Synthesis and bonding in the diamagnetic dinuclear tantalum(IV) hydride species ([P2N2]Ta)2(μ-H)4 and the paramagnetic cationic dinuclear hydride species {([P2N2]Ta)2(μ-H)4}+I- ([P2N2] =

Article abstract of DOI:10.1021/om000359w

The controlled reaction of the Ta(V) trimethyl species [P₂N₂]TaMe₃, where [P₂N₂] = PhP(CH₂SiMe₂NSiMe₂CH₂)₂PPh, under 0.5 atm of hydrogen gas produces a partially hydrogenated Ta(V) species, ([P₂N₂]TaMe₂(H), of unknown structure. Under 4 atm of hydrogen gas, further hydrogenation does not produce the complex [P₂N₂]TaH₃; instead, reduction of the tantalum center occurs to yield the dinuclear Ta(IV) hydride ([P₂N₂]Ta)₂-(μ-H)₄. This diamagnetic tetrahydride fails to react with many reagents, including ethylene and carbon monoxide; however, upon addition of iodomethane, {([P₂N₂]Ta)₂(μ-H)₄}⁺I^- is produced as a paramagnetic green crystalline solid. The number of hydrides in this reaction product was confirmed by a deuterium labeling study. The results of a variable-temperature magnetic susceptibility study of this tetrahydride cation can be partially modeled with the Curie-Weiss law and a large correction for temperature-independent magnetism. Ab initio calculations using density functional theory were performed in an attempt to further understand the influence of the macrocyclic ligand in the bonding in these complexes.

Full text of DOI:10.1021/om000359w

Organometallics 2000, 19, 3931-3941
3931
Syn th esis a n d Bon d in g in th e Dia m a gn etic Din u clea r
Ta n ta lu m (IV) Hyd r id e Sp ecies ([P 2N2]Ta )2(µ-H)4 a n d th e
P a r a m a gn etic Ca tion ic Din u clea r Hyd r id e Sp ecies
{([P 2N2]Ta )2(µ-H)4}+I- ([P 2N2] )
P h P (CH2SiMe2NSiMe2CH2)2P P h ): Th e Red u cin g Ability
of a Meta l-Meta l Bon d
Michael D. Fryzuk,* Samuel A. J ohnson, and Steven J . Rettig†
Department of Chemistry, University of British Columbia, 2036 Main Mall,
Vancouver, British Columbia, Canada V6T 1Z1
Received April 26, 2000
The controlled reaction of the Ta(V) trimethyl species [P2N2]TaMe3, where [P2N2] )
PhP(CH2SiMe2NSiMe2CH2)2PPh, under 0.5 atm of hydrogen gas produces a partially
hydrogenated Ta(V) species, ([P2N2]TaMe2(H), of unknown structure. Under 4 atm of
hydrogen gas, further hydrogenation does not produce the complex [P2N2]TaH3; instead,
reduction of the tantalum center occurs to yield the dinuclear Ta(IV) hydride ([P2N2]Ta)2-
(µ-H)4. This diamagnetic tetrahydride fails to react with many reagents, including ethylene
and carbon monoxide; however, upon addition of iodomethane, {([P2N2]Ta)2(µ-H)4}+I- is
produced as a paramagnetic green crystalline solid. The number of hydrides in this reaction
product was confirmed by a deuterium labeling study. The results of a variable-temperature
magnetic susceptibility study of this tetrahydride cation can be partially modeled with the
Curie-Weiss law and a large correction for temperature-independent magnetism. Ab initio
calculations using density functional theory were performed in an attempt to further
understand the influence of the macrocyclic ligand in the bonding in these complexes.
In tr od u ction
Herein, we describe the use of the [P2N2] ligand
(where [P2N2] ) PhP(CH2SiMe2NSiMe2CH2)2PPh) to
generate the complex ([P2N2]Ta)2(µ-H)4, which is dia-
magnetic and contains a tantalum-tantalum bond.
Unlike the complex ([NPN]Ta)2(µ-H)4, the more coordi-
natively saturated complex ([P2N2]Ta)2(µ-H)4 does not
react with dinitrogen; however, attempts to remove the
hydride ligands utilizing electrophiles demonstrates
that the primary reactivity of this complex is as a
reducing agent.
There have been several reports of metal-metal
bonds storing electrons that were later utilized in the
reduction of dinitrogen. For example, we have reported
that the hydrogenation of the Ta(V) complex [NPN]-
TaMe3, where [NPN] ) PhP(CH2SiMe2NPh)2, generates
the dinuclear Ta(IV) tetrahydride ([NPN]Ta)2(µ-H)4;
because this complex is diamagnetic, it can be described
as containing a tantalum-tantalum single bond. What
is particularly significant about this dinuclear complex
is that it reacts with dinitrogen with the elimination of
1 equiv of hydrogen gas to generate ([NPN]Ta)2(µ-H)2-
(µ-η1:η2-N2). This formally corresponds to an overall
four-electron reduction of dinitrogen, where two elec-
trons were initially stored in a metal-metal bond.1 The
reduction of dinitrogen by electrons stored in a metal-
metal single bond has also recently been described for
a niobium dinitrogen complex,2,3 and metal-metal
bonding interactions have also been implicated in stor-
ing electrons for the reduction of dinitrogen in nitroge-
nase.4
Resu lts a n d Discu ssion
We have previously reported the preparation of [P2N2]-
TaMe3 and its photochemical conversion to the Ta(V)
methylidene complex [P2N2]TadCH2(Me).5 The macro-
cyclic ancillary ligand [P2N2]6 is a flexible framework
of two amido and two phosphine donors that have been
shown to stabilize complexes of the early transition
metals,7-9 the main-group elements,10 and the lan-
(5) Fryzuk, M. D.; J ohnson, S. A.; Rettig, S. J . Organometallics 1999,
18, 4059-4067.
(6) Fryzuk, M. D.; Love, J . B.; Rettig, S. J . J . Chem. Soc., Chem.
Commun. 1996, 2783.
(7) Basch, H.; Musaev, D. G.; Morokuma, K.; Fryzuk, M. D.; Love,
J . B.; Seidel, W. W.; Albinati, A.; Koetzle, T. F.; Klooster, W. T.; Mason,
S. A.; Eckert, J . J . Am. Chem. Soc. 1999, 121, 523-528.
(8) Fryzuk, M. D.; Love, J . B.; Rettig, S. J .; Young, V. G. Science
1997, 275, 1445.
(9) Fryzuk, M. D.; Love, J . B.; Rettig, S. J . Organometallics 1998,
17, 846-853.
(10) Fryzuk, M. D.; Giesbrecht, G. R.; Rettig, S. J . Inorg. Chem.
1998, 37, 6928-6934.
† Professional Officer: UBC X-ray Structural Laboratory (deceased
October 27, 1998).
(1) Fryzuk, M. D.; J ohnson, S. A.; Rettig, S. J . J . Am. Chem. Soc.
1998, 120, 11024-11025.
(2) Caselli, A.; Solari, E.; Scopelliti, R.; Floriani, C.; Re, N.; Rizzoli,
C.; Chiesi-Villa, A. J . Am. Chem. Soc. 2000, 122, 3652-3670.
(3) Zanoti-Gerosa, A.; Solari, E.; Giannini, L.; Floriani, C.; Chiesi-
Villa, A.; Rizzoli, C. J . Am. Chem. Soc. 1998, 120, 437.
(4) Han, J .; Beck, K.; Ockwig, N.; Coucouvanis, D. J . Am. Chem.
Soc. 1999, 121, 10448.
10.1021/om000359w CCC: $19.00 © 2000 American Chemical Society
Publication on Web 08/24/2000

3932 Organometallics, Vol. 19, No. 19, 2000
Fryzuk et al.
thanides11 in a variety of geometries. The ready avail-
ability of the trimethyl complex [P2N2]TaMe3 prompted
us to further examine its reactivity patterns.
Hyd r ogen a tion of [P 2N2]Ta Me3 a n d Str u ctu r e of
([P 2N2]Ta )2(µ-H)4 (1). A solution of [P2N2]TaMe3 under
4 atm of dihydrogen gradually darkens from light yellow
to a dark brick red. As monitored by 31P{1H} NMR
spectroscopy, this reaction takes 3 days to go to comple-
tion. Examining the gases produced by GC-MS demon-
strated that methane was the only product of this
reaction; no ethane was observed. When this reaction
is performed in hexanes, the product precipitates as a
dark red-brown crystalline solid. The dark color of the
product (λmax 455 nm, ꢀ ) 6200) is not consistent with
1
a Ta(V) species; likewise, the H NMR spectrum is not
consistent with the anticipated Ta(V) hydrogenation
product [P2N2]TaH3. The only non-[P2N2] ligand reso-
1
nance in the H NMR spectrum occurs at δ 5.86. This
signal accounts for two hydride ligands per metal center
and is a quintet due to coupling to four equivalent
phosphorus-31 nuclei. From these data, the hydrogena-
tion product can be assigned as the dimeric Ta(IV)
hydride ([P2N2]Ta)2(µ-H)4 (1) as illustrated in eq 1. The
diamagnetism of hydride 1 can be explained by the
formation of a tantalum-tantalum σ-bond.
F igu r e 1. ORTEP diagram of the solid-state molecular
structure of ([P2N2]Ta)2(µ-H)4 (1), as determined by X-ray
crystallography. Silyl methyls have been omitted for clarity,
and only the ipso carbons of the phenyl rings attached to
phosphorus are shown. Selected bond distances (Å), bond
angles (deg), and torsion angles (deg): Ta(1)-Ta(2), 2.6165-
(5); Ta-P(average) 2.615(3); Ta-N(average), 2.217(6); Ta-
(1)-H(1), 1.80(6); Ta(1)-H(2), 1.87(6); Ta(1)-H(3), 1.853-
(11); Ta(1)-H(4), 1.856(11); Ta(2)-H(1), 1.91(6); Ta(2)-
H(2), 1.93(6); Ta(2)-H(3), 1.847(11); Ta(2)-H(4), 1.851(11);
P(1)-Ta(1)-P(2), 139.87(9); Ta(2)-Ta(1)-P(1), 110.02(7);
N(1)-Ta(1)-N(2), 101.7(2); Ta(2)-Ta(1)-N(1), 129.3(2);
P(1)-Ta(1)-Ta(2)-P(4), -3.00(9); N(1)-Ta(1)-Ta(2)-
N(4), 10.0(3); Ta(2)-Ta(1)-N(1)-Si(1), 129.6(4).
structurally characterized [P2N2] complexes;5 however,
unlike in other cases, the twist in this case does not
appear to be fluxional in solution. The room-tempera-
ture 1H NMR spectrum contains four silyl methyl
environments. If the twist in the ligand backbone were
not rigid, only two silyl methyl signals would be
anticipated, corresponding to silyl methyl groups di-
rected toward the front and the back of the [P2N2] ligand
on each [P2N2] unit.
While this locked conformation of the [P2N2] ligand
may arise from electronic effects due to the interaction
of the amide lone pairs with tantalum metal orbitals, it
may also be explained by steric interactions between
the [P2N2] ligands imposed by other electronic factors.
The opposing [P2N2] ligands on Ta(1) and Ta(2) are
eclipsed, with a negligible P(1)-Ta(1)-Ta(2)-P(4) tor-
sion angle of -3.00(9)°. This eclipsed arrangement of
the [P2N2]Ta units is likely of electronic origin, as
sterically it appears less favorable; in this orientation
the phenyl rings attached to phosphorus point away
from each other only because of the twist in the [P2N2]
framework. For this twist in the conformation of the
[P2N2] framework to be fluxional, the [P2N2]Ta units
would have to be able to rotate with respect to each
other, as steric interactions between phenyl groups on
opposing [P2N2] ligands prevent this twisting motion
The solid-state molecular structure as determined by
X-ray crystallography is shown in Figure 1. Crystal-
lographic data are given in Table 1. Although there is
no crystallographically imposed symmetry to the mol-
ecule in the solid-state structure, the molecule has
approximate D2 symmetry, with a C2 axis running
through the Ta-Ta bond and two perpendicular C2 axes
that are oriented through the center of the Ta(1)-Ta-
(2) bond that pass between, rather than through, the
tantalum hydrides.
Additional symmetry elements are lacking, due to a
twist in the [P2N2] ligand backbone such that the Ta-
(2)-Ta(1)-N(1)-Si(1) torsion angle is 129.6(4)°, rather
than near 90°. This twist has been noted before in other
1
with the [P2N2] ligands eclipsed. The solution H NMR
spectrum therefore confirms that the solution structure
is identical with the solid-state structure and that the
[P2N2] conformation and orientation with respect to the
opposing [P2N2] ligand is rigid in solution.
(11) Fryzuk, M. D.; Love, J . B.; Rettig, S. J . J . Am. Chem. Soc. 1997,
119, 9071-9072.

Diamagnetic and Paramagnetic Dinuclear Ta Hydrides
Organometallics, Vol. 19, No. 19, 2000 3933
Ta ble 1. Cr ysta l Da ta a n d Str u ctu r e Refin em en t
for 1 a n d 3
1a
3b
formula
fw
C48H88N4P4Si8Ta2 C48H88IN4P4Si8Ta2
1431.72
1558.63
color, habit
cryst size, mm
cryst syst
space group
a, Å
red, prism
green needle
0.15 × 0.20 ×0.25 0.50 × 0.35 × 0.15
monoclinic
P21/n (No. 14)
12.748(3)
21.186(2)
23.919(2)
90
trigonal
P3121 (No. 152)
19.2489(7)
b, Å
c, Å
14.9628(2)
90
90
120
4801.3(2)
3
F igu r e 2. Depiction of “staggered by 90°” and “eclipsed”
ligand arrangements.
R, deg
â, deg
101.857(11)
90
6321.8(17)
4
γ, deg
compared to those in the closely related [Cl2(PMe3)2Ta]2-
V, Å3
12
13
(µ-H)4 and [(cb)(PMe2Ph)2(H)Ta]2(µ-H)4 may be due
to the steric repulsion between the ligands on the metal
centers, or perhaps due to the competition of the
π-electrons of the amido donors for metal orbitals
utilized in bridging the metal-metal bond, as will be
discussed later in the section on bonding considerations.
Z
T, °C
21.0
1.504
2888.00
Mo
31.41
0.498-1.000
ω-2θ
-93
F
calcd, g/cm3
1.616
F(000)
2322.00
Mo
41.77
0.7594-1.0000
ω
radiation
µ, cm-1
transmissn factors
scan type
The metal hydride atoms in hydride 1 were identified
in electron density difference maps, and their locations
were refined, though with distance constraints for H(3)
and H(4). The bridging hydrides are staggered 45° with
respect to the ligand donors, so that the hydride atoms
lie between the planes described by P(1)-Ta(1)-P(2)
and N(1)-Ta(1)-N(2). Not considering the Ta-Ta bond,
the geometry at each tantalum center is approximately
square-antiprismatic. The tantalum hydride distances
determined from the X-ray data range from 1.80(6) Å
to 1.93(6) Å, similar to those in [Cl2(PMe3)2Ta]2(µ-H)4,
where the Ta-H bond length is 1.81(21) Å as deter-
mined from X-ray crystallography,12 and the related
rhenium complex Re2H8(PEtPh)4, where the Re-H
distance is 1.878(7) Å as determined from neutron
diffraction data.18
scan range in ω, deg
scan speed, deg/min
2θmax, deg
cryst decay, %
total no. of rflns
no. of unique rflns
Rmerge
no. of rflns with
I g nσ(I)
no. of variables
R
1.10 + 0.35 tan θ
16 (up to 9 rescans)
55
1.19
15 507
14 853
0.061
5751
60.1
negligible
42 796
4600
0.072
4481
610
303
0.037a
0.029a
1.65
0.079b
0.058b
1.19
Rw
GOF
max ∆/σ
0.003
0.03
4.41 (near Ta),
-3.25
residual density e/Å3 0.77, -0.85
(both near Ta)
a
Rigaku AFC6S diffractometer: R ) ∑||Fo| - |Fc||/∑|Fo|; Rw
)
(∑w(|Fo| - |Fc|)2/∑w|Fo| )1/2
.
Rigaku/ADSC CCD diffractometer:
2
b
2
2
2
2
2
4
R ) ∑||Fo | - |Fc ||/∑|Fo |; Rw ) (∑w(|Fo | - |Fc |)2/∑w|Fo| )1/2
.
Not surprisingly, the structure of species 1 is similar
to that of the aforementioned compound [Cl2(PMe3)2Ta]2-
(µ-H)4;12 however, the structural differences are intrigu-
ing. While the end groups in [Cl2(PMe3)2Ta]2(µ-H)4 are
eclipsed, as in 1, the phosphine ligands on opposing
tantalum centers are staggered by 90°. Eclipsed and
staggered by 90° could have various meanings in the
context of these complexes; it is intended here to
describe the eclipsed and staggered by 90° geometries
that are depicted in Figure 2. Similar staggered by 90°
ligand arrangements have been observed in the group
6 quadruple metal-metal bonded dimers Mo2Cl4(PMe3)4
and W2Cl4(PMe3)4,19 although examples of the eclipsed
ligand conformation are known both in metal-metal
bonded complexes as well as in a dinuclear tantalum
hydride featuring the 9H-carbazole amido ligand, [(cb)-
(PMe2Ph)2(H)Ta]2(µ-H)4.13 The [P2N2] ligand clearly has
an influence on the bonding in 1, either via electronic
effects caused by the difference in the bonding of the
amides vs chlorides, slightly modified phosphines, or the
geometrical restrictions of the macrocyclic ligand or via
the steric interactions of the [P2N2] ligand. The theoreti-
cal aspects of the bonding in 1 will be addressed in
As expected, the [P2N2] ligand has an influence on
the geometry at each tantalum center. The amide
nitrogens are approximately cis-disposed, with a N(1)-
Ta(1)-N(2) angle of 101.7(2)°. The phosphine donors are
closer to a trans disposition, with a P(1)-Ta(1)-P(2)
angle of 139.87°.
The Ta-Ta distance of 2.6165(5) Å is within that
expected for a metal-metal bonding interaction but
longer than in structurally related tantalum complexes;
shorter Ta-Ta distances of 2.511(2) and 2.5359(4) Å
12
were reported for the complexes [Cl2(PMe3)2Ta]2(µ-H)4
and [(cb)(PMe2Ph)2(H)Ta]2(µ-H)4,13 where cbH is 9H-
carbazole. Longer Ta-Ta interactions of 2.854(1) and
2.8280(4) Å have also been observed in the compounds
(C5Me4Et)2ClTa(µ-H)2TaCl2(C5Me4Et)14 and [(CyN)2-
ClTa]2(µ-H)2,15,16 where the tantalum centers are bridged
by only two hydrides, as well as in the compound [(But -
3
SiO)2(H)2Ta]2,17 where an unbridged Ta-Ta distance of
2.720(4) Å was reported. The long Ta-Ta distance in 1
(12) Scioly, A. J .; Luetkens, M. L. J .; Wilson, R. B.; Huffman, J . C.;
Sattelberger, A. P. Polyhedron 1987, 6, 741-757.
(13) Profilet, R. D.; Fanwick, P. E.; Rothwell, I. P. Polyhedron 1992,
11, 1559-1561.
(14) Belmonte, P. A.; Schrock, R. R.; Day, C. S. J . Am. Chem. Soc.
1982, 104, 3082-3089.
(17) Miller, R. L.; Toreki, R.; LaPointe, R. E.; Wolczanski, P. T.; Van
Duyne, G. D.; Roe, D. C. J . Am. Chem. Soc. 1993, 115, 5570-5588.
(18) Bau, R.; Carroll, W. E.; Teller, R. G.; Koetzle, T. F. J . Am. Chem.
Soc. 1977, 99, 3872-3874.
(19) Cotton, F. A.; Extine, M. W.; Felthouse, T. R.; Kolthammer, B.
W. S.; Lay, D. G. J . Am. Chem. Soc. 1981, 103, 3.
(15) Cotton, F. A.; Daniels, L. M.; Murillo, C. A.; Wang, X. J . Am.
Chem. Soc. 1996, 118, 12449-12450.
(16) Scoles, L.; Ruppa, K. B. P.; Gambarotta, S. J . Am. Chem. Soc.
1996, 118, 2529-2530.

3934 Organometallics, Vol. 19, No. 19, 2000
Fryzuk et al.
greater detail in the section on bonding considerations
and density functional theory calculations.
Various aspects of the data, such as the small coupling
constant of the hydride resonance to phosphorus21,22 and
the long T1 value for the hydride resonance, could be
construed to imply a dimeric structure,23 although the
correlation between structure and T1 relaxation times
of metal hydrides is fraught with uncertainty.24,25
Likewise, the two phosphorus environments observed
in the 31P{1H} NMR spectrum is consistent with an
eight-coordinate dimeric structure rather than the
seven-coordinate monomer, which would likely be flux-
ional and have only one phosphorus environment at
room temperature.5 Unfortunately, no definitive data
exist to confirm either a monomeric or a dimeric
structure for complex 2.
Species 1 is not particularly reactive, as evidenced by
the fact that exposure of solutions of 1 to ethylene or
carbon monoxide resulted in no reaction. This lack of
reactivity is quite different from that reported for the
less coordinatively saturated hydride complex ([NPN]-
Ta)2(µ-H)4, which reacts directly with dinitrogen and
eliminates 1 equiv of dihydrogen to form a dinitrogen
complex.1 Likewise, complex 1 did not react directly with
D2 gas to form the deuterated analogue ([P2N2]Ta)2(µ-
D)4; this latter complex was synthesized from the
reaction of [P2N2]TaMe3 with D2 gas. A comparison of
the IR spectra of the hydride complex 1 and the
deuterated species 1-d4 did not definitively identify the
bridging hydride stretches, as the region in which they
were expected was obscured by other peaks associated
with the [P2N2] ligand. There were no stretches in the
IR spectra of 1 or 1-d4 in the region normally associated
with terminal Ta-H bonds.
This σ-bond metathesis reaction that forms 2 could
also be performed using silanes. The reaction of [P2N2]-
TaMe3 with BunSiH3 also produced 1 and intermediate
2, but not as cleanly as the reactions with dihydrogen.
During the reaction of [P2N2]TaMe3 with BunSiH3,
intermediate 2 accumulates in a larger amount relative
to the concentration of 1 than in the analogous hydro-
genation reaction, consistent with the much slower
reaction of the bulkier BunSiH3 with intermediate 2
than with dihydrogen.
P a r tia l Hyd r ogen a tion of [P 2N2]Ta Me3. The fail-
ure of [P2N2]TaMe3 to hydrogenate to form the mono-
nuclear Ta(V) hydride [P2N2]TaH3 raises the question
as to at what point in the hydrogenation the reduction
to Ta(IV) took place. Monitoring the hydrogenation of
+
-
2
Syn th esis an d Str u ctu r e of {([P N2]Ta)2(µ-H)4} I
(3). The reaction of alkyl halides with transition-metal
hydrides is a well-known reaction; the formation of a
new C-H bond and a metal halide complex has been
used as a diagnostic and quantitative test for metal
hydride complexes.26-28 It was anticipated that this
reaction could also be used in a synthetic manner to
produce a [P2N2]TaIV halide starting material, such as
([P2N2]TaI2)x. The reaction of a toluene solution of 1 with
excess MeI resulted in the precipitation of an insoluble
crystalline paramagnetic green solid, 3. This solid
dissolved in CH2Cl2 to give a dichroic solution from
which transmitted light appears blue-green and re-
flected light appears red. In contrast to our expectations,
X-ray crystallography identified the compound as
{([P2N2]Ta)2(µ-H)3}+I-, a cationic hydride, with three
hydride ligands bridging the tantalum metal centers.
The paramagnetism would be explained if there were
no metal-metal bond between the two Ta(IV) centers.
The overall reaction would then be the result of the
electrophilic abstraction of a hydride ligand by CH3I,
with 1 equiv of CH4 being produced in the reaction.
However, X-ray crystallography is not always a reliable
method for determining hydride ligand locations, and
as one of the hydrides in this solution to the X-ray data
lies on a symmetry element, where errors in electron
density often accumulate, the structure could not be
accepted without further study. A second possible
1
[P2N2]TaMe3 by H NMR spectroscopy in a sealed tube
under 1 atm of dihydrogen revealed that an intermedi-
ate hydrogenation product, 2, is formed, which is
converted entirely to 1 over 3 days. Performing the
hydrogenation of [P2N2]TaMe3 in a sealed NMR tube
under 0.5 atm of dihydrogen resulted in a pale orange
solution after 1 week. Unlike the hydrogenation under
1 atm of H2, the 31P{1H} NMR spectrum indicates that
under these conditions very little of 1 is formed, and
the solution is mostly a mixture of the starting material
and the hydrogenation intermediate 2.
Unfortunately, attempts to isolate and purify this
intermediate have failed. Logical candidates for this
presumed Ta(V) species are the partially hydrogenated
species [P2N2]TaMe2(H) and [P2N2]TaMe(H)2, although
hydride-bridged dimers of these compounds are also a
possibility.20 The 31P{1H} NMR spectrum displays two
phosphorus-31 environments for 2, coupled to each other
1
with a 125 Hz coupling constant. Prominent in the H
NMR spectrum of the hydrogenation intermediate is a
peak at 5.68 ppm, the same region in which the bridging
hydrides in 1 were observed; no coupling to phosphorus-
31 could be resolved. Attempts to obtain an accurate
integration of this peak proved to be difficult. An
inversion-recovery measurement of the T1 relaxation
1
times for the H spectrum indicated that the hydridic
proton relaxed much more slowly than the ligand
protons. From inversion-recovery experiments, the T1
value of this hydride signal is estimated to be ∼20 s at
room temperature in a 500 MHz spectrometer. The
integration of this peak from a 1H NMR spectrum
obtained with an increased delay time, compared with
the integration of the peaks associated with the ortho
protons of the PPh groups and the new tantalum methyl
group at δ 0.88, supports an empirical formula of [P2N2]-
TaMe2(H).
(21) Crabtree, R. H. Acc. Chem. Res. 1979, 12, 331.
(22) Bau, R.; Carroll, W. E.; Hart, D. W.; Teller, R. G.; Koetzle, T.
F. Adv. Chem. Ser. 1978, No. 167, 73-92.
(23) Bakhmutov, V. I.; Vorontsov, E. V.; Nikonov, G. I.; Lemenovskii,
D. A. Inorg. Chem. 1998, 37, 279-282.
(24) Crabtree, R. H.; Segmuller, B. E.; Uriarte, R. J . Inorg. Chem.
1985, 24, 1949-1950.
(25) Hlatky, G. G.; Crabtree, R. H. Coord. Chem. Rev. 1985, 65,
1-48.
(26) Collman, J . P.; Hegedus, L. S.; Norton, J . R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry;
University Science Books: Mill Valley, CA, 1987.
(27) Cotton, F. A.; Wilkinson, G. Advanced Inorganic Chemistry, 5th
ed.; Wiley-Interscience: New York, 1988; p 1105.
(28) Crabtree, R. H. The Organometallic Chemistry of the Transition
Metals, 2nd ed.; Wiley-Interscience: New York, 1994; p 62.
(20) Mayer, J . M.; Wolczanski, P. T.; Santarsiero, B. D.; Olson, W.
A.; Bercaw, J . E. Inorg. Chem. 1983, 22, 1149-1155.

Diamagnetic and Paramagnetic Dinuclear Ta Hydrides
Organometallics, Vol. 19, No. 19, 2000 3935
product is the result of the one-electron oxidation of 1
to produce {([P2N2]Ta)2(µ-H)4}+I-, a cationic tetrahy-
dride in which one of the electrons in the tantalum-
tantalum bond has reduced MeI by one electron, to
produce I- and a methyl radical, which is likely rapidly
converted to methane by reaction with the solvent. This
compound should be paramagnetic regardless of whether
the electron resides in a tantalum-tantalum bonding
orbital, because the formal oxidation state of the two
metal centers is now intermediate between Ta(IV) and
Ta(V).
A variety of approaches were taken to ascertain the
nature of 3. Protonation of a hydride is a common route
to cationic hydrides and proceeds via hydride abstrac-
tion. Similar to the reaction of CH3I with 1, the reaction
of 1 with H+{B[3,5-(CF3)2C6H3]3} also produced a para-
magnetic green solid; this solid dissolves in THF to give
a solution that exhibits the same dichroic behavior
observed for 3. Conversely, the reaction of 3 with the
hydride source K+{BEt3H}- regenerates the neutral
tetrahydride 1. Although both of these observed reac-
tivities would be expected if 3 is the cationic trihydride
{([P2N2]Ta)2(µ-H)3}+I-, they do not exclude the possibil-
ity that 3 is the cationic tetrahydride {([P2N2]Ta)2(µ-
H)4}+I-. A more definitive study was performed using
deuterium-labeled 1-d4, ([P2N2]Ta)2(µ-D)4, prepared from
the reaction of [P2N2]TaMe3 with D2. Reaction of 1-d4
with CH3I did not produce CH3D, as determined by GS-
MS; rather, CH4 was produced. As additional evidence,
the reaction of deuterated 3 with K+{BEt3H}- did not
produce the partially deuterated 1-d3, ([P2N2]Ta)2(µ-H)-
(µ-D)3, but rather the fully deuterated 1-d4, ([P2N2]Ta)2-
(µ-D)4. The reaction of 1 with CH3I or H+{B[3,5-
(CF3)2C6H3]3} is therefore not a hydride abstraction
reaction, and instead, we can assign compound 3 as the
cationic tetrahydride {([P2N2]Ta)2(µ-H)4}+I-. The role of
K+{BEt3H}- in the reaction with 3 is therefore not as
a hydride source but simply as a reducing agent. These
two reactions are depicted in Scheme 1. Dibenzyl
(PhCH2CH2Ph) was observed as a product in both
F igu r e 3. Cyclic voltammogram for ([P2N2]Ta)2(µ-H)4 (1).
The cyclic voltammogram of 3 shows identical one-electron
oxidation waves. Note that where the scan begins, at -0.6
V, species 1 has already been oxidized to 3 at the electrode.
Sch em e 1
1
reactions by H and 13C{1H} NMR spectroscopy.
To add further evidence to the nature of compound 3
and to assess the reducing ability of the tantalum-
tantalum bond in 1, cyclic voltammetry was performed
on both the neutral and the cationic species. The cyclic
voltammograms were performed in CH2Cl2 containing
{NBun4}+{PF6}- as the electrolyte and were referenced
with respect to (C5Me5)2Fe as an internal standard. Both
compounds 1 and 3 exhibit two identical one-electron
oxidations; one occurs at -1.18 V and the second at
-0.33 V relative to the SCE, providing further proof as
to the identity of 3. The cyclic voltammogram for 1 is
shown in Figure 3. Species 3 exhibits additional oxida-
tions at 0.23 and 0.61 V, due to oxidation of the
counteranion. All the oxidations observed were revers-
ible. The first oxidation of 1 is expected to produce the
cation observed in 3; the second oxidation indicates that
the species {([P2N2]Ta)2(µ-H)4}2+, which contains two
Ta(V) metal centers and no electrons in a Ta-Ta
bonding orbital, is also stable. The large separation
between the two oxidation potentials is as expected for
a case where the single electron is delocalized in a
metal-metal bond between the two tantalum atoms,
rather than a structure that contains localized Ta(IV)
and Ta(V) metal centers.
Compound 3 is moderately air stable; dilute solutions
in CH2Cl2 only lose color after several hours when
exposed to atmospheric oxygen. In fact, upon exposure
of CH2Cl2 solutions of 1 to air, a distinctive dichroic
solution results that is indicative of the cation in
complex 3. Solutions of 3 do not react with simple small
molecules such as ethylene or carbon monoxide; this
lack of reactivity provides indirect evidence that the
dinuclear nature of 3 is maintained in solution.
The solid-state molecular structure of 3 as determined
by X-ray crystallography is shown in Figure 4. Crystal-
lographic data are given in Table 1. The molecule has
crystallographic C2 symmetry, with the C2 axis passing

3936 Organometallics, Vol. 19, No. 19, 2000
Fryzuk et al.
F igu r e 5. Measured molar magnetic susceptibilities, ø (∆),
and effective magnetic moments, µeff (O), versus tempera-
ture for {([P2N2]Ta)2(µ-H)4}+I- (3).
F igu r e 4. ORTEP diagram of the solid-state molecular
structure of the cationic fragment of {([P2N2]Ta)2(µ-H)4}+I-
(3), as determined by X-ray crystallography. Silyl methyls
have been omitted for clarity, and only the ipso carbons of
the phenyl rings attached to phosphorus are shown. The
bridging hydrides were not located. Selected bond dis-
tances (Å), bond angles (deg), and torsion angles (deg): Ta-
(1)-Ta(1)*, 2.7721(5); Ta(1)-P(1), 2.602(2); Ta(1)-N(1),
2.145(5); P(1)-Ta(1)-P(2), 143.58(5); Ta(1)*-Ta(1)-P(1),
108.91(4); N(1)-Ta(1)-N(2), 102.4(2); Ta(1)*-Ta(1)-N(1),
127.25(12); P(1)-Ta(1)-Ta(1)*-P(1)*, -2.81(10); N(1)-Ta-
(1)-Ta(1)*-N(1)*, -8.9(5); Ta(1)-Ta(1)*-N(1)*-Si(1)*,
129.6(4).
observed before for other paramagnetic species.29 The
solution and solid-state EPR spectra provide little
structural data, as both contain a very broad signal,
with no hyperfine or superhyperfine coupling to tanta-
lum, phosphorus, or hydrogen resolved. The solid-state
EPR contains a signal at a g value of 1.89 with a very
large half-height width of 800 G. In contrast, the
monomeric Ta(IV) complex TaH4(dmpe)2 reportedly
exhibits hyperfine coupling to 181Ta (99.99%, I ) 7/2, aTa
) 106.3 G) and 31P (100.0%, I ) 1/2, aP ) 32.9 G),30 and
the related dinuclear complex {[(PEt2Ph)2H2Re]2(µ-
H)4}+ contains a complex spectrum due to one unpaired
electron that resides in a Re-Re bonding orbital and
therefore couples to two rhenium nuclei.31
through the center of the molecule and parallel to the
plane containing the four phosphine ligands. The ar-
rangement of the ancillary [P2N2] ligands in the cationic
portion of species 3 is remarkably similar to that in 1.
The phosphine ligands on opposing metal centers are
eclipsed with a P(1)-Ta(1)-Ta(1)*-P(1)* torsion angle
of -2.81(10)°. The P(1)-Ta(1)-P(2) angle of 143.58(5)°
and the N(1)-Ta(1)-N(2) angle of 102.4(2)° are also
nearly identical with those observed in 1. The Ta(1)-
P(1) distance of 2.602(2) Å is only about 0.01 Å shorter
than the Ta-P distances in 1, whereas the Ta(1)-N(1)
distance of 2.145(5) Å is shorter by 0.054 Å than the
Ta(1)-N(1) distance in 1, perhaps due to the increased
positive charge on the tantalum centers strengthening
the bonds to these anionic ligands, or perhaps from the
availability of an extra orbital of correct symmetry to
overlap with occupied amide lone-pair orbitals. The
twist in the ligand conformation present in 1 is also
observed in the solid-state structure of 2, with an
identical Ta(1)-Ta(1)*-N(1)*-Si(1)* torsion angle of
129.6(4)°. The iodide anion has no close contacts with
the metal centers. Despite the remarkable similarity
between the arrangement, bond angles, and bond lengths
in the [P2N2]Ta fragments of the solid-state structures
of 1 and 3, a significantly longer Ta-Ta distance is
observed. The Ta(1)-Ta(1)* distance of 2.7721(5) Å in
the tetrahydride cation 3 is 0.156 Å longer than the
Ta(1)-Ta(2) distance in the neutral tetrahydride 1,
despite the absence of only a single bonding electron.
Ma gn etism of {([P 2N2]Ta )2(µ-H)4}+I- (3). Figure 5
shows a plot of ø, the molar magnetic susceptibility
corrected for diamagnetic contributions, and µeff, the
effective magnetic moment, for the temperature range
of 2-300 K.
Attempts to model the entire data set for 3 with the
Curie-Weiss law, using g, Θ, and a temperature-
independent paramagnetism term (TIP) as parameters,
did not provide a good fit. This is possibly due to spin-
orbit coupling or perhaps due to the error in the
approximation of the diamagnetic correction for species
3. For the third-row transition elements strong spin-
orbit coupling is common, unlike for the first-row
transition elements, and is not trivial to model. While
there are no strictly degenerate orbitals on the indi-
vidual tantalum metal centers that would allow for a
large contribution from first-order spin-orbit coupling,
low-lying orbitals that may be thermally populated
could affect the magnetism, leading to a second-order
spin-orbit coupling effect. Modeling only the low-
temperature data from 2 to 50 K, which should be less
affected by thermally dependent second-order spin-
orbit coupling, as well as by errors in the diamagnetic
correction, the parameters g ) 1.86, TIP ) 1670 × 10-6
emu mol-1, and Θ ) -0.11 K were determined. This
model is shown in Figure 6. The g value is as anticipated
from the EPR data, and is consistent with a d1 metal
center where there are no degenerate orbitals present
The 1H NMR spectrum of species 3 in CD2Cl2 contains
only broad paramagnetically shifted peaks; though over
time CH2Cl2 solutions of 3 decompose to yield colorless
diamagnetic products. As has been noted, solutions of
3 in CH2Cl2 are dichroic; this behavior has been
(29) For example, the ubiquitous ferrocenium ion, [(C5H5)2Fe]+, also
exhibits a blue-green/red dichroism.
(30) Allison, J . D.; Walton, R. A. J . Am. Chem. Soc. 1984, 106, 163.
(31) Allison, J . D.; Walton, R. A. J . Am. Chem. Soc. 1984, 106, 163-
168.

Diamagnetic and Paramagnetic Dinuclear Ta Hydrides
Organometallics, Vol. 19, No. 19, 2000 3937
F igu r e 6. Plot of the effective magnetic moment (µeff)
versus temperature for compound 3 from 2 to 50 K. The
circles represent the data points and the bold line a model
using the Curie-Weiss law with the parameters g ) 1.86,
TIP ) 1670 × 10-6 emu mol-1, and Θ ) -0.11 K.
for a large first-order spin-orbit coupling effect. The TIP
is responsible for the magnetic moment being much
larger than expected for a d1 complex at room temper-
ature. The negative Θ value indicates that small
intermolecular interactions between unpaired electrons
on adjacent molecules do in fact occur, although the
exact nature of these interactions is unclear.32,33
Bon d in g Con sid er a t ion s a n d Den sit y F u n c-
tion a l Th eor y Ca lcu la tion s. The bonding of hydride-
bridged dinuclear metal complexes has been treated at
various levels of theory in the past. A general approach
using fragment molecular orbital analysis and sup-
ported by extended Hu¨ckel theory has been applied to
a variety of metals and metal-ligand fragment geom-
etries.34 A similar analysis supported by a multiple-
scattering XR calculation was performed on [Cl2(PH3)2-
Ta]2(µ-H)4, a model complex for the species [Cl2(PMe3)2-
Ta]2(µ-H)4, which is pertinent to the complexes described
here.12 Both methods relied on arbitrary geometries,
chosen from relevant crystal structure data, rather than
optimized geometries. Both analyses also present a
similar approach to describing the bonding in these
structures, where the available metal-ligand fragment
orbitals are overlapped with the four possible linear
combinations of the four bridging hydride ligands. This
approach as applied to complex 1 is illustrated in Figure
7.
The orbital analysis for 1 in Figure 7 shows the four
linear combinations of the hydride 1s orbitals and the
metal-based orbitals that are of the appropriate sym-
metry for overlap to form three-center σ, π, or δ bonding
orbitals. It has been noted previously that the lowest
energy fully bonding linear combination of the hydride
orbitals is of much lower energy than the metal-based
orbitals of appropriate symmetry to form a σ-bonding
orbital. The combination of the two metal orbitals of
σ-symmetry forms a metal-metal bonding orbital that
includes only a small contribution by the bridging
hydride ligands. Complex 1 contains 10 electrons that
fill these 5 orbitals, and therefore, this complex is
diamagnetic and the metal-metal bonding orbital is
expected to be the highest in energy. It is worth noting
that there is one remaining metal-based d orbital on
each metal center that is not considered in either of the
F igu r e 7. Depiction of the overlap of metal-based orbitals
with the bridging hydride ligands to generate delocalized
molecular orbitals for complex 1 along with the appropriate
symmetry labels for its respective point group (D2).
two aforementioned bonding descriptions, as it has no
significant overlap with the hydride orbitals in either
complex.
The utility of density functional theory (DFT) in
modeling compounds of the second- and third-row
transition metals that contain metal-metal bonds has
recently been demonstrated, even utilizing relatively
small basis sets.35-37 While DFT does not strictly
provide molecular orbitals and their respective energies
in the same manner as Hartree-Fock calculations, most
investigations to date have demonstrated that the
resultant Kohn-Sham orbitals are very similar to those
predicted by Hartree-Fock theory.38-42 Therefore, it has
been suggested that the Kohn-Sham orbitals of DFT
are of equal utility to chemists in the qualitative
analysis of reactivity and structure. Even the energies
of the Kohn-Sham orbitals appear to be empirically
(35) Cotton, F. A.; Feng, X. J . Am. Chem. Soc. 1997, 119, 7514-
7520.
(36) Cotton, F. A.; Feng, X. J . Am. Chem. Soc. 1998, 120, 3387-
3397.
(37) Alonso, E.; Casas, J . M.; Cotton, F. A.; Feng, X.; Fornies, J .;
Fortuno, C.; Milagros, T. Inorg. Chem. 1999, 38, 5034-5040.
(38) Stowasser, R.; Hoffmann, R. J . Am. Chem. Soc. 1999, 121,
3414-3420.
(39) Baerends, E. J .; Gritsenko, O. V.; van Leeuwen, R. In Chemical
Application of Density Functional Theory; Laird, B. B., Ross, R., Ziegler,
T., Eds.; ACS Symposium Series 629; American Chemical Society:
Washington, DC, 1996; p 20.
(32) O’Connor, C. J . Prog. Inorg. Chem. 1982, 29, 203.
(33) Carlin, R. L. Magnetochemistry; Springer-Verlag: Heidelberg,
Germany, 1986.
(34) Dedieu, A.; Albright, T. A.; Hoffmann, R. J . Am. Chem. Soc.
1979, 101, 3141.
(40) Bickelhaupt, F. M.; Baerends, E. J .; Ravenek, W. Inorg. Chem.
1990, 29, 350.
(41) DeKoch, R. L.; Baerends, E. J .; Hengelmolen, R. Organometal-
lics 1984, 3, 289.
(42) Sargent, A. L.; Titus, E. P. Organometallics 1998, 17, 65.

3938 Organometallics, Vol. 19, No. 19, 2000
Fryzuk et al.
Ta ble 2. Selected Bon d Len gth s a n d En er gies for
th e Ab In itio DF T Geom etr y Op tim iza tion s of 1A
compd
tetrahydride model 1A
2.6444 Å
1.8969, 2.0862 Å
2.0766 Å
2.6680 Å
-375.1650764 hartree
Ta-Ta
Ta-H
Ta-N
Ta-P
total energy
related to the actual molecular orbital energies.38 En-
couraged by these results, we decided to attempt to
model complex 1 using ab initio DFT calculations in
order to gain a better understanding of the bonding in
this complex. It appeared unusual that the arrangement
of the ligands in both 1 and 3 has the phosphine ligands
on opposite metal centers eclipsed, rather than stag-
gered by 90° as in the structure of the related complex
[Cl2(PMe3)2Ta]2(µ-H)4, and additionally the Ta-Ta bond
length in these related species differ by 0.105 Å. Also
of interest was the interaction of the amide donors with
the metal-metal bonding orbitals. The interactions of
π-bonding ligands with multiple bonds between metals
are of current interest in generating one-dimensional
conducting polymers,43 and these interactions are of
potential importance in the reactivity of other dinuclear
[P2N2] complexes.7,8
F igu r e 8. Depiction of the metal-metal σ-bond interaction
in the HOMO of tetrahydride 1A. The PH3 and NH2 donors
are labeled by P and N, respectively. The frontmost NH2
ligand on each tantalum center obscures the view of the
rear NH2 ligand.
Rather than using the [P2N2] fragment, complex 1
was modeled using the species [(H2N)2(H3P)2Ta]2(µ-H)4
(1A) to simplify the calculations. Constraints were
added to ensure that this complex simulated the geo-
metrically restricted [P2N2] macrocycle. The Ta-Ta-P
angles were set at 110.0° and the Ta-Ta-N angle at
129.5°, as observed in the solid-state X-ray structure of
1. In an attempt to simulate the “twist” in the [P2N2]
ligand noted in 1, the H-N-Ta-Ta dihedral angle was
restrained to 120.0°. Failure to restrict this angle in
calculations on 1A resulted in a H-N-Ta-Ta dihedral
near 180°; however, the Si-N-Ta-Ta dihedral angle
in the [P2N2] macrocycle cannot be that large, due to
geometrical restraints in the macrocyclic ring.
The geometry of model complex 1A was optimized
(with the exception of the previously stated constraints)
with idealized D2 symmetry using the Gaussian 98
program44 and the hybrid functional B3LYP method.45
The basis functions and effective core potentials (ECP)
used were those in the LANL2DZ basis set46 but with
additional d polarization functions added to P atoms
with the exponent of the d functions set at 0.37. This
level of theory has recently been shown to be adequate
for modeling complexes containing metal-metal bonds
of various orders, although metal-metal bonds were
consistently slightly longer than experimental values.35
Selected optimized bond lengths, bond angles, and
dihedral angles for the model complex 1A are shown in
F igu r e 9. Illustration of the amide lone-pair orbital
overlaps for model 1A when the H-N-Ta-Ta dihedral
angle is 180° (a) and when the H-N-Ta-Ta dihedral angle
is 90° (b), depicted in the N-Ta-N plane. The PH3 donors
and bridging hydrides are omitted.
Table 2. The general features of the optimized geometry
of this model complex are in good agreement with the
structure of complex 1. The arrangement of the ligands
on the opposing metal centers as described by the
P-Ta-Ta-P dihedral angle 2.37° is similar to that in
1; the phosphine ligands on opposing metal centers are
essentially eclipsed. The calculated Ta-Ta distance of
2.644 Å is only slightly longer than the actual Ta-Ta
bond length in complex 1 of 2.6165(5) Å.
The Kohn-Sham orbitals obtained from the geometry
optimization for 1A appear to qualitatively fit the
bonding description for 1 shown in Figure 7; however,
several other interesting features appear, such as the
influence of the amido “lone pairs” (occupied p orbitals
on the amido donors) on the bonding in 1A. The HOMO
of 1A is the Ta-Ta bonding orbital, as anticipated. An
isosurface of this orbital is shown in Figure 8 and clearly
shows good orbital overlap between the two metal
orbitals. There is very little contribution from the
bridging hydride 1s orbitals in the HOMO. The en-
semble of hydride orbitals of appropriate symmetry lies
much lower in energy.
(43) Chisholm, M. H. Acc. Chem. Res. 2000, 33, 53.
(44) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J . R.; Zakrzewski, V. G.; Montgomery, J . A.;
Stratmann, R. E.; Burant, J . C.; Dapprich, S.; Millam, J . M.; Daniels,
A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J .; Petersson,
V. 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. G.; Chen, W.; Wong, M. W.; Andres, J . L.; Head-
Gordon, M.; Repogle, E. S.; Pople, J . A. Gaussian 98, Revision A.7;
Gaussian, Inc., Pittsburgh, PA, 1998.
The next lowest energy orbitals for 1A are different
symmetry combinations of the nonbonding electrons on
the amido ligands. The interaction of the amido lone
pairs with metal-based orbitals was ignored in the
model shown in Figure 7; however, these calculations
demonstrate that they are not negligible. It was men-
tioned previously that attempts to optimize the geom-
(45) Becke, A. D. J . Chem. Phys. 1993, 98, 5648.
(46) Hay, P. J .; Wadt, W. R. J . Chem. Phys. 1985, 82, 270.

Diamagnetic and Paramagnetic Dinuclear Ta Hydrides
Organometallics, Vol. 19, No. 19, 2000 3939
F igu r e 10. Four possible linear combinations of the nitrogen lone-pair orbitals in 1A, in D2 symmetry, depicted in the
N-Ta-N plane, along with their symmetry labels. The PH3 donors and bridging hydrides are not shown.
etry of 1A with no restriction on the H-N-Ta-Ta
dihedral angle resulted in a dihedral angle of 180°. The
amide lone pairs in this situation have good overlap with
the δ symmetry orbitals (labeled b1 in Figure 7), as
illustrated in Figure 9a). As mentioned, the backbone
of the [P2N2] macrocycle renders this conformation
impossible. If the H-N-Ta-Ta dihedral angle is near
90°, the best overlap now occurs with the metal-based
σ-bonding orbitals (the highest energy orbital of a
symmetry in Figure 7: this overlap is illustrated in
Figure 9b).
F igu r e 11. Depictions of the molecular orbitals containing
three of the four possible linear combinations of the amido
“lone-pair” p orbitals for tetrahydride 1A. The front NH2
groups and their π-donor lone-pair orbitals obscure the
view of the back NH2 groups.
In the case of the actual model 1A, the H-N-Ta-Ta
dihedral angle was restricted to 120° in an attempt to
simulate the conformational restrictions of the macro-
cyclic [P2N2] ligand. Therefore, overlap with both the
metal orbitals in Figure 9 is possible, although each
overlap may be poorer than illustrated in Figure 9.
There are four possible linear combinations of the amide
π-donor orbitals, and these are shown in Figure 10.
These have a, b1, b2, and b3 symmetry, respectively.
There are orbitals of each of these symmetries in
model complex 1A, as can be seen from Figure 7;
however, the b2 and b3 symmetry metal-hydride π-bond-
ing orbitals are too low in energy for a large interaction
with the amide lone-pair orbitals. Therefore, in model
complex 1A the two highest energy amide lone-pair
orbitals, HOMO-1 and HOMO-2, are of b2 and b3
symmetry and do not have a significant contribution
from any metal-based orbital. The next lowest energy
symmetry combination of the amido lone pair is of a
symmetry and has some overlap with the metal-metal
orbital of a symmetry similar to the HOMO orbital. This
metal orbital has already been used to form the metal-
metal bond of the HOMO orbital. There is a slight
antibonding contribution from the lone pair orbital to
the metal center in the HOMO, so that overall this
interaction does not greatly strengthen metal-amide
bonding. The first three symmetry combinations of the
amido donors are shown in Figure 11.
F igu r e 12. Depiction of the overlap of the fourth sym-
metry combination of the amido “lone-pair” p orbitals with
the metal orbitals involved in δ bonding through the
hydride ligands for tetrahydride 1A.
π-bonding interactions in 1A, as anticipated. These are
shown in Figure 13. These π-bonding interactions are
likely responsible for the eclipsed phosphines on the
opposite metal centers. In compound 1A it is not clear
that the σ- or δ-bonding interaction should lead to a
preference in the arrangement of the ligands on the
opposing metal centers (i.e. if the phosphines are
eclipsed or staggered by 90°). If the energies of the two
π-interactions are largely disparate, as for 1A, the
eclipsed conformation should be favored, as observed
here.
An interesting aspect of the [P2N2] ligand is that the
orientation in which amido lone-pair orbitals are di-
rected is restricted to a very small range. The bonding
of the amido lone-pair orbitals appears to destabilize
the metal σ-bonding orbital relative to the other re-
maining metal d orbitals. This destabilization may be
responsible for the longer Ta-Ta distance in 1 compared
The fourth symmetry combination of the amido lone
pairs is b1, which is the correct symmetry to overlap
with the δ-bonding interaction between metal centers.
This metal orbital is also occupied and involved in δ
bonding with the hydrides. This overlap results in two
occupied orbitals delocalized over the nitrogen donors,
the tantalum centers, and the hydrides, as shown in
Figure 12.
Along with the σ-bonding and δ-bonding interactions
in the Ta2(µ-H)4 moiety, there are also two strong

3940 Organometallics, Vol. 19, No. 19, 2000
Fryzuk et al.
Bruker AMX-500 instrument operating at 500.1 MHz for 1H
1
spectra. H NMR spectra were referenced to internal C6D5H
(7.15 ppm), CDHCl2 (5.32 ppm), and C7D7H (2.09 ppm) and
31P{1H} NMR spectra to external P(OMe)3 (141.0 ppm with
respect to 85% H3PO4 at 0.0 ppm). Solution and solid-state
EPR spectra were recorded on a Bruker ECS 106. Mr. P. Borda
of this department performed the elemental analyses.
Variable-temperature magnetic susceptibility data were
collected on a Quantum Design (MPMS) SQUID magnetom-
eter. Magnetic susceptibilities were corrected for the back-
ground signal of the sample holder and for diamagnetic
susceptibilities of all atoms (-870 × 10-6 emu‚mol-1). Mea-
surements were made from 2 to 300 K and at a field strength
of 10000 G on a microcrystalline sample that had been ground
into a fine powder.
F igu r e 13. Depiction of the two π interactions between
metal centers mediated by the bridging hydride ligands for
the tetrahydride model complex 1A.
The compound [P2N2]TaMe3 was prepared as previously
described.5 Methyl iodide was purchased from Aldrich, de-
gassed via three freeze-pump-thaw cycles and stored over
copper wire.
to [Cl2(PMe3)2Ta]2(µ-H)4,12 because in 1 the bridging
hydride, amide lone-pair and metal-based electrons
must vie for bonding with the same metal orbitals. A
further result of this competition of the amide donors
with the metal-metal bonding orbital should be an
increased reduction potential as compared to that in the
absence of these strong π-donors.
Syn th esis of ([P 2N2]Ta )2(µ-H)4 (1). A yellow solution of
[P2N2]TaMe3 (1.0 g, 1.3 mmol) in 200 mL of hexanes was
transferred to a 400 mL thick-walled glass reaction vessel
equipped with a Teflon valve and a stir bar. The solution was
degassed by two freeze-pump-thaw cycles. With the solution
frozen in a liquid N2 bath, 1 atm of hydrogen gas was added
and allowed to cool for 20 min. The Teflon valve was then
sealed, and the solution was warmed to room temperature, to
produce a hydrogen pressure of 4 atm. Once at room temper-
ature the solution was stirred rapidly, and within 30 min the
solution turned dark red and a microcrystalline precipitate
formed. After 2 days the hydrogen gas was removed and the
brick red precipitate was collected by filtration, rinsed with
10 mL of hexanes, and dried under vacuum, giving ([P2N2]Ta)2-
(µ-H)4 in 80% yield. The dark brown hexane rinse did not
appear to contain any NMR-active species. Species 1 has only
trace solubility in hexanes and is moderately soluble in
aromatic solvents. Single crystals suitable for X-ray analysis
were obtained by performing the reaction without stirring. 1H
NMR (500 MHz, C7H8, 25 °C): δ 0.230, 0.228, 0.46, and 1.17
(s, 48H total, SiCH3), 0.66 and 0.99 (ABX, 2J HH ) 13.5 Hz, 8H
total, CH2 ring), 1.04 and 1.43 (ABX, 2J HH ) 13.1 Hz, 8H total,
Con clu sion s
The hydrogenation of [P2N2]TaMe3 produces the Ta-
(IV) dimer ([P2N2]Ta)(µ-H)4, which contains a tantalum-
tantalum bond. Attempts to react this species with
electrophiles such as MeI and H+{B[3,5-(CF3)2C6H3]3}
led only to the oxidation of ([P2N2]Ta)(µ-H)4 to produce,
in the case of the reaction with MeI, {([P2N2]Ta)2(µ-
H)4}+I-. This reaction involves the loss of one electron
from the tantalum-tantalum bond. Cyclic voltammetry
demonstrates that both electrons in the tantalum-
tantalum bond can be removed, to generate ([P2N2]Ta)2-
+
(µ-H)4 and ([P2N2]Ta)2(µ-H)42+, respectively, and this
allows for the quantification of the reducing power of
these metal-metal bonding electrons. DFT calculations
indicate that the overlap of the amide lone-pair orbitals
with the metal-metal bonding orbitals has an influence
on the orbital energies, an effect that may have impor-
tance in other related systems where electrons stored
in metal-metal bonds are used to reduce dinitrogen,
such as the closely related system, ([NPN]Ta)2(µ-H)4.1
2
CH2 ring), 5.86 (q, J PH ) 9.9 Hz, 4H, µ-H), 7.13 (m, 12H, m-/
p-H), 7.42 (m, 8H, o-H). 31P{1H} NMR (C7H8, 25 °C): δ 10.75.
UV/vis (toluene; λ (nm), ꢀ): 455, 6200; 348, 8600; 313, 18 000.
Anal. Calcd for C24H44N2P2Si4Ta: C, 40.27; H, 6.19; N, 3.91.
Found: C, 40.37; H, 6.34; N, 3.85.
([P 2N2]Ta )2(µ-D)4 (1-d 4). The deuterated analogue of 1 was
prepared in an identical manner using D2 gas in lieu of H2
gas.
[P 2N2]Ta Me2(H) (2). A yellow solution of [P2N2]TaMe3 (20
mg) in 1 mL of C6D6 was sealed in a NMR tube under 0.5 atm
of hydrogen gas using a vacuum manifold equipped with a
mercury manometer and stirred in the absence of light for 1
week. The resulting solution was identified by 1H and 31P{1H}
NMR as having the empirical formula [P2N2]TaMe2(H). 1H
NMR (500 MHz, C6H6, 25 °C): δ 0.06, 0.22, 0.39, and 0.46 (s,
Exp er im en ta l Section
Unless otherwise stated, all manipulations were performed
under an atmosphere of dry oxygen-free dinitrogen by means
of standard Schlenk or glovebox techniques (Vacuum Atmo-
spheres HE-553-2 glovebox equipped with a MO-40-2H puri-
fication system and a -40 °C freezer). Most solvents were dried
under argon; however, using a dinitrogen atmosphere does not
interfere. Hexanes were predried by refluxing over CaH2 and
then distilled under argon from sodium benzophenone ketyl
with tetraglyme added to solubilize the ketyl. Anhydrous
diethyl ether was stored over 4 Å molecular sieves and distilled
from sodium benzophenone ketyl under argon. Toluene was
predried by refluxing over CaH2 and then distilled from sodium
under argon. Nitrogen was dried and deoxygenated by passing
the gases through a column containing molecular sieves and
MnO. Deuterated benzene and toluene were dried by refluxing
with molten potassium metal and molten sodium metal,
respectively, in a sealed vessel under partial pressure and then
trap-to-trap distilled and freeze-pump-thaw-degassed three
times. Unless otherwise stated, 1H, 31P, 1H{31P}, 13C{1H}, 13C,
and variable-temperature NMR spectra were recorded on a
2
2
24H total, SiCH3), 0.88 (dd, J HP ) 7 Hz, J HP ) 7 Hz, 6H,
TaCH3), 0.93, 1.34, 1.39 and 1.90 (dd, 8H total, SiCH2P), 5.68
(s, w1/2 ) 4.0 Hz, 1H, Ta2(µ-H)), 7.11 (overlapping multiplets,
m-/p-H), 7.47 (m, 2H, PPh o-H), 7.79 (m, 2H, PPh o-H). 31P
2
2
NMR (C6H6, 25 °C): δ 17.2 (d, J PP ) 125 Hz), 24.4 (d, J PP
)
125 Hz).
Syn th esis of {([P 2N2]Ta )2(µ-H)4}+I- (3). A stirred solution
of [P2N2]Ta(µ-H)4 (0.429 g, 0.300 mmol) in 60 mL of toluene
was degassed, and then a large excess of MeI was added by
vacuum transfer. Over 3 h the dark solution went colorless
and a green microcrystalline solid precipitated. The toluene
and excess MeI were removed under vacuum. The solid was
collected on a glass filter and rinsed with 20 mL of toluene
and then 20 mL of hexanes. The remaining green solid was
dried under vacuum, yielding {([P2N2]Ta)2(µ-H)4}+I- (0.420 g,

Diamagnetic and Paramagnetic Dinuclear Ta Hydrides
Organometallics, Vol. 19, No. 19, 2000 3941
90%). This green solid is insoluble in aromatic solvents and
tetrahydofuran but soluble in methylene chloride, though
solutions decompose over the course of 1 week to give a
colorless solution. Methylene chloride solutions of 3 are
dichroic; reflected light appears burgundy and transmitted
light is blue-green. Single crystals suitable for X-ray analysis
were obtained by performing the synthesis in an identical
manner, but without stirring. Anal. Calcd for C48H88N4P4Si8-
Ta2: C, 37.01; H, 5.63; N, 3.60. Found: C, 37.25; H, 5.45; N,
3.47. UV/vis (toluene; λ (nm), ꢀ): 576, 2900; 450, sh; 346,
16 000; 298, 15 000. EPR (solid, 295 K): g ) 1.88, W1/2 ) 800
G. EPR (solution, 295 K): g ) 1.89, broad.
The colorless toluene-soluble reaction products were evapo-
rated to dryness and identified by 1H and 13C{1H} NMR as
dibenzyl (PhCH2CH2Ph). 1H NMR (C6D6, 400 MHz, 298 K): δ
2.73 (s, 4H, PhCH2CH2Ph), 6.98 (m, 4H, o-H), 7.05 (m, 2H p-H),
7.13 (m, 4H, m-H). 13C{1H} NMR (C6D6, 298 K): δ 38.1 (s,
PhCH2CH2Ph), 126.2 (s, p-C), 128.5 (s, m-C), 128.8 (s, o-C),
142.0 (s, ipso-C).
X-r a y Cr ysta llogr a p h ic An a lyses of 1 a n d 3. Crystal-
lographic data for 1 and 3 appear in Table 1. For 1 the final
unit-cell parameters were obtained by least-squares methods
on the setting angles for 25 reflections with 2θ ) 20.1-25.4°.
The intensities of 3 standard reflections were measured every
200 reflections. Throughout the data collections the standards
decreased by 1.19%. A linear correction factor was applied to
the data to account for this phenomenon. The data were
processed and corrected for Lorentz and polarization effects
and absorption.47 The structure of 1 was solved by heavy-atom
Patterson methods and expanded using Fourier techniques.
The non-hydrogen atoms were refined anisotropically. The
metal hydride atoms were refined isotropically, with distance
constraints for H(3) and H(4) and a fixed thermal parameter
for H(3). The remaining hydrogen atoms were fixed in calcu-
lated positions with C-H ) 0.98 Å.
For 3 the final unit-cell parameters were obtained by least-
squares methods on the setting angles for 33 122 reflections
with 2θ ) 4.0-60.1° for 3. The data were processed and
corrected for Lorentz and polarization effects. The structure
of 3 was solved by direct methods and expanded using Fourier
techniques. All the non-hydrogen atoms were refined aniso-
tropically. Metal hydrides were placed in difference map
positions but were not refined, and the remaining hydrogen
atoms were fixed in calculated positions with C-H ) 0.98 Å.
The absolute configuration of 3 was determined by Flack
parameter refinement (0.014(5)).
Neutral atom scattering factors for the non-hydrogen atoms
were taken from refs 48 and 49. Atomic coordinates, aniso-
tropic thermal parameters, complete bond lengths and bond
angles, torsion angles, intermolecular contacts, and least-
squares planes are included as Supporting Information.
Ca lcu la tion s. The ab initio DFT calculations on the model
compound [(H2N)2(H3P)2Ta]2(µ-H)4 were performed with the
hybrid functional B3LYP method45 using the Gaussian 98
package.44 The basis functions and effective core potentials
(ECP) used were those in the LANL2DZ basis set developed
by Hay and Wadt46 and provided with the Gaussian 98
program, but with an additional d polarization functions added
to P atoms. The exponent of the d function was 0.37. This level
of theory has recently been shown to be adequate for modeling
complexes containing metal-metal bonds of various orders.35
Model complex 1A was optimized with D2 symmetry with the
Ta-Ta-P and Ta-Ta-N angles frozen at 110 and 129.5°,
respectively, and the H-N-Ta-Ta dihedral angle frozen at
120°. Orbital depictions were created using the MOLDEN
program.50
Ack n ow led gm en t. We gratefully acknowledge both
the NSERC of Canada and the Petroleum Research
Fund, administered by the American Chemical Society,
for funding in the form of research grants. Financial
support was also provided by the NSERC of Canada in
the form of a postgraduate scholarship for S.A.J ., as well
as the University of British Columbia in the form of a
University Graduate Fellowship. We thank Victor
Sanchez for assistance with the use of the SQUID
magnetometer as well as Dr. R. C. Thompson for
informative discussions on the subject of magnetism.
Su p p or tin g In for m a tion Ava ila ble: Text giving full
experimental details, tables of X-ray parameters, fractional
coordinates and thermal parameters, bond lengths, bond
angles, and torsion angles, and ORTEP diagrams for com-
pounds 1 and 3, variable-temperature magnetic susceptibility
data for compound 3, and full density functional theory results
for 1A. This material is available free of charge via the Internet
at http://pubs.acs.org.
OM000359W
(48) International Tables for X-ray Crystallography; Kynoch Press:
Birmingham, U.K. (present distributor Kluwer Academic: Boston,
MA), 1974; Vol. IV.
(49) International Tables for Crystallography; Kluwer Academic:
Boston, MA, 1992; Vol. C.
(47) teXsan: Crystal Structure Analysis Package; Molecular Struc-
ture Corp., The Woodlands, TX, 1995.
(50) Schaftenaar, G. Molden, 3.5 ed.; CAOS/CAMM Center, Uni-
versity of Nijmegen, Nijmegen, The Netherlands, 1991.