Mechanistic and theoretical analysis of the oxidative addition of h2 to six-coordinate molybdenum and tungsten complexes M(PMe3)4X2 (M = Mo, W; X = F, Cl, Br, I): An inverse equilibrium isotope effect and an unprecedented halide dependence

Article abstract of DOI:10.1021/ja9920009

Experimental observations, together with a theoretical analysis, indicate that the energetics of the oxidative addition of H₂ to the six-coordinate molybdenum and tungsten complexes trans-M(PMe₃)₄X₂ (M = Mo, W; X = F, Cl, Br, I) depend very strongly on the nature of both the metal and the halogen. Specifically, the exothermicity of the reaction increases in the sequences Mo < W and I < Br < Cl < F. Of most interest, this halogen dependence provides a striking contrast to that reported for oxidative addition of H₂ to the Vaska system, trans-Ir(PPh₃)₂(CO)X. A theoretical analysis suggests that the halide dependence for trans-M(PMe₃)₄X₂ is a result of both steric and electronic factors, the components of which serve to reinforce each other. Oxidative addition is thus favored sterically for the fluoride derivatives since the increased steric interactions upon forming the eight-coordinate complexes M(PMe₃)₄H₂X₂ would be minimized for the smallest halogen. The electronic component of the energetics is associated with the extent that π-donation from X raises the energy of the doubly occupied 3e*, π-antibonding, d_xz and d_yz pair of orbitals in trans-M(PMe₃)₄X₂. Consequently, with F as the strongest π-donor, trans-M(PMe₃)₄X₂ is destabilized with respect to M(PMe₃)₄H₂X₂ by p_π-d_π interaction to the greatest extent for the fluoride complex, so that oxidative addition becomes most favored for this derivative. Equilibrium studies of the oxidative addition of H2 to trans-W(PMe₃)₄I₂ have allowed the average W-H bond dissociation energy (BDE) in W(PMe₃)H₂I₂ to be determined [D(W-H) = 62.0(6) kcal mol^-1]. The corresponding average W-D BDE [D(W-D) = 63.8(7) kcal mol^-1] is substantially greater than the W-H BDE, to the extent that the oxidative addition reaction is characterized by an inverse equilibrium deuterium isotope effect [K_H/KD = 0.63(5) at 60C]. The inverse nature of the equilibrium isotope effect is associated with the large number (six) of isotope-sensitive vibrational modes in the product, compared to the single isotope-sensitive vibrational mode in reactant H₂. A mechanistic study reveals that the latter reaction proceeds via initial dissociation of PMe₃, followed by oxidative addition to five-coordinate [W(PMe₃)₃I₂], rather than direct oxidative addition to trans-W(PMe₃)₄I₂. Conversely, reductive elimination of H₂ does not occur directly from W(PMe₃)₄H₂I₂ but rather by a sequence that involves dissociation of PMe₃ and elimination from the seven-coordinate species [W(PMe₃)SH₂I₂].

Full text of DOI:10.1021/ja9920009

11402
J. Am. Chem. Soc. 1999, 121, 11402-11417
Mechanistic and Theoretical Analysis of the Oxidative Addition of H2
to Six-Coordinate Molybdenum and Tungsten Complexes
M(PMe3)4X2 (M ) Mo, W; X ) F, Cl, Br, I): An Inverse
Equilibrium Isotope Effect and an Unprecedented Halide Dependence
Tony Hascall, Daniel Rabinovich, Vincent J. Murphy,† Michael D. Beachy,
Richard A. Friesner,* and Gerard Parkin*
Contribution from the Department of Chemistry, Columbia UniVersity, New York, New York 10027
ReceiVed June 14, 1999
Abstract: Experimental observations, together with a theoretical analysis, indicate that the energetics of the
oxidative addition of H2 to the six-coordinate molybdenum and tungsten complexes trans-M(PMe3)4X2 (M )
Mo, W; X ) F, Cl, Br, I) depend very strongly on the nature of both the metal and the halogen. Specifically,
the exothermicity of the reaction increases in the sequences Mo < W and I < Br < Cl < F. Of most interest,
this halogen dependence provides a striking contrast to that reported for oxidative addition of H2 to the Vaska
system, trans-Ir(PPh3)2(CO)X. A theoretical analysis suggests that the halide dependence for trans-M(PMe3)4X2
is a result of both steric and electronic factors, the components of which serve to reinforce each other. Oxidative
addition is thus favored sterically for the fluoride derivatives since the increased steric interactions upon forming
the eight-coordinate complexes M(PMe3)4H2X2 would be minimized for the smallest halogen. The electronic
component of the energetics is associated with the extent that π-donation from X raises the energy of the
doubly occupied 3e*, π-antibonding, dxz and dyz pair of orbitals in trans-M(PMe3)4X2. Consequently, with F
as the strongest π-donor, trans-M(PMe3)4X2 is destabilized with respect to M(PMe3)4H2X2 by pπ-dπ interaction
to the greatest extent for the fluoride complex, so that oxidative addition becomes most favored for this derivative.
Equilibrium studies of the oxidative addition of H2 to trans-W(PMe3)4I2 have allowed the average W-H bond
dissociation energy (BDE) in W(PMe3)4H2I2 to be determined [D(W-H) ) 62.0(6) kcal mol-1]. The
corresponding average W-D BDE [D(W-D) ) 63.8(7) kcal mol-1] is substantially greater than the W-H
BDE, to the extent that the oxidative addition reaction is characterized by an inVerse equilibrium deuterium
isotope effect [KH/KD ) 0.63(5) at 60 °C]. The inverse nature of the equilibrium isotope effect is associated
with the large number (six) of isotope-sensitive vibrational modes in the product, compared to the single
isotope-sensitive vibrational mode in reactant H2. A mechanistic study reveals that the latter reaction proceeds
via initial dissociation of PMe3, followed by oxidative addition to five-coordinate [W(PMe3)3I2], rather than
direct oxidative addition to trans-W(PMe3)4I2. Conversely, reductive elimination of H2 does not occur directly
from W(PMe3)4H2I2 but rather by a sequence that involves dissociation of PMe3 and elimination from the
seven-coordinate species [W(PMe3)3H2I2].
Introduction
coordinate molybdenum and tungsten complexes M(PMe3)4X2
(M ) Mo, W; X ) F, Cl, Br, I). This study demonstrates that
the oxidative addition of dihydrogen is characterized by an
inVerse deuterium equilibrium isotope effect (i.e., KH/KD < 1)
and also by a halide dependence (i.e., KF > KCl > KBr > KI),
which is counter to that in previously studied transition metal
systems.3
The oxidative addition of dihydrogen to a metal center and
its microscopic reverse, reductive elimination, are two of the
simplest yet most important transformations in organometallic
chemistry.1 For example, oxidative addition of dihydrogen is a
key step in many, if not all, transition metal-catalyzed reactions
involving H2, as exemplified by olefin hydrogenation and
hydroformylation.2 In this paper, we describe a mechanistic and
theoretical analysis of the oxidative addition of H2 to the six-
Results and Discussion
We have recently reported the syntheses and molecular
structures of the series of eight-coordinate tungsten hydride
complexes W(PMe3)4H2X2 (X ) F, Cl, Br, I).4-6 Despite their
similar structures, however, the stabilities of these complexes
are markedly dependent upon the nature of the halogen. For
† Present address: Symyx Technologies, 3100 Central Expressway, Santa
Clara, CA 95051.
(1) (a) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G.
Principles and Applications of Organotransition Metal Chemistry; Univer-
sity Science Books: Mill Valley, CA, 1987. (b) Mondal, J. U.; Blake, D.
M. Coord. Chem. ReV. 1982, 47, 205-238. (c) Hay, P. J. In Transition
Metal Hydrides; Dedieu, A., Ed.; VCH: New York, 1992; pp 127-147.
(2) (a) Parshall, G. W.; Ittel, S. D. Homogeneous Catalysis, 2nd ed.;
Wiley: New York, 1992. (b) James, B. R. In ComprehensiVe Organome-
tallic Chemistry; Wilkinson, G., Stone, F. G. A., Abel, E. W., Eds.;
Pergamon Press: New York, 1982; Vol. 8, Chapter 51, pp 285-363. (c)
James, B. R. Homogeneous Hydrogenation; Wiley: New York, 1973.
(3) A portion of this work concerned with the equilibrium deuterium
isotope effect has been communicated. See: Rabinovich, D.; Parkin, G. J.
Am. Chem. Soc. 1993, 115, 353-354.
(4) Murphy, V. J.; Rabinovich, D.; Hascall, T.; Klooster, W. T.; Koetzle,
T. F.; Parkin, G. J. Am. Chem. Soc. 1998, 120, 4372-4387.
10.1021/ja9920009 CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/30/1999

OxidatiVe Addition of H2 to Mo and W Complexes
J. Am. Chem. Soc., Vol. 121, No. 49, 1999 11403
Table 1. Qualitative Thermal Stabilities of M(PMe3)4H2X2 Derivatives
F
Cl
Br
I
Mo
W
t1/2 > 2 days at 30 °C
stable at 60 °C
t1/2 ≈ 2 h at 30 °C
stable at 60 °C
t1/2 < 30 min at 30 °C
stable at 60 °C
not observed
reversible elimination of H2 at ambient temperature
example, whereas the fluoride, chloride,7 and bromide deriva-
tives are stable under ambient conditions, the iodide complex
slowly reductively eliminates H2 at ambient temperature to give
trans-W(PMe3)4I2.8 The latter reaction is, however, reversible,
and W(PMe3)4I2 reacts with H2 to give W(PMe3)4H2I2 (eq 1).
Together with the fact that W(PMe3)4Cl2 reacts irreversibly with
H2 to give W(PMe3)4H2Cl2,5b these observations indicate that
the W-H bond energies in the complexes W(PMe3)4H2X2 are
strongly influenced by the nature of the halogen, being weakest
for the iodide derivative. A similar trend is also observed in
the molybdenum system, although the complexes Mo(PMe3)4H2X2
(X ) F,9 Cl, Br, I) are much less stable in solution than their
tungsten counterparts (Table 1). Thus, while each of the
complexes Mo(PMe3)4H2X2 may be generated by reaction of
Mo(PMe3)5H2 with HX,10 only the fluoride complex Mo-
(PMe3)4H2F29 exhibits appreciable thermal stability: the chloro
and bromo complexes Mo(PMe3)4H2X2 readily eliminate H2 to
give Mo(PMe3)4X2,11 while the iodo complex Mo(PMe3)4H2I2
is sufficiently unstable that it has not even been spectroscopically
detected (Scheme 1).
Figure 1. van’t Hoff plots for oxidative addition of H2 and D2 to
W(PMe3)4I2.
Scheme 1. Syntheses and Reductive Elimination of
Mo(PMe3)4H2X2
The above observations clearly indicate that the thermody-
namic ability of H2 to oxidatively add to the metal center in
M(PMe3)4X2 is very dependent upon the nature of both the metal
and the halogen. In view of the importance of oxidative addition
and reductive elimination processes in organometallic chemistry,
Table 2. Equilibrium Constants for the Oxidative Addition of H2
(D2) to W(PMe3)4I2
T/°C
KH/M-1
KD/M-1
KH/KD
(5) For the first reports of the chloride derivative, W(PMe3)4H2Cl2, see:
(a) Sharp, P. R. Ph.D. Thesis, Massachusetts Institute of Technology, 1980.
(b) Sharp, P. R.; Frank, K. G. Inorg. Chem. 1985, 24, 1808-1813. (c) Chiu,
K. W.; Lyons, D.; Wilkinson, G.; Thornton-Pett, M.; Hursthouse, M. B.
Polyhedron 1983, 2, 803-810.
(6) For reviews on metal hydrides, see: (a) Transition Metal Hydrides;
Dedieu, A., Ed.; VCH Publishers, Inc.: New York. 1991. (b) Transition
Metal Hydrides; Muetterties, E. L., Ed.; Marcel Dekker, Inc.: New York,
1971. (c) Teller, R. G.; Bau, R. Struct. Bonding (Berlin) 1981, 44, 1-82.
(d) Crabtree, R. H. In Encyclopedia of Inorganic Chemistry; King, R. B.,
Ed.; Wiley: New York, 1994; Vol. 3, pp 1392-1400. (e) Hlatky, G. G.;
Crabtree, R. H. Coord. Chem. ReV. 1985, 65, 1-48. (f) Moore, D. S.;
Robinson, S. D. Chem. Soc. ReV. 1983, 12, 415-452. (g) Kuhlman, R.
Coord. Chem. ReV. 1997, 167, 205-232. (h) Also see the special volume:
Inorg. Chim. Acta 1997, 259, Nos. 1 and 2.
(7) The thermal stability of W(PMe3)4H2Cl2 at 80 °C with respect to
reductive elimination was first noted by Sharp (ref 5b).
(8) W(PMe3)4I2 has also been reported to exist as a diamagnetic cis-
isomer, but we have not observed such a species. See: Chiu, K. W.; Jones,
R. A.; Wilkinson, G.; Galas, A. M. R.; Hursthouse, M. B.; Malik, K. M.
A. J. Chem. Soc., Dalton Trans. 1981, 1204-1211.
40
50
60
70
80
90
6.6(7) × 103
2.5(3) × 103
1.0(1) × 103
4.4(5) × 102
2.0(2) × 102
0.8(1) × 102
1.0(1) × 104
3.5(4) × 103
1.6(2) × 103
6.0(6) × 102
2.2(2) × 102
0.8(1) × 102
0.66(13)
0.71(14)
0.63(5)
0.73(15)
0.91(18)
1.00(20)
we have performed a combination of equilibrium, kinetics, and
computational studies to ascertain the factors responsible for
influencing these fundamental transformations.
Thermodynamics of Oxidative Addition of H2: M-H
Bond Dissociation Energies and the Origin of the Halide
Dependence. (a) Equilibrium Studies. The magnitude of the
equilibrium constant for oxidative addition of H2 to the tungsten
iodide complex trans-W(PMe3)4I2 is such that, under 1 atm of
1
H2; the equilibrium constant can be determined by using H
(9) Murphy, V. J.; Hascall, T.; Chen, J. Y.; Parkin, G. J. Am. Chem.
Soc. 1996, 118, 7428-7429.
(10) It should be noted that the isolation of the fluoro complex actually
requires an additional step, because Mo(PMe3)4H2F2 reacts further with
HF(aq) to give a bifluoride derivative. See ref 9.
NMR spectroscopy, and its temperature dependence (Figure 1
and Table 2) allows determination of the enthalpy and entropy
for the oxidative addition: ∆H° ) -19.7(6) kcal mol-1 and
∆S° ) -45(2) eu12 Since we may define ∆H° ) D(H-H) -
(11) trans-Mo(PMe3)4X2 (X ) Cl,a,b Br,b,d Ib,d) compounds have been
previously reported. (a) Carmona, E.; Marin, J. M.; Poveda, M. L.; Atwood,
J. L.; Rogers, R. D. Polyhedron 1983, 2, 185-193. (b) Carmona, E.; Marin,
J. M.; Poveda, M. L.; Rogers, R. D.; Atwood, J. L. J. Am. Chem. Soc.
1983, 105, 3014-3022. (c) Krueger, S. T.; Poli, R.; Rheingold, A. L.; Staley,
D. L. Inorg. Chem. 1989, 28, 4599-4607. (d) Brookhart, M.; Cox, K.;
Cloke, F. G. N.; Green, J. C.; Green, M. L. H.; Hare, P. M.; Bashkin, J.;
Derome, A. E.; Grebenik, P. D. J. Chem. Soc., Dalton Trans. 1985, 423-
433.
(12) While this value for ∆S° is larger than has been reported for
oxidative addition in other systems, and may possibly reflect a systematic
error in the data, it should be noted that larger entropy changes have been
reported for several other reactions.a For example, the dissociation of
i
{[TpPr Me]Co(O2)}2 is characterized by ∆S° ) 60(3) eu.b (a) Minas da
Piedade, M.; Martino Simo˜es, J. A. J. Organomet. Chem. 1996, 518, 167-
180. (b) Reinaud, O.; Yap, G. P. A.; Rheingold, A. L.; Theopold, K. H.
Angew. Chem., Int. Ed. Engl. 1995, 34, 2051-2052.

11404 J. Am. Chem. Soc., Vol. 121, No. 49, 1999
Hascall et al.
Figure 3. Trends in computed energetics for oxidative addition of H2
to M(PMe3)4X2. For comparison, data for Ir(PH3)2(CO)Cl, taken from
ref 27, are also included.
complexes using Jaguar (Version 3.5).18 Geometry-optimized
bond length and angle data are listed in the Supporting
Information (Tables S2-S5), which also includes experimental
values for those trimethylphosphine complexes that have been
structurally characterized. A comparison of the experimental
and calculated structures is provided in the Experimental
Section, but it is important to note here that there is a significant
difference in the calculated structures of M(PH3)4X2 and
M(PMe3)4X2. Specifically, the calculated geometries of the PH3
complexes M(PH3)4X2 are approximately octahedral (with all
bond angles at M close to 90°), whereas those of M(PMe3)4X2
deviate from idealized octahedral coordination due to an
alternate up-down “ruffling” of the PMe3 ligands about the
equatorial plane (Figure 2). The latter geometries much more
closely resemble the true geometries of M(PMe3)4X2, suggesting
that the PH3 ligand is a poor substitute for PMe3 in theoretical
calculations of this system. The discrepancy is undoubtedly a
consequence of the lower steric demands of PH3 Versus PMe3.
The energetics of the oxidative addition reactions, as defined
by
Figure 2. Comparison of the geometry-minimized structures of
W(PMe3)4I2 and W(PH3)4I2 with the experimental structure for the
former.
2D(W-H),1b,13 the average W-H bond dissociation energy
(BDE) in W(PMe3)4H2I2 has a value of 62.0(6) kcal mol-1 14
.
Measurement of this value is of some interest since W-H BDEs
have been previously reported for only a few classes of
complexes.15,16
A knowledge of the factors that influence metal-hydrogen
BDEs is central to understanding, and predicting, reaction
pathways involving H2. However, since it was not possible to
obtain equilibrium constant data for the other halide systems,
we have performed computational studies in an attempt to
evaluate the metal and halide dependence of M-H BDEs in
the series of molybdenum and tungsten complexes M(PMe3)4H2X2
(M ) Mo, W; X ) F, Cl, Br, I).
(b) Computational Studies: Geometry-Optimized Molec-
ular Structures and Computed Trends in the Thermody-
namics of the Oxidative Addition of H2. Density functional
theory (DFT) calculations were performed on M(PR3)4X217 and
M(PR3)4H2X2 (M ) Mo, W; R ) H, Me; X ) F, Cl, Br, I)
∆E ) E[M(PR3)4H2X2] - {E[M(PR3)4X2] + E[H2]}
(13) For rationalization of the use of this expression in determining bond
energies, see ref 34 in the following: Hartwig, J. F.; Andersen, R. A.;
Bergman, R. G. Organometallics 1991, 10, 1875-1887.
are summarized in Table 3 and Figure 3. The variation in ∆E
would be expected to mirror trends in ∆H°, assuming that zero
point energy and excitation corrections are similar for each pair
of complexes. Support for this assumption is provided by the
fact that the difference between ∆H° and ∆E is very similar
for each of the PH3 derivatives, W(PH3)4F2 (+3.1 kcal mol-1),
W(PH3)4Cl2 (+3.0 kcal mol-1), W(PH3)4Br2 (+3.4 kcal mol-1),
and W(PH3)4I2 (+3.1 kcal mol-1).19 Furthermore, the correction
for addition of H2 to the PMe3 complex W(PMe3)4I2 is only
(14) For dihydrogen, D(H-H) ) 104.2 kcal mol-1 and D(D-D) ) 106.0
kcal mol-1. CRC Handbook of Chemistry and Physics, 70th ed.; Weast, R.
C., Ed.; CRC Press: Boca Raton, FL, 1989-1990; p F-199.
(15) For examples, see Table S1 in the Supporting Information.
(16) For tabulations of bond energies, see: (a) Bonding Energetics in
Organometallic Compounds; Marks, T. J. Ed.; ACS Symposium Series 428;
American Chemical Society: Washington, DC, 1990. (b) Pearson, R. G.
Chem. ReV. 1985, 85, 41-49. (c) Halpern, J. Inorg. Chim. Acta 1985, 100,
41-48. (d) Skinner, H. A.; Connor, J. A. Pure Appl. Chem. 1985, 57, 79-
98. (e) Martinho Simo˜es, J. A.; Beauchamp, J. L. Chem. ReV. 1990, 90,
629-688. (f) Connor, J. A. Top. Curr. Chem. 1977, 71, 71-110. (g) Wang,
K.; Rosini, G. P.; Nolan, S. P.; Goldman, A. S. J. Am. Chem. Soc. 1995,
117, 5082-5088. (h) Wang, D.; Angelici, R. J. J. Am. Chem. Soc. 1996,
118, 935-942. (i) Eisenberg, D. C.; Norton, J. R. Isr. J. Chem. 1991, 31,
55-66. (j) Pilcher, G.; Skinner, H. A. In The Chemistry of the Metal-
Carbon Bond; Hartley, F. R., Patai, S., Eds.; Wiley: New York, 1982;
Vol. 1, Chapter 2, pp 43-90. (k) Labinger, J. A.; Bercaw, J. E.
Organometallics 1988, 7, 926-928. (l) King, W. A.; Di Bella, S.; Gulino,
A.; Lanza, G.; Fragala, I. L.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc.
1999, 121, 355-366. (m) Hoff, C. D. Prog. Inorg. Chem. 1992, 40, 503-
561. (f) Energetics of Organometallic Species; Martinho Simo˜es, J. A., Ed.;
Kluwer Academic Publishers: Dordrecht, The Netherlands, 1992.
(17) For theoretical calculations on related chalcogenido complexes,
M(PR3)4(E)2 (E ) O, S, Se, Te), see: (a) Cotton, F. A.; Feng, X. Inorg.
Chem. 1996, 35, 4921-4925. (b) Kim, W.-S.; Kaltsoyannis, N. Inorg. Chem.
1998, 37, 674-678. (c) Kaltsoyannis, N. J. Chem. Soc., Dalton Trans. 1994,
1391-1400.
+0.2 kcal mol-1
.
Examination of Table 3 indicates that, even though the overall
trend in the energetics of the oxidative addition of H2 to
M(PH3)4X2 is similar to that for the M(PMe3)4X2 system, the
magnitude of the variation in E is substantially greater for the
latter. While this difference in thermodynamics may be a result
of the different electronic properties of PH3 and PMe3, it is more
likely a consequence of the different ground-state geometries
of M(PH3)4X2 and M(PMe3)4X2, due to the smaller size of the
PH3 ligand, as described above. In view of this difference in
computed structures, it is not surprising that the calculated
energetics for oxidative addition to the PH3 system are different
(18) Jaguar 3.5, Schrodinger, Inc., Portland, OR, 1998.
(19) The differences between ∆H° and ∆E are those at 25 °C.

OxidatiVe Addition of H2 to Mo and W Complexes
Table 3. Calculated ∆Ea (kcal mol-1) for the Addition of H2 to M(PR3)4X2 (M ) Mo, W; X ) F, Cl, Br, I; R ) H, Me)
Cl Br
J. Am. Chem. Soc., Vol. 121, No. 49, 1999 11405
F
I
PH3
PMe3
PH3
PMe3
PH3
PMe3
PH3
PMe3
E(Mo)
E(W)
E(W) - E(Mo)
-2.1
-19.6
-17.5
-6.4
-25.1
-18.7
1.2
-13.3
-14.7
5.8
-10.6
-16.4
0.7
-13.9
-14.5
9.3
-7.1
-16.4
1.7
-12.0
-13.7
12.0
-2.7
-14.7
a ∆E ) E[M(PR3)4H2X2] - {E[M(PR3)4X2] + E[H2]}
Table 4. Estimated M-H BDEs (kcal mol-1) for M(PMe3)4H2X2
(M ) Mo, W; X ) F, Cl, Br, I), Relative to BDE(W-H) ) 62.0
kcal mol-1 for W(PMe3)4H2I2
upon passing from the fluoro to iodo derivatives (Table 5 and
Figure 3). To understand the origin of the striking difference
between these two systems, it is essential to understand how,
and why, various substituents influence the susceptibility of a
metal center towards oxidative addition. In this regard, it is
commonly assumed that the ability of a metal center to undergo
an oxidative addition reaction is directly related to its electron
richness, i.e., “factors that tend to increase the electron density
on the metal make for an increase in oxidizability”.28-30 For
example, the equilibrium constants for the oxidative addition
of H2 to trans-Ir(PR3)2(CO)Cl increase as the σ-donor PR3
ligands become more basic.31 Likewise, the equilibrium con-
stants for oxidative addition of PhCO2H to trans-Ir(PR3)2(CO)X
(X ) Cl, Br, I), giving trans-Ir(PR3)2(CO)X(O2CPh)H, also
F
Cl
Br
I
BDE(Mo-H)a
63.8
73.2
9.4
57.7
66.0
8.2
56.0
64.2
8.2
54.6
62.0
7.4
BDE(W-H)a
BDE(W-H) - BDE(Mo-H)
a BDE(MX-H) ) 0.5[EWI - EMX + 124].
from those of the PMe3 system, with the latter presumably
providing a much better estimate of the true values. Jacobsen
and Berke have also recently noted that it is important to use
the real ligand, PMe3, rather than PH3 in calculations to assess
relative energies of isomers of [Fe(PMe3)4H3]+.20
increase with the basicity of PR3 in the sequence PPh3
<
Figure 3 indicates that the exoergicity of the oxidative addition
of H2 to M(PMe3)4X2 strongly increases in the sequences (i)
Mo < W and (ii) I < Br < Cl < F. Since the thermodynamic
differences are directly related to the differences in M-H bond
energies as defined above, the M-H bond energies may be
estimated by appropriate scaling with respect to the experimen-
tally determined value of 62.0(6) kcal mol-1 for the average
W-H bond energy of W(PMe3)4H2I2 (Table 4). Inspection of
Table 4 indicates that the W-H bond energies are ca. 7-10
kcal mol-1 greater than the corresponding Mo-H energies, in
line with other reports of bond energies for third- and second-
row transition metals.15,21,22 For example, the W-H bond in
Cp2WH2 (74.3 kcal mol-1) is 12.9 kcal mol-1 stronger than the
Mo-H bond in Cp2MoH2 (61.4 kcal mol-1).23 Such increases
in M-H bond energies from the first to third series are typically
attributed to more favorable sd hybridization, coupled with a
greater radial extent of the hybrid orbital for the heavier
elements, which results in increased overlap with the hydrogen
1s orbital.24
PMePh2 < PMe2Ph < PMe3.32
While the observed halide dependence for oxidative addition
of H2 to Ir(PR3)2(CO)X is in accord with that expected on the
basis of the variation of the halogen electronegativity,33 it is
important to note that this order is counter to that which would
be predicted by the oVerall electron-donating ability of the
halogens. Specifically, the overall electron-donating ability of
a halogen toward a transition metal decreases in the sequence
F > Cl > Br > I, a consequence of the trend being dictated by
the stronger π-donation of the lighter halogens.34,35 Thus,
although the electron richness of the iridium centers, as judged
by the ν(CO) stretching frequencies of trans-Ir(PPh3)2(CO)X,
decreases in the order F > Cl > Br > I,36,37 the tendency to
oxidatively add H2 increases in the order F < Cl < Br < I
(27) (a) Abu-Hasanayn, F.; Krogh-Jespersen, K.; Goldman, A. S. Inorg.
Chem. 1993, 32, 495-496. (b) Abu-Hasanayn, F.; Goldman, A. S.; Krogh-
Jespersen, K. Inorg. Chem. 1994, 22, 5122-5130.
(28) Cotton, F. A.; Wilkinson, G. AdVanced Inorganic Chemistry, 5th
ed.; Wiley: New York, 1988; p 1191.
Of most interest, the influence of the halide on the thermo-
dynamics of the oxidative addition of H2 is completely counter
to that reported for oxidative addition of H2 to Vaska-type
complexes, trans-Ir(PR3)2(CO)X. Thus, early experimental
studies by Vaska on trans-Ir(PPh3)2(CO)X,25,26 and a compu-
tational study by Goldman and Krogh-Jespersen on trans-Ir-
(PH3)2(CO)X,27 indicate that oxidative addition of H2 to iridium
in these complexes becomes progressively more exothermic
(29) (a) Collman, J. P. Acc. Chem. Res. 1968, 1, 136-141. (b) Collman,
J. P.; Roper, W. R. AdV. Organomet. Chem. 1968, 7, 53-94.
(30) Crabtree has suggested that some oxidative addition reactions of
H2, in which the addition is inhibited in the presence of certain donor ligands
such as halide, should be regarded as “reductive” in nature.a With respect
to such a viewpoint, it is pertinent to quote from the work of Saillard and
Hoffmann: “Formalisms are convenient fictions which contain a piece of
the truthsand it is so sad that people spend a lot of time arguing about the
deductions they draw, often ingeniously and artfully, from formalisms,
without worrying about their underlying assumptions. The ‘complex’ or
dative bonding picture which led to ‘oxidation at metal’ of course is an
exaggeration. The M-H σ bonds are in good part covalent. To the extent
that they are so, the real d electron population at the metal moves back
from dn-2 toward dn. To the extent that it probably never quite gets back to
dn it is still informative to call this an oxidative addition.” b (a) Crabtree,
R. H.; Quirk, J. M. J. Organomet. Chem. 1980, 199, 99-106. (b) Saillard,
J.-Y.; Hoffmann, R. J. Am. Chem. Soc. 1984, 106, 2006-2026.
(31) Specifically, equilibrium constants (at 30 °C) for addition of H2 to
trans-Ir(PR3)2(CO)Cl are 3.2 × 104 M-1 (R ) C6H5), 4.4 × 104 M-1 (R )
C6H4Me), and 5.2 × 104 M-1 (R ) C6H4OMe). Furthermore, the complex
with R ) C6F5 was unreactive toward H2. See ref 25.
(20) Jacobsen, H.; Berke, H. Chem. Eur. J. 1997, 3, 881-886.
(21) Kinetic barriers to reductive elimination of H2 from a metal center
follow a similar trend, with the barriers being greater for the heavier metal.
See, for example: Halpern, J.; Cai, L.; Desrosiers, P. J.; Lin, Z. J. Chem.
Soc., Dalton Trans. 1991, 717-723.
(22) M-C bond energies also increase upon descending a transition metal
group. See, for example: Mancuso, C.; Halpern, J. J. Organomet. Chem.
1992, 428, C8-C11.
(23) Dias, A. R.; Martinho Simo˜es, J. A. Polyhedron 1988, 7, 1531-
1544.
(24) (a) Ohanessian, G.; Goddard, W. A., III. Acc. Chem. Res. 1990, 23,
386-392. (b) Landis, C. R.; Firman, T. K.; Root, D. M.; Cleveland, T. J.
Am. Chem. Soc. 1998, 120, 1842-1854.
(25) Vaska, L.; Werneke, M. F. Ann. N.Y. Acad. Sci. 1971, 172, 546-
562.
(26) It should be noted that the Vaska system had also been studied by
Strohmeier, who obtained results that are not in accord with Vaska’s. See:
Strohmeier, W.; Mu¨ller, F. J. Z. Naturforsch. 1969, 24b, 931-932.
(32) (a) Deeming, A. J.; Shaw, B. L. J. Chem. Soc. (A) 1969, 1802-
1804. (b) van Doorn, J. A.; Masters, C. van der Woude, C. J. Chem. Soc.,
Dalton Trans. 1978, 1213-1220.
(33) Furthermore, oxidative addition of PhCO2H to trans-Ir(PR3)2(CO)X
(X ) Cl, Br, I) becomes more favored in the sequence Cl < Br < I. See
ref 32.
(34) (a) Caulton, K. G. New J. Chem. 1994, 18, 25-41. (b) Caulton, K.
G. ChemtractssInorg. Chem. 1993, 5, 170-172.

11406 J. Am. Chem. Soc., Vol. 121, No. 49, 1999
Hascall et al.
Table 5. Halogen Dependence of the Thermodynamics for Addition of H2 to a Metal Center
reactant
product
∆H° (kcal mol-1
)
∆S° (eu)
refs
Ir(PPh3)2(CO)X
X ) F
Ir(PPh3)2(CO)H2X
25, 27bb
>-10 [-13.6]a
X ) Cl
-14.4(8) [-22.0]a
-17.0(6) [-24.1]a
-18.8(8) [-27.3]a
-26(2.4)
-31(2.0)
-27(3.2)
X ) Br
X ) I
Ir(PBut2Ph)2H2X
X ) Cl
Ir(PBut2Ph)2H2(η2-H2)X
Ir(PPri3)2H2(η2-H2)X
41b
56c
-6.8(2)
-7.9(9)
-9.3(2)
-19.2(7)
-19.7(3.2)
-22.7(1.8)
X ) Br
X ) I
Ir(PPri3)2H2X
X ) Cl
-12.3(3)
-12.0(2)
-10.2(4)
-37(3)
-33(3)
-23(3)
X ) Br
X ) I
a Values in brackets are computed ∆E(MP4) values for addition of H2 to Ir(PH3)2(CO)X. b Reference 62, p 187. c Equilibrium data listed are for
n-hexane solution. Values for toluene and methylcyclohexane were also reported.
(Table 5). Although this observation seems paradoxical, calcula-
tions by Krogh-Jespersen and Goldman on addition of H2 to
trans-Ir(PH3)2(CO)X (X ) Cl, Br, I, NH2, PH2) indicate that
the presence of occupied π-orbitals on X strongly disfavors
oxidative addition.38 Specifically, π-interactions have the ability
to stabilize the “16-electron” trans-Ir(PH3)2(CO)X reactant39 but
destabilize the 18-electron dihydride product trans-Ir(PPh3)2-
(CO)(X)H2 via “dπ-pπ filled-filled repulsions”.34a,40 Such
interactions would be greatest for the fluoride derivative, which
therefore exhibits the least tendency to oxidatively add H2.41
An important issue, therefore, is concerned with why the “16-
electron” molybdenum and tungsten complexes M(PMe3)4X2
do not behave analogously to the Vaska system and instead
exhibit a completely opposite trend with respect to oxidative
addition.
Since oxidative addition of H2 involves an increase in
coordination number, steric interactions would be expected to
be least favorable for the largest halide derivative. The fact that
the largest halide derivative, trans-Ir(PPh3)2(CO)I, oxidatively
adds H2 most exothermically indicates that electronic factors
(as described above) must dominate over the steric factors and
be responsible for the observed trend in the Vaska system. For
the molybdenum and tungsten systems, however, steric effects
may be expected to be more influential in determining the
thermodynamics since the product of oxidative addition is eight-
coordinate, rather than six-coordinate as in Ir(PPh3)2(CO)(X)-
H2. Thus, the observed halide dependence for oxidative addition
of H2 to M(PMe3)4X2 is consistent with the thermodynamics
being dictated by steric effects. However, there is also an
electronic rationalization for the observed trend, which is
concerned with the fact that the four π-symmetry orbitals of
the two trans-X atoms in M(PMe3)4X2 (with D2d symmetry)
belong to two sets of E symmetry orbitals, one pair of which
interacts with the dxz and dyz orbitals (Figure 4). Since the
antibonding component of this interaction (the metal-based 3e*
HOMO) is occupied by two electrons, the energy of the system
is raised as the strength of the π-donor interaction increases
(35) For example: (i) the enthalpy of protonation of a variety of [CpR]M-
(PR3)X (M ) Ru, Os) derivatives indicates that the basicity of the metal
center increases in the sequence I < Br < Cl.a,b (ii) Spectroscopic and
electrochemical studies on trans-M(PPh3)2(CO)X (M ) Rh, Ir),c CpRe-
(NO)(PPh3)X,d and CpRe(NO)(CO)Xd indicate that the fluoro complexes
have the most electron-rich metal centers. Curiously, however, the opposite
trend is observed for [(tacn)Re(NO)(CO)X]+, for which ν(CO) and ν(NO)
1
stretching frequencies decrease in the order F > Cl > Br > I.e (iii) JSi-H
coupling constants and νSi-H stretching frequencies for Cp2Zr(NBut-
SiMe2H)X, which exhibit three-center, two-electron Zr-H-Si interactions,
increase in the sequence H < I < Br < Cl < F, indicating that the zirconium
center of the fluoro complex exhibits the lowest electrophilicity.f (iv) ESR
and electronic spectroscopic studies on Cp*2TiX complexes indicate that
the π-interactions increase in the sequence I < Br < Cl < F.g (v)
Electrochemical studies on Cp2MX2 (M ) Nb, Ta)h and Cp2W(R)Xi indicate
that the ease of oxidation increases in the sequence I < Br < Cl. (a) Rottink,
M. K.; Angelici, R. J. J. Am. Chem. Soc. 1993, 115, 7267-7274. (b) Rottink,
M. K.; Angelici, R. J. J. Am. Chem. Soc. 1992, 114, 8296-8298. (c)
Doherty, N. M.; Hoffman, N. W. Chem. ReV. 1991, 91, 553-573. (d)
Agbossou, S. K.; Roger, C.; Igau, A.; Gladysz, J. A. Inorg. Chem. 1992,
31, 419-424. (e) Pomp, C.; Wieghardt, K. Inorg. Chem. 1988, 27, 3796-
3804. (f) Procopio, L. J.; Carroll, P. J.; Berry, D. H. J. Am. Chem. Soc.
1994, 116, 177-185. (g) Lukens, W. W., Jr.; Smith, M. R., III; Andersen,
R. A. J. Am. Chem. Soc. 1996, 118, 1719-1728. (h) Hunter, J. A.; Lindsell,
W. E. McCullough, K. J.; Parr, R. A.; Scholes, M. L. J. Chem. Soc., Dalton
Trans. 1990, 2145-2153. (i) Asaro, M. F.; Cooper, S. R.; Cooper, N. J. J.
Am. Chem. Soc. 1986, 108, 5187- 5193.
(39) The precise nature of π-donation in square planar d8 complexes
2
2
deserves some comment since the only unoccupied d orbital, i.e. the dx _y ,
lies along the ligand axes and so is incapable of π-overlap with halogen p
orbitals. However, in the presence of a trans π-acceptor ligand such as
CO, interactions with dyz and dxz (with the halogen located on the z axis)
become possible via a “push/pull” mechanism.a It is also possible that
π-interactions may occur with metal p orbitals. For example, calculations
suggest that the PH2 ligand in (H3P)3RhPH2 is coplanar with the RhP3 group
due to favorable P(p)-Rh(p) π-interactions.b In contrast, however, calcula-
tions suggest that there are no substantial π-interactions involving a Pt(p)
orbital in (H3P)2Pt(H)(NH2).c (a) Reference 34a. (b) Dahlenburg, L.; Ho¨ck,
N.; Berke, H. Chem. Ber. 1988, 121, 2083-2093. (c) Cowan, R. L.; Trogler,
W. C. J. Am. Chem. Soc. 1989, 111, 4750-4761.
(40) The overlap of filled orbitals results in a repulsive interaction because
the antibonding orbital is destabilized to a greater extent than the bonding
orbital is stabilized. For example, the overlap of two identical orbitals results
in a bonding orbital that is stabilized by an energy ∆/(1 + S), while the
antibonding combination is destabilized by ∆/(1 - S), where ∆ is the
interaction energy and S is the overlap integral. See: Albright, T. A.; Burdett,
J. K.; Whangbo, M.-H. Orbital Interactions in Chemistry; Wiley: New
York, 1985.
(41) Goldman and Krogh-Jespersen have also used ab initio calculations
to study the transition-state properties for addition of H2 to Vaska-type
complexes, trans-Ir(L)2(CO)X (L ) PH3 and X ) F, Cl, Br, I, CN, H), and
have concluded that strong π-donation increases the barrier to oxidative
addition.a Likewise, Caulton has reported that the activation barriers for
coordinating H2 to Ir(PBut2Me)2H2X increase in the order I < Br < Cl,
with the better π-donor inhibiting binding.b Conversely, the activation barrier
for elimination of H2 follows the opposite trend, with Cl having the lower
barrier since π-donation promotes elimination. (a) Abu-Hasanayn, F.;
Goldman, A. S.; Krogh-Jespersen, K. J. Phys. Chem. 1993, 97, 5890-
5896. (b) Hauger, B. E.; Gusev, D.; Caulton, K. G. J. Am. Chem. Soc.
1994, 116, 208-214.
(36) ν(CO)/cm-1 for trans-Ir(PPh3)2(CO)X: 1957 (F), 1965 (Cl), 1966
(Br), and 1967 (I). See: Vaska, L.; Peone, J., Jr. J. Chem. Soc., Chem.
Commun. 1971, 418-419.
(37) Goldman and Krogh-Jespersen have offered another suggestion to
account for the observed variation of ν(CO) stretching frequencies in the
Vaska system. Specifically, they have attributed the observed ν(CO) trend
to the greater electrostatic field strength of the lighter halides, which raises
the energy of the Ir d orbitals and, as a consequence, promotes π-back-
bonding to CO. See ref 27b.
(38) Whereas occupied π-orbitals on the halide substituents strongly
disfavor oxidative addition to Ir(PR3)2(CO)X, Goldman and Krogh-Jespersen
have calculated that substituents with π-acceptor orbitals, such as BH2,
would promote oxidative addition by stabilizing the 18-electron product.
See ref 27b.

OxidatiVe Addition of H2 to Mo and W Complexes
J. Am. Chem. Soc., Vol. 121, No. 49, 1999 11407
Figure 4. Qualitative MO diagrams for M(PMe3)4X2 (D2d) and M(PMe3)4H2X2 (Cs). The symmetry labels in parentheses for M(PMe3)4H2X2 are
those for idealized C2V symmetry.
(Figure 4).42 Consequently, ground-state destabilization of
M(PMe3)4X2 electronically promotes oxidative addition for the
better π-donor. This marked instability of the fluoride complexes
M(PMe3)4F2 with respect to M(PMe3)4H2F2 has a close parallel
with the observation that terminal oxo complexes of the late
transition metals are very rare.43 Specifically, Mayer has
rationalized the paucity of such complexes to be a result of
repulsive π-interactions between oxygen “lone pairs” and filled
d orbitals.44
In an attempt to differentiate the electronic and steric
components, it is instructive to consider the oxidative addition
as proceeding via a hypothetical sequence involving first a
structural reorganization of trans-M(PMe3)4X2 to a cis geometry
(with the PMe3 and X ligands located in the same positions
that they occupy in eight-coordinate M(PMe3)4H2X2), followed
by addition of H2. The relative energies for these processes are
summarized in Figure 5 for the tungsten system, with the energy
of the hypothetical cis-W(PMe3)4X2 entities arbitrarily fixed at
a common value for comparison purposes.
Figure 5. Structural reorganization and intrinsic bond energy com-
ponents for oxidative addition of H2. The energies of cis-W(PMe3)4X2
are arbitrarily fixed at a common value.
To a first approximation, steric factors would be expected to
play a more important role in the structural reorganization step
than in the oxidative addition step. This statement is not intended
to imply that electronic factors do not play a role in the structural
reorganization step. Indeed, π-interactions would be expected
to destabilize a cis geometry for a d4 octahedral complex of the
type ML4X2, by virtue of the fact that a formally nonbonding
occupied orbital (b2g) in the trans arrangement becomes
antibonding in the cis structure, as illustrated schematically in
Figure 6. As such, the strongest π-donor ligand, fluorine, would
be expected to electronically disfavor structural rearrangements
(42) Calculations on the series of d0 complexes, Cp2TiX2 (X ) F, Cl,
Br, I), also indicate that the orbital energies vary in a similar way to those
in Mo(PMe3)4X2. See: Bruce, M. R. M.; Kenter, A.; Tyler, D. R. J. Am.
Chem. Soc. 1984, 106, 639-644.
(43) See, for example: (a) Nugent, W. A.; Mayer, J. M. Metal-Ligand
Multiple Bonds; Wiley-Interscience: New York, 1988. (b) Trnka, T. M.;
Parkin, G. Polyhedron 1997, 16, 1031-1045. (c) Parkin, G. Prog. Inorg.
Chem. 1998, 47, 1-165.
(44) Mayer, J. M. Comments Inorg. Chem. 1988, 8, 125-135.

11408 J. Am. Chem. Soc., Vol. 121, No. 49, 1999
Hascall et al.
Figure 6. Qualitative MO diagrams for cis and trans ML4X2.
from a trans structure.45 Since electronic factors disfavor a trans-
to-cis rearrangement most for the fluoride derivative, the result
that the first step of the hypothetical transformation is least
endothermic for the fluoride derivative strongly suggests that
the structural reorganization is dominated by steric factors.
The second step of the hypothetical transformation is also
most exothermic for the fluoride derivative; this step corresponds
to the intrinsic W-H bond energy of W(PMe3)4H2X2, i.e., that
without allowing for molecular relaxation. Since the hydride
ligand is small, the energetics of the latter step would be
anticipated to be more a reflection of electronic factors, and
the trend is consistent with the notion that π-donation within
W(PMe3)4X2 destabilizes the system with respect to oxidative
addition by population of antibonding levels. It is therefore
suggested that steric and electronic factors may serve to reinforce
each other in promoting oxidative addition most for the fluoride
derivative W(PMe3)4F2.
Support for the notion that electronic factors contribute to
making oxidative addition most favored for the fluoride deriva-
tives is provided by analyzing the variation in energies of the
highest occupied molecular orbitals of M(PMe3)4X2 and
M(PMe3)4H2X2. Plots of the energies of these orbitals obtained
from the DFT calculations are shown in Figure 7. The HOMOs
for each of the M(PMe3)4X2 species are a pair of half-occupied
degenerate 3e* orbitals, corresponding to the metal-based dxz
and dyz orbitals that are antibonding by virtue of their π-interac-
tion with the halogen p orbitals. Notably, the energy of the
HOMO increases dramatically from the iodo to fluoro deriva-
tives (Figure 7), due to enhanced π-interactions, so that its
occupation strongly destabilizes M(PMe3)4X2 with respect to
M(PMe3)4H2X2. It is important to point out that the 3a′′* HOMO
of M(PMe3)4H2X2, which is principally the metal dyz orbital,46
is less sensitive to the nature of the halogen than is the 3e*
Figure 7. Variation of the highest occupied orbitals in W(PMe3)4X2
and W(PMe3)4H2X2. The Mo system is similar.
(45) Arguments of this type have been used previously to rationalize
the trans (as opposed to cis) arrangement of octahedral d2 dioxo complexes,
[ML4(O)2]. See, for example, ref 43a.
(46) For M(PH3)4H2X2 (C2V), the HOMO is of a2 symmetry and is
principally the dxy orbital.
HOMO in M(PMe3)4X2, so that its occupation does not dictate
the thermodynamics of the interconversion (Figure 7). In

OxidatiVe Addition of H2 to Mo and W Complexes
J. Am. Chem. Soc., Vol. 121, No. 49, 1999 11409
Figure 8. Qualitative MO diagrams for M(PMe3)4(NH2)2 with parallel and orthogonal orientations of the NH2 ligands.
essence, “filled-filled” repulsions are more dominant in
M(PMe3)4X2 than in M(PMe3)4H2X2,47 with the result that
oxidative addition is most favored for the fluoride derivatives.
It is noteworthy that, although the average M-H BDEs for
M(PMe3)4H2X2 increase in the sequence I < Br < Cl < F, the
experimental νW-H stretching frequencies for W(PMe3)4H2X2
exhibit the opposite trend: F (1835 and 1900 cm-1), Cl (1933
cm-1), Br (1934 cm-1), I (1961 cm-1).48,49 Thus, counterintu-
itively, M-H stretching frequencies in a related series of
complexes are not necessarily good indicators of trends in bond
strengths.50 In this regard, it should be noted that stronger bonds
do not always correlate with shorter bonds of the same type.51
A further illustration of the way in which π-donation
influences oxidative addition of H2 is provided by calculations
on two isomers of trans-M(PMe3)4(NH2)2, which differ accord-
ing to whether the NH2 groups are parallel (C2V symmetry) or
orthogonal (D2d symmetry) to each other. Unlike halide sub-
stituents that have two π-donor orbitals, the NH2 group has only
a single π-donor function. Furthermore, depending upon the
relative orientation of the two NH2 groups, the π-interactions
within trans-M(PMe3)4(NH2)2 could be with either a single d
orbital or two different d orbitals, as illustrated in Figure 8.
The former possibility is in marked contrast to the halide
complexes M(PMe3)4X2, for which π-interactions must occur
in a pairwise manner with two d orbitals.52 Significantly,
oxidative addition of H2 to the orthogonal (D2d) isomer of the
tungsten complex W(PMe3)4(NH2)2 is calculated to be substan-
tially more exothermic by 13.9 kcal mol-1 than that for the
parallel (C2V) arrangement.53 Since the product of oxidative
addition, i.e., W(PMe3)4H2(NH2)2, is the same for both reactions,
the difference in exothermicity directly reflects the relative
stabilities of the two isomers of W(PMe3)4(NH2)2. Examination
of Figure 8 indicates that the parallel (C2V) isomer is the more
stable because the p orbitals of both NH2 groups interact with
the same metal d orbital, such that the “d4” electrons occupy
two formally nonbonding d orbitals. In contrast, the p orbitals
(47) Presumably this is a consequence of increased π-overlap for the
trans MX2 fragment as compared to that for a bent geometry (see Figures
4 and 5).
(48) (a) Reference 4. (b) Green, M. L. H.; Parkin, G.; Chen, M.; Prout,
K. J. Chem. Soc., Dalton Trans. 1986, 2227-2236.
(49) Values in toluene solution are F, 1839 and 1882 cm-1; Cl, 1917
cm-1; Br, 1933 cm-1; and I, 1959 cm-1. Computational studies on
W(PH3)4H2X2 reproduce this trend. Thus, calculated νW-H stretching
frequencies (A1 and B2 symmetry) for W(PH3)4H2X2 are F, 1844 and 1835
cm-1; Cl, 1893 and 1883 cm-1; Br, 1897 and 1892 cm-1; and I 1900 and
(51) See, for example: (a) Ernst, R. D.; Freeman, J. W.; Stahl, L.; Wilson,
D. R.; Arif, A. M.; Nuber, B.; Ziegler, M. L. J. Am. Chem. Soc. 1995, 117,
5075-5081. (b) Tudela, D. J. Chem. Educ. 1996, 73, A297. (c) Boyd, S.
L.; Boyd, R. J. J. Am. Chem. Soc. 1997, 119, 4214-4219. (d) Mayer, P.
M.; Glukhovtsev, M. N.; Gauld, J. W.; Radom, L. J. Am. Chem. Soc. 1997,
119, 12889-12895. (e) Bowmaker, G. A.; Schmidbaur, H.; Kru¨ger, S.;
Ro¨sch, N. Inorg. Chem. 1997, 36, 1754-1757. (f) McKean, D. C.; Torto,
I.; Morrison, A. R. J. Phys. Chem. 1982, 86, 307-309. (g) Vollhardt, K. P.
C.; Cammack, J. K.; Matzger, A. J.; Bauer, A.; Capps, K. B.; Hoff, C. D.
Inorg. Chem. 1999, 38, 2624-2631.
1892 cm-1
.
(50) The experimental and calculated W-H stretching frequencies do,
however, qualitatively correlate with the calculated W-H bond lengths listed
in Table S5, in accord with Badger’s rule, k ) [aij(re - dj)]-3, where k is
the force constant, re is the equilibrium bond length, and aij and dij are
constants.a-d Exceptions to Badger’s rule are, nevertheless, known.e (a)
Badger, R. M. J. Chem. Phys. 1934, 2, 128-131. (b) Badger, R. M. J.
Chem. Phys. 1935, 3, 710-714. (c) Herschbach, D. R.; Laurie, V. W. J.
Chem. Phys. 1961, 35, 458-463. (d) Pauling, L. The Nature of The
Chemical Bond, 3rd ed.; Cornell University Press: Ithaca, NY, 1960; p
231. (e) Christen, D.; Gupta, O. D.; Kadel, J.; Kirchmeier, R. L.; Mack, H.
G.; Oberhammer, H.; Shreeve, J. M. J. Am. Chem. Soc. 1991, 113, 9131-
9135.
(52) For an example of the influence of selective π-interactions stabilizing
unusual oxo complexes, see: Bursten, B. E.; Cayton, R. H. Organometallics
1987, 6, 2004-2005.
(53) ∆E for addition to the parallel (C2V) isomer is -3.1 kcal mol-1
while that for addition to the orthogonal (D2d) isomer is -17.0 kcal mol-1
,
.

11410 J. Am. Chem. Soc., Vol. 121, No. 49, 1999
Hascall et al.
of the NH2 groups of the orthogonal (D2d) isomer interact with
two different metal d orbitals, so that only one d orbital remains
formally nonbonding. As a consequence, a pair of electrons is
forced to occupy an antibonding orbital (i.e., a “filled-filled”
repulsion), which destabilizes the structure relative to that of
the parallel (C2V) isomer. The latter situation is closely analogous
to that in the halide M(PMe3)4X2 derivatives. Therefore,
consideration of the stabilities of the two isomers of W(PMe3)4-
(NH2)2, in which differences in steric interactions are likely to
be negligible, emphasizes the important role that π-interactions
have in influencing the energetics of oxidative addition of H2
to M(PMe3)4X2 complexes.
exclusively responsible for modifying the electron richness of
a metal center.
The more facile oxidation of M(PMe3)4X2 to M(PMe3)4H2X2
for the lighter halogens bears an interesting comparison with
the electrochemical oxidation of the quadruply bonded dinuclear
complexes Mo2X4(PR3)4 (X ) Cl, Br, I), for which the ease of
oxidation increases in the sequence I < Br < Cl.54 The
unexpected halide dependence was proposed to be a result of
metal-to-halide back bonding, which serves to lower the energy
of the δ and δ* orbitals, an effect that should be greatest for
the iodide derivative. An alternative possibility, however, which
is more in line with the rationalization for the ease of oxidative
addition of H2 to M(PMe3)4X2, is that the trend reflects the
greater π-donor abilities of the lighter halogens, which serve
to destabilize metal d orbitals more and thereby promote
oxidation.
In addition to the influence of the halide ligand on the
thermodynamics of the oxidative addition of H2, the energetics
of forming dihydrogen complexes has also been briefly de-
scribed in the literature (Table 5). The situation with respect to
Ir(PR3)2H2X derivatives, however, is conflicting. For example,
Caulton has reported that addition of H2 to Ir(PBut2Ph)2H2X,
giving the dihydrogen complexes Ir(PBut2Ph)2H2(η2-H2)X,
becomes more exothermically favored in the sequence Cl < Br
< I.41b,55 As with oxidative addition to the Vaska-type com-
plexes, Caulton has concluded that π-donation by X stabilizes
“16-electron” Ir(PBut2Ph)2H2X, thereby resulting in H2 addition
being less exothermic than those in the absence of π-donation.
In contrast to Caulton’s Ir(PBut2Ph)2H2X system, Jensen has
reported that the exothermicity of the closely related reaction
of Ir(PPri3)2H2X with H2 increases in the opposite sequence,
i.e., I < Br < Cl.56,57 At the same time, however, the
“thermodynamic stabilization of the dihydrogen ligand”, as
judged by the equilibrium ratios of Ir(PPri3)2H2(η2-H2)X:Ir-
(PPri3)2H2X, has been noted to increase in the sequence, Cl <
Br < I,58,59 which is in accord with the equilibrium constant
data for the Ir(PBut2Ph)2H2X system.60,61
A principal distinction between the M(PMe3)4X2 and Vaska
systems is that, because of the molecular symmetry, π-donation
within M(PMe3)4X2 must occur in a pairwise manner; i.e., it is
not possible to partake in a selective π-interaction with a single
metal orbital. In contrast, the molecular symmetry of trans-Ir-
(PPh3)2(CO)X is such that the two p orbitals on X are not
degenerate, so that it is possible for a selective π-interaction
between a p orbital on X and a single metal orbital to occur
without requiring excessive “filled-filled” repulsions with the
other nonbonding X p orbital. An alternative, simplified, view
of the situation is that π-donation within 16-electron trans-Ir-
(PPh3)2(CO)X may stabilize the molecule by transforming it
into an “18-electron” complex,39 whereas π-donation within 16-
electron M(PMe3)4X2 must destabilize the molecule by trans-
forming it into a “20-electron” complex. However, it is essential
to also consider the factors influencing the stabilities of the
products of oxidative addition in order to rationalize fully the
halide dependence. Having 18-electron configurations, both
M(PMe3)4H2X2 and trans-Ir(PPh3)2(CO)(H)2X will exhibit
“filled-filled” repulsions, but such interactions are likely to be
exacerbated for the d6 Vaska system since the octahedral
geometry is ideally suited for π-interactions with both halogen
p orbitals. In contrast, as a result of the less regular symmetry
of the eight-coordinate complexes M(PMe3)4H2X2, not all of
the halogen p orbitals interact efficiently with the metal d
orbitals; thus, unfavorable “side-on” overlap, of the type
illustrated by the 3a′′(a2) HOMO in Figure 4, results in less
destabilization. Consequently, “filled-filled” repulsions most
likely destabilize trans-Ir(PPh3)2(CO)(H)2X to a greater extent
than those for M(PMe3)4H2X2 with respect to reductive elimina-
tion of H2.
An Inverse Equilibrium Isotope Effect for Oxidative
Addition of H2 to W(PMe3)4I2. Rather surprisingly, the
oxidative addition of H2 to W(PMe3)4I2 constitutes the first
(54) Zietlow, T. C.; Hopkins, M. D.; Gray, H. B. J. Am. Chem. Soc.
1986, 108, 8266-8267.
(55) For theoretical calculations on this system, see: Clot, E.; Eisenstein,
O. J. Phys. Chem. A 1998, 102, 3592-3598.
(56) (a) Lee, D. W.; Jensen, C. M. Inorg. Chim. Acta 1997, 259, 359-
362. (b) Lee, D. W.; Jensen, C. M. J. Am. Chem. Soc. 1996, 118, 8749-
8750.
The electronic components of the oxidative addition of H2
to M(PMe3)4X2 and trans-Ir(PPh3)2(CO)X may, therefore, be
rationalized according to the influence of “filled-filled” repul-
sions on the reactants and products. In essence, π-antibonding
interactions are proposed to be (i) greater in M(PMe3)4X2 than
those in Ir(PPh3)2(CO)X and (ii) greater in Ir(PPh3)2(CO)(H)2X
than those in M(PMe3)4H2X2. As a result, π-donor ligands exert
an opposite influence on the oxidative addition of H2 to
M(PMe3)4X2 and trans-Ir(PPh3)2(CO)X: strong π-donors favor
addition to M(PMe3)4X2 but disfavor addition to trans-Ir(PPh3)2-
(CO)X. The strikingly different effects that result from π-dona-
tion in these “16-electron” systems, M(PMe3)4X2 and Ir(PPh3)2-
(CO)X, serve to emphasize the importance of considering the
detailed nature of metal-ligand interactions in predicting and
understanding the reactivity of a metal center. Thus, ancillary
ligand variations do not necessarily influence a given reaction
type (oxidative addition, in the present case) in the same way
for different systems. The aforementioned common notion that
electron-donating ligands promote oxidative addition reactions
is, therefore, only likely to be universally true if σ effects are
(57) Jensen has also noted that the exothermicity of the addition of H2
to Ir(PPri3)2H2X in toluene solution is significantly less than that of the
corresponding reaction in an alkane. Likewise, the entropy change for
addition of H2 is less negative in toluene than in an alkane. These
observations have been proposed to be a result of arene solvation stabilizing
each of the five-coordinate Ir(PPri3)2H2X species. Furthermore, the lower
exothermicity for addition of H2 to the iodide complex in alkane solvent
with respect to that for the chloride and bromide has been suggested to
indicate preferential alkane coordination to the five-coordinate iodide (ref
57).
(58) Le-Husebo, T.; Jensen, C. M. Inorg. Chem. 1993, 32, 3797-3798.
(59) Specifically, under H2 (0.5 atm) at -80 °C, the reported IrXH2(η2-
H2)(PPri3)2:IrXH2(PPri3)2 ratios are 0.56 (Cl), 0.91 (Br), and 1.67 (I).
(60) Furthermore, the [Ir(η2-H2)] interaction in Ir(PPri3)2H2(η2-H2)X (X
) Cl, Br, I) has been suggested to be strongest for the iodide derivative on
the basis that the latter complex exhibits the highest barrier for rotation of
the H2 ligand [i.e., 0.98(5) kcal mol-1 for Ir(PPri3)2H2(η2-H2)I, as compared
to 0.51(2) kcal mol-1 for the chloride derivative]. See: Eckert, J.; Jensen,
C. M.; Koetzle, T. F.; Husebo, T. L.; Nicol, J.; Wu, P. J. Am. Chem. Soc.
1995, 117, 7271-7272.
(61) A possible explanation for the discrepency of the ∆H° values for
the two Ir(PR3)2H2(η2-H2)X systems may be associated with the equilibrium
constant data having being measured over a relatively small temperature
range.

OxidatiVe Addition of H2 to Mo and W Complexes
J. Am. Chem. Soc., Vol. 121, No. 49, 1999 11411
bond energy.68 With this perspective, the inverse nature of the
deuterium equilibrium isotope effect for oxidative addition of
H2 to W(PMe3)4I2 must be regarded as counterintuitive.
However, it is now apparent that the occurrence of an inverse
deuterium equilibrium isotope effect for addition of H2 to a
single transition metal center (be it the formation of a dihydride
or dihydrogen complex) is a rather general phenomenon (Table
6).69
The temperature dependence of K reveals that the origin of
the inverse equilibrium deuterium isotope effect is enthalpic,
with oxidative addition of D2 being more exothermic than
oxidative addition of H2 [∆H°D ) -21.6(7) kcal mol-1 versus
∆H°H ) -19.7(6) kcal mol-1]. As a consequence, the W-D
bond is substantially stronger than the W-H bond [D(W-D)
) 63.8(7) kcal mol-1 versus D(W-H) ) 62.0(6) kcal mol-1].
The entropic contribution to the equilibrium isotope effect is
small [∆S°D ) -51(3) eu versus ∆S°H ) -45(2) eu] but
actually attempts to counter the inverse nature. The source of
this small difference in ∆S° for the oxidative addition of H2
and D2 is presumably the greater entropy of D2 compared to
that of H2 (39.0 and 34.0 eu, respectively, at 300 K).70
Examination of the data in Table 6 indicates that the inverse
equilibrium isotope effects observed in other systems are also
due to an enthalpic influence (i.e., ∆H°H - ∆H°D > 0), with
the small entropic contributions attempting to counter the effect
and favor addition of H2, i.e., ∆S°H - ∆S°D > 0.
Figure 9. Normal and inverse equilibrium isotope effects.
literature report3 of a system for which the equilibrium isotope
effect for oxidative addition of H2 to a transition metal complex
was determined.62,63 It is, therefore, noteworthy that the oxidative
addition of D2 to W(PMe3)4I2 is characterized by a substantial
inVerse equilibrium deuterium isotope effect, with KH/KD
)
0.63(5) at 60 °C (Table 2). Prior to this example, the only other
literature report of which we are aware is the composite primary
and secondary equilibrium deuterium isotope for the oxidative
addition of dihydrogen to Cp2Ta(µ-CX2)2Ir(CO)(PPh3) (X )
H, D).64 For this latter system, however, rapid deuterium
exchange into the [Ta(µ-CH2)2Ir] methylene groups prevented
determination of the primary effect.65
The interpretation of equilibrium isotope effects is normally
discussed in terms of four factors, namely SYM (a symmetry
number factored out of the rotational partition function ratio),
MMI (a mass and moment of inertia contribution due to changes
in rotational and translational motion), EXC (an excitation
contribution accounting for population of vibrational levels),
and ZPE (a zero point energy term).71,72 Of these terms, SYM,
The inverse nature of the deuterium equilibrium isotope effect
(i.e., KH/KD < 1) for oxidative addition of H2 is particularly
interesting since it is completely counter to that which would
have been predicted on the basis of the simple notion concerning
primary isotope effects, namely that deuterium prefers to reside
in the stronger bond (i.e., the higher energy oscillator). Thus,
with a D-D bond which is 1.8 kcal mol-1 stronger than the
H-H bond,14 a normal (i.e., KH/KD > 1) equilibrium isotope
effect may have been predicted for the oxidative addition of
H2 to a transition metal complex (Figure 9). For example, the
equilibrium between a dihydride [MH2] and a dihydrogen [M(η2-
H2)] complex was argued to be inverse on the basis of the notion
that M-H (M-D) bonds are weaker than H-H (D-D)
bonds.66,67 Furthermore, solid-state metal deuterides are normally
less stable than the corresponding hydrides with respect to
elimination of D2 (H2) due to the greater D-D versus H-H
EIE ) SYM‚MMI‚EXC‚ZPE
MMI, and EXE relate to the entropy of the EIE, while ZPE
provides the major contribution for the enthalpy. Most com-
monly, the ZPE term is considered to dominate the EIE, which
(67) Other examples which support this notion include the observations
that the heavier isotope (2H or 3H) favors the dihydrogen site in (i) [Re-
(PMe2Ph)3(CO)(η2-H2)H2]+ a and (ii) CpNb(CO)3(η2-H2) versus CpNb-
(CO)3H2.b Furthermore, an inverse kinetic isotope effect for reductive
elimination of H2 from cis-Pt(PMe3)2H2 has been interpreted in terms of a
preequilibrium with cis-Pt(PMe3)2(η2-H2) in which deuterium prefers to be
located in the dihydrogen species.c However, more recent work indicates
that this prediction is oversimplified, with several counter examples having
been reported. Thus, the heavier isotope favors the hydride sites in [TpRR′]-
Ir(PR3)(η2-H2)H]+ d,e and [Cp2W(η2-H2)H]+.f Furthermore, using assumed
values for the EIE for coordination of H2 and for oxidative addition of H2
in a hypothetical system, the heavier isotopomer has been predicted to favor
the terminal hydride site.g It is, therefore, evident that the factors which
influence the preference for deuterium (or tritium) to occupy either a hydride
or dihydrogen site are complex. (a) Luo, X-L.; Crabtree, R. H. J. Am. Chem.
Soc. 1990, 112, 6912-6918. (b) Haward, M. T.; George, M. W.; Hamley,
P.; Poliakoff, M. J. Chem. Soc., Chem. Commun. 1991, 1101-1103. (c)
Packett, D. L.; Trogler, W. C. Inorg. Chem. 1988, 27, 1768-1775. (d)
Heinekey, D. M.; Oldham, W. J., Jr. J. Am. Chem. Soc. 1994, 116, 3137-
3138. (e) Oldham, W. J., Jr.; Hinkle, A. S.; Heinekey, D. M. J. Am. Chem.
Soc. 1997, 119, 11028-11036. (f) Henderson, R. A.; Oglieve, K. E. J. Chem.
Soc., Dalton Trans. 1993, 3431-3439. (g) Bender, B. R.; Kubas, G. J.;
Jones, L. H.; Swanson, B. I.; Eckert, J.; Capps, K. B.; Hoff, C. D. J. Am.
Chem. Soc. 1997, 119, 9179-9190.
(62) The equilibrium deuterium isotope effect for the oxidative addition
of H2 to trans-Ir(PPh3)2(CO)Cl has been reported: Werneke, M. F. Ph.D
Thesis, Clarkson College of Technology, Potsdam, NY, 1971.
(63) For reviews of isotope effects in reactions of transition metal
hydrides, see: (a) Bullock, R. M. In Transition Metal Hydrides; Dedieu,
A., Ed.; VCH: New York, 1992; pp 263-307. (b) Rosenberg, E. Polyhedron
1991, 8, 383-405.
(64) Hostetler, M. J.; Bergman, R. G. J. Am. Chem. Soc. 1992, 114,
7629-7636.
(65) The deuterium equilibrium isotope effect for addition of Et3SiH to
Cp2Ta(µ-CH2)2Ir(CO)(PPh3) has also been shown to be inverse, ranging
from 0.53 ( 0.04 at 0 °C to 0.77 ( 0.06 at 60 °C. As with addition of H2,
these values also are a composite of primary and secondary effects. However,
for this particular system, the primary and secondary isotope contributions
were estimated to be 0.73 and 0.74, respectively, on the basis of IR
calculations for the primary effect. See ref 64.
(68) However, in some instances, e.g., VH2, NbH2, (V,Nb)H2, and
LaNi5H6, the deuterides are more stable. See: (a) Wiswall, R. H., Jr.; Reilly,
J. J. Inorg. Chem. 1972, 11, 1691-1696. (b) Luo, W.; Clewley, J. D.;
Flanagan, T. B. J. Phys. Chem. 1990, 93, 6710-6722.
(69) An exception appears to be binuclear oxidative addition of H2, which
has recently been reported to exhibit a normal equilibrium isotope effect.
See ref 51g.
(66) (a) Packett, D. L.; Trogler, W. C. J. Am. Chem. Soc. 1986, 108,
5036-5038. (b) Packett, D. L.; Trogler, W. C. Inorg. Chem. 1988, 27,
1768-1775. (c) Howarth, O. W.; McAteer, C. H.; Moore, P.; Morris, G.
E. J. Chem. Soc., Dalton Trans. 1984, 1171-1180. (d) Reference 1a, p
331.

11412 J. Am. Chem. Soc., Vol. 121, No. 49, 1999
Hascall et al.
Table 6. Deuterium Equilibrium Isotope Effects for Addition of H2 to Transition Metal Complexesa
reactant
W(PMe3)4I2
product
W(PMe3)4H2I2
KH/KD
∆H°H, ∆H°D ∆S°H, ∆S°D ∆H°H - ∆H°D ∆S°H - ∆S°D
refs
0.63(5) at 60 °C -19.7(6),
-21.6(7)
0.50 at -13 °C
0.37 at -13 °C
-45(2),
-51(3)
1.9(1.3)
6(5)
this work
Ir(PCy3)2HCl2
Ir(PCy3)2H(η2-H2)Cl
b
Ir(PBut2Me)2H2Cl
Ir(PBut2Me)2H2(η2-H2)Cl
-6.8(2),
-7.7(5)
-19.2(7),
0.9
1.8
1.5
5.3
41b
-20.7(1.8)
-24.7(2.0),
-30.0(2.0)
Cr(CO)3(PCy3)2
Cr(CO)3(PCy3)2(η2-H2)
0.65(15) at 22 °C -6.8(5),
-8.6(5)
67g
W(CO)3(PCy3)2(N2)
Os(PPri3)2(CO)(Cl)H
W(CO)3(PCy3)2(η2-H2)
Os(PPri3)2(CO)-
0.70(15) at 22 °C
67g
c
0.35 at 85°C
-14.1(5),
-14.8
-7.7(2)
-14.4(8)
-14.8(8)
-30(1),
0.7
0
(Cl)H(η2-H2)
-30
Ru(PPri3)2(CO)(Cl)H
Ir(PPh3)2(CO)Cl
Ru(PPri3)2(CO)(Cl)H(η2-H2)
Ir(PPh3)2(CO)ClH2
-23.2(1.0)
-26.0(2.4)
-28.0(2.3)
-23.7(6),
-25.9(1.2)
d
e
0.47 at 30 °C
0.8
2.0
Cp2Ta(µ-CH2)2Ir(CO)- Cp2Ta(µ-CH2)2IrH2(CO)-
(PPh3) (PPh3)
0.54(4) at 22 °C -12.0(2),
-13.0(4)
1.0(6)
2.2(1.8)
64
a The equilibrium isotope effects for reactions involving Ir(PCy3)2HCl2, Ir(PBut2Me)2H2Cl, Os(PPri3)2(CO)(Cl)H, and Cp2Ta(µ-CH2)2Ir(CO)(PPh3)
are complicated by secondary isotope effects due to deuterium exchange into the hydride sites. Furthermore, the value for W(CO)3(PCy3)(N2) is for
net binding and does not distinguish between dihydride and dihydrogen forms. b Gusev, D. G.; Bakhmutov, V. I.; Grushin, V. V.; Vol’pin, M. E.
Inorg. Chim. Acta 1990, 177, 115-120. c Bakhmutov, V. I.; Bertra´n, J.; Esteruelas, M. A.; Lledo´s, A.; Maseras, F.; Modrego, J.; Oro, L. A.; Sola,
E. Chem. Eur. J. 1996, 2, 815-825. Values for deuterated system calculated from the data presented in this reference. d Gusev, D. G.; Vymenits,
A. B.; Bakhmutov, V. I. Inorg. Chem. 1992, 31, 2-4. e Reference 62, pp 187 and 239. A value of 0.55(6) at 22 °C is reported in Abu-Hasanayn,
F.; Krogh-Jespersen, K.; Goldman, A. S. J. Am. Chem. Soc. 1993, 115, 8019-8023.
is the reason why it is normally expected that the isotopomer
with deuterium located in the site with the highest vibrational
frequency is the most stable. Specifically, for a diatomic
molecule, the difference in zero point energies, ∆E° ) ꢀ°H -
ꢀ°D, is directly proportional to the square of the force constant
(k),73 such that ∆ꢀ° increases with vibrational frequency. As a
consequence, the zero point energy stabilization for a system
is greatest when deuterium resides in the highest energy
oscillator; i.e., deuterium prefers to reside in the stronger bond
(Figure 9). As is evident from the above discussion, an inverse
equilibrium deuterium isotope effect for oxidative addition to
W(PMe3)4I2 would not have been predicted by consideration
of the zero point energy differences associated with the observed
W-H and W-D stretching frequencies alone. Specifically, the
observed νW-H and νW-D stretching frequencies of 1961 and
1416 cm-1 in W(PMe3)4H2I2 and W(PMe3)4D2I2, respectively,
result in a zero point energy difference of 273 cm-1 for this
vibrational mode. If the symmetric and asymmetric stretches
have similar frequencies, then the combined zero point energy
lowering would be less than the difference between D2 and H2
zero point energies (630 cm-1),74 so that a normal equilibrium
isotope effect would have been predicted. To rationalize the
enthalpic origin of the inverse primary equilibrium isotope
effect, it is important to consider bending modes associated with
the dihydride moiety, modes that are not normally invoked when
discussing primary75 isotope effects due to their low energy.76
However, even though bending modes are of low energy, the
fact that there are four77 such modes associated with the [WH2]
moiety (Figure 10) means that, in combination, they may provide
an important contribution. Thus, inclusion of the bending modes
results in a significant lowering of the zero point energy for
W(PMe3)4D2I2 with respect to W(PMe3)4H2I2, to the extent that
Figure 10. Vibrational modes associated with a [MH2] moiety.
an inverse equilibrium isotope effect may result. For example,
assuming a value of 600 cm-1 for each of the four [WH2]
deformations,76 and a corresponding value of ca. 425 cm-1 for
[WD2], each of the bending modes would contribute a difference
of ca. 175 cm-1 to the difference in zero point energies of
W(PMe3)4H2I2 and W(PMe3)4D2I2, so that the total difference
would be ca. 923 cm-1, rather than the 223 cm-1 predicted by
consideration of the W-H(D) stretching frequencies alone.
Since this difference of 923 cm-1 in zero point energies is
greater than that between D2 and H2 (630 cm-1), an inverse
deuterium equilibrium isotope effect will result (Figure 9).
Theoretical calculations support such a notion, with the ZPE
level of W(PMe3)4D2I2 calculated to be 1049 cm-1 (3.0 kcal
mol-1 78
lower than that of W(PMe3)4H2I2. At 60 °C, the
)
oxidative addition of D2 to W(PMe3)4I2 is calculated to be 0.86
kcal mol-1 more exothermic than that of H2,79 and the EIE is
predicted to be 0.73, in favorable agreement with the experi-
mental value of 0.63(5).
(76) Relatively little information is available concerning the bending
modes associated with transition metal hydride complexes, although the
bending modes in Mo(PR3)4H4 derivatives are observed in the range 600-
800 cm-1.a Similar values have also been suggested and calculated for other
systems.b-e (a) Polyakova, V. B.; Borisov, A. P.; Makhaev, V. D.;
Semenenko, K. D. Koord. Khim. 1980, 6, 743-752. (b) Girling, R. B.;
Grebenik, P.; Perutz, R. N. Inorg. Chem. 1986, 25, 31-36. (c) Macgregor,
S. A.; Eisenstein, O.; Whittlesey, M. K.; Perutz, R. N. J. Chem. Soc., Dalton
Trans. 1998, 291-300. (d) Reference 67g. (e) Jonas, V.; Thiel, W. J. Chem.
Phys. 1996, 105, 3636-3648. (f) Egdell, W. F.; Fisher, J. W.; Asato, G.;
Risen, W. M., Jr. Inorg. Chem. 1969, 8, 1103-1108. (g) Sweany, R. L.;
Russell, F. N. Organometallics 1988, 7, 719-727.
(70) Wooley, H. W.; Scott, R. B.; Brickwedde, F. G. J. Res. Nat. Bur.
Stand. 1948, 41, 379-475.
(71) Abu-Hasanayn, F.; Krogh-Jespersen, K.; Goldman, A. S. J. Am.
Chem. Soc. 1993, 115, 8019-8023.
(72) McLennan, D. J. In Isotopes in Organic Chemistry; Buncel, E., Lee,
C. C., Eds.; Elsevier: New York, 1987; Vol 7, Chapter 6, pp 393-480.
(73) Specifically, ∆ꢀ° ) A(k)1/2, where A ) [(h/4π)(1 - 0.51/2)]. The
latter equation assumes an X-H harmonic oscillator [i.e., ꢀ° ) (1/2)hν; ν
) (1/2π)(k/µ)1/2], where X is heavy, such that µ ) mH and mD. See ref 72.
(74) νH-H ) 4395 cm-1; νD-D ) 3118 cm-1. See: Herzberg, G.
Molecular Spectra and Molecular Structure; Van Nostrand Reinhold Co.:
New York, 1950; Vol. 1, pp 532-533.
(77) These modes are derived from the lost translational and rotational
degrees of freedom for H2.
(78) ZPE of W(PMe3)4H2I2 ) 296.141 kcal mol-1; ZPE of W(PMe3)4D2I2
) 293.129 kcal mol-1
.
(79) At 60 °C, H[W(PMe3)4I2] ) -1 214 192.67 kcal mol-1, H[W(PMe3)4-
H2I2] ) -1 214 926.98 kcal mol-1, H[W(PMe3)4D2I2] ) -1 214 929.68
kcal mol-1, H[H2] ) -731.79 kcal mol-1, H[D2] ) -733.63 kcal mol-1
;
(75) It is, however, important to consider bending modes in evaluating
secondary isotope effects.
S[W(PMe3)4I2] ) 184.17 eu, S[W(PMe3)4H2I2] ) 205.17 eu, S[W(PMe3)4D2I2]
) 206.62 eu, S[H2] ) 31.90 eu, S[D2] ) 35.31 eu

OxidatiVe Addition of H2 to Mo and W Complexes
J. Am. Chem. Soc., Vol. 121, No. 49, 1999 11413
In essence, the occurrence of an inverse deuterium equilibrium
isotope effect is a consequence of there being a single isotope-
sensitive vibrational mode in the reactant (H2) yet six (albeit
lower energy) isotope-sensitive modes in the product. It is
because of this peculiar situation with respect to the large
difference in number of isotope-sensitive modes for addition
of H2 that the observed isotope effect becomes counterintui-
tive: in all other examples of primary equilibrium isotope effects
there is not such a large difference in the number of isotope-
sensitive modes, and so the magnitude and direction of the
isotope effect is governed principally by the relative strengths
of the bonds broken and formed.80
A more detailed analysis of the origin of the inverse
equilibrium isotope effect has recently been provided by Krogh-
Jespersen and Goldman for addition of H2 to Vaska’s complex.71
Specifically, Krogh-Jespersen and Goldman have calculated that
the combination of the three entropy-determining contributions
of SYM (1.0), MMI (5.66), and EXC (0.84) result in a normal
contribution to the EIE (i.e., a positive value of ∆S°H - ∆S°D),
and that the inverse nature of the EIE is essentially a
consequence of a dominant zero point energy term, with ZPE
) 0.10, resulting in a positive value of ∆H°H - ∆H°D.81 That
the four bending modes are responsible for the inverse nature
of the EIE is further indicated by the fact that eliminating them
from the calculation results in the prediction of a normal EIE
of 4.6. Using a similar approach, we have used the calculated
isotope-sensitive vibrational frequencies of W(PMe3)4H2I2 to
predict an EIE of 0.73 at 60 °C,82,83 which is identical to the
value of 0.73 predicted by the complete energy calculation
described above. As with the Vaska system, the inverse nature
of the EIE is the result of a dominant zero point energy term:
ZPE (0.17), MMI (5.25), and EXC (0.80).83
Figure 11. Variation of koa and kre with [PMe3].
Scheme 2. Proposed Mechanism for Oxidative Addition of
H2 to W(PMe3)4I2
Bender, Kubas, Hoff, and co-workers have also considered
the origin of inverse deuterium equilibrium isotope effects in
the formation of dihydrogen complexes. Specifically, calcula-
tions analogous to those of Krogh-Jespersen and Goldman
described above revealed that the observed isotope effect [KH/
KD ) 0.70(15) at 22 °C] for the formation of W(CO)3(PCy3)2-
(η2-H2) from W(CO)3(PCy3)2(N2) is primarily attributable to an
inverse zero point energy factor (0.20), with a smaller contribu-
tion from the vibrational excitation factor (0.67); as with
oxidative addition of H2, the MMI factor is large and normal
(5.77). Furthermore, as with full oxidative addition, analysis of
the zero point energy components indicates that the inclusion
of all six [WH2] vibrational modes is essential to obtaining an
inverse deuterium EIE.
transformation to be elucidated. In particular, these studies
demonstrate that W(PMe3)4H2I2 is not obtained by the direct
oxidative addition of H2 to six-coordinate trans-W(PMe3)4I2 but
rather occurs via initial dissociation of PMe3. For example, the
rate of oxidative addition is strongly inhibited by addition of
PMe3, so that a mechanism involving PMe3 dissociation and a
five-coordinate [W(PMe3)3I2] intermediate is implied (Scheme
2).84 Further support for a mechanism involving rate-determining
dissociation of PMe3 is provided by the observation that there
is no deuterium kinetic isotope effect for addition of D2 to trans-
W(PMe3)4I2; however, a deuterium kinetic isotope effect is
observed in the presence of PMe3 (Figure 11), under which
conditions addition of PMe3 to the five-coordinate [W(PMe3)3I2]
intermediate becomes competitive with oxidative addition of
H2.
Mechanism of Oxidative Addition and Reductive Elimina-
tion of H2. Kinetics studies have allowed details of the
mechanism of the oxidative addition/reductive elimination
(80) For example, the deuterium equilibrium isotope effect for oxidative
addition of Me-H versus Me-D is calculated to be normal since the
reactant and product have the same number of isotope-sensitive modes and
C-H bonds are stronger than M-H bonds. See: Abu-Hasanayn, F.; Krogh-
Jespersen, K.; Goldman, A. S. J. Am. Chem. Soc. 1993, 115, 8019-8023.
(81) As a result of the difference in symmetry number between [MH3]
and [cis-HMD2] moieties, Goldman and Krogh-Jespersen have predicted
an even more pronounced inverse equilibrium isotope effect of 0.33 for
oxidative addition to M(PR3)2(CO)H. The difference in symmetry number
arises from the fact that, for statistical reasons, elimination of H2 from [MH3]
is twice as fast as elimination of D2 from [cis-HMD2]. Such a symmetry
effect would not, however, be operative for an equilibrium involving [trans-
HMD2]. See ref 80.
(84) On its own, the observation of inhibition by PMe3 can also be
rationalized by formation of a seven-coordinate species W(PMe3)5I2, which
serves to reduce the concentration of six-coordinate W(PMe3)4I2. However,
(82) Calculated vibrational frequencies (cm-1) for W(PMe3)4H2I2: 1991,
1987, 1102, 837, 759, 649. Calculated vibrational frequencies (cm-1) for
W(PMe3)4D2I2: 1419, 1411, 782, 609, 549, 464. The experimental values
of 4395 and 3118 cm-1 were used for H2 and D2.
(83) MMI, EXC, and ZPE terms were determined from the calculated
vibrational frequencies using a program provided by Dr. Bruce Bender.
1
such a species is not observed by H NMR spectroscopy. Furthermore, if
inhibition were a result of forming W(PMe3)5I2, a kinetic isotope effect in
the absence of added PMe3 would have been expected. In contrast, the
kinetic isotope effect is observed only in the presence of PMe3 (see text).

11414 J. Am. Chem. Soc., Vol. 121, No. 49, 1999
Hascall et al.
As implied by microscopic reversibility, the reductive elimi-
nation of H2 from W(PMe3)4H2I2 must also proceed via initial
PMe3 dissociation. Accordingly, the reductive elimination of
H2 from W(PMe3)4H2I2 is also inhibited by the presence of
PMe3, and a deuterium kinetic isotope effect is observed only
in the presence of added PMe3 (Figure 11).
Prior dissociation of PMe3 in the oxidative addition and
reductive elimination reactions of W(PMe3)4I2 and W(PMe3)4H2I2
has literature precedence.85 For example, the principal pathway
for hydrogenation of Wilkinson’s catalyst, Rh(PPh3)3Cl, giving
Rh(PPh3)3H2Cl, involves initial dissociation of PPh3.86 Thus,
under normal conditions, the rate of the reaction is limited by
phosphine dissociation and is consequently independent of H2
concentration; only in the presence of excess PPh3 does the
direct pathway become significant. Despite these observations,
however, it is important to emphasize that initial ligand
dissociation is not a required step for all oxidative addition and
reductive elimination reactions involving H2. Indeed, oxidative
addition and reductive elimination reactions involving H-H,
C-H, and C-C bonds may proceed either directly or via
sequences involving either (i) prior ligand loss or (ii) prior ligand
addition.85
Figure 12. Free energy surface for oxidative addition of H2 to
W(PMe3)4I2.
Table 7. Kinetic Isotope Effects for Oxidative Addition of H2 and
D2 to W(PMe3)4I2
It is, therefore, evident there must be a rather subtle interplay
of factors which influence whether oxidative addition and
reductive elimination occur directly or via prior ligand dissocia-
tion or addition. Nevertheless, a simple (perhaps oversimplified)
rationalization as to why W(PMe3)4H2I2 undergoes PMe3
dissociation prior to reductive elimination of H2 is associated
with the notion that the strong σ-donor PMe3 ligand stabilizes
high valence states. As such, reductive elimination of H2 would
be promoted by prior loss of PMe3. It is also possible that the
lower coordination number may allow for more facile structural
reorganization that accompanies the cis-to-trans rearrangement.
k1 (s-1
)
k2/k-1
k-2/k3 (M-1
)
k-3 (s-1
)
kH
kD
3.91(40) × 10-4 4.3(1.7) 5.4(1.7) × 10-3 3.38(40) × 10-4
3.68(40) × 10-4 3.7(1.3) 2.7(10) × 10-3 3.25(40) × 10-4
kH/kD 1.00a
1.2(2)
2.0(2)
1.0(1)
a Since neither product nor reactant incorporates deuterium, an
isotope effect does not exist.
Table 8. Kinetic Deuterium Isotope Effects for Addition and
Elimination of H2 to Transition Metal Complexes
reactant
product
kH/kD
ref
Kinetic Isotope Effects for Oxidative Addition and Reduc-
tive Elimination. A detailed study of the kinetics of the
oxidative addition and reductive elimination transformations has
allowed the free energy surface illustrated in Figure 12 to be
established. It is likely that dihydrogen species,87 for example
[W(PMe3)3(η2-H2)I2] and [W(PMe3)4(η2-H2)I2], may also be
intermediates on the energy surface, but since our data are not
capable of providing support either for or against such species
as intermediates (as opposed to transition states), we have
excluded them from the analysis. The primary kinetic deuterium
isotope effect for the oxidative addition step of H2 to [W(PMe3)3I2]
is k2(H)/k2(D) ) 1.2(2) at 60 °C (Table 7).63 While such a value
is small, it is in accord with previous studies, as illustrated in
Table 8.88,89 In view of the large difference in zero point energies
of H2 versus D2 (due to the strong nature of the bonds), low
Addition of H2
[W(PMe3)3I2]
[Cr(CO)5(C6H12)]
[Rh(PPh3)2Cl]
[Fe(CO)4]
[W(PMe3)3H2I2]
1.2(2) at 60 °C this work
Cr(CO)5(η2-H2)
Rh(PPh3)2H2Cl
[Fe(CO)4H2]
1.9 at 23 °C
1.5 at 23 °C
1.1(1) at 24 °C
a
b
c
d
Ir(PPh3)2(CO)Cl
Ir(PPh3)2(CO)ClH2 1.09 at 25 °C
Elimination of H2
[W(PMe3)3I2]
[W(PMe3)3H2I2]
2.0 at 60 °C
this work
c
W(CO)3(PPh3)2(H2) W(CO)3(PPh3)2(py) 1.7 (15-35 °C)
Cr(CO)5(η2-H2)
cis-Pt(PMe3)2H2
[Cr(CO)5]
[cis-Pt(PMe3)2]
5.0 at 23 °C
a
0.45(1) at 21 °C f
a [Cr(CO)5(C6H12)] was generated by photolysis of Cr(CO)6 in
cyclohexane. See: Church, S. P.; Grevels, F.-W.; Hermann, H.;
Schaffner, K. J. Chem. Soc., Chem. Commun. 1985, 30-32. b [Rh(PPh3)2-
Cl] was generated by photolysis of Rh(PPh3)2(CO)Cl. See: Wink, D.
A.; Ford, P. C. J. Am. Chem. Soc. 1987, 109, 436-442. c Wang, W.;
Narducci, A. A.; House, P. G.; Weitz, E. J. Am. Chem. Soc. 1996,
118, 8654-8657. d (a) Zhou, P.; Vitale, A. A.; San Filippo, J., Jr.;
Saunders, W. H., Jr. J. Am. Chem. Soc. 1985, 107, 8049-8054 and
references therein. (b) Abu-Hasanayn, F.; Krogh-Jespersen, K.; Gold-
(85) For a listing of oxidative addition and reductive elimination reactions
that involve either direct addition or elimination, or a mechanism involving
initial dissociation or addition of a ligand, see Table S6 in the Supporting
Information.
(86) (a) Halpern, J.; Wong, C. S. J. Chem. Soc., Chem. Commun. 1973,
629-630. (b) Halpern, J.; Okamoto, T.; Zakhariev, A. J. Mol. Catal. 1976,
2, 65-68. (c) Halpern, J. Inorg. Chim. Acta 1981, 50, 11-19.
(87) (a) Heinekey, D. M.; Oldham, W. J., Jr. Chem. ReV. 1993, 93, 913-
926. (b) Crabtree, R. H. Acc. Chem. Res. 1990, 23, 95-101. (c) Jessop, P.
G.; Morris, R. H. Coord. Chem. ReV. 1992, 121, 155-284. (d) Crabtree,
R. H. Angew. Chem., Int. Ed. Engl. 1993, 32, 789-805. (e) Kubas, G. J.
Acc. Chem. Res. 1988, 21, 120-128.
(88) Kinetic deuterium isotope effects for a three-centered transition state
are greatest for a symmetric transition state and decrease for transition states
that are either product-like or reactant-like; indeed, calculations indicate
that the isotope effect could become inverse (<1) in certain situations.
See: (a) Bigeleisen, J. Pure Appl. Chem. 1964, 8, 217-222. (b) Melander,
L. Acta Chem. Scand. 1971, 25, 3821-3826. (c) More O’Ferrall, R. A. J.
Chem. Soc. (B) 1970, 787-790. (d) Bigeleisen, J.; Wolfsberg, M. AdV.
Chem. Phys. 1958, 1, 15-76.
man, A. S. J. Am. Chem. Soc. 1993, 115, 8019-8023. (c) ∆Hq
)
)
H
10.6 kcal mol-1, ∆HqD ) 11.0 kcal mol-1; ∆SqH ) -24.4 eu, ∆Sq
D
-23.2 eu. e Zhang, K.; Gonzalez, A. A.; Hoff, C. D. J. Am. Chem.
Soc. 1989, 111, 3627-3632. f Packett, D. L.; Trogler, W. C. Inorg.
Chem. 1988, 27, 1768-1775.
values for deuterium kinetic isotope effects have previously been
invoked as an indication of an early transition state (i.e., little
H-H bond cleavage);29b,90 however, recent work suggests that
such interpretation may be greatly oversimplified. Kinetic
isotope effects may be considered to be a product of three terms,
(89) Hydride transfer from a transition metal hydride to [Ph3C]+ is known
to be characterized by both normal and inverse kinetic isotope effects. See:
Cheng, T.-Y.; Bullock, R. M. J. Am. Chem. Soc. 1999, 121, 3150-3155.

OxidatiVe Addition of H2 to Mo and W Complexes
J. Am. Chem. Soc., Vol. 121, No. 49, 1999 11415
i.e., KIE ) MMI‚EXC‚ZPE, and Goldman and Krogh-Jespersen
have demonstrated that the observed small deuterium kinetic
isotope effect for addition of H2 to Vaska’s complex, Ir(PPh3)2-
(CO)Cl, is a result of a large normal MMI term (5.66) being
compensated by inverse values for EXC (0.57) and ZPE (0.38),
which yield a calculated KIE of 1.23.41a,90 As with the
equilibrium isotope effects described above, the origin of the
inverse values for EXC and ZPE may be attributed to the
difference between both the number and magnitude of isotope-
sensitive modes in the reactants and transition state. Thus, the
activated complex has fiVe isotope-sensitive modes that are
orthogonal to the reaction coordinate, whereas the reactant (H2)
has only a single (albeit higher energy) vibration, thereby
resulting in an inverse ZPE term. Furthermore, since the isotope-
sensitive modes in the transition state are of low energy, there
will be a significant difference in the occupation of excited
vibrational states for the deuterio versus protio activated
complex. By comparison, the difference in occupation of excited
vibrational states for H2 and D2 is small due to the high
vibrational frequency of their bonds, and so an inverse value
for EXC results.
orbitals, and, as a consequence, π-donation of a single halogen
lone pair may stabilize the system. The significant differences
between M(PMe3)4X2 and Ir(PPh3)2(CO)X toward oxidative
addition of H2 provide an important example which indicates
that ancillary ligand variations may not necessarily exert the
same influence on a given reaction in different systems.
Equilibrium studies of the reaction of H2 with W(PMe3)4I2
indicate that the oxidative addition is characterized by an inVerse
equilibrium deuterium isotope effect [KH/KD ) 0.63(5) at 60
°C]. The inverse nature of the equilibrium isotope effect is
associated with the large number (six) of isotope-sensitive
vibrational modes in the product, compared to the single isotope-
sensitive vibrational mode in H2.
Finally, a mechanistic study reveals that oxidative addition
of H2 does not proceed directly to M(PMe3)4I2. Rather, the
reaction proceeds via initial dissociation of PMe3 and oxidative
addition to five-coordinate [W(PMe3)3I2]. Conversely, reductive
elimination of H2 does not occur directly from W(PMe3)4H2I2
but rather by a sequence that involves dissociation of PMe3 and
reductive elimination from seven-coordinate [W(PMe3)3H2I2].
In contrast to the kinetic isotope effect for oxidative addition,
Experimental Section
that for reductive elimination of H2 from [W(PMe3)3H2I2] (k-2
)
General Considerations. All manipulations were performed using
a combination of glovebox, high-vacuum, and Schlenk techniques.91
Solvents were purified and degassed by standard procedures. NMR
spectra were measured on Varian VXR 200, 300, and 400 spectrometers
and Bruker Avance DRX300WB and Bruker DPX300 spectrometers.
31P NMR spectra are referenced relative to 85% H3PO4 (δ ) 0) using
P(OMe)3 as an external reference (δ ) 141.0); all “J” values are given
cannot be determined directly, since only the ratio k-2/k3 can
be measured. However, on the basis that the secondary isotope
effect for addition of PMe3 to [W(PMe3)3H2I2] is expected to
be negligible (i.e., k3(H)/k3(D) ≈ 1), the kinetic isotope effect
for the reductive elimination step may be estimated as k-2(H)
/
k-2(D) ≈ 2 at 60 °C. Such a value is comparable to other values
for reductive elimination (Table 8) and is completely in accord
with the inverse equilibrium isotope effect; i.e., an inverse
equilibrium isotope effect for addition of H2 merely implies that
the kinetic isotope effect for reductive elimination is greater
than that for oxidative addition.
4
in hertz. Mo(PMe3)5H2,92 Mo(PMe3)4H2F2,9 and W(PMe3)4H2I2 were
prepared as previously reported.
Synthesis of Mo(PMe3)4H2Cl2. A solution of Mo(PMe3)5H2 (0.48
g, 1.00 mmol) in pentane (30 mL) was treated dropwise with HClaq
(12 M), giving a yellow precipitate. The addition was continued until
no further precipitation occurred (ca. 0.2 mL). The solid was isolated
by filtration, washed with pentane (10 mL), and dried in vacuo
overnight. Yield of Mo(PMe3)4H2Cl2: 0.36 g (76%). IR data (cm-1):
2971 (m), 2904 (m), 1897 (w) [νMo-H], 1423 (m), 1297 (m), 1280 (m),
940 (s), 846 (m), 716 (m), 665 (m). 1H NMR data (C6D6): δ 1.30 [18
H, vt, “JP-H” ) 3, 2 PMe3], 1.24 [18 H, vt, “JP-H” ) 3, 2 PMe3],
Summary
In conclusion, experimental and theoretical studies indicate
that the exothermicity of the oxidative addition of H2 to the
six-coordinate molybdenum and tungsten complexes M(PMe3)4X2
(M ) Mo, W; X ) F, Cl, Br, I) increases in the sequences Mo
< W and I < Br < Cl < F. The observed halogen dependence
is most interesting since it provides a striking contrast to that
reported for oxidative addition of H2 to trans-Ir(PPh3)2(CO)X.
Theoretical studies suggest that the halide dependence for
M(PMe3)4X2 is a result of both steric and electronic factors.
Thus, sterically, oxidative addition is favored most for the
fluoride derivatives, since the increased steric interactions upon
forming the eight-coordinate complexes M(PMe3)4H2X2 would
be minimized for the smaller halogen. Electronically, oxidative
addition is favored most for the fluoride derivatives since
π-donation destabilizes M(PMe3)4X2 by raising the energy of
the antibonding HOMO. The distinction with the Vaska system
arises from the fact that the dxz and dyz orbitals in M(PMe3)4X2
are degenerate and interact with a pair of halogen pπ orbitals.
As a result, pπ-dπ, donation destabilizes M(PMe3)4X2 due to
“filled-filled” repulsions. In contrast, symmetry considerations
do not dictate that both halogen lone pairs in trans-Ir(PPh3)2-
(CO)X must interact in a pairwise fashion with the iridium
2
2
2
-5.85 [2 H, d, JP-H ) 42: d, JP-H ) 47; t, JP-H ) 72, MoH2]. 13C
NMR (C6D6): δ 25.6 [vt, “JP-C” ) 14; q, 1JC-H ) 128, 2 PMe3], 19.2
[vt, “JP-C” ) 11; q, JC-H ) 127; JC-H ) 129, 2 PMe3]. 31P NMR
data (C6D6): δ 13.8 [JP-P ) 25, 2 PMe3], -8.3 [t, 2JP-P ) 25, 2 PMe3].
Synthesis of Mo(PMe3)4H2Br2. A solution of Mo(PMe3)5H2 (0.38
g, 0.79 mmol) in pentane (20 mL) was treated with HBr (48% aqueous)
in a dropwise manner, resulting in the formation of a yellow precipitate.
The addition was continued until no further precipitation was observed
(ca. 0.3 mL). The mixture was filtered, and the solid was washed with
pentane (10 mL) and dried in vacuo overnight. Yield of Mo(PMe3)4H2-
Br2: 0.35g (79%). IR data (cm-1): 2971 (m), 2903 (m), 1907 (w)
[νMo-H], 1422 (m), 1280 (m), 939 (s), 845 (m), 779 (w), 714 (m), 664
1
1
1
(m). H NMR data (C6D6): δ 1.38 [18 H, vt, “JP-H” ) 3, 2 PMe3],
2
1.26 [18 H, vt, “JP-H” ) 4, 2 PMe3], -6.91 [2 H, t, JP-H ) 45; t,
2JP-H ) 72, MoH2]. 31P NMR data (C6D6): δ 7.0 [t, JP-P ) 26, 2
2
2
PMe3], -17.8 [t, JP-P ) 26, 2 PMe3].
Synthesis of Mo(PMe3)4I2. A solution of Mo(PMe3)5H2 (2.0 g, 4.2
mmol) in pentane (150 mL) was treated with HIaq (57 wt %) in a
dropwise manner, resulting in the formation of a tan precipitate. The
addition was continued until no further precipitation was observed. The
mixture was filtered, and the solid was dried in vacuo, extracted into
(91) (a) McNally, J. P.; Leong, V. S.; Cooper, N. J. In Experimental
Organometallic Chemistry; Wayda, A. L., Darensbourg, M. Y., Eds.;
American Chemical Society: Washington, DC, 1987; chapter 2, pp 6-23.
(b) Burger, B. J.; Bercaw, J. E. In Experimental Organometallic Chemistry;
Wayda, A. L., Darensbourg, M. Y., Eds.; American Chemical Society:
Washington, DC, 1987; Chapter 4, pp 79-98.
(90) Earlier theoretical studies calculated that the kinetic isotope should
be inverse on the basis of the MMI, EXC, and EXP(ZPE) terms (ca. 0.5)
and therefore introduced a tunneling correction to achieve a normal kinetic
isotope effect.a The more recent study, however, indicates that tunneling is
not a required correction.b (a) Zhou, P.; Vitale, A. A.; San Filippo, J., Jr.;
Saunders, W. H., Jr. J. Am. Chem. Soc. 1985, 107, 8049-8054. (b)
Reference 41b.
(92) Lyons, D.; Wilkinson, G.; Thornton-Pett, M.; Hursthouse, M. B. J.
Chem. Soc., Dalton Trans. 1984, 695-700.

11416 J. Am. Chem. Soc., Vol. 121, No. 49, 1999
Hascall et al.
hot toluene (50 mL), and filtered. The filtrate was allowed to cool to
room temperature, depositing a red microcrystalline solid which was
isolated by filtration, washed with pentane (5 mL), and dried in vacuo.
Yield of Mo(PMe3)4I2: 1.30 g (48%).
Stability of Mo(PMe3)4H2F2 with Respect to H2 Loss. A solution
of Mo(PMe3)4H2F2 (ca. 15 mg) in C6D6 (0.7 mL) was heated at 30 °C
for 2 days. The mixture was monitored by 1H NMR spectroscopy, which
indicated that Mo(PMe3)4H2F2 was stable to loss of H2 under these
conditions.
rates. The pseudo-first-order rate constants, koa and kre, as a function
of PMe3 concentration are listed in Tables S7 and S8 in the Supporting
Information. For the reductive elimination reaction, a plot of 1/kre versus
[PMe3] (Figure 11) allows k3 to be established from the intercept, since
[H2] ) 0 during the initial stages of the reaction. The ratio k-2/k3 may
likewise be determined from the slope, using the aforementioned value
of k-3. For the oxidative addition reaction, a plot of 1/koa versus [PMe3]
(Figure 11) allows k1 to be determined from the intercept, since k-3
,
K, and [H2] are known. Correspondingly, the ratio k-1/k2 may be
determined from the slope, since ki and [H2] are known. ∆Gq was
calculated from the Eyring equation, ∆Gq ) RT ln(κkBT/kh), assuming
a transmission coefficient (κ) of 1.
Interconversion of Mo(PMe3)4H2Cl2 and Mo(PMe3)4Cl2. (a) A
solution of Mo(PMe3)4H2Cl2 (ca. 15 mg) in C6D6 (0.7 mL) was heated
at 60 °C for 5 min. The complete conversion to Mo(PMe3)4Cl2 was
1
Computational Details. All DFT calculations were performed using
Jaguar Version 3.5. Geometries were optimized at the BLYP level using
the LACVP** basis set. D2d symmetry and triplet multiplicity were
used for the M(PR3)4X2 (R ) H, Me; X ) F, Cl, Br, I) complexes; C2V
symmetry and singlet multiplicity were used for M(PH3)4H2X2 com-
plexes; Cs symmetry and singlet multiplicity were used for M(PMe3)4H2X2
complexes.96 For W(PMe3)4(NH2)2, the structure was optimized using
both singlet and triplet multiplicities. The singlet refinement converged
to a structure with parallel NH2 groups and C2V symmetry, while the
triplet refinement converged to a structure with orthogonal NH2 groups
and D2d symmetry. For W(PMe3)4H2(NH2)2, the structure was optimized
with singlet multiplicity and no symmetry constraints. Single-point
energies were then calculated on all optimized structures (with
appropriate symmetry constraints) at the B3LYP level using triple-ú
basis sets (LACV3P** for Mo, W, Br, and I and cc-pvtz(-f) for all
other elements). H2 was optimized at the B3LYP level using the cc-
pvtz(-f) basis set, giving a calculated bond length of 0.74 Å.
Comparisons of experimental and calculated structural parameters
for the compounds studied are presented in Tables S2-S5 in the
Supporting Information. In general, the angular values obtained from
the DFT calculations agree quite well with those determined via
crystallography. However, there are substantial and systematic differ-
ences that can be observed for some of the bond lengths. In particular,
the metal-halide and metal-phosphorus bond lengths are systemati-
cally overestimated. In the most extreme cases, for example the iodide
species, the deviations would appear to be outside of the range normally
expected for DFT calculations as compared to experiment, even for
metal-containing systems, based on the data available in the literature.97
We have investigated this issue in detail as follows. First, we ran
calculations using the fully analytical integral (as opposed to pseu-
dospectral) methodology available in Jaguar for the entire series of
six-coordinate complexes M(PMe3)4X2 (M ) Mo, W; X ) F, Cl, Br,
I), as illustrated in Table S9 of the Supporting Information. For all test
cases, differences in bond lengths between pseudospectral and analytical
results were less than 0.005 Å when the default DFT grid was used.
Use of a fine DFT grid in the analytical calculations resulted in little
change. Second, we carried out calculations on W(PMe3)4H2I2 using a
hybrid functional incorporating exact HF exchange (B3LYP) rather than
the BLYP functional that was used as a standard for the geometry
optimizations in the present work. This produced a small improvement,
but on the scale of the deviation from experiment, the change was not
significantsa rather surprising result as gradient-corrected and hybrid
functionals often give structures with significant differences. Finally,
we ran an LMP2 rather than DFT geometry optimization for
W(PMe3)4H2I2. The LMP2 results are closer to the experimental results
than are the DFT results but still differ from them significantly.98
In addition, the series of Mo complexes Mo(PH3)4X2 and Mo-
(PH3)4H2X2 (X ) F, Cl, Br, I) were optimized at the B3LYP level
using the LACVP** basis set, and then single-point energy calculations
were performed on the optimized structures using triple-ú basis sets as
described above. This procedure resulted in minor differences in the
Mo-X bond lengths, as shown in Table S10 of the Supporting
Information. Of more importance to the present study, it can be seen
demonstrated by H NMR spectroscopy. At 30 °C, the half-life is ca.
2 h.
(b) A solution of Mo(PMe3)4Cl2 (ca. 15 mg) in C6D6 (0.7 mL) was
treated with H2 (1 atm) and heated at 60 °C for 4 days. The mixture
1
was monitored by H NMR spectroscopy, which indicated that Mo-
(PMe3)4Cl2 was stable with respect to Mo(PMe3)4H2Cl2 under these
conditions.
Conversion of Mo(PMe3)4H2Br2 to Mo(PMe3)4Br2. (a) A solution
of Mo(PMe3)4H2Br2 (ca. 15 mg) in C6D6 (0.7 mL) was heated at 30 °C
for 3 h. The complete conversion to Mo(PMe3)4Br2 over this period
1
was demonstrated by H NMR spectroscopy (t1/2 < 30 min).
(b) A solution of Mo(PMe3)4Br2 (ca. 10 mg) in C6D6 (0.7 mL) was
treated with H2 (1 atm) and heated at 55 °C for 3 h. The mixture was
monitored by 1H NMR spectroscopy, which indicated that Mo(PMe3)4-
Br2 was stable with respect to Mo(PMe3)4H2Br2 under these conditions.
Stability of Mo(PMe3)4I2 with Respect to Addition of H2. A
solution of Mo(PMe3)4I2 (ca. 10 mg) in C6D6 (0.7 mL) was treated
with H2 (1 atm) and heated at 80 °C for 4 days. The mixture was
monitored by 1H NMR spectroscopy, which indicated that Mo(PMe3)4I2
was stable with respect to Mo(PMe3)4H2I2 under these conditions.
Thermodynamics and Kinetics of the Equilibration of W(PMe3)4I2
and H2 with W(PMe3)4H2I2. (a) Equilibrium Studies. In a typical
experiment, a solution of W(PMe3)4H2I2 in C6D6 (0.8 mL of 10.1 mM)
in a gas-tight NMR tube was saturated with H2 (1 atm at 23 °C). The
sample was placed in a constant temperature oil bath (( 1 °C) and
removed periodically to monitor (by 1H NMR spectroscopy) the
formation of the equilibrium mixture with W(PMe3)4I2. The molar ratio
of W(PMe3)4H2I2 to W(PMe3)4I2 was determined directly from the 1H
NMR spectra, and the molar concentration of dihydrogen in solution
at equilibrium was estimated by a method similar to that described
previously,93 using a combination of Henry’s law with the mole fraction
solubilities of H2 calculated directly or extrapolated from the equation
given by Clever.94 The equilibrium constant (K) was measured over
the temperature range 40-90 °C (Table 2), and a van’t Hoff plot of ln
K vs 1/T (Figure 1) yielded the values ∆H° and ∆S° from the slope
and the intercept of the resulting straight line, respectively. An
analogous procedure was used for the equilibrium studies between
W(PMe3)4I2 and W(PMe3)4D2I2, for which the mole fraction solubilities
of D2 were calculated directly or extrapolated from the equation given
by Clever.95
(b) Kinetics Studies. Samples of W(PMe3)4I2 for kinetics studies
at 60 °C were prepared by a method analogous to that described above,
but variable amounts of PMe3 were added prior to saturation with H2
at 23 °C. The concentration of H2 at 60 °C was determined as described
above, while the concentration of PMe3 was determined by integration
with respect to mesitylene as an internal standard. Samples of
W(PMe3)4H2I2 were prepared similarly, except H2 was not added. The
result of application of the steady-state approximation for the intermedi-
ates [W(PMe3)3I2] and [W(PMe3)3H2I2] for the mechanism of Scheme
2 is given in Chart 1 in the Supporting Information. For convenience,
the kinetics were analyzed by separately examining the oxidative
addition and reductive elimination reactions using the method of initial
(93) Thompson, M. E.; Baxter, S. M.; Bulls, A. R.; Burger, B. J.; Nolan,
M. C.; Santarsiero, B. D.; Schaefer, W. P.; Bercaw, J. E. J. Am. Chem.
Soc. 1987, 109, 203-219.
(94) ln x ) -5.5284 - 813.90/(T/K). See: Clever, H. L. Hydrogen and
Deuterium. In Solubility Data Series, Young, C. L., Ed.; Pergamon Press:
Oxford, 1981; Volume 5/6, p 159.
(95) ln x ) -5.7399 - 743.44/(T/K). See: Clever, H. L. Hydrogen and
Deuterium. In Solubility Data Series, Young, C. L., Ed.; Pergamon Press:
Oxford, 1981; Volume 5/6, p 287.
(96) Crystallographic studies indicate that M(PMe3)4H2X2 complexes
exist as several rotamers in the solid state, depending upon the orientation
of the PMe3 methyl groups. For consistency, all calculations were performed
for a common rotameric structure.
(97) Frenking, G.; Pidun, U. J. Chem. Soc., Dalton Trans. 1997, 1653-
1662.
(98) W-I bond lengths (Å): BLYP, 3.040, 3.106; B3LYP, 3.014, 3.067;
LMP2, 3.003, 3.037; expt, 2.903, 2.922.

OxidatiVe Addition of H2 to Mo and W Complexes
J. Am. Chem. Soc., Vol. 121, No. 49, 1999 11417
that performing the geometry optimizations at the B3LYP level has a
negligible effect on the calculated energies for the addition of H2 to
Mo(PH3)4X2 (Table S11 of the Supporting Information).
for support of this research. This work was also supported in
part by a grant to R.A.F. from the NIH (GM-40526). Dr. Bruce
Bender is gratefully thanked for providing us with a copy of
his program to calculate MMI, EXC, and ZPE terms.
The sum total of the results suggests that the deviations in bond
length are, in fact, a consequence of the DFT treatment of electron
correlation for this particular system, which appears to be considerably
more challenging than other, simpler, transition metal systems for which
results have been reported in the literature. It is also possible that there
is some contribution to the deviation from crystal packing effects (not
modeled in our calculations), but the magnitude of the deviation is
larger than can plausibly be explained by this. Clearly, this problem
warrants further investigations, using higher levels of electron correla-
tion such as multireference perturbation theory, and we intend to pursue
such calculations in future work. For the present paper, however, the
observed deviations in bond length from the experimental data are not
incompatible with reasonable predictions for bond energies or other
relative properties of the various molecules, and we judge that the
calculations are still quite useful in this regard.
Supporting Information Available: Tables S1-S11 of
selected W-H bond energies, comparisons of calculated and
experimentally determined structures, a compilation of mech-
anisms for oxidative addition and reductive elimination reac-
tions, and experimentally determined rate constants for oxidative
addition of H2 to W(PMe3)4I2 and for reductive elimination of
H2 from W(PMe3)4I2H2; Chart 1, giving kinetic expressions for
oxidative addition of H2 to W(PMe3)4I2 and for reductive
elimination of H2 from W(PMe3)4I2H2 (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
Acknowledgment. G.P. thanks the U.S. Department of
Energy, Office of Basic Energy Sciences (DE-FG02-93ER14339),
JA9920009