Synthesis and spectroscopic properties of dihydrogen isocyanide niobocene [Nb(η5-C5H4SiMe3)2(η2-H2)(CNR)]+ complexes. Experimental and theoretical study of the blocked rotation of a coordinated dihydrogen

Article abstract of DOI:10.1021/ja9640354

Synthesis of stable hydride isocyanide derivatives Nb(η⁵- C₅H₄SiMe₃)₂(H)(CNR) has been achieved through the formation of coordinatively unsaturated 16-electron species Nb(η⁵-C₅H₄SiMe₃)₂H by thermolytic loss of H₂ followed by the coordination of an isocyanide ligand. Low-temperature protonation with a slight excess of CF₃COOH leads to the η²-dihydrogen complexes [Nb(η⁵-C₅H₄SiMe₃)₂(η²-H₂)(CNR)]⁺. NMR spectra of these H-H complexes and their monodeuterated H-D isotopomers present a single high-field resonance at room temperature. By lowering the temperature to 178 K, decoalescence of the signal was observed for the H-D complexes but not for the H-H ones. By combining DFT electronic structure calculations with a monodimensional rotational tunneling model, it has been shown that the absence of decoalescence of the H-H signal is due to the existence of a very large exchange coupling. Conversely, for the H-D isotopomer, the difference in zero point energy corresponding to two nonequivalent (H-D and D-H) positions leads to a slight asymmetry which dramatically reduces the exchange coupling, allowing decoalescence to be observed. Therefore, the H-D classical rotation and the quantum exchange processes will not be practically observed for this complex, whereas only the classical process for the H-H species is quenched out on the NMR time scale.

Full text of DOI:10.1021/ja9640354

J. Am. Chem. Soc. 1997, 119, 6107-6114
6107
Synthesis and Spectroscopic Properties of Dihydrogen
Isocyanide Niobocene [Nb(η5-C5H4SiMe3)2(η2-H2)(CNR)]+
Complexes. Experimental and Theoretical Study of the
Blocked Rotation of a Coordinated Dihydrogen
Antonio Antin˜olo,† Fernando Carrillo-Hermosilla,† Mariano Fajardo,‡
Santiago Garcia-Yuste,† Antonio Otero,*,† Santiago Camanyes,§ Feliu Maseras,|
Miquel Moreno,§ Agust´ı Lledo´s,*,§ and Jose´ M. Lluch*,§
Contribution from the Departamento de Qu´ımica Inorga´nica y Bioqu´ımica, UniVersidad de
Castilla-La Mancha, Campus UniVersitario de Ciudad Real, 13071 Ciudad Real, Spain,
Departamento de Qu´ımica Inorga´nica, UniVersidad de Alcala´, 28871 Alcala´ de Henares, Spain,
Departament de Qu´ımica, UniVersitat Auto`noma de Barcelona, 08193 Bellaterra,
Barcelona, Spain, and Laboratoire de Structure et Dynamique des Syste`mes Moleculaires et
Solides, U.M.R. 5636, UniVersite´ de Montpellier II, 34095 Montpellier Cedex 5, France
ReceiVed NoVember 21, 1996X
Abstract: Synthesis of stable hydride isocyanide derivatives Nb(η5-C5H4SiMe3)2(H)(CNR) has been achieved through
the formation of coordinatively unsaturated 16-electron species Nb(η5-C5H4SiMe3)2H by thermolytic loss of H2 followed
by the coordination of an isocyanide ligand. Low-temperature protonation with a slight excess of CF3COOH leads
to the η2-dihydrogen complexes [Nb(η5-C5H4SiMe3)2(η2-H2)(CNR)]+. NMR spectra of these H-H complexes and
their monodeuterated H-D isotopomers present a single high-field resonance at room temperature. By lowering the
temperature to 178 K, decoalescence of the signal was observed for the H-D complexes but not for the H-H ones.
By combining DFT electronic structure calculations with a monodimensional rotational tunneling model, it has been
shown that the absence of decoalescence of the H-H signal is due to the existence of a very large exchange coupling.
Conversely, for the H-D isotopomer, the difference in zero point energy corresponding to two nonequivalent (H-D
and D-H) positions leads to a slight asymmetry which dramatically reduces the exchange coupling, allowing
decoalescence to be observed. Therefore, the H-D classical rotation and the quantum exchange processes will not
be practically observed for this complex, whereas only the classical process for the H-H species is quenched out on
the NMR time scale.
1. Introduction
being unusual.5 Although the two types of phenomena appear
to be very different, they actually both come from the same
origin: the exchange of a pair of hydrogens through an energy
barrier. Apart from the length of the tunneling path, the
difference between the two processes is the different magnitude
of the energy barrier through which tunneling takes place: less
than 3.5 kcal/mol for the dihydrogen rotational process4,6 and
more than 10 kcal/mol for the hydride exchange.7,8
Recently the first thermally stable group 5 dihydrogen
complexes were described. Chaudret and co-workers9 reported
the tantalocene complex [Ta(η5- C5H5)2(η2-H2)(CO)]+ and some
of us10 the niobocene complexes [Nb(η5- C5H4SiMe3)2(η2-
Since the discovery of the first η2-dihydrogen complex by
Kubas,1 several transition-metal complexes containing dihy-
drogen as a ligand have been described for a wide variety of
metals.2 Until the seminal discovery by Kubas, the only way
in which hydrogen was known to coordinate to a wide range of
metals was in the polyhydride form.3 Both types of complexes
present very interesting spectroscopic properties. Many dihy-
drogen complexes exhibit the phenomenon of rotational tun-
neling splitting, which is observed by inelastic neutron scattering
in the microwave zone of the electromagnetic spectrum.4 On
the other hand, the proton NMR spectra of several polyhydride
complexes exhibit extremely large and temperature-dependent
scalar couplings, the isotope dependence of these couplings also
(5) (a) Heinekey, D. M.; Millar, J. M.; Koetzle, T. F.; Payne, N. G.;
Zilm, K. W. J. Am. Chem. Soc. 1990, 112, 909. (b) Antin˜olo, A.; Chaudret,
B.; Commenges, G.; Fajardo, M., Jalo´n, F.; Morris, R. H.; Otero, A.;
Schweitzer, C. T. J. Chem. Soc., Chem. Commun. 1988, 1210. (c) Gusev,
D. G.; Kuhlman, R.; Sini, G.; Eisenstein, O.; Caulton, K. G. J. Am. Chem.
Soc. 1994, 116, 2685. (d) Chaudret, B.; Limbach, H. H.; Moise, C. C. R.
Hebd. Se´ances Acad. Sci. 1992, 315-II, 533.
† Universidad de Castilla-La Mancha.
‡ Universidad de Alcala´.
§ Universitat Auto`noma de Barcelona.
| Universite´ de Montpellier.
X Abstract published in AdVance ACS Abstracts, June 15, 1997.
(1) Kubas, G. J.; Ryan, R. R.; Swanson, B. I.; Vergamini, P. J.;
Wasserman, H. J. J. Am. Chem. Soc. 1984, 106, 451.
(2) (a) Kubas, G. J. Acc. Chem. Res. 1988, 21, 120. (b) Crabtree, R. H.
Acc. Chem. ReV. 1990, 23, 95. (c) Jessop, P. J.; Morris, R. H. Coord. Chem.
ReV. 1992, 121, 155. (d) Heinekey, D. M.; Oldham, W. J., Jr. Chem. ReV.
1993, 93, 913. (e) Crabtree, R. H. Angew. Chem., Int. Ed. Engl. 1993, 32,
789.
(3) (a) Hlatky, G. G.; Crabtree, R. H. Coord. Chem. ReV. 1985, 65, 1.
(b) Dedieu, A. Transition Metal Hydrides; VCH: Weinheim, Germany,
1991. (c) Lin, Z.; Hall, M. B. Coord. Chem. ReV. 1994, 135, 845.
(4) Eckert, J.; Kubas, G. J. J. Phys. Chem. 1993, 97, 2378 and references
therein.
(6) (a) Kubas, G. J.; Burns, C. J.; Eckert, J.; Johnson, S. W.; Larson, A.
C. Vergamini, P. J.; Unkefer, C. J.; Khalser, G. R. K.; Jackson, S. A.;
Eisenstein, O. J. Am Chem. Soc. 1993, 115, 569. (b) Eckert, J.; Jensen, C.
M.; Jones, G.; Clot, E.; Eisenstein, O. J. Am Chem. Soc. 1993, 115, 11056.
(c) Li, J.; Ziegler, T. Organometallics 1996, 15, 3844.
(7) (a) Jarid, A.; Moreno, M.; Lledo´s, A.; Lluch, J. M.; Bertra´n, J. J.
Am. Chem. Soc. 1993, 115, 5861. (b) Jarid, A.; Moreno, M.; Lledo´s, A.;
Lluch, J. M.; Bertra´n, J. J. Am. Chem. Soc. 1995, 117, 1069.
(8) (a) Barthelat, J. C.; Chaudret, B.; Daudey, J. P.; DeLoth, P.; Poilblanc,
R. J. Am. Chem. Soc. 1991, 113, 9896. (b) Limbach, H. H.; Scherer, G.;
Maurer, M.; Chaudret, B. Angew. Chem., Int. Ed. Engl. 1992, 31, 1369. (c)
Clot, E.; Leforestier, C.; Eisenstein, O.; Pe´lissier, M. J. Am. Chem. Soc.
1995, 117, 1797.
S0002-7863(96)04035-8 CCC: $14.00 © 1997 American Chemical Society

6108 J. Am. Chem. Soc., Vol. 119, No. 26, 1997
Antin˜olo et al.
Table 1. IR Data (cm-1) for Complexes 2-4 in Comparison with
H2)L]+, (L ) P(OMe)3, P(OEt)3, P(OPh)3, PMe2Ph) by proto-
nation at low temperature of the corresponding neutral hydride
niobium(III) complex. One of the interesting points of these
dihydrogen complexes lies in the fact that the activation barriers
for the rotation are surprisingly high (ca. 10 kcal/mol), therefore
providing an intermediate case between the normal dihydrogen
complexes and the polyhydride ones. As a consequence of the
high barrier, rotation of the dihydrogen molecule seems to be
blocked at the NMR time scale.9b,10 For these complexes a
single resonance is obtained for the NMR spectrum at high
temperatures. When the temperature is lowered, decoalescence
is not observed for the nonisotopically substituted H2 isomer,
but it is observed for the partially deuterated HD species. This
has allowed estimation of the activation barriers for these
complexes. This surprising result has been explained through
the postulation of a large exchange coupling in the H2 case. It
is known9b that for J/∆δ ratios higher than 10 the expected AB-
type spectrum becomes a single line. Although it is expected
that H-H and H-D have very similar barriers for the rotation,
there is no direct experimental evidence of the H-H rotation
being also practically blocked. From a theoretical point of view,
the rotational barrier for the H-H rotation has been calculated
for the simplified model complexes [Cl2Ta(η2-H2)CO]+ and
[Cl2Ta(η2-H2)(PH3)]+ to be 9.7 and 11.3 kcal/mol, respectively.9b
The purpose of this paper is twofold: first, to obtain a new
experimental example of a nonrotating coordinated H-D
molecule and, second, to analyze its structure and the hydrogen
exchange process to obtain theoretical evidence that for the
corresponding H-H molecule this process is also classically
blocked at the NMR time scale and to estimate the order of
magnitude of the exchange coupling. To this aim we decided
to explore the reactivity of the Nb(η5-C5H4SiMe3)2(H)3 toward
isocyanides, which has allowed the preparation of stable
hydride-isocyanide derivatives, Nb(η5-C5H4SiMe3)2(H)(CNR),
as well as their protonation process to give [Nb(η5-C5H4SiMe3)2-
(η2-H2)(CNR)]+, a new niobocene containing dihydrogen as a
ligand that is likely to have the dihydrogen rotation practically
blocked.
the Corresponding Free Ligand
complex ν(NbH) (cm-1
)
ν(CN) (cm-1
)
ν(CN) (cm-1) (free ligand)
2
3
4
1717
1715
1676
2077; 1825
2254; 1813
2000
2129
2138
2113
THF. In addition, Bercaw and co-workers15 have suggested the
formation of an undetected intermediate Nb(η5-C5H5)2(H)-
(CNMe) in the synthesis of an amide-isocyanide derivative.
Taking all of this into account, the synthesis of the Nb(η5-
C5H5)2(H)(CNR) complexes is an additional challenge we will
consider in this work.
2. Results and Discussion
Preparation of Hydride-Isocyanide Niobocene Com-
plexes. The complexes Nb(η5-C5H4SiMe3)2(H)(CNR) (R )
tBu, 2; Cy, 3; 2,6-Me2C6H3dXylyl, 4) are readily prepared in
high yields (95%) by thermal treatment in THF of Nb(η5-C5H4-
SiMe3)2(H)3 (1) in the presence of the corresponding isocyanide
(eq 1). In the preparation of hydride-isocyanide complexes,
the initial step would be the formation of the coordinatively
unsaturated 16-electron species Nb(η5-C5H4SiMe3)2H by ther-
molytic loss of H2 followed by the coordination of isocyanide
ligand (eq 1). The reaction is similar to that previously
reported16 for the preparation of several families of niobocene
complexes, Nb(η5-C5H4SiMe3)2(H)(L) (L ) π-acid ligand). The
complexes were isolated as air-sensitive red oily materials of
spectroscopic purity. It is noteworthy that no subsequent
insertion process was observed even when the process was
carried out with an excess of ligand in refluxing THF for a long
time. This surprising behavior has been previously reported14
for the complex [Mo(η5-C5H5)2(H)(CNMe)]+, where the inser-
tion was not observed either under similar reaction conditions.
This behavior contrasts with the one found for hydride group 4
metallocenes where the insertion process to give iminoacyl
derivatives is especially favored.17
Isocyanide complexes of a wide variety of transition metals
are known, although insertion reactions of this ligand into
transition-metal-carbon, -hydride, -halide, -nitrogen, -sul-
fur, and -oxygen bonds are found to readily occur in organo-
metallic chemistry.11 In particular, the insertion process into
metal-hydride bonds to give stable iminoacyls (-CHN) is
specially noteworthy.12 In this field, early transition metal-
locenes containing isocyanides as ligands have been extensively
studied11 and some stable group 5 metallocenes with halide and
alkyl as ancillary ligands were reported.13 However, because
of the aforementioned easy insertion of isocyanides into the
metal-hydride bond, only one hydride-isocyanide metallocene
species has been described,14 [Mo(η5-C5H5)2(H)(CNMe)]+I-,
which was prepared by the reaction of Mo(η5-C5H5)2HI with
CNMe, which did not undergo insertion even under refluxing
Spectroscopic Properties of Hydride-Isocyanide Niobo-
cene Complexes. A terminal hydride stretching mode, ν(Nb-
H), is detected by infrared spectroscopy for the complexes 2-4
at ca. 1700 cm-1 (Table 1). However the most significant IR
band corresponds to isocyanide stretching mode, ν(CN) at ca.
2000 cm-1. In addition, a band in the spectra of 2 and 3 at ca.
1800 cm-1 could be a combination band (Table 1). The data
agree with the presence of a linear isocyanide ligand.
(9) (a) Sabo-Etienne, S.; Chaudret, B.; Abou el Makarim, H.; Barthelat,
J-C.; Daudey, J.-P.; Mo¨ıse, C.; Leblanc, J.-C. J. Am. Chem. Soc. 1994,
116, 9335. (b) Sabo-Etienne, S.; Chaudret, B.; Abou el Makarim, H.;
Barthelat, J.-C.; Daudey, J.-P.; Ulrich, S.; Limbach, H. H.; Mo¨ıse, C. J.
Am. Chem. Soc. 1995, 117, 11602.
(10) Jalo´n, F.; Otero, A.; Manzano; B.; Villasen˜or, E.; Chaudret, B. J.
Am. Chem. Soc. 1995, 117, 10123.
1H NMR spectra show a high-field resonance at ca. -6.0
ppm for hydride ligands and four multiplets, corresponding to
(11) (a) Singleton, E.; Oosthuizen, H. E. AdV. Organomet. Chem. 1983,
22, 209; (b) Treichel, P. M. AdV. Organomet. Chem. 1983, 11, 21.
(12) Durfee, L. D.; Rothwell, I. A. Chem. ReV. 1988, 88, 1059.
(13) (a) Serrano, R.; Royo, P. J. Organomet. Chem. 1983, 247, 33. (b)
Klazinga, A. H.; Teuben, J. J. Organomet. Chem. 1980, 192, 75. (c) Martinez
de Ilarduya, J. M.; Otero, A.; Royo, P. J. Organomet. Chem. 1988, 340,
187.
(15) Burger, B. J.; Santarsiero, B. D.; Trimmer, M. S.; Bercaw, J. E. J.
Am. Chem. Soc. 1988, 110, 3134.
(16) (a) Antin˜olo, A.; Fajardo, M.; Jalo´n, F.; Lo´pez-Mardomingo, C.;
Otero, A.; Sanz-Bernabe´, C. J. Organomet. Chem. 1989, 369, 187. (b)
Antin˜olo, A.; Carrillo, F.; Fajardo, M.; Garcia-Yuste, S.; Otero, A. J.
Organomet. Chem. 1994, 482, 93. (c) Antin˜olo, A.; Carrillo, F.; Garcia-
Yuste, S.; Otero, A. Organometallics 1994, 13, 2761.
(14) Calhorda, M. J.; Dias, A. R.; Duarte, M. T.; Martins, A. M.; Matias,
P. M.; Romao,C. C. J. Organomet. Chem. 1992, 440, 119.
(17) See, for example: Wolczanski, P. T.; Bercaw, J. E. J. Am. Chem.
Soc. 1979, 101, 6450.

Dihydrogen Isocyanide Niobocene
J. Am. Chem. Soc., Vol. 119, No. 26, 1997 6109
Table 2. Observed JH-D and Activation Parameter for Complexes
5-7-d1, Observed Minimum T1 Values for the Dihydrogen Ligand
in Complexes 5-7, and Estimated rH-H Distances, Assuming Slow
H2 Spinning
an AA′BB′ spin system, for the cyclopentadienyl rings in
accordance with an asymmetrical environment for the niobium
center (see Experimental Section).
In the 13C NMR spectra (see Experimental Section), the most
interesting feature is the extremely low-field resonance which
has been observed for the carbon of isocyanide bonded to the
niobium center, at δ 267.0, 263.0, and 240.0 ppm for 2-4,
respectively, far from the δ values (154-165 ppm) obtained in
the free isocyanides. This resonance, which should appear as
an 1:1:1 triplet from 13C-14N spin coupling, is usually
observed11a as a broad singlet, like in the present case, or a
triplet with a more intense central peak due to the quadrupolar
relaxation of the coupling.
complex JH-D (Hz) Tc (K) ∆G (kcal/mol)
T1 (ms)
rH-H (Å)
5
6
7
30
30
30
193
198
183a
8.9
9.1
8.4a
9.4 (203 K)
7.9 (193 K)
10.8 (204 K)
1.01
1.03
1.06
a See text.
1J(H-D) ) 30 Hz (at 243 K) (Table 2) verify the presence of
an H-D bond and are an effective proof of the presence of a
dihydrogen complex.2 Moreover, the observation of low
longitudinal relaxation times, T1, is particularly useful to
diagnose the presence of dihydrogen ligand in fluxional poly-
hydrides.23 The T1 values of the dihydrogen nuclei, T1(H2), of
5-7 and the hydride ligand, T1(H), of 4 were determined over
the temperature range 178-273 K in acetone-d6 (Table 2). The
T1(min) were found to be 9.4 ms at 193 K for 5, 7.9 ms at 198
K for 6, 10.8 ms at 204 K for 7, and 500 ms at 154 K for 4. In
accordance with the Crabtree’s proposal,23 the H-H distance
rH-H in the coordinated H2 ligand can be determined from the
T1 measurements, but although the H-H dipole interaction is
the most important factor in the relaxation mechanism, we have
considered that additional contributions must be introduced.
Thus, the observed relaxation (Robs) for 5-7 was considered as
the sum of what is due to H-H dipole interaction (RH-H) and
what is due to all other effects (Ro). The value of Ro was
approximated as the relaxation observed for 4, as a reasonable
indication of the niobium and solvent contributions. Then, the
following equation is obtained:
Preparation and Spectroscopic and Theoretical Charac-
terization of η2-Dihydrogen Complexes. Low-temperature
protonation of neutral hydride complexes to give cationic
dihydrogen species was reported for the first time by Crabtree
et al. in 1985.18 Following this method, several cationic
dihydrogen complexes have been prepared for group 8 metals
and iridium in the last years.19,20
Complexes 2-4 were easily protonated at low temperature
to give the corresponding dihydrogen-containing complexes
according to eq 2. These processes were carried out by adding
Nb(η5-C5H4SiMe3)2(H)(RNC) + CF3COOH f
[Nb(η5-C5H4SiMe3)2(η2-H2)(CNR)]CF3CO2
R ) tBu, 5; Cy, 6; 2,6-Me2C6H3dXylyl, 7
(2)
a little excess of CH3COOH to an acetone-d6 solution of these
complexes at 183 K in a NMR tube.
The dihydrogen complexes 5-7 constitute, as far as we know,
the third example of dihydrogen-isocyanide complexes. Simp-
son and co-workers20 prepared the first complex, [Ru(η5-
C5H5)(PPh3)(CNtBu)(η2-H2)]+PF6-, by protonation of the cor-
responding neutral hydride isocyanide species. Recently,
Heinekey et al.21 also obtained several rhenium complexes by
using the same procedure.
Robs(5-7) ) RH-H + Ro; Ro ) Robs(4)
From this equation we have calculated RH-H values of 104.38,
124.58, and 90.59 (µs)-1, for complexes 5-7, respectively. rH-H
values between 1.01 and 1.06 Å were found in accordance with
the Morris approximations assuming slow H2 spinning (Table
2).24
The 1H NMR spectra show a broad dihydrogen resonance at
213 K at ca. -7.0 ppm and a large line width, 60 Hz, explained
by the rapid dipolar relaxation of these nuclei. At low-
temperature decoalescence was not observed, and at 178 K, a
single high-field resonance was present. When the temperature
is raised near 283 K, decomposition of the dihydrogen com-
plexes 5 and 6 takes place to give a complex mixture and H2.
Nevertheless, 7 evolves with elimination of H2 and formation
of the neutral complex Nb(η5-C5H4SiMe3)2(OOCCF3)(CNXylyl)
(8).22
The calculated JHD of 30 Hz for 5-7-d1 agree very well with
the presence of an unstretched dihydrogen ligand (as a com-
parison, the value of JHD in the HD gas molecule is 43.2
Hz25), and it is comparable with the value of JHD ) 27.5 Hz
found previously in [Ta(η5-C5H5)2(η2-HD)(CO)]+. The data are
indicative of a similarity in the back-donation of both metal
centers to the σ* orbital of HD, in accordance with a comparable
behavior as π-acid ligands of both carbonyl and isocyanide
ligands. In contrast, in the analogous niobocene complexes10
[Nb(η5-C5H4SiMe3)2(η2-HD)(L)]+ (L ) P(OEt)3, PMe2Ph), the
reported JH-D values (18.2 and 15.0 Hz, respectively) indicate
the presence of a stretched H-D ligand. This is due to the
fact that a more effective back-donation on the σ* orbital of
HD may occur in the presence of the less π-acceptors phosphite
or phosphine ligands.
The monodeuterated complexes 5-7-d1 have been pre-
pared by adding CF3COOD to an acetone-d6 solution of 2-4
at 183 K in a NMR tube. 1H NMR spectra show an 1:1:1 triplet
which broadens at low temperature (with cooling). Couplings
(18) Crabtree, R. H.; Lavin, M. J. Chem. Soc., Chem. Commun. 1985,
1661.
Experimental characterization of species 5-7 as dihydrogen
complexes is only spectroscopical. X-ray or neutron diffraction
experiments have not been feasible for any dihydrogen metal-
locene complex of group 5 metal, nor for any dihydrogen
complex having an isocyanide ligand. At this point, ab initio
calculations have proven to be very useful providing quite
(19) Some selected references: (a) Morris, R. M.; Sawyer, J. F.; Shiralian,
M.; Zubkowski, J. D. J. Am. Chem. Soc. 1985, 107, 5581. (b) Chinn, M.
S.; Heinekey, D. M. J. Am Chem. Soc. 1987, 109, 5865. (c) Albertin, G.;
Antoniutti, S.; Bordignon, E. J. Am Chem. Soc. 1989, 111, 2072. (d) Chinn,
M. S.; Heinekey, D. M.; Paune, N. E.; Sofield, C. D. Organometallics 1989,
8, 1824. (e) Antoniutti, S.; Albertin, G.; Amendola, P.; Bordignan, E. J.
Chem. Soc., Chem.Commun. 1989, 229. (f) Chinn, M. S.; Heinekey, D. M.
J. Am Chem. Soc. 1990, 112, 5166.
(20) Conroy-Lewis, F. M.; Simpson, S. O. J. Chem. Soc., Chem.
Commun. 1986, 506.
(21) Heinekey, D. M.; Voges, M. M.; Barnhart, D. M. J. Am Chem. Soc.
1996, 118, 10792.
(22) Antin˜olo, A.; Carrillo-Hermosilla, F.; Fajardo, M.; Freitas, M.;
Garc´ıa-Yuste, S.; Otero, A.; Prashar, S.; Villasen˜or, E. Inorg. Chim. Acta,
in press.
(23) (a) Crabtree, R. H.; Lavin, M.; Bonneviot, L. J. Am. Chem. Soc.
1986, 108, 4032. (b) Hamilton, D. G.; Crabtree, R. H. J. Am. Chem. Soc.
1988, 110, 4126.
(24) Earl, K. A.; Jia, G.; Maltby, P. A.; Morris, R. H. J. Am. Chem. Soc.
1991, 113, 3027.
(25) Bloyce, P. E.; Rest, A. J.; Whitwell, L.; Graham, W. A. G.; Holmes-
Smith, R. J. Chem. Soc., Chem. Commun. 1985, 846.

6110 J. Am. Chem. Soc., Vol. 119, No. 26, 1997
Antin˜olo et al.
Table 3. Main Geometrical Parametersa of Structures 9-11
strongly bonded to the metal than the dihydrogen ligand in 9
provokes a weakening of the Nb-CNR bond which clearly
enlarges when going from the dihydrogen 9 to the dihydride
10 structures.
geometrical
parameters
dihydrogen trans complex rotational transition
complex 9
10
structure 11
Nb-H1
Nb-H2
H1‚‚‚H2
Nb-C1
C1-N
Nb‚‚‚XCp
H1NbH2
NbC1N
1.853
1.880
0.871
2.149
1.178
2.135
27.0
179.5
139.3
101.0
178.6
1.739
1.739
3.166
2.192
1.172
2.135
131.1
180.0
139.4
65.5
2.093
2.092
0.765
2.134
1.181
2.127
21.1
178.4
138.5
86.6
Rotational Process. In metallocene complexes 5-7 with
two different substituents, namely isocyanide and dihydrogen
ligands, the two hydrogen atoms might be observed as chemi-
cally inequivalent if rotation could be frozen out30 or decoa-
lescence might occur if intramolecular rotation of the η2-
dihydrogen ligand could be slowed to a rate comparable to the
chemical shift difference (estimated at ca. 0.2 ppm at 183 K
for 5-7) between the two ends of the H2 ligand. Upon lowering
of the temperature (Figure 2), the experiments were carried out
at 300 MHz in acetone-d6 as solvent and the triplet signals
observed in 5-d1 and 6-d1 decoalesced below 193 and 198 K,
respectively, and were transformed in two unresolved broad
singlets (line width of 100 Hz) which are due to the nonrotating
H-D molecule in the two different rotamers endo H 5-d1 and
6-d1 and exo H 5-d1 and 6-d1, with the H atom located either
next or opposite to the isocyanide ligand, respectively. For
complex 7-d1 the triplet signal coalesced at 183 K, but the
splitting in two signals corresponding to the two rotamers was
not observed in the range of temperatures used (293-178 K).
The processes were reversible; thus, when the temperature was
raised, the rotamers endo H 5-d1 and 6-d1 and exo H 5-d1 and
6-d1 were interconverted by rotation of the H-D molecule.
A similar behavior, which involves a slowly rotating HD
molecule coordinated to a metal center, has been recently
described for the first time for the [Nb(η5-C5H4SiMe3)2(η2_HD)-
(PMe2Ph)]CF3CO210 and [Ta(η5-C5H5)2(η2-HD)(CO)]BF49b com-
plexes. Inelastic neutron scattering studies have been carried
out on several dihydrogen complexes, such as W(CO)3(η2-H2)-
(PCy3)2,31a IrClH2(η2-H2)(PiPr3)2,6b MoCO(η2-H2)(dppe)2,6a [FeH-
(η2-H2)(dppe)2]+,31b and TpMe2Rh(η2-H2)H2.31c Rotational en-
ergy barriers below 3.5 kcal/mol were found in all cases.4,6,31
Free energy of activation of our processes was found to be ca.
8.4-9.1 kcal/mol (Table 2) from a line-shape analysis, although
we have not clearly observed the exchange limit of the
resonances of both rotamers in the NMR spectra.32 These values
are close to the 11.0 and 9.6 kcal/mol free energies of activation
obtained respectively for the aforementioned niobocene10 and
tantalocene9b complexes. Although decoalescence for 7-d1 was
not observed, taking into account a similar line separation as
for 5-d1 and 6-d1, a free energy of activation of 8.4 kcal/mol
for the process can be estimated. Moreover, the single high-
field resonance observed for 5-7 at low temperature could be
attributed to either a kinetic isotope effect or a very large
exchange coupling. For complex [Ta(η5-C5H5)2(η2-H2)(CO)]-
BF4, no kinetic isotope effect was observed for the classical
rotation of the dihydrogen molecule and the presence of a large
exchange coupling constant in the signal was proposed.9b
To theoretically evaluate the energy barrier for the rotational
process, we have located the transition state for the rotation. It
has been calculated by a complete optimization of the geometry
(as for 9 and 10, the only restriction applies to the internal C5H5
coordinates which are restricted to a local C5v symmetry; see
Details of the DFT Calculations). However, now, the H-H
X
CpNbXCp
H1NbC1
C1NC2
179.8
179.9
a Distances in angstroms and angles in degrees.
accurate data in both classical polyhydride3c,26 and nonclassical
dihydrogen complexes.27 Up to now, the only theoretical
calculations of a group 5 metal dihydrogen complexes are
restricted to very simplified models: [Cl2Ta(H2)CO]+ and cis-
[Cl2Ta(H2)(PH3)]+.9b We have performed, for the first time,
such a study for the [Nb(η5-C5H5)2(η2-H2)(CNCH3)]+ complex
9, which is a realistic model of the 5-7 complexes where the
SiMe3 group of the C5H4SiMe3 ligand has been substituted by
a hydrogen atom and the R group of the isocyanide ligand is a
methyl group.
Density functional theory (DFT) calculations lead to the
characterization of 9 as a minimum energy structure. Its main
geometrical parameters are shown in the second column of Table
3, whereas Figure 1a depicts the geometry and the numeration
used in Table 3 to label the atoms. It can be clearly seen that
a dihydrogen ligand is present. The H-H distance of 0.871 Å
is smaller than the values obtained from NMR relaxation times
assuming a slow rotation (see Table 2). However, it is well-
known that the T1 method give ambiguous values of H-H
distances which depend on the interpretation of the motion of
the H2 ligand.28 The best estimate of rH-H in solution for a
dihydrogen complex when JH-D is known is provided by an
equation obtained by Morris and co-workers.28 According to
that equation, a value of JH-D of 30 Hz for 5-7-d1 corresponds
to a distance of 0.919 Å. In a previous study of compounds
very similar to the ones considered here, the direct protonation
leads to a dihydride transoid structure instead of the dihydrogen
complexes 5-7 obtained when a little excess of CF3COOH is
added.10 Therefore, we have also calculated the dihydride
transoid 10 structure which is also shown in Table 3 (third
column) and Figure 1b. 10 is also a minimum energy structure,
and it is slightly more stable (0.30 kcal/mol) than the dihydrogen
one at our level of calculation. This very small difference
explains the fact that slight modifications of the experimental
conditions can lead to different tautomers. It is quite clear that
10 possesses two hydride ligands. Nb-H distances are, as
expected, shorter for the dihydride 10 structure; these distances
are also very similar to the equivalent Nb-H distances in the
trihydride Nb(η5-C5H5)2(H)3 complex which has been previously
reported.26d,29 The fact that in 10 the hydride ligands are more
(26) (a) Lin, Z.; Hall, M. B. Inorg. Chem. 1991, 30, 2569. (b) Esteruelas,
M. A.; Jean, Y.; Lledo´s, A.; Oro, L. A.; Ruiz, N.; Volatron, F. Inorg. Chem.
1994, 33, 3609. (c) Buil, M. L.; Espinet, P.; Esteruelas, M. A.; Lahoz, F.
J.; Lledo´s, A.; Mart´ınez-Ilarduya, J. M.; Maseras, F.; Modrego, J.; On˜ate,
E.; Oro, L. A.; Sola, E.; Valero, C. Inorg. Chem. 1996, 35, 1250. (d)
Camanyes, S.; Maseras, F.; Moreno, M.; Lledo´s, A.; Lluch, J. M.; Bertra´n,
J. J. Am. Chem. Soc. 1996, 118, 4617.
(27) (a) Dapprich, S.; Frenking, G Angew Chem., Int. Ed. Engl. 1995,
34, 354. (b) Maseras, F.; Lledo´s, A.; Costas, M.; Poblet, J. M. Organome-
tallics 1996, 15, 2947.
(29) Wilson, R. D.; Koetzle, T. F.; Hart, D. W.; Kvick, Å.; Tipton, D.
L.; Bau, R. J. Am. Chem. Soc. 1977, 99, 1775.
(30) Only for the rotamer with the η2-dihydrogen ligand perpendicular
to the molecular symmetry plane, the two hydrogen atoms would be
equivalents even in the absence of rotation.
(31) (a) Eckert, J.; Kubas, G. J.; Hall, J. H.; Hay, P. J.; Boyle, C. M. J.
Am. Chem. Soc. 1990, 112, 2324. (b) Eckert, J.; Blank, H.; Bautista, M. T.;
Morris, R. H. Inorg. Chem. 1990, 29, 747. (c) Eckert, J.; Albinati, A.;
Bucher, U. E.; Venanzi, L. M. Inorg. Chem. 1996, 35, 1292.
(32) The maximum separation found was used in the calculation.
(28) Maltby, P. A.; Schlaf, M.; Steinbeck, M.; Lough, A. J.; Morris, R.
H.; Klooster, W. T.; Koetzle, T. F.; Srivastava, R. C. J. Am. Chem. Soc.
1996, 118, 5396.

Dihydrogen Isocyanide Niobocene
J. Am. Chem. Soc., Vol. 119, No. 26, 1997 6111
Figure 1. Structures corresponding to (a) 9, (b) 10, and (c) 11.
transition state. The energy barrier of the rotational process is
10.85 kcal/mol at our level of calculation. This value compares
well with the free energy barriers obtained from the analysis of
NMR spectra at different temperatures for 5-7 complexes (see
Table 2). These high barriers to rotation of coordinated
dihydrogen contrast with other complexes previously studied
where the rotational barriers obtained were always below 3.5
kcal/mol.4,6,31
The high rotational barrier can be explained through molec-
ular orbital analysis. The occupied d orbital in a d2 [Nb(η5-
C5H5)2(L)] fragment lies in a plane that contains the Nb-L bond
and is perpendicular to the straight line joining the two Cp
centers.33 This orbital can back-donate to the σ* of the H2
fragment when the dihydrogen is placed in that plane. This is
the case of minimum 9. However, when the H2 is rotated 90°,
the σ* orbital does not find any occupied d orbital in the metal
to interact so that the back-donation is completely lost in
transition state 11, which therefore has a remarkably high
energy. The energy of the rotational barrier can be attributed
then to the complete loss of back-donation in the transition state.
In this way, the contribution of backdonation to the binding
energy of the H2 to the [Nb(η5-C5H5)2(CNCH3)]+ fragment is
10.85 kcal/mol. We have also calculated this binding energy:
its value is 20.7 kcal/mol, indicating a strong bonding of H2 to
the fragment, comparable to the values experimentally deter-
mined34 and theoretically predicted27a by high-level calculations
in complexes M(CO)5(η2-H2), (M ) Cr, Mo, W). The differ-
ence (9.8 kcal/mol) between the binding energy and the
rotational barrier can be taken as an approximate value of the
σH f Nb donation contribution to the metal H-H bond energy.
2
Rotational Tunneling. A very interesting feature of com-
pounds 5-7 is the non-decoalescence of the NMR signal of
the dihydrogen upon lowering the temperature. In a similar
case, [Ta(η5-C5H5)2(η2-H2)(CO)]+, the presence of a very large
exchange coupling (J/∆δ > 10) has been tentatively proposed.9b
In this subsection, we will use the DFT results to analyze the
rotational tunneling of the dihydrogen ligand which will allow
us to obtain the order of magnitude of the exchange coupling
in complex 9.
Figure 2. Variable-temperature 300 MHz 1H NMR spectra in the high-
field region of [Nb(η5-C5H4SiMe3)2(η2-HD)(CNtBu)](CF3COO)(5-d1)
in acetone-d6. The asterisk denotes the presence of residual [Nb(η5-
C5H4SiMe3)2(η2-H2)(CNtBu)](CF3COO).
The first step to achieve this goal is to obtain a reasonable
monodimensional tunneling path through the whole potential
energy hypersurface. To do so, we have considered four internal
coordinates as the only ones relevant to the rotation: the H1-
H2 distance RHH, the distance between the Nb and the center of
the H1-H2 bond RMX, the angle between the H1-H2 bond and
the Nb-X vector æ, and the rotational angle θ of the H1-H2
bond (see Scheme 1).
bond is twisted 90° with respect to the position in the minimum
9. In this way structure 11 is found. Figure 1c and the last
column of Table 3 give, respectively, the geometry and the most
relevant parameters of this transition state. It is worthwhile to
note a diminution of the H1-H2 bond as compared to the
minimum dihydrogen 9 structure which is accompanied by a
clear increase of Nb-H distances. The values of 0.765 Å for
the H-H distance and ca. 2.09 Å for Nb-H distances are very
close to the values previously reported by some of us26d for the
transition state corresponding to the exchange of two hydrides
in the Nb(η5-C5H5)2(H)3 system. This similarity validates the
mechanism proposed for the hydrogen exchange in trihydrides
which also involved a twisted dihydrogen structure as a
The energy profile has been obtained by considering different
values of the angle θ and optimizing the other relevant
geometrical parameters. The rest of the parameters are kept
(33) Lauer, J. W.; Hoffmann, R. J. Am. Chem. Soc. 1976, 98, 1729.
(34) Gonzalez, A. A.; Zhang, K.; Nolan, S. P.; Lopez de la Vega, R.;
Mukerjee, S. L.; Hoff, C. D.; Kubas, G. J. Organometallics 1988, 7, 2429.

6112 J. Am. Chem. Soc., Vol. 119, No. 26, 1997
Antin˜olo et al.
Table 4. Calculated Exchange Couplings at Various Temperatures
for the Different Isotopomers
Scheme 1
T (K)
JH-H (Hz)
JD-D (Hz)
JT-T (Hz)
0
50
100
150
200
9.9 × 105
9.9 × 105
9.9 × 105
1.0 × 106
1.2 × 106
2.7 × 102
2.7 × 102
2.8 × 102
4.3 × 102
2.1 × 103
2.3 × 100
2.3 × 100
2.3 × 100
3.9 × 100
1.4 × 102
frozen at the minimum energy values. The path length is
evaluated by obtaining the distance between two consecutive
points in mass-weighted Cartesian coordinates. This implies
that each point must be rotated and the center of mass of each
pair of geometries must coincide in order to generate neither
linear nor angular momenta along the path.35 In this manner,
we obtain a symmetric double-well profile with a maximum
energy at θ ) 90°. The energy barrier, 11.02 kcal/mol, is only
slightly above the 10.85 kcal/mol of transition state 11 found
in the previous subsection. This validates the election of the
geometrical parameters used to analyze the rotational process.
It has to be remarked that the hydrogen exchange in a η2-
dihydrogen complex involves a shorter tunneling path than in
a polyhydride complex. In effect, the exchange of two hydrides
takes place through the formation of a hydrogen molecule
bonded to the transition metal, which is already able to rotate.
The evolution from the dihydride structure to the hydrogen
molecule originates a long tunneling path. On the contrary,
the hydrogen molecule already exists in a η2-dihydrogen
complex, in such a way that the tunneling path length in this
case is essentially determined by the rotation motion.
The second column in Table 4 shows the exchange coupling
for the original, nonisotopically substituted complex 9. The
value of the exchange at the usual experimental temperatures
(ca. 200 K) is on the order of 106 Hz. This high value would
clearly indicate that decoalescence cannot occur for the H-H
complex. Even at 0 K the exchange would still be too fast to
be observed by NMR experiments as seen in Table 4. There-
fore, the classical H-H rotation is dramatically slowed down
for compound 9 but a significant quantum exchange (i.e.,
tunneling) still occurs.
A very different matter happens for the monodeuterated H-D
isotopomer. In this case, given the different isotopic substitution
of both hydrogens, the double-well energy profile for the
rotational process is not symmetric. We have evaluated the
asymmetry by assuming that it comes only from the different
contribution to the zero point energy of the H-H/H-D
stretching frequency. To achieve that, given the high frequency
of the hydrogen stretching, we have considered that the rest of
the motions in the molecule do not noticeably affect H-H
frequency. Therefore, a numerical evaluation of the second
derivatives of the energy through small displacements of the
Cartesian coordinates restricted to the two hydrogen atoms has
been carried out. The 6 × 6 matrix obtained is then diagonal-
ized and the highest eigenvalue assigned to H-H stretching.
In this manner we have evaluated H-H stretching in 9 to be
2592 cm-1. This value is clearly smaller than the one for the
free dihydrogen (4470 cm-1 at our level of calculation),
indicating a considerable degree of cleavage for the H-H bond
in the complex. This frequency depends on the mass of the
atoms. Upon deuteration of one of the hydrogens, the system
is no longer symmetric and two different frequencies can be
found for the H-D isotopomer. This difference in frequency
is very small and implies a difference of zero point energy of
only 0.037 kcal/mol. However, it is known that only a very
small difference in energy between the two wells breaks the
symmetry of the system and the tunneling dramatically de-
creases.37 A new calculation of the energy levels of the double
well by introducing this very small asymmetry shows that
tunneling is totally quenched. Then, the presence of nonidenti-
cal particles, H and D, prevents permutational symmetry on the
wave function, in such a way that, for the H-D isotopomer,
the quantum exchange is not possible and decoalescence should
be obtained as it is observed experimentally.
Once the energy profile was obtained by DFT calculations,
a symmetric double well was built by means of a cubic spline
function which passes through each consecutive pair of points
along the reaction coordinate. Because we are interested in the
evaluation of the exchange coupling, we must obtain the
vibrational states of the double well. For this purpose, we have
used a basis set methodology by taking a set of localized
Gaussian functions which have the form36
1/4
R
π
-R
2
øi )
exp
(s - si)2
( )
[
]
where R is an optimizable parameter and si values are equally
spaced points along the coordinate space. Then, a variational
calculation by using n functions provides the lowest n eigen-
values and eigenfunctions of the one-dimensional system. In
particular, we have used 99 Gaussian functions throughout the
calculations. We have observed that a further increase in the
number of basis functions does not appreciably modify the
energies of the levels below the barrier.
This calculation provides the energy of each pair of near-
degenerate levels. The difference in energy between the two
levels of the lowest lying of this pairs provides the tunneling
splitting at 0 K which is a direct measure of the quantum
exchange coupling. At higher temperatures, additional pairs
of levels have to be considered. Since each consecutive pair is
widely separated in energy, it can be safely assumed that the
exchange process can be considered separately in each pair and
a Boltzmann distribution over the thermally accessible energy
levels has been performed in order to obtain the exchange
coupling at different temperatures. Results are summarized in
Table 4.
A different behavior is predicted for the D-D and T-T
isotopomers respectively shown in the third and fourth column
of Table 4. In these cases, the symmetry of the energy profile
is preserved so that quantum exchange is not quenched. The
main difference from the initial H-H complex comes now from
the different mass of the rotating atoms which greatly increases
the tunneling path which is measured in mass-weighted Carte-
sian coordinates. Therefore, the quantum exchange is smaller
for the D-D isotopomer. Anyway, the values, even at very
low temperatures, are still very large compared with the NMR
displacements of the deuterium so that, in this case, decoales-
(35) (a) Miller, W. H.; Ruf, B. A.; Chang, Y. J. Chem. Phys. 1988, 89,
6298. (b) Bosch, E.; Moreno, M.; Lluch, J. M. J. Am Chem. Soc. 1992,
114, 2072.
(36) (a) Hamilton, I. P.; Light, J. J. Chem. Phys. 1986, 84, 306. (b) Makri,
N.; Miller, W. H. J. Chem. Phys. 1987, 86, 1451.
(37) Gelabert, R.; Moreno, M.; Lluch, J. M. J. Comput. Chem. 1994,
15, 125.

Dihydrogen Isocyanide Niobocene
J. Am. Chem. Soc., Vol. 119, No. 26, 1997 6113
cence should not be observed either. This fact has been experi-
mentally observed for the [Cp2Ta(D2)(CO)]+ complex.9b Fi-
nally, for the T-T complex, the additional diminution of the
quantum exchange down to values lower than 4 Hz at 150 K
indicates that this hypothetical species should show decoales-
cence. In other words, at very low temperature, the classical
rotation as well as the quantum exchange of the T-T species
seem to be blocked at the NMR time scale.
both factors diminish the splitting, so these processes can be
observed in the radiofrequency zone of the electromagnetic
spectrum, around 103 Hz. For the case studied here, the energy
barrier is high but the tunneling path is short so that exchange
couplings are on the order of 106 Hz. These couplings are
detected in the NMR spectra, preventing the decoalescence of
the H2 signal. A similar effect would occur if a case of low
barrier but long tunneling path for the H-H exchange could
be found. Additional theoretical research on this field is now
in progress.
3. Concluding Remarks
In this paper, we have prepared stable hydride isocyanide
derivatives Nb(η5-C5H4SiMe3)2(H)(CNR) through formation of
coordinatively unsaturated 16-electron species Nb(η5-C5H4-
SiMe3)2H by thermolytic loss of H2 followed by the coordination
of isocyanide ligand. Contrary to the behavior observed in other
hydride-isocyanide metallocenes, no evidence of isocyanide
insertion on the niobium-hydride bond is found. Low-
temperature protonation with a little excess of HCF3CO2 leads
to the η2-dihydrogen complexes [Nb(η5-C5H4SiMe3)2(η2-H2)-
(CNR)]+.
These H-H complexes and their monodeuterated H-D
isotopomers show very interesting spectroscopic properties.
Their NMR spectra present a single high-field resonance at room
temperature. By lowering the temperature to 178 K, decoales-
cence of the signal was observed for the H-D complexes but
not for the H-H ones. The temperature of decoalescence has
allowed the estimation of free energy of activation of the
dihydrogen internal rotation which ranges between 8.4 and 9.1
kcal/mol. A similar behavior had been previously observed for
related complexes. Non-decoalescence of the H-H signal was
there tentatively attributed to a large exchange coupling.
In this paper, for the first time, we have theoretically demon-
strated the existence of exchange coupling in the [Nb(η5-C5H5)2-
(η2-H2)(CNCH3)]+ complex that is a realistic model of the
experimental cases. This has been achieved by first determining
the structure of the minima and the transition state for the
dihydrogen rotation. The energy barrier obtained from these
DFT calculations compares very well with the experimental
estimations. The magnitude of the rotational tunneling of the
dihydrogen ligand has been evaluated by obtaining the vibra-
tional quantum states of a monodimensional double-well energy
profile for the H-H rotation.
4. Experimental Section
Oxygen and water were excluded by the use of vacuum lines
supplied with purified N2. Tetrahydrofuran (THF) was dried over
and distilled from sodium-benzophenone. Deuterated solvents were
dried over 4 Å molecular sieves and degassed prior to use. Isocyanide
ligands were used as purchased from Aldrich. NMR spectra were
recorded on a Varian Unity 300 (300 MHz for H, 75 MHz for 13C)
1
spectrometer. Chemical shifts were measured related to partially
deuterated solvent peaks and reported relative to TMS. T1 measure-
ments were made at 300 MHz using the inversion recovery method,
and probe temperatures were calibrated by comparison to the ob-
served chemical shifts differences in the spectrum of pure methanol
with use of data reported by Van Geet.38 IR spectra were recorded on
a Perkin Elmer 883 spectrometer in Nujol mulls over CsI windows.
Complex Nb(η5-C5H4SiMe3)2(H)3 was prepared as described in the
literature.5b
Preparation of [Nb(C5H4SiMe3)2(H)(CNtBu)] (2). Nb(C5H4SiMe3)2-
(H)3 (1) (0.300 g, 0.800 mmol) was dissolved in 40 mL of THF to
form a tan solution. To this solution was added 0.091 mL (0.067 g,
0.800 mmol) of CNtBu by syringe.
The mixture was stirred at 343 K for 2 h. The resulting red-brown
solution was filtered and evaporated to dryness. Complex 2 was
isolated as a red oily material after maintaining it under vacuum for a
lengthy period (yield: 95%). 1H-NMR (300 MHz, C6D6, δ): -5.94
(s, 1H, Nb-H), 0.22 (s, 18H, SiMe3), 1.20 (s, 9H, CN(C(CH3)3)), 4.30
(m, 2H), 4.50 (m, 2H), 4.90 (m, 2H), 5.20 (m, 2H) (AA′BB′, C5H4,
exact assignment not possible). 13C-NMR (75 MHz, C6D6, δ): 0.7
(SiMe3), 30.5 (CN(C(CH3)3)), 58.0 (CN(C(CH3)3)), 88.2, 91.2, 91.8,
92.6 (C2-C5, C5H4), 94.6 (C1, C5H4), 267.0 (CNtBu). IR (Nujol, cm-1):
1717 (Nb-H), 2077, 1825 (CdN, see the text), 1245 (SiMe3).
Preparation of [Nb(C5H4SiMe3)2(H)(CNCy)](3) and [Nb-
(C5H4SiMe3)2(H)(CN-2,3-Me2C5H3)] (4). Complexes 3 and 4 were
isolated as red oily materials in a manner similar to that for 2, in 95%
yield.
We have obtained exchange couplings of ca. 106 Hz for the
H-H case even at 0 K, so that decoalescence should not be
observed. Conversely, for the H-D isotopomer, the difference
in zero point energy corresponding to two nonequivalent (H-D
and D-H) positions leads to a slight asymmetry which
dramatically reduces the coupling, allowing decoalescence to
be observed. Therefore, for this complex, the H-D classical
rotation and the quantum exchange processes will not be
practically observed, whereas for the H-H isomer, only the
classical process is quenched out on the NMR time scale.
Additionally, our theoretical results also predict that the D-D
isotopomer would not show decoalescence, whereas for the T-T
isotopomer decoalescence could be seen.
1
3: H-NMR (300 MHz, C6D6, δ): -5.94 (s, 1H, Nb-H), 0.22 (s,
18H, SiMe3), 1-1.95 (m, CNC6H11), 4.30 (m, 2H), 4.50 (m, 2H), 4.90
(m, 2H), 5.20 (m, 2H) (AA′BB′, C5H4, exact assignment not possible).
13C-NMR (75 MHz, C6D6, δ): 0.7 (SiMe3), 24.6, 25.8, 34.0 (Cy), 91.6,
91.7, 92.2, 93.9 (C2-C5, C5H4), 87.8 (C1, C5H4), 263.0 (CNCy). IR
(Nujol, cm-1): 1715 (Nb-H), 2254, 1813 (CdN, see the text), 1246
(SiMe3).
1
4: H-NMR (300 MHz, C6D6, δ): -6.20 (s, 1H, Nb-H), 0.63 (s,
18H, SiMe3), 2.65 (s, 6H, CH3), 4.82 (m, 2H), 4.94 (m, 2H), 5.36
(m, 2H), 5.60 (m, 2H) (AA′BB′, C5H4), exact assignment not pos-
sible), 7.13-7.14 (m, 3H, C6H3). 13C-NMR (75 MHz, C6D6, δ): 0.4
(SiMe3), 19.6 (CH3), 86.7, 91.6, 91.7 (C5H4), 124.8, 127.8, 128.2, 130.6
(CN(2,6-Me2C6H3)), 240.0 (CN(2,6-Me2C6H3)). IR (Nujol, cm-1):
1676 (Nb-H), 2000 (CdN, see the text), 1245 (SiMe3).
The complex studied here is then an intermediate case
between the hydrogen rotation of dihydrogen complexes and
the exchange of a pair of hydrides in polyhydride complexes.
The former implies the exchange of hydrogens through a low
barrier and a relatively short tunneling path, leading to a splitting
of the energy levels for the double-well potential which can be
found by inelastic neutron scattering in the microwave zone of
the electromagnetic spectrum (frequencies on the order of ca.
1010 Hz). On the other hand, for polyhydrides, the exchange
implies a high energy barrier and a very long tunneling path;
Protonation of 2-4. CF3COOH, or CF3COOD, was added to an
acetone-d6 solution of 2, 3, or 4, in a 5 mm NMR tube at 183 K, to
give the dihydrogen complexes [Nb(C5H4SiMe3)2(H2)(CNR)]+(CF3CO2)-
(5-7) and their (HD) isotopomers (5-d1, 6-d1, and 7-d1).
Details of the DFT Calculations. All calculations have been carried
out with the GAUSSIAN 9439 series of programs. Electronic structure
calculations have been performed with the same methodology employed
in the previous study of metallocene trihydride complexes [Nb(η5-
C5H5)2(H)3]n+ (M ) Mo, W, n ) 1; M ) Nb, Ta, n ) 0).26d The
(38) Van Geet, A. L. Anal. Chem. 1968, 40, 2227.

6114 J. Am. Chem. Soc., Vol. 119, No. 26, 1997
Antin˜olo et al.
Density functional theory (DFT)40 was applied. The particular func-
tional used was the Becke’s three-parameter hybrid method41a using
the LYP correlation functional (Becke3LYP).41b
and H atoms of the isocyanide ligand, the valence double-ú 6-31G basis
set was used.43a A polarization p shell was added to the two hydrogen
atoms directly attached to the metal.43
An effective core potential operator has been used to replace core
electrons of metal atom.42 This involved 28 electrons. The basis set
used for the metal was that associated with the pseudopotential42 with
a standard valence double-ú LANL2DZ contraction.39 For the N, C,
All along the exploratory process of the potential energy surface,
the (C5H5) fragments were restricted to a local C5V symmetry.
Acknowledgment. A.A., F.C.-H., M.F., S.G.-Y., and A.O.
gratefully acknowledge financial support from the DGICYT
(Grant. No. PB92-0715) of Spain. M.M., A.Ll. and J.M.Ll. also
acknowledge financial support from the DGES through projects
Nos. PB95-0637 and PB95-0639-C02-01. The use of compu-
tational facilities of the Centre de Computacio´ i de Comunica-
cions de Catalunya are gratefully appreciated as well.
(39) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B; Gill, P. M. W.; Johnson,
B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T. A.; Petersson, G. A.;
Montgomery, J. A.; Raghavachari, K.; Al-Laham, M. A.; Zakrzewski, V.
G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stevanov, B. B.;
Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, P. Y.; Chen, W.;
Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-
Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian 94; Gaussian, Inc.:
Pittsburgh, PA, 1995.
JA9640354
(40) (a) Parr, R. G.; Yang, W. Density-Functional Theory of Atoms and
Molecules; Oxford University Press: Oxford, U.K., 1989. (b) Ziegler, T.
Chem. ReV. 1991, 91, 651.
(41) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Lee, C.; Yang,
W.; Parr, R. G. Phys. ReV. B 1988, 37, 785.
(42) Hay, P. J.; Wadt, W. R. J. Chem. Phys. 1985, 82, 299.
(43) (a) Hehre, W. J.; Ditchfield, R.; Pople, J. A. J. Chem. Phys. 1972,
56, 2257. (b) Hariharan, P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28,
213.