Synthesis and characterization of mixed-phosphine osmium polyhydrides: Hydrogen delocalization in [OsH5P3]+ systems

Article abstract of DOI:10.1021/om010173c

The hexahydrido complex OsH₆(PⁱPr₃)₂ (1) reacts with PHPh₂ to give molecular hydrogen and the tetrahydride OsH₄(PHPh₂)(PⁱPr₃)₂ (2). However, the formation of OsH₂(PHPh₂)₂-(PⁱPr₃) ₂ (3) as a consequence of the substitution of a second hydrogen molecule from 2 by PHPh₂ does not occur. The treatment of 2 with 1.0 equiv of PHPh₂ in toluene at 80 °C leads after 3 days to OsH₂(PHPh₂)₃(PⁱPr₃) (4). The preparation of 3 requires the previous acidolysis of 2 with HBF₄, which gives [OsH₅(PHPh₂)(PⁱPr₃)₂] BF₄ (5). In contrast 2, the addition of PHPh₂ to 5 affords [OsH₃(PHPh₂)₂(PⁱPr₃) ₂]BF₄ (6) and molecular hydrogen. Deprotonation of 6 with Et₃N yields 3. The skeleton of the cation of 5 has been determined by X-ray diffraction. The configuration is consistent with a Y-shaped OsP₃ disposition with the osmium atom in the common vertex. Complex 5 also reacts with methanol and water to give [OsH₅-{P(OMe)Ph₂}(PⁱPr₃)₂ ]BF₄ (7) and [OsH₅{P(OH)Ph₂}(PⁱPr₃)₂]B F₄ (8), respectively. The addition of Et₃N to 7 affords OsH₄{P(OMe)Ph₂}(PⁱPr₃)₂ (9). A theoretical study on the OsH₅(PH₃)₃⁺ model complex reveals that although a static description is fully consistent with a classical pentahydride assignment, the formation of a dihydrogen is a very low energy costing process, on both thermodynamic and kinetic grounds. Thus, these polyhydride systems might be better described as possessing delocalized hydrogen atoms. A further QM/MM IMOMM study on the actual OsH₅(PR₃)⁺ systems indicates that the inclusion of bulky phosphine substituents plays a role against the stability of dihydrogen forms, because of the higher steric congestion of lower coordination number complexes arising from repulsions between bulky phosphines. Although IMOMM calculations improve significantly the agreement with experimental structures, they do not change the validity of the aforementioned statement concerning delocalization.

Full text of DOI:10.1021/om010173c

Organometallics 2001, 20, 5297-5309
5297
Syn th esis a n d Ch a r a cter iza tion of Mixed -P h osp h in e
Osm iu m P olyh yd r id es: Hyd r ogen Deloca liza tion in
[OsH5P 3]+ System s
Miguel A. Esteruelas,*,‡ Agust´ı Lledo´s,*,† Marta Mart´ın,‡ Feliu Maseras,†
Raquel Ose´s,‡ Natividad Ruiz,‡ and J aume Toma`s†
Departament de Quı´mica, Edifici Cn, Universitat Auto`noma de Barcelona, 08193 Bellaterra,
Barcelona, Spain, and Departamento de Qu´ımica Inorga´nica, Instituto de Ciencia de
Materiales de Arago´n, Universidad de Zaragoza, CSIC, E-50009 Zaragoza, Spain
Received March 2, 2001
The hexahydrido complex OsH6(PiPr3)2 (1) reacts with PHPh2 to give molecular hydrogen
and the tetrahydride OsH4(PHPh2)(PiPr3)2 (2). However, the formation of OsH2(PHPh2)2-
(PiPr3)2 (3) as a consequence of the substitution of a second hydrogen molecule from 2 by
PHPh2 does not occur. The treatment of 2 with 1.0 equiv of PHPh2 in toluene at 80 °C leads
after 3 days to OsH2(PHPh2)3(PiPr3) (4). The preparation of 3 requires the previous acidolysis
of 2 with HBF4, which gives [OsH5(PHPh2)(PiPr3)2]BF4 (5). In contrast to 2, the addition of
PHPh2 to 5 affords [OsH3(PHPh2)2(PiPr3)2]BF4 (6) and molecular hydrogen. Deprotonation
of 6 with Et3N yields 3. The skeleton of the cation of 5 has been determined by X-ray
diffraction. The configuration is consistent with a Y-shaped OsP3 disposition with the osmium
atom in the common vertex. Complex 5 also reacts with methanol and water to give [OsH5-
{P(OMe)Ph2}(PiPr3)2]BF4 (7) and [OsH5{P(OH)Ph2}(PiPr3)2]BF4 (8), respectively. The addition
+
of Et3N to 7 affords OsH4{P(OMe)Ph2}(PiPr3)2 (9). A theoretical study on the OsH5(PH3)3
model complex reveals that although a static description is fully consistent with a classical
pentahydride assignment, the formation of a dihydrogen is a very low energy costing process,
on both thermodynamic and kinetic grounds. Thus, these polyhydride systems might be better
described as possessing delocalized hydrogen atoms. A further QM/MM IMOMM study on
the actual OsH5(PR3)+ systems indicates that the inclusion of bulky phosphine substituents
plays a role against the stability of dihydrogen forms, because of the higher steric congestion
of lower coordination number complexes arising from repulsions between bulky phosphines.
Although IMOMM calculations improve significantly the agreement with experimental
structures, they do not change the validity of the aforementioned statement concerning
delocalization.
In tr od u ction
H-H distances fall between the dihydrogen (0.8-1.0 Å)
and dihydride (>1.6 Å) limits. This third class of
complexes has been called elongated or stretched dihy-
drogen complexes. J HD values are useful indicators of
the presence of an elongated dihydrogen ligand,4,5 and
some of them have been characterized by neutron
diffraction techniques, but their chemical properties are
not fully understood with purely structural arguments.6
LMHn polyhydrides with n > 4 have attracted the
interest of theoreticians.3,7,8 The [OsH5(PMe2Ph)3]+
After the discovery that, in addition to the classical
hydride form, hydrogen atoms can be present in the
coordination sphere of a transition metal as a dihydro-
gen ligand, much effort has been devoted to the struc-
tural characterization of transition metal hydrides.1
Although experimental location of H atoms presents
significant difficulties, neutron diffraction determina-
tions2 and NMR (T1(min) and J HD) measurements have
led to accurate structural assignments for a large series
of compounds. Moreover, since computational chemistry
is nowadays a low-cost high-quality technique for pre-
cisely locating H nuclei and predicting classical versus
nonclassical structures, theoretical calculations are
being widely used to this purpose.3 However there is
an increasing number of species that are difficult to
classify from a structural point of view because their
(3) Maseras, F.; Lledo´s, A.; Clot, E.; Eisenstein, O. Chem. Rev. 2000,
100, 601.
(4) (a) Klooster, W. T.; Koetzle, T. F.; J ia, G.; Fong, T. P.; Morris,
R. H.; Albinati, A. J . Am. Chem. Soc. 1994, 116, 7677. (b) 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. (c) Heinekey, D. M.; Luther, T. A. Inorg. Chem. 1996, 35,
4396. (d) Luther, T. A.; Heinekey, D. M. J . Am. Chem. Soc. 1997, 119,
6688.
(5) (a) Bacskay, G. B.; Bytheway, I.; Hush, N. S. J . Am. Chem. Soc.
1996, 118, 3753. (b) Hush, N. S. J . Am. Chem. Soc. 1997, 119, 1717.
(6) (a) Gelabert, R.; Moreno, M.; Lluch, J . M.; Lledo´s, A. J . Am.
Chem. Soc. 1997, 119, 9840. (b) Gelabert, R.; Moreno, M.; Lluch, J .
M.; Lledo´s, A. J . Am. Chem. Soc. 1998, 120, 8168. (c) Barea, G.;
Esteruelas, M. A.; Lledo´s, A.; Lo´pez, A. M.; Tolosa, J . I. Inorg. Chem.
1998, 37, 5033. (d) Barea, G.; Esteruelas, M. A.; Lledo´s, A.; Lo´pez, A.
M.; On˜ate, E.; Tolosa, J . I. Organometallics 1998, 17, 4065.
† Universitat Auto`noma de Barcelona.
‡ Universidad de Zaragoza.
(1) (a) Kubas, G. J . Acc. Chem. Res. 1988, 21, 120. (b) Crabtree, R.
H. Acc. Chem. Res. 1990, 23, 95. (c) J essop, P. G.; Morris, R. H. Coord.
Chem. Rev. 1992, 121, 155. (d) Heinekey, D. M.; Oldham, W. J ., J r.
Chem. Rev. 1993, 93, 913.
(2) Bau, R.; Drabnis, M. H. Inorg. Chim. Acta 1997, 259, 27.
10.1021/om010173c CCC: $20.00 © 2001 American Chemical Society
Publication on Web 11/10/2001

5298 Organometallics, Vol. 20, No. 25, 2001
Esteruelas et al.
complex, extensively studied by Caulton et al.,9 shows
interesting features concerning hydride arrangement.
An early paper suggested that it is a dihydrogen
trihydride complex.9a This characterization was based
on a short T1(min) measured by solution NMR tech-
niques. Although this has often been considered a valid
criterion for the detection of a dihydrogen moiety, the
structure was later reevaluated through neutron dif-
fraction techniques, and finally it was best described
as a dodecahedral pentahydride. This assignment,
however, posed the problem of the anomalously short
T1(min), which might indicate that some residual
H‚‚‚H interaction is present between a pair of hydrides.
Indeed, the neutron diffraction structure located two
hydrides at 1.49 Å from each other,9b a large distance
for a typical dihydrogen unit, but shorter than the sum
of their van der Waals radii. This indicates that the
complex [OsH5(PMe2Ph)3]+ may belong to the rather
diffuse range of elongated dihydrogen complexes. It
shows also remarkable reactivity patterns, hydrogenat-
ing ethylene under mild conditions (25 °C, 1 atm) to
generate [OsH(C2H4)2(PMe2Ph)3]+ and alkane.9b The
saturated character of the pentahydride requires the
dissociation of at least a hydrogen molecule before the
coordination of the hydrogenated olefin. This step of the
hydrogenation process can be formally viewed as the
substitution of a π-acceptor H2 ligand by ethylene. Thus,
although the neutron diffraction study indicates the
absence of H2 ligands in the solid at 11 K, the whole of
the experimental data is compatible with the forma-
tion in solution of dihydrogen species in the [OsH5-
(PMe2Ph)3]+ systems at a very low energy cost.
Early calculations for the [OsH5(PH3)3]+ model system
from Lin and Hall predicted the pentahydride complex
to be the most stable species.10a Later Maseras et al.
analyzed exhaustively the structural features of this
model complex.10b In their systematic approach, 22
different initial structures were carefully examined
through ab initio calculations. This analysis led to the
description of five low-lying isomers, the lowest of them
being a dodecahedral pentahydride complex, in very
good agreement with the actual neutron diffraction of
[OsH5(PMe2Ph)3]+, published one year later.9b
Despite the success of theoretical calculations in the
determination of the most stable structures in several
polyhydride complexes, an accurate description of these
systems is still lacking. On one hand, most of the
calculations have been carried out in model systems,
replacing the actual PR3 ligands by PH3, and this
simplification may lead to a significant error in some
cases. For instance, experimental results on ReH7(PR3)2
derivatives have shown an extreme dependence of one
H-H distance on the nature of the PR3 ligands.11 On
the other hand, high-level theoretical methods that give
a balanced description in all the regions of the potential
energy surface (polyhydride, dihydrogen, bisdihydro-
gen) are required to study the energetics of the H motion
in polyhydrides. The experimental values of the dihy-
drogen/dihydride equilibrium in the Kubas complex
have been reproduced only theoretically by using
CCSD(T) calculations.12
As a part of our study on the chemical properties of
the hexahydride OsH6(PiPr3)2,6d,13 we have observed
that, in addition to [OsH5(PMe2Ph)3]+, mixed-phosphine
[OsH5(PR3)2(PR3′)]+ complexes can be prepared. In this
paper we report (i) the synthesis of [OsH5(PHPh2)-
(PiPr3)2]BF4 and [OsH5{P(OMe)Ph2}(PiPr3)2]BF4 and
related tetra-, tri-, and dihydride compounds; (ii) the
X-ray-determined structure of [OsH5(PHPh2)(PiPr3)2]-
BF4; and (iii) theoretical studies on [OsH5(PH3)3]+,
[OsH5(PHPh2)(PiPr3)2]+, [OsH5{P(OMe)Ph2}(PiPr3)2]+,
and [OsH5(PMe2Ph)3]+ by means of high-level quan-
tum mechanical (QM) and hybrid quantum mechanics/
molecular mechanics (QM/MM) IMOMM calculations.
Resu lts a n d Discu ssion
1. Rea ction s of OsH6(P iP r 3)2 w ith P HP h 2. OsH6-
(PiPr2Ph)2 has been characterized by neutron diffraction
as a classical hexahydride with a typical dodecahedral
coordination geometry.14 Although these OsH6(PR3)2
complexes do not contain any dihydrogen ligand, B3LYP
calculations gave a low value for the dissociation energy
of molecular hydrogen (15.4 kcal/mol),13i while our high-
level CCSD(T)//B3LYP calculations gave an even smaller
value (10.0 kcal/mol). This suggests that the elimination
of molecular hydrogen from the hexahydride is easy. In
agreement with this, the treatment of OsH6(PiPr3)2 (1)
with 1.0 equiv of diphenylphosphine in toluene at 80
°C leads after 1 h to the tetrahydride OsH4(PHPh2)-
(PiPr3)2 (2), which was isolated as a white solid in 71%
yield, according to eq 1.
(11) (a) Hamilton, D. G.; Crabtree, R. H. J . Am. Chem. Soc. 1988,
110, 4126. (b) Costello, M. T.; Walton, R. A. Inorg. Chem. 1988, 27,
2563. (c) Howard, J . A. K.; Mason, S. A.; J ohnson, O.; Diamond, I. C.;
Crennell, S.; Keller, P. A.; Spencer, J . L. J . Chem. Soc., Chem.
Commun. 1988, 1502. (d) Luo, X.-L.; Crabtree, R. H. Inorg. Chem. 1989,
28, 3775. (e) Luo, X.-L.; Crabtree, R. H. J . Am. Chem. Soc. 1990, 112,
4813. (f) Luo, X.-L.; Baudry, D.; Boydell, P.; Charpin, P.; Nierlich, M.;
Ephritikhine, M.; Crabtree, R. H. Inorg. Chem. 1990, 29, 1511. (g)
Brammer, L.; Howard, J . A. K.; J ohnson, O.; Koetzle, T. F.; Spencer,
J . L.; Stringer, A. M. J . Chem. Soc., Chem. Commun. 1991, 241. (h)
Haynes, G. R.; Martin, R. L.; Hay, P. J . J . Am. Chem. Soc. 1992, 114,
28. (i) Michos, D.; Luo, X.-L.; Howard, J . A. K.; Crabtree, R. H. Inorg.
Chem. 1992, 31, 3914.
(12) Toma`s, J .; Lledo´s, A.; J ean, Y. Organometallics 1998, 17, 190.
(13) (a) Aracama, M.; Esteruelas, M. A.; Lahoz, F. J .; Lo´pez, J . A.;
Meyer, U.; Oro, L. A.; Werner, H. Inorg. Chem. 1991, 30, 288. (b)
Esteruelas, M. A.; Lahoz, F. J .; Lo´pez, J . A.; Oro, L. A., Schlu¨nken, C.;
Valero, C., Werner, H. Organometallics 1992, 11, 2034. (c) Esteruelas,
M. A.; J ean, Y.; Lledo´s, A.; Oro, L. A.; Ruiz, N.; Volatro´n, F. Inorg.
Chem. 1994, 33, 3609. (d) Atencio, A.; Esteruelas, M. A.; Lahoz, F. J .;
Oro, L. A., Ruiz, N. Inorg. Chem. 1995, 34, 1004. (e) Demachy, I.;
Esteruelas, M. A.; J ean, Y.; Lledo´s, A.; Maseras, F.; Oro, L. A.; Valero,
C.; Volatron, F. J . Am. Chem. Soc. 1996, 118, 8388. (f) Esteruelas, M.
A.; Lahoz, F. J .; Lo´pez, A. M.; On˜ate, E.; Oro, L. A.; Ruiz, N.; Sola, E.;
Tolosa, J . I. Inorg. Chem. 1996, 35, 7811. (g) Castillo, A.; Esteruelas,
M. A.; On˜ate, E.; Ruiz, N. J . Am. Chem. Soc. 1997, 119, 9691. (h)
Castillo, A.; Barea, G.; Esteruelas, M. A.; Lahoz, F. J .; Lledo´s, A.;
Maseras, F.; Modrego, J .; On˜ate, E.; Oro, L. A.; Ruiz, N.; Sola, E. Inorg.
Chem. 1999, 38, 1814. (i) Barrio, P.; Castarlenas, R.; Esteruelas, M.
A.; Lledo´s, A.; Maseras, F.; On˜ate, E.; Toma´s, J . Organometallics 2001,
20, 442.
(7) (a) Hlatky, G. G.; Crabtree, R. H. Coord. Chem. Rev. 1985, 65,
1. (b) Lin, Z.; Hall, M. B. Coord. Chem. Rev. 1994, 135/ 136, 845.
(8) (a) Maseras, F.; Li, X.-K.; Koga, N.; Morokuma, K. J . Am. Chem.
Soc. 1993, 115, 10974. (b) Barea, G.; Maseras, F.; J ean, Y.; Lledo´s, A.
Inorg. Chem. 1996, 35, 6401. (c) Bayse, C. A.; Couty, M.; Hall, M. B.
J . Am. Chem. Soc. 1996, 118, 8916. (d) Bosque, R.; Maseras, F.;
Eisenstein, O.; Patel, B. P.; Yao, W.; Crabtree, R. H. Inorg. Chem. 1997,
36, 5505. (e) Bayse, C. A.; Hall, M. B.; Pleune, B.; Poli, R. Organome-
tallics 1998, 17, 4309.
(9) (a) J ohnson, T. J .; Huffman, J . C.; Caulton, K. G.; J ackson, S.
A.; Eisenstein, O. Organometallics 1989, 8, 2073. (b) J ohnson, T. J .;
Albinati, A.; Koetzle, T. F.; Ricci, J .; Eisenstein, O.; Huffman, J . C.;
Caulton, K. G. Inorg. Chem. 1994, 33, 4966.
(10) (a) Lin, Z.; Hall, M. B. J . Am. Chem. Soc. 1992, 114, 6102. (b)
Maseras, F.; Koga, N.; Morokuma, K. J . Am. Chem. Soc. 1993, 115,
8313.
(14) Howard, J . A. K.; J ohnson, O.; Koetzle, T. F.; Spencer, J . L.
Inorg. Chem. 1987, 26, 2930.

Mixed-Phosphine Osmium Polyhydrides
Organometallics, Vol. 20, No. 25, 2001 5299
membered rings of the type EH‚‚‚HM, a weak hydrogen-
hydrogen interaction has been proposed to exist.
The T1 values of the hydrogen nuclei of the OsH4 of 2
were determined over the temperature range 293-213
K. The T1(min) value (137 ms) was obtained at 233 K
and supports the tetrahydride character of this complex.
In contrast to 2, the carbonyl derivative OsH4(CO)-
19
(PiPr3)2 has a T1(min) of 32 ms, which suggests non-
The presence of a diphenylphosphine ligand in 2 is
1
strongly supported by the IR and 31P{1H} and H NMR
classical interaction between two of the four hydrogens
bonded to the osmium atom.20 DFT calculations on the
model compound OsH4(CO)(PH3)2 indicate the presence
of a dihydrogen ligand trans to a hydride, with an H-H
separation of 0.87 Å. An Os-H2 dissociation energy of
18.9 kcal/mol, similar to that found for 1, was calculated
for OsH2(η2-H2)(CO)(PH3)2.20 In agreement with this,
the carbonyl derivative and 1 show the same reactivity
toward phosphines. The reactions of both compounds
afford molecular hydrogen and OsHn-2 species.
spectra of the complex. The IR spectrum in KBr shows
the ν(P-H) absorption at 2292 cm-1, along with three
bands at 2052, 2013, and 1886 cm-1, corresponding to
the Os-H vibrations. At room temperature, the 31P{1H}
NMR spectrum contains at 39.2 ppm a doublet due to
the triisopropylphosphine ligands and at -4.6 ppm a
triplet corresponding to the triphenylphosphine group.
The value of the P-P coupling constant is 17.1 Hz. This
spectrum is temperature-invariant between 25 and -80
°C. This, along with the equivalence of the triisopropyl-
phosphine ligands and the value of the P-P coupling
constant, suggests a T-shaped OsP2P′ skeleton, which
is rigid in solution. This disposition agrees well with
that found for the OsP3 skeleton of the complex
OsH4(PMe2Ph)3 by neutron and X-ray diffraction stud-
ies.15 The structure of the latter complex is a distorted
pentagonal bipyramid, like that shown in eq 1 for 2,
with the four hydride ligands, osmium, and one phos-
phorus atom being essentially coplanar. A theoretical
study of OsH4(PH3)3 gave this pentagonal bipyramid as
the most stable isomer.8a In contrast to 2, the OsP3
skeletons of OsH4(PMe2Ph)3 and related OsH4(PR3)3
In contrast to 1 and the carbonyl derivative, the
replacement of a hydrogen molecule from 2 by phos-
phine is a complex process (Scheme 1). The treatment
of 2 with 1.0 equiv of diphenylphosphine in toluene at
80 °C does not produce a significant amount of reaction
after 1 h. The disappearance of about 50% of 2 is
observed after 3 days. However, the expected OsH2-
(PHPh2)2(PiPr3)2 (3) complex is not formed, instead
OsH2(PHPh2)3(PiPr3) (4) is obtained in about 50% yield,
as a white solid. The formation of 4 suggests that the
dissociation energy of a hydrogen molecule from 2 is
similar to the energy necessary to break a Os-PiPr3
bond. CCSD(T)//B3LYP calculations on the OsH4(PH3)3,
OsH2(PH3)3, and OsH4(PH3)2 model complexes have
allowed a good estimation of these dissociation energies.
The strong hydridic character of the H ligands in 2,
which agrees with the high energy required for the H2
elimination, has already been pointed out in a theoreti-
cal study.21 Our results fully agree with this previous
result, giving a De(M-H2) of 40.9 kcal/mol. The dis-
sociation energy for the PH3 phosphine is somewhat
higher (50.1 kcal/mol). However, with PHPh2 instead
of PH3, one can expect that the formation of Os-
(PHPh2)3(PiPr3) might be further prevented by steric
congestion around the Os atom. This fact, along with
the high dissociation energy for the H2 moiety, might
well explain the reactivity found for complex 2.
1
complexes appear to be nonrigid in solution.16 The H
NMR spectrum of 2 is temperature-dependent. At 80
°C, it shows in the hydrido region (-10.19 ppm) a single
resonance. This observation is consistent with the
operation of some thermally activated exchange process,
which proceeds at a rate sufficient to lead to the single
hydride resonance. Consistent with this, lowering the
sample temperature leads to broadening of the reso-
nance, although decoalescence is not observed at -80
°C. At 80 °C, the hydrido resonance is observed as a
triplet of doublets of doublets, by spin coupling with the
phosphorus atoms of the triisopropylphosphine (J (HP)
) 10.5 Hz) and diphenylphosphine (J (HP) ) 3.0 Hz)
ligands, and the PH hydrogen atom of the diphen-
ylphosphine group (J (HH) ) 3.0 Hz). The latter coupling
was confirmed by 1H-COSY and selective heteronu-
The IR spectrum of 4 in KBr contains three ν(P-H)
absorptions at 2374, 2361, and 2348 cm-1, along with
two bands at 2000 and 1927 cm-1 corresponding to the
Os-H vibrations, which indicate a mutually cis disposi-
1
clear H{31P} NMR spectra. In the low-field region the
most noticeable resonance is that due to the PH
hydrogen, which appears at 7.45 ppm as a doublet of
triplets of doublets with H-P coupling constants of
346.0 and 8.4 Hz. The value of the H-H coupling
constant (3.0 Hz) is similar to the values found for
the H-H(E) coupling constants (E ) P, N) in the 1H
NMR spectra of the complexes MH2(PHPh2)(PiPr3)2
1
tion for the hydrido ligands. The 31P{1H} and H NMR
spectra are temperature-invariant, suggesting that the
complex has a rigid structure in solution. The 31P{1H}
NMR spectrum shows an XYZ2 (X ) PiPr3, Y ) PHPh2
trans to PiPr3, Z ) PHPh2 trans to hydrides) simplified
spin system, which is defined by δx ) 33.1, δy ) 13.9, δz
) δz′ ) -8.0, J (PxPy) ) 209.3 Hz, J (PxPz) ) J (PxPz′) )
18.4 Hz, and J (PyPz) ) J (PyPz′) ) 11.5 Hz. In the 1H
(M ) Ru (4.1 Hz), Os (4.5 Hz))17 and [OsH(η5-C5H5)-
{NHdC(Ph)C6H4}(PiPr3)]BF4 (4.1 Hz),18 where, via four-
(18) Esteruelas, M. A.; Gutie´rrez-Puebla, E.; Lo´pez, A: M.; On˜ate,
E.; Tolosa, J . I. Organometallics 2000, 19, 275.
(19) Werner, H.; Esteruelas, M. A.; Meyer, U.; Wrackmeyer, B.
Chem. Ber. 1987, 120, 11.
(20) Gusev, D. G.; Kuhlman, R. L.; Renkema, K. B.; Eisenstein, O.;
Caulton, K. G. Inorg. Chem. 1996, 35, 6775.
(15) Hart, D. W.; Bau, R.; Koetzle, T. F. J . Am. Chem. Soc. 1977,
99, 7557.
(16) (a) Douglas, P. G.; Shaw, B. L. Chem. Commun. 1969, 624. (b)
Douglas, P. G.; Shaw, B. L. J . Am. Chem. Soc. A 1970, 334. (c) Bell,
B.; Chatt, J .; Leigh, J . J . Chem. Soc., Dalton Trans. 1973, 997.
(17) Buil, M. L.; Esteruelas, M. A.; On˜ate, E.; Ruiz, N. Organome-
tallics 1998, 17, 3346.
(21) Maseras, F.; Lledo´s, A.; Costas, M.; Poblet, J . M. Organome-
tallics 1996, 15, 2947.

5300 Organometallics, Vol. 20, No. 25, 2001
Esteruelas et al.
Sch em e 1
NMR spectrum, the most noticeable resonances are
those due to hydride ligands and the PH hydrogen
atoms of the diphenylphosphine groups, which appear
at -10.00, and between 5.91 and 7.00 ppm as the parts
CC′ and AA′B, respectively, of an AA′BCC′XYZZ′ (A, A′,
and B ) PHPh2, C and C′ ) H-Os, X ) PiPr3, Y )
PHPh2 trans to PiPr3, Z and Z′ ) PHPh2 trans to
hydrides) spin system. The spin coupling between the
hydride ligands (CC′) and the PH hydrogen atoms of
the diphenylphosphine groups (B and AA′) was con-
1
firmed by a H-COSY NMR spectrum.
The preparation of 3 requires the protonation of 2, in
agreement with the tendency of the acydolysis to
facilitate the elimination of molecular hydrogen from
polyhydrides.22 Treatment of diethyl ether solutions of
2 with 1.0 equiv of HBF4‚OEt2 (Scheme 1) leads to the
pentahydride derivative [OsH5(PHPh2)(PiPr3)2]BF4 (5),
which was isolated as a white solid in 97% yield. The
reaction is reversible. Thus, the addition of 1.0 equiv of
Et3N to dichloromethane solutions of 5 regenerates 2.
In contrast to the tetrahydride 2, the addition at room
temperature of 1.0 equiv of diphenylphosphine to dichlo-
romethane solutions of 5 produces the elimination of a
hydrogen molecule and the formation of the trihydride
[OsH3(PHPh2)2(PiPr3)2]BF4 (6), which was isolated as a
white solid in 86% yield. The deprotonation of 6 with
Et3N yields 3.
F igu r e 1. Molecular diagram of the cation of complex
[OsH5(PHPh2)(PiPr3)2]BF4 (5).
Ta ble 1. Selected Bon d Dista n ces (Å) a n d An gles
(d eg) for [OsH5(P HP h 2)(P iP r 3)2]BF 4 (5)
Os-P(1)
Os-P(2)
Os-P(3)
2.3870(16)
2.3867(16)
2.3891(16)
P(3)-H(06)
P(3)-C(19)
P(3)-C(25)
1.27(5)
1.818(6)
1.820(7)
P(1)-Os-P(2)
P(1)-Os-P(3)
P(2)-Os-P(3)
149.43(6)
105.72(6)
94.80(6)
Os-P(3)-C(19)
Os-P(3)-C(25)
Os-P(3)-H(06)
C(19)-P(3)-C(25)
C(19)-P(3)-H(06)
C(25)-P(3)-H(06)
120.3(2)
117.8(2)
110(2)
106.6(3)
101(2)
Figure 1 shows a view of the skeleton of the cation of
5. Selected bond distances and angles are listed in
Table 1.
The configuration is consistent with a Y-shaped
OsP2P′ skeleton, with the osmium atom situated in
the common vertex. The three Os-P distances are
statistically identical, about 2.39 Å, while the three
P-Os-P angles are different. The P(1)-Os-P(2)
angle that involving both triisopropylphosphine ligands
(149.43(6)°) is bigger than the other two, P(1)-Os-P(3)
(105.72(6)°) and P(2)-Os-P(3) (94.80(6)°). Despite the
asymmetry, at room temperature, the 31P{1H} NMR
spectrum of 5 shows only one resonance for the triiso-
propylphosphine ligands, which appears at 42.2 ppm
as a doublet with a P-P coupling constant of 10.1 Hz.
The resonance corresponding to the diphenylphos-
phine ligand is observed at -24.4 ppm, as a triplet.
Lowering the sample temperature produces a very light
broadening of the resonances. Decoalescence is not
98(2)
observed upon -80 °C. This suggests that in solution
the P(1)-Os-P(3) and P(2)-Os-P(3) angles average
rapidly and that the Y-shaped OsP2P′ skeleton does not
undergo an important distortion.
Unfortunately, from the X-ray diffraction study, it was
not possible to locate the hydride ligands. However their
1
presence is strongly supported by the IR and H NMR
spectrum. The IR spectrum in KBr shows the ν(P-H)
absorption at 2362 cm-1, along with three bands at
2096, 2086, and 1984 cm-1 corresponding to the Os-H
vibrations, and the absorption due to the [BF4]- anion
with Td symmetry centered at 1058 cm-1. In the 1H
NMR spectrum in dichloromethane at room tempera-
ture, the most noticeable resonances are a triplet at
-8.26 ppm with a H-P coupling constant of 8.2 Hz,
corresponding to the hydride ligands, and at 6.97 ppm
(22) Bruno, J . W.; Huffman, J . C.; Caulton, K. G. J . Am. Chem. Soc.
1984, 106, 1663, and references therein.

Mixed-Phosphine Osmium Polyhydrides
Organometallics, Vol. 20, No. 25, 2001 5301
a double triplet with H-P coupling constants of 373.7
and 3.9 Hz, due to the PH hydrogen of the diphen-
ylphosphine group. In this spectrum, spin coupling be-
tween the hydride and the PH hydrogen atom is not
observed. However the 1H-COSY NMR spectrum shows
signals of crossing between the hydride and PH reso-
nances. The two spectra together suggest that the H-H
coupling constant between the hydrides and PH hydro-
gen atom is smaller than 1 Hz. The decrease of the value
of this coupling constant in 5 with regard to 2 appears
to indicate a decrease of the fortress of the PH-hydride
interaction in 5 with regard to 2, mainly as a result of
the increase in the electrophilicity of the hydride
ligands. Although the net positive charge on 5 should
increase the acidity of the PH group, it should be noted
that the deprotonation of 5 occurs at the osmium atom
instead of at the diphenylphosphine group. A relation-
ship similar to the above-mentioned for 2 and 5 has been
previously found between the complexes RuH2(CO)-
(PHPh2)(PiPr3)2 and [RuH(CO)2(PHPh2)(PiPr3)2]+.17
293-213 K. A T1(min) value of 157 ms was found at
233 K. This value suggests a low contribution of non-
classical species to the structure of 6. In agreement with
this, a H-D coupling constant smaller than 1 Hz has
been found in the partially deuterated complex [OsH2D-
(PHPh2)2(PiPr3)2]BF4, which was prepared by addition
of 1.0 equiv of DBF4‚D2O to dichloromethane-d2 solu-
tions of 3. These values agree well with those pre-
viously reported for the complexes OsH3(Hbiim)(PiPr3)2,
(PiPr3)2H3Os(µ-biim)M(COD) (M ) Rh, Ir; H2biim ) 2,2′-
biimidazole),13f [OsH3(η4-diolefin)(PiPr3)2]+ (diolefin )
tetrafluorobenzobarrelene, 2,5-norbornadiene),13g OsH3-
{κ-N,κ-S-(2-Spy)}(PiPr3)2,
OsH3{κ-N,κ-O-OC(O)CH-
[CH(CH3)2]NH2}(PiPr3)2, OsH3{κ-N,κ-O-(2-Opy)}(PiPr3)2,
and [(PiPr3)2H3Os(µ-biim)M(TFB)]2 (M ) Rh, Ir).13h
However, they differ from those found for derivatives
of the types [OsH3(R2PCH2CH2PR2)2]+ 23 and [OsH3-
{P(OR)3}4]+,24 which have been described as hydride-
elongated dihydrogen species.
On the basis of the closely related OsH3(PPh3)4
+
,
OsH3(PMe3)4+, and OsH3(PEt3)4 complexes25 along
with the spectroscopical data, a distorted tetrahedral
P4 coordination environment with three of the four
trigonal P3 faces occupied by the hydride ligand is
suggested for this complex 6. Furthermore, as has been
previously reported in similar cases, the exchange
process that averages the 31P chemical shifts would be
the hydride migration process between the two PiPr3-
PiPr3-PHPh2 faces.26
+
The T1 values of the hydrogen nuclei of the OsH5 unit
of 5 were determined over the temperature range 293-
213 K. The T1(min) value (112 ms) was obtained at 233
K and is 44 ms higher than that reported for [OsH5(PMe2-
Ph)3]+ (68 ms).9a At first glance, this suggests that the
contribution of nonclassical species to the structure of
5 is smaller than that to the Caulton’s pentahydride.
The protonation of 2 with DBF4‚D2O affords [OsH4D-
(PHPh2)(PiPr3)2]BF4, which shows a H-D coupling
constant of 1.7 Hz. Exchange between the deuterium
and PH positions is not observed.
The IR spectrum of 3 in KBr shows a ν(P-H)
absorption at 2265 cm-1 and, in agreement with the
mutually cis disposition of the hydride ligands, a Os-H
The IR spectrum of 6 in KBr shows two ν(P-H)
absorptions at 2358 and 2326 cm-1 along with three
bands at 2068, 2058, and 1978 cm-1, corresponding to
the Os-H vibrations, and the absorption due to the
1
band at 1993 cm-1. The 31P{1H} and H NMR spectra
are temperature-invariant and consistent with the
structure shown for this compound in Scheme 1. Thus,
the 31P{1H} shows the X2YY′ part of a simplified
AA′CC′X2YY′ spin system (X ) PiPr3, Y ) Y′ ) PHPh2)
[BF4]- anion with Td symmetry centered at 1057 cm-1
.
The 31P{1H} and H NMR spectra in dichloromethane-
d2 are temperature-dependent. At room temperature the
31P{1H} NMR spectrum shows a broad signal at 28.2
ppm corresponding to the diphenylphosphine ligands
and at -23.6 ppm a triplet with a P-P coupling
constant of 7.3 Hz, due to the triisopropylphosphine
groups.
1
defined by the parameters δX ) 25.5 ppm, δY ) δY′
)
-17.8 ppm, J (PXPY) ) J (PXPY′) ) 14.0 Hz, J (PYPY′) )
1
1.7 Hz). In the H NMR spectrum the PH and hydride
resonances appear at 7.88 and -11.24 ppm as the parts
AA′ and CC′, respectively, of the AA′CC′X2YY′ spin
system. According to selective heteronuclear and homo-
nuclear spin decoupling spectra, the coupling constants
of the spin system are J (HAHA′) ) 0, J (HAHC) ≈
J (HA′HC′) < 2 Hz, J (HA′PY′) ) J (HAPY) ) 309.2 Hz,
J (HA′PY) ) J (HAPY′) ) 0, J (HAPX) ) J (HA′PX) ) 6.4 Hz,
J (HCHC′) ) 4.2 Hz, J (HCPX) ) J (HC′PX) ) 26.6 Hz,
J (HCPY′) ) J (HC′PY) ) 55.2 Hz, and J (HCPY) ) J (HC′PY′)
) -16.4 Hz. These values were confirmed by simulation
of the signal corresponding to the CC′ part of the spin
system (Figure 2).
Lowering the sample temperature leads to the broad-
ening of both resonances. At about -40 °C, decoales-
cence of the diphenylphosphine resonance is observed,
to give a triplet at 26 ppm and a multiplet at -26 ppm.
Upon -80 °C, decoalescence of the triisopropylphos-
1
phine resonance is not observed. In the H NMR spec-
trum at room temperature, the most noticeable reso-
nances are those due to the hydrides and PH hydrogen
atoms. The hydride resonance appears at -10.68 ppm
as a triplet with a H-P coupling constant of 16.5 Hz,
whereas the PH resonance is observed at 6.98 ppm as
a broad doublet with a H-P coupling constant of about
350 Hz. The 1H-COSY NMR spectrum shows weak
signals of crossing between the hydride and PH reso-
nances, suggesting that in this case, as for 5, there is a
small spin coupling between the hydride and PH nuclei.
Lowering the sample temperature leads to the broaden-
ing of the hydride and PH resonances, although decoa-
lescence is not observed at -80 °C.
2. Rea ction of [OsH5(P HP h 2)(P iP r 3)2]BF 4 w ith
Meth a n ol a n d Wa ter . We have recently shown that
the treatment of the cyclopentadienyl complex Os(η5-
C5H5)Cl(PHPh2)(PiPr3) with TlPF6 in the presence of
methanol or water leads to the complexes [Os(η5-C5H5)-
H2{P(OR)Ph2}(PiPr3)]PF6 (R ) Me, H).27 The formation
(23) Earl, K. A.; J ia, G.; Maltby, P. A.; Morris, R. H. J . Am. Chem.
Soc. 1991, 113, 3027.
(24) Amendola, P.; Antonintti, S.; Albertin, G.; Bordignon, E. Inorg.
Chem. 1990, 29, 318.
(25) (a) Siedle, A. R.; Newmark, R. A.; Pignolet, L. H. Inorg. Chem.
1986, 25, 3412. (b) Gusev, D. G.; Hu¨bener, R.; Burger, P.; Orama, O.;
Berke, H. J . Am. Chem. Soc. 1997, 119, 3716.
The T1 values of the hydrogen nuclei of the OsH3
unit of 6 were determined over the temperature range
(26) Gusev, D. G.; Berke, H. Chem. Ber. 1996, 129, 1143.

5302 Organometallics, Vol. 20, No. 25, 2001
Esteruelas et al.
Sch em e 2
1917 cm-1, corresponding to the Os-H vibrations, and
the absorption due to the [BF4]- anion with Td sym-
metry centered at 1049 cm-1. At room temperature, the
31P{1H} NMR spectrum in dichloromethane-d2 shows
at 102.4 ppm a triplet with a P-P coupling constant of
8.8 Hz and at 43.3 ppm a doublet with the same
coupling constant, which were assigned to methoxy-
diphenylphosphine and triisopropylphosphine, respec-
1
tively. In the H NMR spectrum at room temperature
the methoxy group of the methoxydiphenylphosphine
ligand gives rise to a doublet with a H-P coupling
constant of 12.0 Hz, at 3.23 ppm, whereas the resonance
corresponding to the hydride ligands appears at -8.20
ppm as a doublet of triplets with H-P coupling con-
stants of 9.2 and 5.7 Hz. The 31P{1H} and 1H NMR
spectra are temperature-dependent and show a behavior
with the temperature similar to that previously men-
tioned for 5.
T1 measurements for the hydride resonance gave a
value of 308 ms at 294 K, which decreases to 90 ms at
178 K. In this temperature range, a T1(min) value was
not found. However, it is clear that the value of this
parameter must be lower than that of 5 (112 ms). This
suggests that the contribution of nonclassical species
to the structure of 7 is higher than the contribution of
this type of species to the structure of 5. In agreement
with this, the H-D coupling constant in [OsH4D-
{P(OMe)Ph2}(PiPr3)2]+ (2.5 Hz) is higher than that in
[OsH4D(PHPh2)(PiPr3)2]+ (1.7 Hz). Complex [OsH4D-
{P(OMe)Ph2}(PiPr3)2]+ was formed by protonation of
OsH4{P(OMe)Ph2}(PiPr3)2 (9) with DBF4‚D2O in dichlo-
romethane-d2 as solvent. The tetrahydride was obtained
by deprotonation of 7 with Et3N. The higher contribu-
tion of nonclassical species in 7 could be related with a
higher π-acceptor power of the methoxydiphenylphos-
phine with regard to the diphenylphosphine.
F igu r e 2. 1H NMR spectra (300 MHz) in C6D6 in the high-
field region of OsH2(PHPh2)(PiPr3)2 (3) at room tempera-
ture: experimental UP and simulated DOWN.
of these species involves the release of the chloro ligand
followed by the intramolecular P-H oxidative addition
of diphenylphosphine in the unsaturated [Os(η5-C5H5)-
(PHPh2)(PiPr3)]+ metallic fragment. The oxidative ad-
dition of the P-H bond of secondary phosphines to
unsaturated transition metal complexes is a well-known
process of interest in connection with the catalytic phos-
phination and hydrophosphination of olefins.28 Once the
hydride-phosphido intermediate [Os(η5-C5H5)H(PPh2)-
(PiPr3)]+ is formed, the RO-H addition to the Os-
phosphino bond affords [Os(η5-C5H5)H2{P(OR)Ph2}-
(PiPr3)]+. Related additions of methanol and water
across WdP bonds have also been reported.29
The previously mentioned tendency of 5 to lose a
hydrogen molecule prompted us to carry out the reac-
tions of 5 with methanol and water, and as expected,
the complexes [OsH5{P(OMe)Ph2}(PiPr3)2]BF4 (7) and
[OsH5{P(OH)Ph2}(PiPr3)2]BF4 (8) were formed (Scheme
2). Complex 7 was obtained in 74% yield by stirring 5
in methanol for 6 days at room temperature, whereas
complex 8 was obtained in 69% yield by stirring a
dichloromethane solution of 5 with water for 5 days,
also, at room temperature.
The presence of a P(OH)Ph2 ligand in 8 is strongly
supported by the IR and 1H NMR spectra. The IR
spectrum in KBr shows ν(OH) and ν(P-O) bands at
3223 and 1034 cm-1, respectively. In addition, the
spectrum contains three ν(Os-H) absorptions at 2132,
2084, and 2015 cm-1 and the absorption due to the
In the IR spectrum of 7 in KBr, the most noticeable
feature is the absence of any ν(P-H) absorption and the
presence at 1101 cm-1 of a P-O vibration. In addition
the spectrum contains three bands at 2085, 2012, and
[BF4]- anion with Td symmetry centered at 1099 cm-1
.
(27) Esteruelas, M. A.; Lo´pez, A. M.; Tolosa, J . I.; Vela, N. Orga-
nometallics 2000, 19, 4650.
1
At room temperature, the H NMR spectrum in dichlo-
romethane-d2 shows at 1.63 ppm a broad singlet corre-
sponding to the OH proton and at -8.27 ppm a doublet
of triplets with H-P coupling constants of 9.0 and 4.8
Hz, due to the hydride ligands. At this temperature, the
31P{1H} NMR spectrum contains at 86.6 (P(OH)Ph2)
ppm a triplet with a P-P coupling constant of 8.9 Hz
and at 42.7 (PiPr3) ppm a doublet with the same P-P
coupling constant. The 1H and 31P{1H} NMR spectra are
temperature-dependent and show a behavior with the
temperature similar to those of 5 and 7.
(28) See for example: (a) Wicht, D. K.; Kourkine, I. V.; Lew, B. M.;
Nthenge, J . M.; Glueck, D. S. J . Am. Chem. Soc. 1997, 119, 5039. (b)
Kourkine, I. V.; Sargent, M. D:; Glueck, D. S. Organometallics 1998,
17, 125. (c) Wicht, D. K.; Paisner, S. N.; Lew, B. M.; Glueck, D. S.;
Yap, G. P. A.; Liable-Sands, L. M.; Rheingold, A. L.; Haar, C. M.; Nolan,
S. P. Organometallics 1998, 17, 652. (d) Wicht, D. K.; Kovacile, I.;
Glueck, D. S.; Liable-Sands, L. M.; Incarvito, C. D.; Reingold, A. L.
Organometallics 1999, 18, 5141. (e) Wicht, D. K.; Kourkine, I. V.;
Kovacik, I.; Glueck, D. S.; Concolino, T. E.; Glenn, P. A.; Yap, G. P. A.;
Incarvito, C. D.; Rheingold, A. L. Organometallics 1999, 18, 5381.
(29) (a) J o¨rg, K.; Malisch, W.; Reich, W.; Meyer, A.; Schubert, U.
Angew. Chem., Int. Ed. Engl. 1986, 25, 92. (b) Malisch, W.; Hirth, U.-
A.; Gru¨n, K.; Schmeusser, M.; Fey, O.; Weis, U. Angew. Chem., Int.
Ed. Engl. 1995, 34, 2500.

Mixed-Phosphine Osmium Polyhydrides
Organometallics, Vol. 20, No. 25, 2001 5303
Sch em e 3. Liga n d Nu m ber in g for [Os′′H5′′(P F 3)+]
Ta ble 2. Ma in Geom etr ica l P a r a m eter s of
P en ta h yd r id e (10a ), Dih yd r ogen Tr ih yd r id e (10b
a n d 10c), a n d Bis-d ih yd r ogen Hyd r id e (10d )
[OsH5(P H3)+] Mod el Com p lexes (d ista n ces in Å,
a n gles in d eg)
Com p lexes
10a
10b
10c
10d
expta
Os-P(2)
Os-P(3)
Os-P(4)
Os-H(5)
Os-H(6)
Os-H(7)
Os-H(8)
Os-H(9)
H(6)-H(7)
H(8)-H(9)
H(5)-H(8)
2.454
2.392
2.392
1.672
1.645
1.645
1.624
1.626
1.598
1.636
1.904
2.458
2.384
2.384
1.633
1.800
1.800
1.606
1.633
0.863
1.605
1.626
95.25
95.24
2.392
2.398
2.398
1.676
1.650
1.651
1.706
1.698
1.634
0.955
2.188
93.56
93.57
2.381
2.384
2.384
1.639
1.811
1.811
1.706
1.729
0.855
0.921
2.002
2.389(2)
2.387(2)
2.387(2)
P(2)-Os-P(3) 96.32
92.07 105.72(6)
92.07 94.80(6)
P(2)-Os-P(4) 95.32
P(3)-Os-P(4) 154.46 165.32 152.77 165.72 149.43(6)
a
X-ray diffraction data for the OsH5(PiPr3)2(PHPh2)+ complex.
stable isomer in the previous theoretical study.10b It
resembles the neutron diffraction structure of [OsH5-
(PMe2Ph)3]+ (11). If two adjacent H atoms approach
each other, structures 10b and 10c are attained. The
simultaneous shortening of two H-H distances leads
to the 10d structure.
A B3LYP optimization has been performed for the
four structures. All of them are actual minima in the
B3LYP potential energy surface. Their main geometrical
parameters are listed in Table 2. For the ligand num-
bering, see Scheme 3.
The T1 values of the hydrogen nuclei of the OsH5 unit
of 8 were determined over the temperature range 293-
193 K. In this case, a T1(min) value of 106 ms was found
as 213 K. This value agrees well with that found in 5.
The tetrahydride complex 9 was obtained as a brown
oil in nearly quantitative yield and was characterized
1
by MS and 31P{1H} and H NMR spectroscopy. At room
temperature, the 31P{1H} NMR spectrum shows at 121.2
(P(OMe)Ph2) ppm a triplet with a P-P coupling con-
stant of 11.5 Hz and at 41.6 (PiPr3) ppm a doublet with
the same coupling constant. At this temperature, the
B3LYP distances and angles in the pentahydride are
very similar to the MP2-optimized geometry of this
species.10b The shortening of the H(6)-H(7) distance
leads to a dihydrogen heptacoordinate complex (10b)
which can be described as a distorted pentagonal
bipyramid with P(3) and P(4) occupying its axial posi-
tions. 10c is also a pentagonal bipyramid, but with a
phosphine (P(2)) and a dihydrogen (H(8)-H(9)) as axial
ligands. The bis-dihydrogen (10d ) adopts a distorted
octahedral geometry. The P(3)-Os-P(4) angle is very
sensitive to the decrease of the coordination number. It
is to be expected that P-Os-P bond angles will be
affected considerably by steric effects, but this cannot
be accounted for by calculations in the model system
10 with PH3 ligands. The P(3)-Os-P(4) angle in 10a
(154.5°) is slightly different from that experimentally
determined in [OsH5(PMe2Ph)3]+(11) (146.6°) and 5
(149.43(6)°). As is to be expected, although C1 geometry
optimizations were carried out, the asymmetry in the
phosphine arrangement caused by the pattern of asym-
metrical substituents in 5 is not reproduced. Thus,
P(2)-Os-P(3) and P(2)-Os-P(4) optimized angles in
the model system have nearly the same value. As a
consequence, the P(2)-Os-P(3) angle is far away from
that experimentally determined, indicating clearly that
both the asymmetry and the P(2)-Os-P(3) angle open-
ing in the real system are caused by the substitution
pattern of the actual phosphines. Despite all discrep-
ancies between model and experiment, the P(3)-Os-
P(4) angle suggests that from a geometrical point of view
the pentahydride form is the one that fits the experi-
mental data in a better way. This geometrical compari-
son will be later recalled with a theoretical discussion
on the real systems.
1
most noticeable resonance in the H NMR spectrum is
a doublet of triplets with H-P coupling constants of 15.0
and 7.2 Hz, which appears at -10.05 ppm and corre-
1
sponds to the four hydride ligands. The 31P{1H} and H
NMR spectra show a behavior with the temperature
similar to that of 2.
3. Th eor etica l Stu d y of [OsH 5P 3]+ Com p lexes.
[OsH5(PR3)3]+ complexes have received a lot of attention
in the recent past. The possibility of isomerism between
a wide array of classical and nonclassical forms, along
with the availability of an unambiguous structural
+
characterization of the OsH5(PMePh2)3 complex by
neutron diffraction techniques, has made them an
appealing target to theoretical chemistry. Thus, we have
chosen them within the whole series of reported OsHn-
(PR3)m (n ) 2, 3, 4, 5; m ) 2, 3) complexes in order to
get a deeper insight into the structure and dynamic
processes involved in such polyhydride systems. In the
next section, there follows a high-level ab initio study,
supplemented by QM/MM IMOMM calculations.
i. [OsH5(P H3)3]+ Mod el System . On the basis of a
previous theoretical study and the available X-ray and
neutron diffraction data, four isomeric forms have been
initially considered for the [OsH5(PH3)3]+ model complex
10: eight-coordinate pentahydride (10a ), seven-coordi-
nate dihydrogen/trihydride (10b and 10c), and six-
coordinate bisdihydrogen/hydride (10d ) (Scheme 3). 10a
corresponds to the dodecahedral structure with the
three phosphine ligands in B sites and one hydride
occupying the last B site available (H9 according to our
labeling, see Scheme 3), which was found to be the most

5304 Organometallics, Vol. 20, No. 25, 2001
Esteruelas et al.
Ta ble 3. Rela tive En er gies a t Differ en t Levels of
Ca lcu la tion for P en ta h yd r id e (10a ), Dih yd r ogen
Tr ih yd r id e (10b a n d 10c), a n d Bis-d ih yd r ogen
+
(10d ) Str u ctu r es of OsH5(P H3)3 Mod el Com p lexes
Ta k en fr om B3LYP Geom etr y Op tim iza tion s
B3LYP
HF
MP2 MP4SDQ CCSD CCSD(T)
10a
10b
10c
10d
0.0
-3.4
1.3
0.0
0.0
6.3
9.7
0.0
1.6
6.3
0.0
1.0
5.6
6.0
0.0
1.9
6.2
7.5
-10.2
-5.1
-16.2
-3.6
14.4
15.3
Relative energies of the four isomers of the [OsH5-
(PH3)3]+ system are shown in Table 3. The energetic
order at the B3LYP//B3LYP level (both geometry opti-
mization and energy at the B3LYP level of therory)
agrees neither with previous calculations for this sys-
tem10b nor with the neutron diffraction assignment of a
pentahydride structure to [OsH5(PMe2Ph)3]+ (11).9b
Surprisingly the most stable form is not the pentahy-
dride, but the bis-dihydrogen isomer 10d , which lies 3.6
kcal/mol below the pentahydride. Even the dihydrogen/
trihydride 10b is 3.4 kcal/mol more stable than the
pentahydride. Single-point energy only calculations
have been performed at different levels using the
B3LYP optimized geometries (Table 3).
Data in Table 3 suggest that B3LYP overestimates
the stability of dihydrogen species, whereas MP2 un-
derestimates it. This conclusion has already been pointed
out by other authors.30 The bis-dihydrogen species is the
most affected by the methodology chosen. Relative
energies at the higher computational levels (MP4SDQ,
CCSD, and CCSD(T)) are very similar, indicating that
the quantum mechanical description is practically con-
verged. CCSD(T) calculations reverse the B3LYP order-
ing, giving the pentahydride as the most stable isomer,
in better agreement with neutron diffraction data of
[OsH5(PMe2Ph)3]+ (11). From now on we will present
only CCSD(T) energies.
Our CCSD(T)//B3LYP calculations (CCSD(T) single-
point energy calculations on B3LYP optimized geom-
etries) give the pentahydride and the dihydrogen 10b
very close in energy. From a structural, static, point of
view the existence of both is possible. Can their inter-
conversion occur easily? This question is related to the
dynamic behavior of the system and can be answered
looking at the potential energy surface (PES) for the
interconversion. A bidimensional potential energy sur-
face has been calculated at the CCSD(T)//B3LYP level
in order to estimate the energetic cost of the pentahy-
dride/dihydrogen trihydride/bis-dihydrogenhydride in-
terconversion processes. This bidimensional PES was
built by varying the H(6)-H(7) and H(8)-H(9) dis-
tances between 0.8 and 1.8 Å. At each fixed value of
the H(6)-H(7) and H(8)-H(9) distances all the geo-
metrical parameters were optimized. The obtained
potential energy surface is depicted in Figure 3.
F igu r e 3. CCSD(T)//B3LYP potential energy surface for
H6-H7 and H8-H9 elongation in OsH5(PH3)+ model
complex.
ence between both isomers is kept along the whole inter-
conversion process. The energy of the [OsH5(PH3)3]+
system is practically independent of the H(6)-H(7)
separation between 0.8 and 1.8 Å. The pentahydride f
dihydrogen/trihydride isomerization takes place with an
energy cost of no more than 2 kcal/mol. The energy
barrier for the breaking/formation of such a strong bond
is lower than or the same order as a rotational barrier.
This behavior is very similar to that found in theoretical
studies of elongated dihydrogen complexes: two hydro-
gen atoms moving almost freely in a large region within
the coordination sphere of the metal.6 This result opens
the door to a new explanation for the system under
study: a certain extent of delocalization concerning the
H(6)-H(7) moiety could explain the hybrid behavior
between classical and nonclassical forms of this kind of
complex. In some way 10 is simultaneously a polyhy-
dride and a dihydrogen hydride.
In the next section there follows an analysis of the
real systems (5, 7, and [OsH5(PMe2Ph)3]+ (11)) through
IMOMM calculations in order to establish how the
inclusion of the real phosphine substituents modifies the
structural parameters and affects the relative stabilities
of the four studied isomers (a -d ).
ii. [OsH5P 3]+ Rea l System s. Geometry optimiza-
tions with the IMOMM (B3LYP:MM3) methodology
were carried out on the pentahydride (a ), dihydrogen
trihydride (b and c), and bis-dihydrogen hydride (d )
structures of the reported [OsH5(PHPh2)(PiPr3)2]+ (5)
and [OsH5{P(OMe)Ph2}(PiPr3)2]+ (7) complexes, as well
as for the neutron-characterized [OsH5(PMePh2)3]+ (11).
A first and very remarkable fact is that not all of the
four minima found in the model system can be found in
the real systems: complexes 5c, 7c, and 11c revert to
5b, 7b, and 11b, respectively. Thus the less stable
dihydrogen isomer is not a minimum in the real
systems. Similarly, the existence of a bis-dihydrogen
form, although reliable in energy terms for the model
system, depends strongly on the phosphine substituents.
When the less bulky phosphines are present, it does
exist (11d ). It could be thought that bulky phospines
At first glance, it should be noted that the less stable
isomers (10c and 10d ) found as minima in the B3LYP
potential energy surface are no longer minima in the
CCSD(T) potential energy surface. The most striking
finding is the extreme flatness of the surface along the
H(6)-H(7) axis, in the direction corresponding to the
10a f 10b rearrangement. The very low energy differ-
(30) Bytheway, I.; Backsay, G. B.; Hush, N. S. J . Phys. Chem. 1996,
100, 6023. (b) Clot, E.; Eisenstein, O. J . Phys. Chem. A 1998, 102, 3592.

Mixed-Phosphine Osmium Polyhydrides
Organometallics, Vol. 20, No. 25, 2001 5305
Ta ble 4. IMOMM (B3LYP :MM3) Op tim ized Dista n ces a n d An gles of [OsH5(P Me2P h )3]+ (11a , 11b, a n d 11d ),
[OsH5(P iP r 3)2(P HP h 2)]+ (5a a n d 5b), a n d [OsH5(P iP r 3)2(P (OMe)P h 2)]+ (7a a n d 7b) Com p lexes (d ista n ces in Å,
a n gles in d eg)
11a a
11b
11d
5a
5b
7a
7b
expta
Os-P(2)
2.475
2.414
2.401
1.667
1.641
1.642
1.622
1.623
1.576
1.605
1.656
98.62
97.80
151.92
2.474
2.398
2.396
1.630
1.789
1.790
1.609
1.629
0.872
1.605
1.656
97.30
95.53
163.32
2.412
2.399
2.399
1.633
1.806
1.807
1.692
1.713
0.859
0.943
1.986
94.33
93.98
164.82
2.479
2.448
2.437
1.664
1.634
1.639
1.617
1.615
1.576
1.650
1.919
102.27
95.40
150.81
2.495
2.435
2.429
1.629
1.774
1.777
1.602
1.618
0.881
1.589
1.675
101.23
93.72
160.21
2.457
2.460
2.439
1.660
1.635
1.636
1.619
1.617
1.605
1.684
1.903
103.86
96.78
149.84
2.478
2.452
2.432
1.627
1.771
1.772
1.602
1.620
0.885
1.638
1.641
103.22
95.79
157.95
2.389(2)
2.387(2)
2.387(2)
Os-P(3)
Os-P(4)
Os-H(5)
Os-H(6)
Os-H(7)
Os-H(8)
Os-H(9)
H(6)-H(7)
H(8)-H(9)
H(5)-H(8)
P(2)-Os-P(3)
P(2)-Os-P(4)
P(3)-Os-P(4)
105.72(6)
94.80(6)
149.43(6)
a
X-ray diffraction data for the OsH5(PiPr3)2(PHPh2)+ complex.
F igu r e 5. IMOMM(B3LYP:MM3) optimized structures of
OsH5(PiPr3)2(PHPh2)+ complexes 5a and 5b.
is taken into account. Moreover, the pattern for the
P(2)-Os-P(3) and P(2)-Os-P(4) is also reproduced, the
optimized value of the former being slightly larger than
the latter and thus leading to the distorted T-shape
reported in the X-ray diffraction structure. The reli-
ability of our IMOMM geometries is also confirmed by
the optimization of OsH5(PMe2Ph)3+ (11), as this allows
us to consider a different case where there is a sym-
metric phosphine substituents pattern. For this com-
plex, P(2)-Os-P(3) and P(2)-Os-P(4) angles are rather
similar, in agreement with neutron diffraction data
available for complex 11. Moreover, being that these
angles concerning the OsP3 backbone are nearly the
same as that found for the OsH5(PH3)+ model complex,
we should conclude that PMe2Ph causes a lesser steric
hindrance in the system than phosphines used in
complexes 5 and 7, a quite logical result taking into
account the smaller cone angle for PMe2Ph.31 Another
interesting geometrical feature of IMOMM calculations
is related to the OsH5 moiety. Although the trends in
Os-H distances are maintained when 5 and 7 are
compared to model complex 10, in general IMOMM
F igu r e 4. IMOMM(B3LYP:MM3) optimized structures of
OsH5(PMe2Ph)3 complexes 11a , 11b, and 11d .
+
should favor the presence of two dihydrogen ligands,
occupying less space in the coordination sphere than
four hydride ligands. But this is not the case due to the
octahedral geometry of the bis-dihydrogen isomer. The
P-Os-P angles imposed by an octahedral arrangement
(around 90° and 180°) are hardly possible when bulky
phosphines (as PiPr3) are present.
The main geometrical parameters of the minima are
listed in Table 4, whereas optimized geometries for
systems 11, 5, and 7 are depicted in Figures 4, 5, and
6, respectively.
The main structural modifications caused by changing
the phosphines are found in the P-Os-P angles.
Comparisons with the experimental structures are
promising: with the inclusion of real phosphine sub-
stituents, calculations nicely take account of the asym-
metry in the phosphine backbone of the complexes. The
P(3)-Os-P(4) angle fits almost exactly the X-ray-
determined angle in 5 if the pentahydride structure 5a
(31) Tolman, C. A. Chem. Rev. 1977, 77, 313.

5306 Organometallics, Vol. 20, No. 25, 2001
Esteruelas et al.
number complexes arising from repulsions between
bulky phosphines. This is nicely shown when complexes
5, 7, and 11 are considered: for the latter, with the
comparatively small P(Me2Ph)3 phosphines, a lesser
distortion of the ab initio part of the system is to be
expected. As a consequence, the same relative energy
difference than for calculations in the model system
+
OsH5(PH3)5 is found between 11a and 11b (1.9 kcal/
mol). For complexes 5 and 7, this difference is somewhat
higher (2.6 and 3.0 kcal/mol, respectively) but still small
enough to keep valid the definition of the [OsH5(P3)3]+
systems as having two delocalized hydrogen atoms when
the real phosphines are included in the calculations.
Con clu d in g Rem a r k s
The investigations on the reactivity of the hexahy-
dride complex OsH6(PiPr3)2 have been focused on the
tendency of this compound to lose molecular hydrogen.
However, the dominant point of view has been that the
dissociation energy of molecular hydrogen was high.
Thus, because of the saturated character of OsH6-
(PiPr3)2, it has been assumed that this complex had a
relative kinetic inertia toward the hydrogen substitution
and that its activation should involve the previous
acidolysis33 or alternatively the use of ligands containing
relatively acidic hydrogen atoms.13f,h Recently we have
shown that the complex OsH6(PiPr3)2 also activates
C-H and C-F bonds of aromatic ketones. On the basis
of a low energy for the dissociation of molecular hydro-
gen from OsH6(PiPr3)2 we have proposed that the first
step of these activation processes involves the loss of
molecular hydrogen to give the unsaturated species
OsH2(η2-H2)(PiPr3)2.13i This study reports the first ex-
perimental observation of the low energy necessary to
dissociate molecular hydrogen from OsH6(PiPr3)2.
The dissociation energy of a hydrogen molecule from
OsH6(PiPr3)2 is similar to that reported for the dihy-
dride-dihydrogen OsH2(η2-H2)(CO)(PiPr3)2,20 and as the
former compound, OsH6(PiPr3)2 reacts with diphenylphos-
phine to give molecular hydrogen and the tetrahydride
OsH4(PHPh2)(PiPr3)2. In contrast to the hexahydride,
this tetrahydride has a high dissociation energy of
molecular hydrogen, which is lower than or of the same
order as the energy necessary to break an Os-P bond.
The reaction of OsH4(PHPh2)(PiPr3)2 with diphenylphos-
phine requires drastic conditions and does not give rise
to the selective substitution of molecular hydrogen, but
the replacement of this small molecule is accompanied
with the substitution of a triisopropylphosphine ligand.
The selective replacement of molecular hydrogen in
OsH4(PHPh2)(PiPr3)2 requires its previous acidolysis,
which leads to [OsH5(PHPh2)(PiPr3)]+.
F igu r e 6. IMOMM(B3LYP:MM3) optimized structures of
OsH5(PiPr3)2(P(OMe)Ph2)+ complexes 7a and 7b.
Ta ble 5. Rela tive En er gies of [OsH5(P Me2P h )3]+
(11a , 11b, a n d 11d ), [OsH5(P iP r 3)2(P HP h 2)]+ (5a a n d
5b), a n d [OsH5(P iP r 3)2(P (OMe)P h 2)]+ (7a a n d 7b)
Com p lexes (en er gies in k ca l/m ol)
11a
11b
11d
5a
5b
7a
7b
B3YLP
CCSD(T)
MM3
B3LYP+MM3
CCSD(T)+MM3
0.0
0.0
0.0
0.0
0.0
-3.8 -3.6 0.0 -2.6 0.0 -2.3
1.4
0.5
7.2 0.0
1.2 0.0
2.4 0.0
0.2 0.0
2.8
0.2
-3.3 -2.4 0.0 -2.4 0.0 -2.1
1.9 8.4 0.0 2.6 0.0 3.0
distances are slightly shorter. This may not be impor-
tant, but one has to recall that if a dihydrogen form is
accessible, the ease of H2 elimination is related to
Os-H2 bond strength. Thus, when we compare 5b and
7b with their model counterpart 10b, there is a small
but noticeable difference in the Os-H(6)-H(7) dihydro-
gen unit: as in the real systems Os-H distances are
shorter and H-H distances are longer, and we might
expect that the actual substitution pattern would render
H2 elimination less favorable. Finally, a word about
Os-P distances is probably to the point: if it is true
that the calculated distances are in all cases longer than
the experimental ones, this seems to be a common
feature of the IMOMM method; this fact should not
affect the validity of the results, as many prior examples
have proved.32
Relative energies of the different isomers are shown
in Table 5. The inclusion of the real phosphine substit-
uents does not change significantly the patterns found
in the ab initio calculations on the model system, since
the steric energy differences are small in comparing the
pentahydride and dihydrogen trihydride isomers (0.2
kcal/mol in favor of pentahydride for complexes 5 and
7). For the three compounds the energy of the pentahy-
dride structure is the lowest one, but the dihydrogen
forms b are very close in energy. However, the presence
of bulky phosphines also induces a distortion in the ab
initio part of the system. Thus, it remains clear that
increasing the size of phosphine substituents plays a
role against the stability of dihydrogen forms, because
of the higher steric congestion of lower coordination
This compound, containing five hydrogen atoms bonded
to the osmium atom, shows a behavior similar to that
of the hexahydride. The reaction with diphenylphos-
phine produces the selective loss of molecular hydrogen
and the formation of [OsH3(PHPh2)2(PiPr3)2]+. Moreover,
the treatment of [OsH5(PHPh2)(PiPr3)2]+ with methanol
and water yields [OsH5{P(OR)Ph2}(PiPr3)2]+ (R ) CH3,
H). According to a recent study carried out by our
(32) (a) Maseras, F. Top. Organomet. Chem. 1999, 4, 165. (b)
Maseras, F. Chem. Commun. 2000, 1821.
(33) Smith, K. T.; Tilset, M.; Kuhlman, R.; Caulton, K. G. J . Am.
Chem. Soc. 1995, 117, 9473.

Mixed-Phosphine Osmium Polyhydrides
Organometallics, Vol. 20, No. 25, 2001 5307
group,27 the formation of the latter compounds involves
not only the loss of molecular hydrogen but also the
intramolecular P-H oxidative addition in unsaturated
[OsH3(PHPh2)(PiPr3)2]+ intermediates.
PHPh2), 7.06 (m, 4H; Hmeta-Ph), 6.98 (m, 2H; Hpara-Ph), 1.67 (m,
6H; PCH(CH3)2), 1.08 (dvt, 36 H, N ) 12.6 Hz, J (HH) ) 5.9
Hz; PCH(CH3)2), -10.19 (tdd, 4H, J (HP) ) 15.0 Hz, J (HP) )
3.0 Hz, J (HH) ) 3.0 Hz; OsH). 31P{1H} NMR (121.4 MHz,
toluene-d8, 293 K): δ 39.2 (d, J (PP) ) 17.1 Hz; PiPr3), -4.6 (t,
J (PP) ) 17.1 Hz; PHPh2). T1 (ms, “OsH4”, 300 MHz, toluene-
d8): 383 (293 K), 249 (273 K), 168 (253 K), 137 (223 K), 183
(213 K).
Geometry optimizations with the IMOMM(B3LYP:
MM3) methodology on the species [OsH5(PHPh2)-
(PiPr3)2]+ and [OsH5{P(OMe)Ph2}(PiPr3)2]+ and that
previously reported by Caulton, [OsH5(PMe2Ph)3]+,9
indicate, in the three cases, that a dodecahedral pen-
tahydride structure, similar to that found by neutron
diffraction for [OsH5(PMe2Ph)3]+, and a pentagonal
bipyramid trihydride-dihydrogen structure, with the
triisopropylphosphine ligands (or dimethylphenylphos-
phine, in the case of the Caulton’s complex) in the axial
positions are minima in the potential energy surface.
Not only are these structures very close in energy, but
furthermore the energy of the systems is practically
independent of the separation between the hydrogen
atoms involved in the nonclassical interaction, in the
range 0.8-1.8 Å. The energy barrier for the breaking/
formation of the hydrogen bond is lower than or the
same order as a rotation barrier. In this context, the
description of these OsH5 species as pentahydride or
trihydride-dihydrogen compounds loses significance; it
appears to be more appropriate to describe them as
species containing two hydrogen atoms moving freely
in a wide region of the coordination sphere of the
osmium atom.
P r ep a r a tion of OsH2(P HP h 2)2(P iP r 3)2 (3). A solution of
[OsH3(PiPr3)2(PHPh2)2]BF4 (6) (391.0 mg, 0.40 mmol) in 10 mL
of dichloromethane was treated with triethylamine (56 µL, 0.40
mmol) and stirred for 1 day to give a dark brown solution.
The solvent was removed in vacuo. Toluene (10 mL) was
added, and the suspension was filtered to eliminate [HNEt3]-
BF4. Solvent was evaporated, and the residue was washed
with methanol to afford a white solid. Yield: 132 mg (37%).
Anal. Calcd for C42H66OsP4: C, 57.00; H, 7.52. Found: C, 56.73;
H, 7.98. IR (KBr, cm-1): ν(PH) 2265(m); ν(OsH) 1993(m);
ν(CdC)aromatic 1587(w), 1574(w).
1H NMR (300 MHz, C6D6, 293 K): δ 7.88 (AA′ part of an
AA′CC′X2YY′ spin system, 2H, J (HAHC) ≈ J (HA′HC′) < 2 Hz,
J (HA′PY′) ) J (HAPY) ) 309.0 Hz, J (HAPX) ) J (HA′PX) ) 6.4 Hz;
PHPh2), 7.73 (m, 8H; Hortho-Ph), 7.10 (m, 8H; Hmeta-Ph), 7.00
(m, 4H, Hpara-Ph), 1.91 (m, 6H; PCH(CH3)2), 1.10 (dvt, 36H, N
) 12.2 Hz, J (HH) ) 5.3 Hz; PCH(CH3)2), -11.24 (CC′ part of
an AA′CC′X2YY′ spin system, 2H, J (HAHC) ≈ J (HA′HC′) < 2
Hz, J (HCHC′) ) 4.2 Hz, J (HCPX) ) J (HC′PX) ) 26.6 Hz, J (HCPY′)
) J (HC′PY) ) 55.2 Hz, J (HCPY) ) J (HC′PY′) ) -16.4 Hz; OsH).
31P{1H} NMR (121.4 MHz, C6D6, 293 K): δ 25.5 (X part of a
simplified AA′CC′X2YY′ spin system, 2P, J (PXPY) ) J (PXPY′)
) 14.0 Hz; PiPr3), -17.8 (YY′ part of a simplified AA′CC′X2-
YY′ spin system, 2P, J (PXPY) ) J (PXPY′) ) 14.0 Hz, J (PYPY′) )
1.7 Hz; PHPh2).
Exp er im en ta l Section
1
1
P h ysica l Mea su r em en ts. H, H{31P}, and 31P{1H} NMR
spectra were recorded on either a Varian Gemini 2000, a
Varian Unity 300, or a Bruker 300 ARX spectrometer. The
probe temperature of the NMR spectrometers was calibrated
at each temperature against a methanol standard. For the T1
measurements the 180° pulses were calibrated at each tem-
perature. The conventional inversion-recovery method (180-
τ-90) was used to determine T1. Chemical shifts are expressed
in ppm upfield from Me4Si (1H) or 85% H3PO4 (31P). Coupling
constants (J and N (N ) J (HP) + J (HP′))) are given in hertz.
IR data were recorded on a Perkin-Elmer 883 or Nicolet 550
spectrophotometer. Elemental analyses were carried out with
a Perkin-Elmer 2400 CHNS/O microanalyzer. Mass spectra
analyses were performed with a VG Autospec instrument. In
FAB+ mode, ions were produced with the standard C+ gun at
ca. 30 kV, and 3-nitrobenzyl alcohol (NBA) was used as the
matrix. The simulation of the hydride signal of 3 was carried
out with the gNMR v 3.6 for Macintosh program.34
P r ep a r a tion of OsH2(P HP h 2)3(P iP r 3) (4). A solution of
2 (208.0 mg, 0.30 mmol) in 13 mL of toluene was treated with
diphenylphosphine (50 µL, 0.30 mmol) and heated at 80 °C.
After 3 days, the solvent was removed in vacuo. The residue
was washed with methanol to afford a white solid. Yield: 127
mg (47%). Anal. Calcd for C45H56OsP4: C, 59.30; H, 6.20.
Found: C, 59.00; H, 6.51. IR (KBr, cm-1): ν(PH) 2374(m),
2361(m), 2348(m); ν(OsH) 2000(m), 1927(m); ν(CdC)aromatic
1586(w), 1573(w).
Syn th esis. All reactions were carried out under an argon
atmosphere using standard Schlenk techniques. Solvents were
dried using appropriate drying agents and freshly distilled
under argon before use. The starting complex OsH6(PiPr3)2 was
prepared by the published method.13a
P r ep a r a tion of OsH4(P HP h 2)(P iP r 3)2 (2). A colorless
solution of OsH6(PiPr3)2 (1) (892.0 mg, 1.73 mmol) in 10 mL of
toluene was treated with diphenylphosphine (296 µL, 1.73
mmol) and heated at 80 °C for 1 h to obtain a yellow solution.
The solvent was removed in vacuo. The residue was washed
with methanol to afford a white solid. Yield: 859 mg (71%).
Anal. Calcd for C30H57OsP3: C, 51.41; H, 8.20. Found: C, 51.05;
H, 7.76. IR (KBr, cm-1): ν(PH) 2292(m); ν(OsH) 2052(m),
2013(m), 1886(s); ν(CdC)aromatic 1587(w), 1575(w). 1H NMR (300
MHz, toluene-d8, 353 K): δ 7.75 (m, 4H; Hortho-Ph), 7.45 (dtd,
1H, J (HP) ) 346.0 Hz, J (HP) ) 8.4 Hz, J (HH) ) 3.0 Hz;
1H NMR (300 MHz, C6D6, 293 K): δ 7.80-6.90 (m, 30H; Ph),
7.00-5.91 (AA′B part of an AA′BCC′XYZZ′ spin system, 3H;
PHPh2), 1.95 (m, 3H; PCH(CH3)2), 1.12 (dd, 18H, J (HP) ) 12.3
Hz, J (HH) ) 6.9 Hz; PCH(CH3)2), -10.00 (CC′ part of an
AA′BCC′XYZZ′ spin system, 2H; OsH). 31P{1H} (121.4 MHz,
C6D6, 293 K): δ 33.1 (X part of a simplified AA′BCC′XYZZ′
spin system, 1P, J (PXPY) ) 209.3 Hz, J (PXPZ) ) J (PXPZ′) ) 18.4
Hz; PiPr3), 13.9 (Y part of a simplified AA′BCC′XYZZ′ spin
system, 1P, J (PYPX) ) 209.3 Hz, J (PYPZ) ) J (PYPZ′) ) 11.5 Hz;
PHPh2), -8.0 (ZZ′ part of a simplified AA′BCC′XYZZ′ spin
system, 2P, J (PZPY) ) J (PZ′PY) ) 18.4 Hz, J (PZPY) ) J (PZ′PY)
(34) gNMR v. 3.6; Cherwell Scientific Publishing Limited: The
Magdalen Center Oxford Science Park, Oxford OX446A, U.K., 1992-
1995.

5308 Organometallics, Vol. 20, No. 25, 2001
Esteruelas et al.
) 11.5 Hz; PHPh2). MS(FAB+): m/z 910 (M+ - 1H), 722 (M+
- 2H - PHPh2).
Ta ble 6. Cr ysta l Da ta a n d Da ta Collection a n d
Refin em en t for [OsH5(P HP h 2)(P iP r 3)2]BF 4 (5)
P r ep a r a tion of [OsH5(P HP h 2)(P iP r 3)2]BF 4 (5). A solution
of 2 (342.0 mg, 0.49 mmol) in 10 mL of diethyl ether was
treated with HBF4‚OEt2 (68 µL, 0.49 mmol), and a white solid
was formed. After 40 min, the solid was filtered off, washed
with diethyl ether, and dried in vacuo. Yield: 373 mg (97%).
Anal. Calcd for C30H58OsP3BF4: C, 45.68; H, 7.41. Found: C,
45.69; H, 7.02. IR (KBr, cm-1): ν(PH) 2362(m); ν(OsH) 2096-
(m), 2086(m), 1984(m); ν(CdC)aromatic 1576(m); ν(BF4-) 1058-
formula
mol wt
C30H58BF4OsP3
788.68
color, habit
space group
a, Å
yellow, prismatic
monoclinic, P21/n
8.463(1)
b, Å
c, Å
25.996(2)
16.114(2)
98.072(8)
3509.2(7)
4
â, deg
V, Å3
1
(vs). H NMR (300 MHz, CD2Cl2, 293 K): δ 7.60-7.50 (m, 10
Z
D(calcd), g cm-3
µ, mm-1
1.493
3.81
H; Ph), 6.97 (dt, 1H, J (HP) ) 373.7 Hz, J (HP) ) 3.9 Hz;
PHPh2), 1.92 (m, 6H; PCH(CH3)2), 1.10 (dvt, 36H, N ) 15.0
Hz, J (HH) ) 8.1 Hz; PCH(CH3)2), -8.26 (t, 5H, J (HP) ) 8,2
Hz; OsH). 31P{1H} NMR (121.4 MHz, CD2Cl2, 293 K): δ 42.2
(d, J (PP) ) 10.1 Hz; PiPr3), -24.4 (t, J (PP) ) 10.1 Hz; PHPh2).
MS(FAB+): m/z 695 (M+ - 6H). T1 (ms, “OsH5”, 300 MHz,
CD2Cl2): 383 (293 K), 260 (273 K), 175 (253 K), 112 (233 K),
124 (213 K).
scan type
θ range, deg
temp, K
ω scans
2 < θ < 26
173.0(2)
no. of data collected
no. of unique data
no. of params refined
8425
6842 (Rint ) 0.0345)
369
0.0383
0.0832
R1 (F2>2σ(F2) (4766 ref)
a
P r ep a r a tion of [OsH3(P HP h 2)2(P iP r 3)2]BF 4 (6). A solu-
tion of 5 (294.0 mg, 0.37 mmol) in 7 mL of dichloromethane
was treated with diphenylphosphine (63 mL, 0.37 mmol). After
1 day at room temperature, the solvent was removed in vacuo.
The residue was washed with diethyl ether to afford a white
solid. Yield: 310 mg (86%). Anal. Calcd for C42H67OsP4BF4:
C, 51.85; H, 6.94. Found: C, 51.44; H, 6.71. IR (KBr, cm-1):
ν(PH) 2358(m), 2326(m); ν(OsH) 2068(m), 2058(m), 1978(m);
ν(CdC)aromatic 1588(m), 1576(m); ν(BF4-) 1057(vs). 1H NMR
(300 MHz, CD2Cl2, 293 K): δ 7.50-7.40 (m, 20H; Ph), 6.98
(brd, 2H, J (HP) ≈ 350 Hz; PHPh2), 2.03 (m, 6H; PCH(CH3)2),
1.05 (dvt, 36H, N ) 14.4 Hz, J (HH) ) 7.2 Hz; PCH(CH3)2),
-10.68 (t, 3H, J (HP) ) 16.5 Hz; OsH). 31P{1H} NMR (121.4
MHz, CD2Cl2, 293 K): δ 28.2 (br, PHPh2), -23.6 (t, J (PP) )
7.3 Hz; PiPr3). MS(FAB+): m/z 886 (M+), 722 (M+ - 4H -
PiPr3), 696 (M+ - 4H - PHPh2). T1 (ms, “OsH3”, 300 MHz,
CD2Cl2): 425 (293 K), 294 (273 K), 204 (253 K), 167 (233 K),
157 (213 K), 184 (193 K).
b
wR2 (all data)
Sc (all data)
0.952 (a ) 0.0377, b ) 0)
R1(F) ) ∑||Fo| - |Fc||/∑|Fo|. wR2(F2) ) [∑[w(Fo2- Fc2)2]/
a
b
∑[w(Fo2)2]1/2; w-1 ) [σ2(Fo2) + (aP)2 + bP], where P ) [max(Fo2,0)
+ 2Fc2)]/3. c S ) [∑[w(Fo2 - Fc2)2/(n - p)]1/2, where n is the number
of reflections and p the number of refined parameters.
P r ep a r a tion of OsH4{P (OMe)P h 2}(P iP r 3)2 (9). A solution
of 7 (85.0 mg, 0.10 mmol) in 6 mL of dichloromethane was
treated with triethylamine (15 µL, 0.10 mmol). After 1 day at
room temperature, the solvent was removed in vacuo. Toluene
(10 mL) was added, and the suspension was filtered to
eliminate [HNEt3]BF4. Solvent was evaporated, and the
1
residue was washed with methanol to afford a brown oil. H
NMR (300 MHz, C6H6, 293 K): δ 8.03 (m, 4H; Hortho-Ph), 7.17
(m, 4H; Hmeta-Ph), 7.04 (m, 2H; Hpara-Ph), 2.94 (d, 3H, J (PH) )
11.1 Hz; OCH3), 1.79 (m, 6H; PCH(CH3)2), 1.10 (dvt, 36H, N
) 12.9 Hz, J (HH) ) 6.0 Hz; PCH(CH3)2), -10.05 (dt, 4H; J (HP)
) 15.0 Hz, J (HP) ) 7.2 Hz; OsH). 31P{1H} NMR (121.4 MHz,
C6D6, 293 K): δ 121.2 (t, J (PP) ) 11.5 Hz; P(OMe)Ph2), 41.6
(d, J (PP) ) 11.5 Hz; PiPr3). MS(FAB+): m/z 727 (M+ - 4H).
Cr ysta l Da ta for 5. Crystals suitable for X-ray diffraction
analysis were mounted onto a glass fiber and transferred to a
Bruker-Siemens P4 (T ) 173.0(2) K) automatic diffratometer
(Mo KR radiation, graphite monochromator, λ ) 0.71073 Å).
Accurate unit cell parameters were determined by least-
squares fitting from the setting of high-angle reflections. Data
were collected by the ω scan method. Lorentz and polarization
corrections were applied. Decay was monitored by measuring
three standard reflections throughout data collection. Correc-
tions for decay and absorption (semiempirical ψ scan method)
were also applied.
The structures were solved by the Patterson method and
refined by full-matrix least-squares on F2.35 Non-hydrogen
atoms were anisotropically refined, and the hydrogen atoms
were observed or included at idealized positions. Although
peaks with adequate geometry to be assigned to the hydrido
ligands were observed in the difference Fourier maps, the
presence of residuals close to the Os atom prevented their
proper refinement. Crystal data and details of the data
collection and refinement are given in Table 6.
P r epar ation of [OsH5{P (OMe)P h 2}(P iP r 3)2]BF4 (7). Com-
plex 5 (111.0 mg, 0.14 mmol) was stirred in 6 mL of methanol.
After 6 days at room temperature, the solvent was removed
in vacuo. The residue was washed with diethyl ether to afford
an ivory solid. Yield: 85 mg (74%). Anal. Calcd for C31H60
-
OOsP3BF4: C, 45.48; H, 7.39. Found: C, 45.11; H, 7.25. IR
(KBr, cm-1): ν(OsH) 2085(m), 2012(m), 1917(w); ν(CdC)aromatic
1589(w), 1579(w); ν(PO) 1101(s); ν(BF4-) 1049(vs). 1H NMR
(300 MHz, CD2Cl2, 293 K): δ 7.80-7.50 (m, 10H; Ph), 3.23 (d,
3H, J (PH) ) 12.0 Hz; OCH3), 2.02 (m, 6H; PCH(CH3)2), 1.07
(dvt, 36H, N ) 15.0 Hz, J (HH) ) 8.1 Hz; PCH(CH3)2), -8.20
(td, 5H, J (HP) ) 9.2 Hz, J (HP) ) 5.7 Hz; OsH). 31P{1H} NMR
(121.4 MHz, CD2Cl2, 293 K): δ 102.4 (t, J (PP) ) 8.8 Hz;
P(OMe)Ph2), 43.3 (d, J (PP) ) 8.8 Hz; PiPr3). MS(FAB+): m/z
727 (M+ - 5H). T1 (ms, OsH5, 300 MHz, CD2Cl2): 308 (294
K), 251 (273 K), 188 (253 K), 134 (233 K), 105 (213 K), 98 (193
K), 90 (178 K).
P r ep a r a tion of [OsH5{P (OH)P h 2}(P iP r 3)2]BF 4 (8). A
solution of 5 (165.0 mg, 0.21 mmol) in 4 mL of dichlorometh-
ane was treated with water (15 mL, 0.84 mmol). After 5
days at room temperature, the solvent was removed in vacuo.
The residue was washed with diethyl ether to afford an
ivory solid. Yield: 117 mg (69%). Anal. Calcd for C30H58OOsP3-
BF4: C, 44.78; H, 7.27. Found: C, 44.73; H, 7.38. IR (KBr,
cm-1): ν(OH) 3223(s), ν(OsH) 2132(m), 2084(m), 2015(m);
ν(CdC)aromatic 1580(vw); ν(BF4-) 1099(vs); ν(PO) 1034(s). 1H
NMR (300 MHz, CD2Cl2, 293 K): δ 7.80-7.40 (m, 10H; Ph),
1.96 (m, 6H; PCH(CH3)2), 1.63 (br, 1H; OH), 1.09 (dvt, 36H, N
) 15.0 Hz, J (HH) ) 7.2 Hz; PCH(CH3)2), -8.27 (td, 5H, J (HP)
) 4.8 Hz, J (HP) ) 9.0 Hz; OsH). 31P{1H} (121.4 MHz, CD2Cl2,
293 K): δ 86.6 (t, J (PP) ) 8.9 Hz; P(OH)Ph2), 42.7 (d, J (PP) )
8.9 Hz; PiPr3). MS (FAB+): m/z 713 (M+ - 5H). T1 (ms, “OsH5”,
300 MHz, CD2Cl2): 268 (293 K), 198 (273 K), 146 (253 K), 114
(233 K), 106 (213 K), 118 (193 K).
Com p u ta tion a l Deta ils. DFT optimizations on model
complexes were carried out with the Gaussian9436 series of
programs using the B3LYP functional.37 Furthermore,
CCSD(T)38 single-point calculations on the DFT optimized
(35) Sheldrick, G. M. SHELX-97; University of Go¨ttingen: Go¨ttin-
gen, Germany, 1997.
(36) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
J ohnson, 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 .;
Stefanov, B. B.; Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala,

Mixed-Phosphine Osmium Polyhydrides
Organometallics, Vol. 20, No. 25, 2001 5309
structures were performed in order to obtain more reliable
energies. A quasi-relativistic effective core potential operator
was used to represent the 60 innermost electrons of the
osmium atom.39 The basis set for the osmium atom was that
associated with the pseudopotential with a standard valence
double-ú LANL2DZ contraction.36 The 6-31G(d) basis set was
used for the phosphorus atoms.40 Hydrogens directly attached
to the metal were described using a 6-311G(p) basis set,40 while
a 6-31G basis set was used for the rest of hydrogens.40
Calculations on the real systems were performed using the
IMOMM method,41 with a program built from modified ver-
(P(OMe)Ph2)+ complex, the whole OMe group has also been
treated with QM methods. The QM part of the calculations
was done at the B3LYP37 level for the optimizations and
CCSD(T)38 level for single-point energy calculations. The same
basis sets were used for complexes 7 and 11, while for complex
5 the hydrogen of PHPh2 was treated with a 6-311G(p) basis
set.36,40 The MM part of the calculations used the mm3(92)
force field.44 Van der Waals parameters for the osmium atom
were taken from the UFF force field.45 Torsional contributions
involving dihedral angles with the metal were set to zero. All
geometrical parameters were optimized, except the bond
distances between the QM and the MM regions of the molecule.
The frozen values were 1.41 Å for the P-H bonds in the QM
part and the crystallographic values for P-C bonds in the MM
part.
42
sions of two standard programs: Gaussian 92/DFT for the
quantum mechanics part and mm3 (92)43 for the molecular
mechanics part. The OsH5P3 backbone has been described at
the QM level. In the OsH5(PiPr3)2(PHPh2)+ complex, the
hydrogen atom of the PHPh2 phosphine has also been included
in the QM part of the calculations, while in the OsH5(PiPr3)2-
Ackn owledgm en t. Financial support from the DGES
of Spain (Projects PB98-0916-02-01 and PB98-1591) is
acknowledged. J .T. acknowledges the “Direccio´ General
de Recerca de la Generalitat de Catalunya” for a grant.
The use of computational facilities of the Centre de
Supercomputacio´ i Comunicacions de Catalunya is
gratefully appreciated as well. We also acknowledge Dr.
Dmitri G. Gusev for the careful reading of the manu-
script and his useful advice.
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.
(37) (a) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785.
(b) Becke, A. D. J . Chem. Phys. 1993, 98, 5648. (c) Stephens, P. J .;
Devlin, F. J .; Chabalowski, C. F.; Frisch, M. J . J . Phys. Chem. 1994,
98, 11623.
(38) (a) Bartlett, R. J . J . Phys. Chem. 1989, 93, 1697. (b) Bartlett,
R. J .; Watts, J . D.; Kucharski, S. A.; Noga, J . Chem. Phys. Lett. 1990,
165, 513.
(39) Hay, P. J .; Wadt, W. R. J . Chem. Phys. 1985, 82, 299.
(40) (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. (c) Francl, M. M.; Pietro, W. J .; Hehre, W. J .; Binkley, J . S.;
Gordon, M. S.; DeFrees, D. J .; Pople, J . A. J . Chem. Phys. 1982, 77,
3654.
(41) Maseras, F.; Morokuma, K. J . Comput. Chem. 1995, 16, 1170.
(42) Frisch, M. J .; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
J ohnson, B. G.; Wong, M. W.; Foresman, J . B.; Robb, M. A.; Head-
Gordon, M.; Replogle, E. S.; Gomperts, R.; Andre´s, J . L.; Raghavachari,
K.; Binkley, J . S.; Gonza´lez, C.; Martin, R. L.; Fox, D. J .; Defrees, D.
J .; Baker, J .; Stewart, J . J . P.; Pople, J . A. Gaussian 92/DFT; Gaussian
Inc.: Pittsburgh, PA, 1993.
Su p p or tin g In for m a tion Ava ila ble: Tables of atomic
coordinates for the non-hydrogen atoms, anisotropic displace-
ment coefficients, atomic coordinates for hydrogen atoms, full
experimental details for the X-ray analysis, and bond lengths
and angles. This material is available free of charge via the
Internet at http://pubs.acs.org.
OM010173C
(44) Allinger, N. L.; Yuh, Y. H.; Lii, J . H. J . Am. Chem. Soc, 1989,
111, 8551.
(45) Rappe´, A. K.; Casewit, C. J .; Cowell, K. S.; Goddard, W. A., III;
Skiff, W. M. J . Am. Chem. Soc. 1992, 114, 10024.
(43) Allinger, N. L. mm3(92); QCPE: Bloomington, IN, 1992.