
(C(16)-N(1))*-Ti(1)-C5*
etry is also characterized by the torsion angles
C5*-Ti-(CN)*-C(10), C(11)-C(10)-N(1)-C(16), and
C(15)-C(10)-N1-C(16) of 117.5(6)°, -59.5(6)°, and
124.4(6)°, respectively.
In 3, the zirconium coordination sphere can also be
described as a distorted piano stool geometry, with η2-
CdN units and indenyl occupying single coordination
sites. The two molecules in the asymmetric unit exhibit
identical overall geometry, and only one is discussed.
For this complex, due to the bulkiness of the two
iminocarbamoyl ligands, the geometry is more dis-
torted than in 2. The angles C5*-M-X vary between
106.4(9)° and 119.1(9)°, while the X-M-X angles are
101.2(9)°, 104.8(9)°, and 108.1(9)°. The angle between
the indenyl plane and the one defined by (CN)*, Cl(1),
and (CN)* is 8.3(4)°. The dihedral angle between the
planes of the two iminocarbamoyl ligands (which are
planar, with rmsd of 0.04(4) Å) is 63.9(6)°.
The distances between the metal and the indenyl
carbon atoms and the Zr-Cl bonds are similar to those
reported for other Zr(IV) complexes.8b,18 Both ligands
have similar metal-iminocarbamoyl carbon bond lengths
(2.195(8) and 2.222(7) Å), in the range found in zirco-
nium iminoacyl4b,18 and iminocarbamoyl silyl complexes.6e
The Zr-N bond lengths (2.135(6) and 2.152(5) Å) are
also in agreement with values reported for similar
Zr(IV)-N bonds.19 As found in complex 2 and [Zr{η2-
C(NMe2)dNAr}3{η2-C(Si(SiMe3)3)dNAr}] (Ar ) 2,6-
Me2Ph),6e the planarity of the ligand and the short
NdC (1.295(10); 1.304(10) Å) and C-NMe2 bonds
(1.332(11) and 1.337(10) Å) are indicative of carbon-
nitrogen multiple-bond character. Accordingly, the sums
F igu r e 2. Molecular diagram of one of the molecules of 3
found in the asymmetric unit. H atoms were omitted for
clarity. Ellipsoids were drawn at the 40% probability level
with ORTEP3.
The general features of the molecular structure
concerning distances between the metal and the indenyl
carbon atoms and the Ti-Cl bonds are similar to those
reported for other Ti(IV) half-sandwich complexes.8b,16
The iminocarbamoyl ligand shows a η2-coordination,
with Ti(1)-C(16) and Ti(1)-N(1) distances (2.023(5)
and 1.982(4) Å) comparable to other Ti-C(sp2) and
Ti-N(sp2) bond lengths, respectively.6f,17 Consistent
with the NMR spectra, the distances between C(16) and
the two nitrogen atoms (1.308(6) and 1.294(5) Å) reveal
double-bond character. The iminocarbamoyl fragment
(C(17)-N(2)-C(18)-C(16)-N(1)) is planar (rmsd of
0.019(4) Å), with a relative orientation to the indenyl
ring given by the angle C5*-Ti-(CN)* of 119.2(6)°. This
arrangement points the N(2), C(16), and N(1) p orbitals
in an adequate position to extend the π system through
the three atoms. The xylyl ring defines with the imi-
nocarbamoyl moiety an angle of 70.8(6)°, a position that
minimizes the stereochemical hindrance between the
endo NMe2 methyl (C(18)) and the xylyl methyl (C(111))
that are directed toward the same region (C‚‚‚C; H‚‚‚H
distances of 3.55 and 2.18 Å, respectively). This geom-
(16) Ready, T. E.; Day, R. O.; Chien, J . C. W.; Rausch, M. D.
Macromolecules 1993, 26, 5822.
(17) Riley, P. N.; Fanwick, P. E.; Rothwell, I. P. J . Chem. Soc., Chem.
Commun. 1997, 1109.
(18) Segerer, U.; Blaurock, S.; Sieler, J .; Hey-Hawkins, E. J .
Organomet. Chem. 2000, 608, 21.
(19) Chisholm, M. H.; Hammond, C. E.; Huffman, J . C. Polyhedron
2002, 7, 2515.

4222 Organometallics, Vol. 22, No. 21, 2003
Martins et al.
Ta ble 2. Selected Bon d Len gth s (Å) a n d An gles
(d eg) for On e of th e Molecu les P r esen t in th e
Asym m etr ic Un it for 3
Distances, Å
Zr(1)-N(1)
Zr(1)-N(3)
Zr(1)-C(16)
Zr(1)-C(26)
Zr(1)-Cl(1)
N(3)-C(26)
N(3)-C(20)
N(4)-C(26)
N(4)-C(27)
N(4)-C(28)
N(1)-C(16)
N(1)-C(10)
N(2)-C(16)
N(2)-C(17)
N(2)-C(18)
2.135(6)
2.152(5)
2.195(8)
2.222(7)
2.538(2)
1.304(10)
1.430(9)
1.337(10)
1.473(12)
1.405(12)
1.295(10)
1.440(11)
1.332(11)
1.465(14)
1.459(14)
Angles, deg
N(1)-Zr(1)-C(16)
C(16)-Zr(1)-C(26)
C(16)-Zr(1)-N(3)
C(16)-Zr(1)-Cl(1)
N(3)-Zr(1)-C(26)
C(26)-Zr(1)-N(1)
C(26)-Zr(1)-Cl(1)
Zr(1)-N(1)-C(16)
Zr(1)-N(3)-C(26)
C(16)-N(1)-C(10)
C(26)-N(3)-C(20)
Zr(1)-C(16)-N(2)
Zr(1)-C(26)-N(4)
Zr(1)-C(16)-N(1)
Zr(1)-C(26)-N(3)
N(1)-C(16)-N(2)
N(3)-C(26)-N(4)
C(16)-N(2)-C(18)
C(26)-N(4)-C(28)
C(16)-N(2)-C(17)
C(26)-N(4)-C(27)
C(18)-N(2)-C(17)
C(28)-N(4)-C(27)
34.8(3)
94.2(3)
108.8(3)
118.3(2)
34.6(2)
114.6(3)
121.7(2)
75.2(5)
F igu r e 3. B3LYP optimized geometries of [Ti(η5-Ind)-
(H2CdNH)Cl2] (left) and [Zr(η5-Ind)(H2CdNH)2Cl] (right)
with the more relevant distances (Å) and angles (deg)
(bottom). The experimental values are presented in italics
(for the Zr species these were taken as the mean of the
two independent molecules of the asymmetric unit).
75.6(4)
132.7(7)
132.7(6)
160.7(7)
160.2(6)
70.1(4)
molecular orbital calculations. To evaluate the method’s
performance in the description of the systems studied,
the optimization of the geometries of model complexes,
[Ti(η5-Ind){C(NH2)dNH}Cl2] and [Zr(η5-Ind){C(NH2)d
NH}2Cl], was carried out. The results are presented in
Figure 3.
69.7(4)
128.9(8)
129.6(7)
126.5(8)
126.8(7)
117.7(9)
117.8(8)
115.7(8)
115.3(8)
108.1(9)
119.1(9)
115.2(9)
106.4(9)
101.2(9)
104.8(9)
70.2(9)
Both complexes have piano stool geometries consider-
ing that the iminocarbamoyl ligands occupy one coor-
dination position. The agreement between the calculated
and the experimental geometries is very good, as shown
by the comparison of bond distances about the metal
and the C-N distances in the iminocarbamoyl ligands.
In fact, the mean and maximum absolute deviations for
these distances are 0.021 and 0.058 Å for the Zr complex
and 0.014 and 0.048 Å for the Ti species, respectively.
The indenyl coordination mode and the relative orienta-
tion of the ligands in each complex are also well
described by the model. The indenyl coordinates in a
slightly distorted η5 mode, as shown by the small values
of the hinge angle, Ω.20 The experimental values of Ω
are 1° and 4° for the Ti and the Zr complexes, respec-
tively, and the corresponding calculated angles are 6°
and 2°. The relative orientation of the indenyl with
respect to the other three ligands can be measured by
the torsion angles C6*-C5*-M-X, where X ) Cl, N,
or C and C5* and C6* represent the centroids of the C5
and the C6 rings of the indenyl ligand, respectively.
Thus, for the Ti complex, the benzene moiety of the
indenyl ligand lies opposite the iminocarbamoyl and
closer to one of the chloride ligands with C6*-C5*-
Ti-Cl angles of 24° and 88° (experimental) and 22° and
92° (calculated). In the Zr species the indenyl benzene
ring lies over one of the C(NR2)NR′ ligands, and the
C6*-C5*-Zr-Cl values are 75° (calculated) and 62°
(experimental, mean of the two independent molecules
of the asymmetric unit). The good performance of the
(C(16)-N(1))*-Zr(1)-(C(26)-N(3))*
(C(26)-N(3))*-Zr(1)-C5*
(C(16)-N(1))*-Zr(1)-C5*
C5*-Zr-Cl(1)
(C(16)-N(1))*-Zr(1)-Cl(1)
(C(26)-N(3))*-Zr(1)-Cl(1)
(N(1)-Zr(1)-C(16))-(N(3)-Zr(1)-C(26))
of the angles around the NMe2 groups are close to 360°
and the ZrCN and NMe2 units are nearly coplanar
with dihedral angles between the planes defined by
Zr(1)-C(16)-N(1) and
C(17)-N(2)-C(18),
and
Zr(1)-C(26)-N(3) and C(27)-N(4)-C(28) of 7.6(9)° and
7.8(9)°, respectively. As in the previous complex, the
xylyl rings adopt an optimal relative conformation to
minimize possible stereochemical interactions, present-
ing angles of 80.5(9)° and 71.5(9)° with the planes of
the iminocarbamoyl to which they are bonded. The two
xylyl rings are almost perpendicular to the indenyl
(angles of 79.2(9)° and 85.2(9)°) and define an angle of
75.9(9)° relative to each other. This optimal geometry
is also characterized by the torsion angles C5*-M-
(CN)*-C of 110.4° and -130.1° and N-C-N-Cxylyl of
9.5° and 3.3°. Short interactions between the endo
NMe and Me-xylyl groups, with H‚‚‚H distances of
2.01 (C(28)-C(60)) and 2.86 Å (C(40)-C(18)) are ob-
served.
Molecu la r Or bita l Stu d ies. The mechanism of
isocyanide insertion CNH in the Zr-N bonds of [Zr(η5-
Ind)(NR2)2Cl] (R ) H, Me) was studied by DFT/B3LYP
(20) Faller, J . W.; Crabtree, R. H.; Habib, A. Organometallics 1985,
4, 929.

Insertion of Isocyanides
Organometallics, Vol. 22, No. 21, 2003 4223
F igu r e 4. Calculated mechanism (B3LYP) for the consecutive addition (a and c) and insertion (b and d) of one isocyanide
molecule, CNR′, into each Zr-N bond of [Zr(η5-Ind)(NR2)2Cl]. The relative energies (kcal/mol, bold) and more relevant
distances (Å, italic) are presented for each step.
theoretical method in the description of the relative
orientation of the ligands in complexes 2 and 3 is
especially relevant considering that bulky substituents
in the real molecules (R ) Me and R′ ) 2,6-Me2Ph) were
replaced by hydrogen atoms in the calculations (R ) R′
) H) and points out that the major factors determining
the geometry around the metals are of electronic rather
than stereochemical nature. This is further reinforced
by the calculated N-C-N and C-N-R′ angles, which,
despite the simplicity of the models, present a mean and
maximum absolute deviation of 2° and 4° with respect
to the experimental values observed for the two com-
plexes.
incoming isocyanide in the metal sphere (steps a and c)
followed by insertion in a Zr-N bond (steps b and d),
forming an addition(a)-insertion(b)-addition(c)-inser-
tion(d) sequence. Figure 4 shows the energy profile for
each step and the optimized geometries of the reactants,
products, and the corresponding transition states, to-
gether with the most relevant structural parameters.
Interestingly the indenyl ligand remains essentially
η5-coordinated along the reaction, the maximum Ω value
(8°) being found for I1.
The models used to calculate the addition steps were
[Zr(η5-Ind)(NMe2)2Cl] and [Zr(η5-Ind){η2-C(NH2)dNH}-
(NMe2)Cl] because the replacement of NMe2 by NH2
led to a false hydrogen interaction between the in-
coming isonitrile and the hydrogens of the NH2 ligands
(N-H‚‚‚C-NH) in the corresponding transition states
(TS1 and TS3). However, due to computational limita-
The insertion of CNH in the Zr-N bonds of [Zr(η5-
Ind)(NR2)2Cl] (R ) Me, H) was explored. The proposed
mechanism is presented in Figure 4 and is based on four
individual steps representing the addition of each

4224 Organometallics, Vol. 22, No. 21, 2003
Martins et al.
tions, both the amido and the isonitrile substituents
were replaced by hydrogen atoms (R ) R′ ) H) in the
remaining calculations.
CNH, as reflected by the Zr charges (NPA) that vary
from 1.69 in R1 to 1.66 in I1 and from 1.61 in I2 to
1.54 in I3. Some conclusions can be drawn based on
those values. The comparison between the Zr charges
in I1 and I2 reveals a more electron rich metal in the
later species, showing that in this environment an
iminocarbamoyl ligand is a better electron donor than
NMe2 and CNH together. Additionally, the comparison
of the Zr charges in R1 and I2, which differ uniquely
in the replacement of an amido by an iminocarbamoyl
ligand, confirms the later ligand as a better electron
donor.
The first step represents the addition of one CNH to
[Zr(η5-Ind)(NMe2)2Cl], R 1. The isonitrile approaches
the complex opposite the indenyl ligand, passing
through a relatively early transition state (TS1) with a
Zr-C(CNH) distance of 3.225 Å, which leads to a metal-
coordinated isonitrile intermediate, [Zr(η5-Ind)(NMe2)2-
(CNH)Cl], I1, with a Zr-C bond length of 2.45 Å. The
major geometrical changes on the Zr coordination
sphere are the opening of the N-Zr-N and N-Zr-Cl
angles, which increase an average of 5° from R1 to the
transition state TS1 and 10° in the intermediate I1. The
resulting Zr-C bond is already incipient in the transi-
tion state, as shown by the corresponding Wiberg
indexes, WI, 0.0776 and 0.3857 in TS1 and I1, respec-
tively.
The last step of the mechanism is the insertion of
the coordinated isonitrile in I3′ into the remaining
Zr-NMe2 bond, leading to the product P 1, which has
two iminocarbamoyl ligands resulting from the overall
insertion of two isonitriles in the Zr-N(amido) bonds
of the original diamido complex, R1. As for the second
step, the major geometrical and electronic changes are
again associated with the formation of the C-NR2 bond.
The distance between the two atoms shortens from 2.76
Å in I3′ to 1.33 Å in P 1 (WI{C-N} ) 1.3134), and the
interaction is already present in the transition state
(TS4), with a 1.92 Å bond distance and a 0.5029 Wiberg
index.
The second step corresponds to the insertion of the
coordinated isocyanide of intermediate I1′ into one
Zr-NH2 bond, leading to the intermediate [Zr(η5-Ind)-
(NH2){C(NH2)dNH}Cl], I2′, which contains amido and
iminocarbamoyl ligands. The major change associated
with this step is the formation of the new C-N bond of
HNC-NH2, as pointed out by the C-N distance varia-
tions from 2.78 Å in I1′ to 1.85 Å in TS2 and 1.32 Å in
I2′. A strong C-N interaction is already present in the
transition state TS2, as emphasized by the Wiberg
index (WI ) 0.5647) that results in bond formation in
I2′, with WI{C-N} ) 1.3075. Other minor changes
associated with this step are the weakening of the
Zr-NH2 and C-NH bonds (d{Zr-N} ) 2.02, 2.19,
3.59 Å; WI{Zr-N} ) 0.9963, 0.5212, 0.0330, for I1′, TS2,
and I2′, respectively; d{C-N} ) 1.17, 1.23, 1.32 Å;
WI{C-N} ) 2.5131, 2.2073, 1.5211, for I1′, TS2, and
I2′, respectively) and the strengthening of the Zr-C
bond (d{Zr-C} ) 2.45, 2.27, 2.28 Å; WI{Zr-C} )
0.3742, 0.5568, 0.5244, for I1′, TS2, and I2′, respec-
tively).
An interesting aspect of P 1 is the delocalized double-
bond character along the N-C-N frame of the imi-
nocarbamoyl ligands (R′NCNR2). This is suggested by
the planarity of the R2NC fragment of the coordinated
iminocarbamoyl ligand in the optimized structure of P 1,
with all the R-N-C-N torsion angles below 5°, and
further confirmed by the calculated C-N bond lengths
(d{R2N-C} ) 1.33, 1.34 Å, and d{C-NR′} ) 1.31, 1.32
Å) and Wiberg indices (WI{R2N-C} ) 1.289, 1.313, and
WI{C-NR′} ) 1.550, 1.512). These values are interme-
diate between a typical CtN triple bond (d{CtN} )
1.180 Å and WI ) 2.429, in CNH) and a single C-N
bond (d{C-N} ) 1.475 Å and WI ) 0.989, in NMe3),
denoting, thus a CdN double-bond character in the
coordinated iminocarbamoyl.
The third step describes the addition of a second
isocyanide to [Zr(η5-Ind)(NMe2){C(NH2)dNH}Cl], I2.
This step is similar to the first one both on geometrical
and on electronic grounds, with a stepwise increase of
the L-Zr-L angles that vary a mean value of 4° from
I2 to TS3 and 8° from I2 to I3 (L represents the chloride,
the amido nitrogen, and the midpoint of the C-N
iminocarbamoyl bond). However, contrary to what hap-
pened in the first addition step, where an even increase
was found for the three L-Zr-L angles, in this case the
C(CNR′)-Zr-iminocarbamoyl angle is that presenting a
greater increase, 19°, on going from I2 to I3. An odd
widening of the X-Zr-X angles is observed for I3, due
to the shift of the chloride ligand away from the
iminocarbamoyl (Cl-Zr-iminocarbamoyl ) 129°) and
closer to the amido ligand (Cl-Zr-N ) 102°). The
Zr-C(CNH) distances in the transition state TS3 and
in the resulting intermediate, I3, are 3.25 and 2.45 Å,
respectively, and the corresponding Wiberg indexes are
0.0767 and 0.3976, showing that the Zr-CNH bond
formed in I3 is present in the transition state as a weak
interaction, in close resemblance to what was obtained
for the first addition.
The energy profile corresponding to the mechanism
shows two rather flat addition steps with low activation
energies (3.7 and 1.4 kcal/mol, respectively) and small
energy variations (1.8 and -1.0 kcal/mol, in the same
order), suggesting that the isonitrile additions cor-
respond to a rapid exchange equilibrium. The insertion
steps present significantly larger activation energies
(12.5 and 11.5 kcal/mol, respectively), reflecting the
greater geometrical and electronic transformations oc-
curring in these steps. On the other hand, each isonitrile
insertion is noticeably exothermic (∆E ) -18.9 and
-20.6 kcal/mol), pointing out that the transformation
of an amido/isonitrile complex into an iminocarbamoyl
species is a favorable process. The increasing stability
of the complexes along the reaction path is directly
related with an enhanced metal electronic richness
shown by the Zr charges in the reactant, R1 (1.69), and
the product, P 1 (1.50). The overall energy variation
should not be directly taken from Figure 4 individual
steps given the different models used for the addition
and the insertion elementary steps. However, it can be
derived from the following equation, where a R1′
reactant (with NH2 instead of NMe2) is used, suggesting
a strongly exothermic reaction:
Another aspect of both addition steps is the electronic
enrichment of the metal atom due to the addition of

Insertion of Isocyanides
Organometallics, Vol. 22, No. 21, 2003 4225
[Zr(η5-Ind)(NH2)2Cl] (R1′) + 2 CNH f
N-C-Zr-C5* torsion angles, where N and C are the
iminocarbamoyl coordinating atoms and C5* the indenyl
C5 ring centroid. Along the two steps, those angles are
100° in P 3, 158° in TS5, 138° in P 1′, and 80°, 169°, and
142° in P 1, TS7, and P 2, respectively. The weakening
of the Zr-C bonds in the transition states are the major
electronic changes in the iminocarbamoyl rotation steps,
as pointed out by the corresponding Wiberg indexes (WI
) 0.557 in P 3, 0.545 in TS5, and 0.556 in P 1′; 0.577 in
P 1, 0.522 in TS7, and 0.569 in P 2). This effect is
partially compensated by the reinforcement of the Zr-N
bonds, for which the corresponding Wiberg indexes rise
from 0.548 in P 3 to 0.573 in TS5 and from 0.581 in P 1
to 0.598 in TS7. Nevertheless, this compensation is not
enough to overcome the loss of electron density in the
metal center due to the weakening of the Zr-C bond,
as shown by the metal charges (1.51 in P 3, 1.54 in TS5,
and 1.49 in P 1′; and 1.50 in P 1, 1.52 in TS7, and 1.50
in P 2). The depletion of electron density is the main
reason for the destabilization associated with the tran-
sition states.
Another interesting result, based on the NMR obser-
vations, is the preferential insersion of an incoming
isonitrile into the Ti-C bond rather than in one of the
Ti-N bonds of complex [Ti(η5-Ind)(NMe2)2Me]. In fact,
this is a key step in a recently proposed catalytic cicle
for the coupling of amines, alkynes, and isocyanides
with a Ti catalyst.21 The behavior of our system was
investigated by means of a comparative study of the
insertion of a CNH into a Ti-C and a Ti-N bond of the
model Ti complex [Ti(η5-Ind)(NH2)2(CH3)] (R2), per-
formed at the same level of theory as the one used for
the Zr complex discussed above.
The addition of a CNH molecule to [Ti(η5-Ind)(NH2)2-
(CH3)] is a slightly endothermic reaction (∆E ) 6.1 kcal/
mol), yielding the corresponding intermediate, [Ti(η5-
Ind)(NH2)2(CH3)(CNH)] (I4), similarly to the Zr complex
reaction (see Figure 4) and, thus, not deserving any
further discussion. The insersion of the CNH ligand into
the Ti-C bond of I4 results in an iminoacyl complex,
[Ti(η5-Ind)(NH2)2(CH3CNH)] (P 4). If, however, the CNH
ligand inserts into a Ti-N bond of one NH2 ligand, an
iminocarbamoyl complex is produced, [Ti(η5-Ind)(NH2)-
(CH3){C(NH2)NH}] (P 5), in a reaction equivalent to
what was discussed for the Zr complex. The energetic
profile for these two reactions is depicted in Figure 6,
with the optimized geometries of the intervening species
and the more relevant geometrical and electronic pa-
rameters.
The two reactions presented are completely equiva-
lent to the insertion steps (2 and 4) of the Zr complex
reaction. In all cases the M-L (CH3 or NH2) bond
breaking and the L-CNH bond forming associated with
the insersion are already clearly present in the corre-
sponding transition states (TS8 and TS9), as shown by
the bond distances and Wiberg indices of Figure 6. From
an energetic point of view the two Ti reactions are,
however, quite different. In fact, the insersion of the
isonitrile ligand into the Ti-C bond is largely more
favorable from a kinetic (Ea ) 1.0 kcal/mol) as well as
from a thermodynamic point of view (∆E ) -38.2 kcal/
mol) than the insertion of CNH into a Ti-N bond (Ea )
7.9 kcal/mol and ∆E ) -26.6 kcal/mol), in absolute
[Zr(η5-Ind){C(NH2)NH}2Cl] (P 1)
∆E ) -48.1 kcal/mol
The experimental observation of a kinetic and a
thermodynamic product identified by NMR (see Scheme
2 and its discussion) led us to a comprehensive study of
the different product conformers and their interconver-
sion, resulting from the rotation around the indenyl-
metal and the iminocarbamoyl-metal bonds. The re-
sults are shown in Figure 5, which depicts the transition
states and the minima optimized geometries, as well as
the relative energies. It should be noticed that the
indenyl ligand coordination remains distorted η5 along
the entire path, the maximum Ω value (7°) being found
for TS7.
Three conformers result from the different relative
orientations of the two iminocarbamoyl and the chloride
ligands. The iminocarbamoyl ligands may be coordi-
nated in a symmetric fashion, with the NR2 and NR′
moieties facing each other, leading to a pseudo-Cs
symmetric product. In this case, two possibilities arise
for the chloride location: between the NR2 or between
the NR′ moieties, leading respectively to P 3 and P 2,
which corresponds to the crystallographic structure
obtained for 3. Alternatively, the iminocarbamoyl ligands
can be asymmetrically coordinated, with each NR2
fragment facing the NR′ moiety of the other ligand. In
this case the product P 1 has C1 symmetry and results
directly from the addition/insertion mechanism dis-
cussed above. The interconversion mechanism gave rise
to a second conformer with C1 symmetry, P 1′, which is
related to P 1 by the relative orientation of the indenyl
and the NR2CNR′ ligands: the indenyl benzene ring lies
over one NR2CNR′ in P 1 and over the other in P 1′.
The first stage of the interconversion mechanism
represented in Figure 5 starts with the rotation of one
iminocarbamoyl ligand (on the right side of the mol-
ecule) in P 3 to give P 1′. The relaxation of this structure
toward a C1 symmetry is accompanied by the rotation
of the indenyl ligand. The conversion to P 2 requires the
rotation of the other iminocarbamoyl ligand. However,
due to its location below the indenyl benzene ring, the
direct rotation of the second iminocarbamoyl ligand (on
the left of the molecule) is disfavored by the proximity
of the NR2 moiety and the indenyl. This is overcome
through an intermediate step that places the indenyl
C6 ring over the first carbamoyl ligand, generating a
second C1 symmetric conformer, P 1. The major geo-
metrical change that goes along with the indenyl
rotation through the transition state TS6 can be mea-
sured by the C6*-C5*-Zr-Cl torsion angle that varies
from 77° in P 1′ to 1° in TS6 and 80° in P 1. No electronic
or geometric parameter, other than the dihedral angle,
suffers a significant variation along this path, meaning
that the destabilization associated with TS6 is es-
sentially due to the stereochemical repulsion between
the indenyl and the chloride ligands. The last step is
the rotation of the second iminocarbamoyl ligand to
give rise to the second pseudo-Cs product, P 2. The
rotation of the iminocarbamoyl ligands goes through the
transition sates TS5 and TS7, where the NR2CNR′
ligands coordinate in an “upright” geometry. The as-
sociated geometrical changes can be measured by the
(21) Cao, C.; Shi, Y.; Odom, A. L. J . Am. Chem. Soc. 2003, 125, 2880.

4226 Organometallics, Vol. 22, No. 21, 2003
Martins et al.
F igu r e 5. Calculated mechanism (B3LYP) for the interconversion of different [Zr(η5-Ind)(H2CdNH)2Cl] conformers,
corresponding to the rotation of the iminocarbamoyl and the indenyl ligands. The relative energies (kcal/mol) are presented.
F igu r e 6. Calculated mechanism (B3LYP) for the insersion of one CNH into a Ti-C (left) or a Ti-N (right) bond of
[Ti(η5-Ind)(NH2)2(CH3)(CNH)], with the more relevant distances (Å) and the corresponding Wiberg indices (italics). The
relative energies (kcal/mol) are presented in bold.
agreement with the NMR data discussed above. This
result is caused, essentially, by the difference between
the Ti-CH3 and Ti-NH2 bond strengths, the former
being considerably weaker than the latter (see Figure
6 parameters). Since a weaker bond (Ti-CH3) is easier
to break, the corresponding transition state is easier to
reach, and, thus, a faster reaction results. Additionally,
a product with two Ti-NH2 bonds (strong) is more
stable by 11.6 kcal/mol than one with one Ti-NH2
(stronger) and one Ti-CH3 bond (weaker).
It should be noticed that, given the simplicity of the
model complexes used in the calculations, the energy
values associated with the mechanisms depicted in Fig-
ures 4, 5, and 6 should not be taken in absolute terms,
but only their relative values are meaningful. Although
the theoretical method employed provides good geom-
etries, at least for the minima (see Figure 3, and its
discussion), a good match between the calculated abso-
lute energy values and the real ones cannot be expected
since bulky substituents (R ) Me, R′ ) 2,6-Me2Ph, t-Bu)
were replaced by hydrogen atoms in the calculations.
nium-indenyl bond, prevented its complete character-
ization. However, the preferencial insertion of an isoni-
trile into a Ti-C rather than a Ti-N bond was confirmed
by DFT calculations. This preference is caused by the
relative bond strength of Ti-CH3 and Ti-NR2; that is,
the insertion in the weaker bond (Ti-CH3) yields a
faster reaction and a more stable final product.
The reactions of CN(2,6-Me2Ph) with [Ti(η5-Ind)-
(NMe2)Cl2] and [Zr(η5-Ind)(NMe2)2Cl] give [Ti(η5-
Ind){C(NMe2)dN(2,6-Me2Ph)}Cl2], 2, and [Zr(η5-Ind)-
{C(NMe2)dN(2,6-Me2Ph)}2Cl], 3, by insertion of the
isocyanide in the metal-nitrogen bonds. The iminocar-
bamoyl ligands of 2 and 3 present extensive delocaliza-
tion of electron density through the N-C-N core.
The insertion of CN-t-Bu in the Zr-N bonds of
[Zr(η5-Ind)(NMe2)2Cl] was studied by NMR. The reac-
tion proceeds slowly on the NMR time scale to give
[Zr(η5-Ind){C(NMe2)dN-t-Bu)(NMe2)Cl], 4, and [Zr(η5-
Ind){C(NMe2)dN-t-Bu)2Cl], 5, for which two conformers,
corresponding to the kinetic and thermodynamic prod-
ucts, have been characterized.
The experimental results agree with those obtained
by DFT calculations of the insertion reaction mecha-
nism. The similarity between the insertion activation
energies and the progressive stability of the insertion
products accounts for the simultaneous formation of the
complexes resulting from one and two isocyanide inser-
tions, even when only 1 equiv of isonitrile is used. The
formation of a kinetic product with C1 symmetry (5a )
Con clu sion s
Complex [Ti(η5-Ind)(NMe2)2{C(Me)dN-t-Bu}], 1, was
identified by NMR as the product of the insertion of
CN-t-Bu in the Ti-CH3 bond of [Ti(η5-Ind)(NMe2)2Me].
The fact that complex 1 is unstable in the presence of
excess isocyanide, leading to the cleavage of the tita-

Insertion of Isocyanides
Organometallics, Vol. 22, No. 21, 2003 4227
C20). IR (KBr, ν): 1670, vs, 1588, w, (νCdN). MS (EI, m/z): 408
(M+), 373 ({M - Cl}+), 293 ({M - (C9H7)}+), 233 ({Ti(C9H7)-
Cl2}+), 198 ({Ti(C9H7)Cl}+), 175 ({C(NMe2)dN(Me2Ph)}+), 145
({C(NMe2)dNC6H3}+), 115 ({C9H7}+), 105 ({Me2Ph}+), 89
({CH3C6H3}+). Anal. Calcd for C20H22N2Cl2Ti: C, 58.71; H,
5.42; N, 6.85. Found: C, 58.36; H, 5.37; N, 6.61.
slowly evolving to a Cs symmetric conformer (5b) was
endorsed by the calculated interconversion mechanism
depicted in Figure 5.
Due to stereochemical constraints that the models do
not take in account, the actual energy values are
expected to be higher than those calculated. Neverthe-
less, the calculated energy profiles reflect the reaction
mechanism and establish that the reaction is essentially
determined by electronic grounds.
[Zr (η5-In d ){C(Me2N)dN(2,6-Me2P h )}2Cl] (3). A solution
of 2,6-Me2PhNC (0.31 g, 2.38 mmol) in toluene (10 mL) was
added to a yellow solution of [Zr(η5-Ind)(NMe2)2Cl] (0.39 g, 1.19
mmol) in the same solvent (20 mL) at -30 °C. The resulting
solution was stirred and warmed to room temperature for 18
h. After filtration of a white solid precipitate, redisolution in
toluene, and reprecipitation at -15 °C, compound 3 was
obtained pure as white in 77% yield (0.54 g). Yellow crystals
were obtained by cooling a toluene solution to -20 °C. 1H NMR
(C6D6): δ 7.69 (m, 3J H5H6 ) 6.3, 4J H5H7 ) 3.0, 2H, H5, H8), 7.03
Exp er im en ta l Section
Gen er a l P r oced u r es. All manipulations were carried out
under a dry nitrogen atmosphere with standard Schlenk
techniques. Solvents were distilled from the indicated drying
agents: pentane (Na), hexane (CaH2), C6D6 (Na), toluene (Na),
and ether (Na). CN-t-Bu was dried with anhydrous Na2SO4
and distilled at room temperature and 10-2 mbar. CN(2,6-
Me2Ph) was used as received. Literature methods were used
to prepare [Ti(η5-Ind)(NMe2)2Me], [Ti(η5-Ind)(NMe2)Cl2], and
3
4
(m, J H6H5 ) 6.3, J H6H8 ) 3.0, 2H, H6, H7), 6.99-6.90 (m, 3H,
15, H16, H17), 6.53 (m, 3J H1H2 ) 2.4, 3H, H1, H2, H3), 2.84, 2.30,
2.06, 2.02 (s, 12H, H11, H12, H19, H20). 13C{1H} NMR (C6D6): δ
205.8 (C10), 152.7 (C14, C18), 147.4 (C13), 131.6, 130.8 (C15, C16
17), 125.5 (C4, C9), 124.3 (C5, C8), 123.2 (C6, C7), 117.3 (C2),
H
,
C
[Zr(η5-Ind)(NMe2)2Cl].8 NMR spectra were recorded on
a
96.2 (C1, C3), 45.5, 35.9 (C11, C12), 20.0, 19.5 (C19, C20). IR (KBr,
ν): 1699, s, 1602, m, (νCdN). MS (EI, m/z): 591 (M+), 476 ({M
- (C9H7)}+), 417 ({M - [C(NMe2)dN(Me2Ph)] + H}+), 370 ({M
- (C9H7) - (Me2C6H4)}+), 346 ({Zr(C9H7)[(Me2N)CN]2}+), 320
({M - (C9H7) - (Me2C6H4) - MeCl}+), 311 ({Zr(C9H7)[(Me2N)-
CN]Cl}+), 308 ({(Me2Ph)N(H)C(NMe2)CN(H)(Me2Ph)}+), 206
({Zr(C9H7)}+), 176 ({(Me2N)CHdN(Me2Ph)}+), 161 ({MeNCHd
N(Me2Ph)}+), 145 ({(Me2N)CdNPh}+; {NCN(Me2Ph)}+), 132
({HCN(Me2Ph)}+), 115 ({C9H7}+), 105 ({ZrN}+), 91 (Zr+), 44
({NMe2}+). Anal. Calcd for C31H37N4ClZr: C, 57.30; H, 6.12;
N, 9.11. Found: C, 57.40; H, 6.03; N, 8.99.
Varian Unity 300 instrument at room temperature. 1H and
13C NMR spectra are referenced to solvent. Infrared spectra
were obtained with a Perkin-Elmer 577 spectrophotometer.
Mass spectra were recorded on a Finnegan MAT System 8200
spectrometer. Elemental analyses were performed on a Fisons
Instruments 1108 device.
[Zr (η5-In d )(NMe2)Cl{C(NMe2)dN-t-Bu }] (4) a n d [Zr (η5-
In d )Cl{C(NMe2)dN-t-Bu }2] (5). A 4-fold excess of tert-butyl
isocyanide (1.34 mL, 11.90 mmol) was added to 25 mL of a
yellow solution of [Zr(η5-Ind)(NMe2)2Cl] (0.98 g, 2.96 mmol)
in toluene, at room temperature. The resulting solution was
heated at 45 °C for 15 h and then evaporated to dryness. A
dark orange oil was obtained and characterized by NMR (1H,
13C, and 15N) as a mixture of compounds 4 and 5a . The addition
of pentane, hexane, or ether did not allow the separation of
the products. However, by stirring the toluene solution for a
month at room temperature, it was possible to isolate 5b in
85% yield (1.25 g). Data for 4: 1H NMR (C6D6): δ 7.63, 6.90-
6.70 (m, 4H, H5, H6, H7, H8), 5.85, 5.81 (m, 2H, H1, H3), 5.68
[Ti(η5-In d )(NMe2)2{C(Me)dN-t-Bu }] (1). tert-Butyl iso-
cyanide (0.14 mL, 1.2 mmol) was added to 0.50 mL of an
orange C6D6 solution of [Ti(η5-Ind)(NMe2)2Me] (0.16 g, 0.60
mmol) at room temperature. The reaction was followed by
NMR, and the quantitative formation of 1 was observed almost
immediately (ca. 4-5 min). After 15 min new unidentifiable
broad resonances appeared. Signals attributed to indene are
3
(t, J H2H1 ) 3.0, 1H, H2), 2.70 (s, 6H, N(CH3)2)), 2.79 and/or
2.73 (s, 6H, CNMe2), 1.18 or 1.09 (s, 9H, t-Bu). 13C{1H} NMR
(C6D6): δ 201.7 (CdN). Data for 5a : 1H NMR (C6D6): δ 7.48,
7.28, 7.15, 6.81 (m, 4H, H5, H6, H7, H8), 6.51, 6.28 (m, 2H,
3J H1/H2/3J H3/H2 ) 3.0, H1, H3), 6.38 (t, 3J H2H1 ) 3.0, 1H, H2), 2.79
and/or 2.73 (s, 12H, CNMe2) 1.18 and/or 1.09 (s, 18H, t-Bu).
13C{1H} RMN (C6D6): δ 205.4 (CdN). 5b: 1H NMR (C6D6): δ
1
3
also visible in the spectra. H NMR (C6D6): δ 7.21 (m, J H5H6
4
3
4
) 6.3, J H5H7 ) 3.0, 2H, H5, H8), 6.75 (m, J H6H5 ) 6.3, J H6H8
) 3.0, 2H, H6, H7), 6.18 (m, 3H, H1, H2, H3), 3.12 (s, 12H,
N(CH3)2), 2.32 (s, 3H, CH3), 0.90 (s, 9H, t-Bu). 13C{1H} RMN
(C6D6): δ 240.9 (MeCdN-t-Bu), 125.3 (C4, C9), 123.4 (C5, C8),
122.4 (C6, C7), 116.7 (C2), 98.1 (C1, C3), 58.9 (C(CH3)3), 50.8
(N(CH3)2), 30.0 (C(CH3)3), 20.5 (CH3).
3
3
7.64 (m, J H5/H6 ) 6.9, 2H, H5, H8), 6.92 (m, J H6/H5 ) 6.9, 3H,
3
H2, H6, H7), 6.40 (d, J H1H2 ) 3.0, 2H, H1, H3), 2.77 (s, 12H,
CNMe2), 1.22 (s, 18H, t-Bu). 13C{1H} RMN (C6D6): δ 200.3
(C10), 129.3 (C4, C9), 125.0 (C5, C8), 122.0 (C6, C7), 114.6 (C2),
96.8 (C1, C3), 55.0 (C(CH3)3), 44.8 (NMe2), 32.5 (C(CH3)3). 15N
NMR (C6D6): δ 448 (CdN), 304 (NMe2). IR (C6D6): 1685 (s,
[Ti(η5-In d ){C(NMe2)dN(2,6-Me2P h ))Cl2] (2). A solution
of 2,6-Me2PhNC (0.21 g, 1.59 mmol) in toluene (10 mL) was
added to an orange solution of [Ti(η5-Ind)(NMe2)Cl2] (0.44 g,
1.59 mmol) in the same solvent (20 mL) at room temperature.
The resulting solution was stirred for 24 h and then evaporated
to dryness. Compound 2 was isolated as an orange solid in
86% yield (0.56 g). Dark red crystals were obtained by cooling
a toluene solution to -20 °C. 1H NMR (CD3CN): δ 7.74 (m,
νCdN).
Cr ysta l Str u ctu r e Deter m in a tion of Com p ou n d s 2 a n d
3. Crystallographic and experimental details of the crystal
structure determinations are given in Table 3. Suitable
crystals of complexes 2 and 3 were mounted on a capillary
under nitrogen. Data were collected at room temperature on
an Enraf Nonius MACH3 diffractometer equipped with Mo
radiation (λ ) 0.71069 Å). Solution and refinement were made
using SIR9722 and SHELXL23 included in the package of
programs WINGX-Version 1.64.03b.24
4
3
3J H5H6 ) 6.4, J H5H7 ) 3.1, 2H, H5, H8), 7.34 (m, J H6H5 ) 6.4,
4J H6H8 ) 3.1, 2H, H6, H7), 7.04 (br, 3H, H15, H16, H17), 6.98 (d,
3J H1H2 ) 3.4, 2H, H1, H3), 6.40 (t, J H2H1 ) 3.4, 1H, H2), 3.33
3
(s, 3H, H11), 2.68 (s, 3H, H12), 2.01 (s, 6H, H19, H20). 13C{1H}
NMR (CD3CN): δ 199.5 (C10), 146.5 (C13), 132.6 (C14, C18), 129.1
(C4, C9), 128.6 (C15, C17), 127.6 (C6, C7), 127.0 (C16), 126.8 (C5,
For compound 3 two molecules (exhibiting slight differences
in the geometric parameters) were found in the asymmetric
unit and refined. The attribution of the space group was made
C8), 118.2 (C2), 109.1 (C1, C3), 44.4 (C11), 37.8 C12), 19.0 (C19
,

4228 Organometallics, Vol. 22, No. 21, 2003
Martins et al.
Ta ble 3. Cr ysta l Da ta a n d Str u ctu r e Refin em en t
Deta ils for 2 a n d 3
coordinated amido (NR2), R, and the entering isonitriles
(CNR′), R′, replaced by hydrogen atoms (R ) R′ ) H) in all
cases except for the two addition steps (Figure 4a and c). In
these cases, a methyl substituent was used for the amido
ligands (R ) Me), since with R ) H the transition state search
was masked by an artificial hydrogen interaction between
the incoming isonitrile carbon and the coordinated amido,
NH‚‚‚CNH. Computational limitations prevented us from
using R ) Me in all calculations.
2
3
empirical formula
fw
C
20H22Cl2N2Ti
C62H74Cl2N8Zr2
1184.63
409.20
cryst syst
triclinic
monoclinic
space group
unit cell dimens
P1h
P21
a ) 7.211(2) Å
b ) 10.872(2) Å
c ) 14.069(2) Å
R ) 98.41(2) deg.
a ) 12.505(2) Å
b ) 15.839(3) Å
c ) 17.342(3) Å
The basis set used for the geometry optimizations (B1)
consisted of a standard 3-21G(*)29 with
a d-polarization
function added for Cl30 and an f-polarization function added
for the metal atoms.31 Transition state optimizations were
performed with the synchronous transit-guided quasi-Newton
method (STQN) developed by Schlegel et al.32 All stationary
points were confirmed by frequency calculations. Single-point
energy calculations were run on the optimized structures at
the same theory level and a basis set (B2) that consisted of
the previous one for Zr and a standard 6-31G*33 for the
remaining elements (Ti, C, H, N, and Cl). The energies were
zero-point corrected with the zero-point energies obtained at
the B3LYP/B1 level.
â ) 102.31(2) deg â ) 90.113(1) deg
γ ) 106.73(2) deg
volume
Z
calcd density
abs coeff
F(000)
1006.4(4) Å3
2
3434.9(9) Å3
2
1.350 Mg/m3
0.695 mm-1
424
1.145 Mg/m3
0.420 mm-1
1232
Flack param
no. of reflns
no. collected/unique
completeness to
θ ) 26.03°
-0.04(6)
4160/3939
7288/6961
[R(int) ) 0.0262] [R(int) ) 0.0158]
99.3%
99.5%
no. of data/restraints/
params
3939/0/226
6961/1/663
A natural population analysis (NPA)34 was performed in
order to evaluate the charge distribution on the optimized
complexes, and the Wiberg indexes (WI)35 obtained are used
as a measure of bond strength.
goodness-of-fit on F2
1.052
1.035
final R indices [I > 2σ(I)] R1 ) 0.0611
R1 ) 0.0525
wR2 ) 0.1262
R1 ) 0.0760
wR2 ) 0.1373
0.591 and
wR2 ) 0.1404
R indices (all data)
R1 ) 0.1054
wR2 ) 0.1674
0.449 and
Ack n ow led gm en t. This work was supported by
Fundac¸a˜o para a Cieˆncia e a Tecnologia, Portugal
(PBICT/P/QUI/2163/95). S.S.R. is grateful for a FCT
scholarship.
largest diff peak
and hole
-0.438 e Å-3
-0.384 e Å-3
unambiguously, the data being acentric. For both complexes
all non-hydrogen atoms were refined anisotropically, and the
hydrogen atoms were inserted in idealized positions riding on
the parent C atom. Figures were made with ORTEP3.25
Data were deposited in the CCDC under the deposit
numbers CCDC 205751 for 2 and CCDC 205752 for 3.
Com p u ta tion a l Deta ils. The geometry optimizations were
accomplished by means of ab initio and DFT calculations
performed with the Gaussian 98 program.26 The B3LYP hybrid
functional was used in all optimizations. That functional
includes a mixture of Hartree-Fock9 exchange with DFT10
exchange-correlation, given by Becke’s three-parameter func-
tional27 with the Lee, Yang, and Parr correlation functional,
which includes both local and nonlocal terms.28 All the
optimized geometries are the result of full optimizations
without any symmetry constraints. The calculations were
performed on model complexes with the substituents of the
Su p p or tin g In for m a tion Ava ila ble: Tables of atomic
coordinates for all the optimized structures. This material is
available free of charge via the Internet at http://pubs.acs.org.
OM030211N
(27) Becke, A. D. J . Chem. Phys. 1993, 98, 5648.
(28) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
(b) Miehlich, B.; Savin, A.; Stoll, H.; Preuss, H. Chem. Phys. Lett. 1989,
157, 200.
(29) (a) Binkley, J . S.; Pople, J . A.; Hehre, W. J . J . Am. Chem. Soc.
1980, 102, 939. (b) Gordon, M. S.; Binkley, J . S.; Pople, J . A.; Pietro,
W. J .; Hehre, W. J . J . Am. Chem. Soc. 1982, 10, 2797. (c) Pietro, W.
J .; Francl, M. M.; Hehre, W. J .; Defrees, D. J .; Pople, J . A.; Binkley, J .
S. J . Am. Chem. Soc. 1982, 104, 5039. (c) Dobbs, K. D.; Hehre, W. J .
J . Comput. Chem. 1986, 7, 359. (d) Dobbs, K. D.; Hehre, W. J . J .
Comput. Chem. 1987, 8, 861. (e) Dobbs, K. D.; Hehre, W. J . J . Comput.
Chem. 1987, 8, 880.
(30) Ho¨llwarth, A.; Bo¨hme, M.; Dapprich, S.; Ehlers, A. W.; Gobbi,
A.; J onas, V.; Ko¨hler, K. F.; Stegmann, R.; Veldkamp, A.; Frenking,
G. Chem. Phys. Lett. 1993, 208, 237.
(31) Ehlers, A. W.; Bo¨hme, M.; Dapprich, S.; Gobbi, A.; Ho¨llwarth,
A.; J onas, V.; Ko¨hler, K. F.; Stegmann, R.; Veldkamp A.; Frenking, G.
Chem. Phys. Lett. 1993, 208, 111.
(32) (a) Peng, C.; Ayala, P. Y.; Schlegel, H. B.; Frisch, M. J . J .
Comput. Chem. 1996, 17, 49. (b) Peng, C.; Schlegel, H. B. Isr. J . Chem.
1994, 33, 449.
(22) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G.; Giaco-
vazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.; Spagna, R.
Sir97: A new tool for crystal structure determination and refinement.
J . Appl. Crystallogr. 1998, 32, 115.
(23) Sheldrick, G. M. SHELXL-97: A program for refining crystal
structures; University of Go¨ttingen: Germany, 1997.
(24) Farrugia, L. J . WingX (v1.64.03b). J . Appl. Crystallogr. 1999,
32, 837.
(25) Farrugia, L. J . J . Appl. Crystallogr. 1997, 30, 565.
(33) (a) Ditchfield, R.; Hehre W. J .; Pople, J . A. J . Chem. Phys. 1971,
54, 724. (b) Hehre, W. J .; Ditchfield R.; Pople, J . A. J . Chem. Phys.
1972, 56, 2257. (c) Hariharan, P. C.; Pople, J . A. Mol. Phys. 1974, 27,
209. (d) Gordon, M. S. Chem. Phys. Lett. 1980, 76, 163. (e) Hariharan,
P. C.; Pople, J . A. Theor. Chim. Acta 1973, 28, 213.
(34) (a) Carpenter, J . E.; Weinhold, F. J . Mol. Struct. (THEOCHEM)
1988, 169, 41. (b) Carpenter, J . E. Ph.D. Thesis, University of
Wisconsin: Madison, WI, 1987. (c) Foster, J . P.; Weinhold, F. J . Am.
Chem. Soc. 1980, 102, 7211. (d) Reed, A. E.; Weinhold, F. J . Chem.
Phys. 1983, 78, 4066. (e) Reed, A. E.; Weinhold, F. J . Chem. Phys. 1983,
78, 1736. (f) Reed, A. E.; Weinstock, R. B.; Weinhold, F. J . Chem. Phys.
1985, 83, 735. (g) Reed, A. E.; Curtiss, L. A.; Weinhold, F. Chem. Rev.
1988, 88, 899. (h) Weinhold, F.; Carpenter, J . E. The Structure of Small
Molecules and Ions; Plenum: New York, 1988; p 227.
(26) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J . A.,
J r.; Stratmann, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.;
Daniels, A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .;
Barone, V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo,
C.; Clifford, S.; Ochterski, J .; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J . B.; Cioslowski, J .; Ortiz, J . V.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts, R.; Martin, R. L.;
Fox, D. J .; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.;
Gonzalez, C.; Challacombe, M.; Gill, P. M. W.; J ohnson, B.; Chen, W.;
Wong, M. W.; Andres, J . L.; Gonzalez, C.; Head-Gordon, M.; Replogle,
E. S.; Pople, J . A. Gaussian 98, Revision A.6; Gaussian, Inc.: Pitts-
burgh, PA, 1998.
(35) Wiberg, K. B. Tetrahedron 1968, 24, 1083.