Evidence for ligand non-innocence in a formally ruthenium(I) hydride complex

Article abstract of DOI:10.1021/ja100894h

(Figure Presented) Formally zerovalent, dlnltrogen-brldged ruthenium complex, {[N₃^Xyl]Ru}₂(-n¹:n ^1-N₂) (1), where [N₃W] = 2,6-(XyLN=CMe) ₂C₅H₃N, reacts with excess H2 to give the blnuclear hydride species, {[N₃]Ru(H)}₂(-n ¹:n¹-N₂) (2) bearing a single hydrogen per ruthenium. Complex 2 is an unusual example of a structurally characterized paramagnetic transition metal hydride, and the first such example for ruthenium. Structural data and DFT calculations suggest unpaired electron density is strongly delocalized onto the non-innocent [N₃] Hgand, with a relatively small degree of the metalloradical character implied by the Ru(I) formal oxidation state, and that the [N₃]VRu(II) formalism may be more informative. Consistent with an effective oxidation state greater than Ru(I), further reaction of 2 with excess H₂ to give metal dihydride species ([N₃]RuH₂(L)) is not observed. The magnetic moment of 2 (3.50 _B) in solution is consistent with one unpaired electron per [N₃]Ru moiety; however, 2 is diamagnetic in the solid due to close (3.26 A) head-to-tall contact between Ru pyridine planes of neighboring molecules. Although the geometry is reminiscent of the weak π-stacking observed for closed-shell aromatic ring systems, DFT calculations indicate the structure and associated spin pairing result from in-phase overlap of the delocallzed SOMOs on neighboring molecules-that is the interaction is best viewed as a weak covalent bond delocalized over 22 atoms. Copyright

Full text of DOI:10.1021/ja100894h

Published on Web 03/03/2010
Evidence for Ligand Non-innocence in a Formally Ruthenium(I) Hydride
Complex
Noah L. Wieder, Michelle Gallagher, Patrick J. Carroll, and Donald H. Berry*
Department of Chemistry and Laboratory for Research on the Structure of Matter, UniVersity of PennsylVania,
Philadelphia, PennsylVania 19104-6323
Received February 1, 2010; E-mail: berry@sas.upenn.edu
Late transition metal complexes of 2,6-bis(imino)pyridyl
ligands initially garnered attention as effective catalysts for olefin
polymerization and other reactions.1 More recently, however,
there has been an increased effort to elucidate the “non-innocent”
character of this remarkably versatile class of multidentate
ligands, and there is clear evidence that [N3R] ligands ([N3R] )
2,6-(RNdCMe)2C5H3N) can serve as acceptors for a variable
number of electrons.2 In certain cases, low-valent [N3R]M
complexes are best described by canonical forms representing
whereas only two xylyl methyl environments would be expected
for a mononuclear [N3]Ru(H)(N2) structure (Cs symmetry at Ru).
Resonances for the hydride ligand, p-pyridine CH and imine
methyl groups are not observed in the 1H spectrum, presumably
due to paramagnetic broadening; furthermore, no signals were
observed in the 13C spectrum. Although the metal hydride in 2
is not directly observed by NMR, it is strongly implied by the
significant shift of the ligand proton resonances in 2-d2, which
is rapidly formed from 1 and D2. Ligand shifts of ∆δ ≈ 0.2 are
appreciably larger than seen in isotopically substituted diamag-
netic molecules,9 but would be consistent with an Ru-H(D) on
a paramagnetic metal center. Theopold and co-workers coined
the term PIECS (paramagnetic isotope effect on chemical shift)
for the phenomenon.10 In addition, the IR spectrum of 2 exhibits
an Ru-H stretch at 2003 cm-1 which is not observed in 2-d2.
The X-ray crystal structure of 2 confirms a binuclear,
dinitrogen-bridged structure. As illustrated in Figure 1, the solid-
state structure of 2 consists of two square pyramidal centers
bridged by an essentially linear dinitrogen ligand (RuNN )
177.3(4)°), with the hydride ligands located trans to the vacant
coordination sites. The two [N3]Ru planes are perpendicular.
Elongation of the imine C-N bond and contraction of the
Cpyr-Cimine bonds have been attributed to delocalization of metal-
based electron density onto primarily ligand-based orbitals The
imine C-N distances (1.333(5), 1.334(6) Å) are slightly shorter,
and the Cpyr-Cimine distances (1.445(6), 1.433(7) Å) are longer
than in 1, suggesting that the ligand in 2 is somewhat less
reduced. The N-N distance of the bridging N2 ligand is also
shorter in 2 (1.132(6) Å) than 1 (1.161(5) Å), again consistent
with a less electron-rich metal in the former.
full transfer of one or more electrons from the metal to relatively
remote ligand orbitals, e.g. [N3R]-/M+ or [N3
]
2-/M2+. Several
R
groups have shown that surprisingly complex electronic struc-
tures and magnetic behavior can result in these formally low-
valent complexes.3 In some instances, such metal-to-ligand
charge transfer character in the ground-state is manifest as ligand-
centered chemical reactivity.4 In many other cases, however,
the reaction chemistry remains metal-centered and is not
appreciably different from what one would predict on the basis
of the formal oxidation state.
Our group has previously described the highly reactive,
formally zerovalent ruthenium complex, {[N3Xyl]Ru}2(µ-η1:η1-
N2) (1), where [N3Xyl] ) 2,6-(XylNdCMe)2C5H3N and Xyl )
2,6-xylyl.5 Unlike many first-row transition metal complexes of
non-innocent ligands that exhibit magnetic properties resulting
from interaction of high-spin metal centers with ligand radicals,
1 is diamagnetic in solution and the solid state. In addition, the
reactivity of 1 reported previously points to nothing more
complicated than a formally Ru(0) center with a neutral [N3]
ligand.5,6 It was noted, however, that the unusual and temperature
1
independent shift of the imine methyl groups in the H NMR
spectrum of 1 may reflect possible ligand non-innocence.5 We
now report the synthesis and properties of a binuclear, formally
Ru(I) dinitrogen-bridged hydride species, {[N3]Ru(H)}2(µ-η1:
η1-N2) (2), and evidence for the role of ligand non-innocence in
the stability and properties of this complex.
Treatment of 1 with excess dihydrogen (∼4 atm) in ether or
aliphatic solvents at ambient temperature leads to the rapid
formation of 2, isolated as a purple crystalline solid (77%, eq
1). Compound 2 appears to be the first formally Ru(I) hydride
complex to have been isolated and structurally characterized.
Many paramagnetic hydride complexes are unstable, although
several odd-electron complexes of other metals bearing terminal7
and bridging8 hydride ligands have been structurally character-
ized.
The 1H NMR spectrum of 2 exhibits isotropically shifted
resonances between δ 2 to 20, consistent with an open-shell
electronic structure. The low symmetry of 2 evident by 1H NMR
is consistent with a binuclear structure in solution, e.g. four
different resonances are observed for the eight xylyl methyl
groups (C2 symmetry overall, C1 local symmetry at each Ru),
Linear plots of the 1H chemical shifts of 2 versus 1/T confirm
Curie-Weiss behavior11 (Supporting Information), and the
solution magnetic moment of 3.50 µB per {[N3]Ru(H)}2(N2)
(Evans Method12) corresponds to one unpaired electron per
[N3]Ru(H) unit. Although clearly paramagnetic in solution, 2
was found to be diamagnetic in the solid state between 2.4 to
130.0 K.13 The origin of this inconsistency between solid and
solution states is found in the crystal structure of 2. As shown
in Figure 2, there are close intermolecular contacts between
neighboring molecules of 2. In an arrangement reminiscent of
aromatic ring “π-stacking”, Ru-pyridine units of two molecules
are positioned parallel and head-to-tail, offset such that each
metal is 3.29 Å from the para carbon of the adjacent pyridine.
9
10.1021/ja100894h  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 4107–4109 4107

C O M M U N I C A T I O N S
Figure 3. Calculated molecular orbital surface plot for the SOMO of
2-mod (left). Mulliken spin densities are indicated for selected atoms
(right).
The calculated molecular orbitals are consistent with a large
contribution of the Ru(II)/[N3]1- formalism. The SOMO of
2-mod is substantially delocalized onto the [N3] ligand (73%),
with only 19% on the metal, and 8% on the N2 ligand, and is
antibonding with respect to the imine CdN bond and bonding
between Cortho-Cimine (Figure 2.) The SOMO exhibits some
density at the pyridine para-carbon, but has nodes at the meta-
positions; this may explain why the meta, but not para, pyridine
Figure 1. ORTEP drawing of complex 2 (30% probability ellipsoids).
Ligand hydrogen atoms and xylyl groups on the [N3]Ru(1a) unit are
omitted for clarity. Selected bond lengths (Å): Ru1-N4, 1.953(3);
N1-C1, 1.333(5); N3-C7, 1.334(6); C1-C2, 1.445(6); C6-C6,
1.433(7); N4-N4a, 1.132(6).
1
protons are observed in the (paramagnetically shifted) H NMR
spectrum. Some delocalization of unpaired spin density on the
imine methyl groups may also explain the absence of the Cimine
-
methyls from the NMR spectrum. Although not a perfect measure
of spin delocalization, the Mulliken spin densities (Figure 3)
indicate substantially lower unpaired electron density on ruthe-
nium (0.11) than on the pyridine and imine atoms of the [N3]
ligand (>0.8).
DFT calculations were also performed for dinitrogen-bridged,
binuclear complexes, both with methyl groups as a model
({[N3Me]Ru(H)}2(N2)), and the full xylyl-substituted 2. These
results will be discussed in detail in a subsequent report, but
the optimized geometrical parameters, molecular orbital descrip-
tions, and distribution of spin density are not qualitatively
different from those for 2-mod; interaction of the unpaired spins
on the two perpendicular [N3]Ru fragments via the dinitrogen
bridge appears to be very modest.
DFT calculations were also undertaken to probe whether
pyridine stacking in the solid state of 2 is merely a vestige of
weak crystal packing forces or a stronger interaction resulting
from the observed spin pairing in the solid state. Dimeric model
complex {2-mod}2 contains two subunits of 2-mod related by
stacking of the pyridyl rings as observed in the crystal structure
of the N-xylyl derivative. The geometry of 2-mod was optimized
as an unrestricted singlet, resulting in pyridyl Cpara-Ru distances
of 3.042 Å. This somewhat shorter contact than in 2 is consistent
with the reduced steric demands of the model. The close contact
between the dimer halves is a clear indication of a bonding
interaction. The HOMO of {2-mod}2 (Figure 4) is clearly the
in-phase combination of SOMOs from the two 2-mod monomers
(Figure 3) and is bonding with respect to the Ru-Cpara and
interpyridine C-C interactions. The imine carbon atoms also
appear to be involved in bonding between the two planar
fragments, which is not unreasonable given the unpaired electron
density found on these atoms in the monomer. This process is
not merely antiferromagnetic coupling of spins on isolated
centers but is more comparable to covalent bond formation from
the combination of two radicals. The homolytic cleavage back
into radicals in solution does indicate the bond is quite weak
Figure 2. Crystal structure of 2 highlighting one set of intermolecular
Ru-pyridine contacts. The Ru-Cpara contacts are 3.29 Å, and the
distance between the mean pyridine planes is 3.26 Å.
This Ru-C separation is longer than in η6-arene ruthenium
complexes (average D(Ru-C) ) 2.21 Å,)14 but much closer than
van der Waals contacts. Furthermore, the plane-plane separation
between the mean pyridine rings is ∼3.26 Å, considerably closer
than the ∼3.4 Å separation seen in organic π-stacks.15 A search
of the Cambridge Structural Database14 revealed a single
previous instance of intermolecular M · · · pyridine stacking in
bis(imino)pyridyl transition metal complexes, the crystal struc-
ture of [N3DiP]Mn(CH3) (DiP ) 2,6-(iPr)2C6H3). Although not
noted in the original report by Gambarotta and co-workers, this
complex exhibits a ∼3.35 Å separation between Mn-pyridine
units of neighboring molecules.16 The magnetic susceptibility
of the manganese complex was reported as 4.82 µB; it is likely
this complex will exhibit a discrepancy between solid- and
solution-state magnetic properties similar to that seen for 2.
The electronic structure of [N3Me]Ru(H)N2 (2-mod), a model
for the Ru(I) hydride complex (2), was investigated computa-
tionally at the B3LYP17 level of theory employing the
6-31G(d,p)18 basis set for nonruthenium atoms and the SDD19
basis set for ruthenium. Mononuclear model hydride complex
2-mod consists of a single open shell [N3]Ru(H)(N2) unit with
methyl groups on the imine nitrogen atoms. The optimized
geometry is in good agreement with the solid state structure; all
calculated N-Ru, CdN, and Cortho-Cimine bond lengths are within
0.02 Å of the experimental values, except for N-N distance in
the N2 ligand, which is terminal in the model, but bridging in 2.
but is no more surprising than the equilibrium dimerization of
20b
trityl20a or Cp*Cr(CO)3
radicals. One largely superficial
difference in the case of 2, of course, is the delocalization of
9
4108 J. AM. CHEM. SOC. VOL. 132, NO. 12, 2010

C O M M U N I C A T I O N S
Supporting Information Available: Details of synthetic proce-
dures, spectroscopic details, X-ray crystallographic data for 2 in
CIF format, computational details, and coordinates of optimized
geometries. This material is available free of charge via the Internet
at http://pubs.acs.org.
References
(1) Gibson, V. C.; Redshaw, C.; Solan, G. A. Chem. ReV. 2007, 107, 1745–
1776.
(2) Enright, D.; Gambarotta, S.; Yap, G.; Budzelaar, P. H. M. Angew. Chem.,
Int. Ed. 2002, 41, 3873–3876.
(3) (a) Budzelaar, P. H. M.; de Bruin, B.; Gal, A.; Wieghardt, K.; van Lenthe,
J. H. Inorg. Chem. 2001, 40, 4649–4655. (b) Knijnenburg, Q.; Hetterscheid,
D.; Kooistra, T. M.; Budzelaar, P H. M. Eur. J. Inorg. Chem. 2004, 2004,
1204–1211. (c) Bart, S. C.; Chlopek, K.; Bill, E.; Bouwkamp, M. W.;
Lobkovsky, E.; Neese, F.; Wieghardt, K.; Chirik, P. J. J. Am. Chem. Soc.
2006, 128, 13901–13912. (d) Bart, S. C.; Lobkovsky, E.; Bill, E.; Wieghardt,
K.; Chirik, P. J. Inorg. Chem. 2007, 46, 7055–7063.
Figure 4. Molecular orbital surface (left) and contour (right) plots of
the HOMO of {2-mod}2.
(4) Knijnenburg, Q.; Gambarotta, S.; Budzelaar, P. H. M. Dalton Trans. 2006,
5442–5448.
(5) Gallagher, M.; Wieder, N. L.; Dioumaev, V. K.; Carroll, P. J.; Berry, D. H.
Organometallics 2010, 29, 591–603.
the bond over 22 atoms. The “bond enthalpy” of this interaction
can be very roughly estimated from the DFT calculations as 3.53
kcal/mol.21
(6) Yoo, H.; Carroll, P. J.; Berry, D. H. J. Am. Chem. Soc. 2006, 128, 6038–
6039.
(7) (a) McQueen, J.; Nagao, N.; Eberspacher, T.; Li, Z.; Taube, H. Inorg. Chem.
2003, 42, 3815–3821. (b) Pleune, B.; Morales, D.; Meunier-Prest, R.;
Richard, P.; Collange, E.; Fettinger, J. C.; Poli, R. J. Am. Chem. Soc. 1999,
121, 2209–2225. (c) Jewson, J. D.; Liable-Sands, L. M.; Yap, G. P. A.;
Rheingold, A. L.; Theopold, K. H. Organometallics 1999, 18, 300–305.
(d) Pleune, B.; Poli, R.; Fettinger, J. C. J. Am. Chem. Soc. 1998, 120, 3257–
3258. (e) Hamon, P.; Toupet, L.; Hamon, J. R.; Lapinte, C. Organometallics
1992, 11, 1429–1431. (f) Mura, P.; Segre, A.; Sostero, S. Inorg. Chem.
1989, 28, 2853–2858. (g) Leutkens, M. L.; Elcessere, W. L.; Huffman,
J. C.; Sattelberger, A. P. Inorg. Chem. 1984, 23, 1718–1726. (h) Luetkens,
M. L.; Elcesser, W. L.; Huffman, J. C.; Sattelberger, A. P. J. Chem.
Commun. 1983, 1072–1074.
(8) (a) MacAdams, L. A.; Buffone, G. P.; Incarvito, C. D.; Golen, J. A.;
Rheingold, A. L.; Theopold, K. H. Chem. Commun. 2003, 1164–1165. (b)
Fryzuk, M. D.; Johnson, S. A.; Rettig, S. J. Organometallics 2000, 19,
3931–3941. (c) Heintz, R.; Ostrander, R.; Rheingold, A.; Theopold, K. H.
J. Am. Chem. Soc. 1994, 116, 11387–11396.
Ironically, it is the inherent limitations of most DFT methods
that highlight the significance of electron pairing in bonding the
two fragments. Commonly employed DFT functionals such as
B3LYP grossly underestimate long-range forces; thus, DFT
calculations do not generally reproduce attractive “π-stacking”
interactions due solely to van der Waals forces. In the absence
of additional stabilization, the pyridine/pyridine interaction would
be shown as repulsiVe by these DFT methods. Indeed, attempts
to optimize the stacked pyridine geometry of {2-mod}2 as a
triplet (one unpaired electron on each monomer) were unsuc-
cessful, as was optimization of the stacked dimer of (open-shell
singlet) 1 subunits.
In conclusion, the DFT calculations and observed dimerization
of 2 in the solid state are consistent with substantial delocal-
ization of unpaired e-density from Ru to the [N3] ligand; i.e. 2
maybe better described as a Ru(II) hydride complex containing
a radical anion [N3] ligand. This higher than “expected” metal
oxidation state may also explain why [N3]RuH2 complexes are
not formed under excess hydrogen. There are also potential
implications regarding the electronic structure and reactivity of
the parent complex, 1. An analogous Ru(I)/[N3]- canonical form
for 1 suggests metalloradical character in a formally closed shell
complex and raises the possibility that the reaction to produce
2, with a single hydrogen atom on each metal, could proceed
via the type of bimetallic, four-centered transition state previ-
ously only observed for odd-electron complexes.22,23 Experi-
mental and computational results on the structure of 1 and the
mechanism of formation of 2 are currently under investigation.
(9) Hansen, P. E. Prog. Nucl. Magn. Reson. Spectrosc. 1988, 20, 207–255.
(10) Heintz, R. A.; Neiss, T. H.; Theopold, K. H. Angew. Chem., Int. Ed. 1994,
33, 2326–2328.
(11) Bertini, I.; Luchiant, C. NMR of Paramagnetic Molecules; Benjamin/
Cummings: Menlo Park, CA, 1986.
(12) Sur, S. K. J. Magn. Reson. 1989, 82, 169–173.
(13) Only a very small paramagnetic response (<5%) introduced by sample
handling was observed.
(14) Allen, F. H. Acta Crystallogr. 2002, B58, 380–388.
(15) Ishida, T.; Doi, M.; Ueda, H.; Inoue, M.; Scheldrick, G. M. J. Am. Chem.
Soc. 1988, 110, 2286–2294.
(16) Reardon, D.; Aharonian, G.; Gambarotta, S.; Yap, G. P. A. Organometallics
2002, 21, 786–788.
(17) Becke, A. D. J. Chem. Phys. 1986, 84, 4524–4529.
(18) Rassolov, V. A.; Ratner, M. A.; Pople, J. A.; Redfern, P. C.; Curtiss, L. A.
J. Comput. Chem. 2001, 22, 976–984.
(19) Andrae, D.; Haeussermann, U.; Dolg, M.; Stoll, H.; Preuss, H. Theor. Chim.
Acta 1990, 77, 123–141.
(20) (a) McBride, J. M. Tetrahedron 1974, 30, 2009–2022. (b) Woska, D. C.;
Ni, Y.; Wayland, B. B. Inorg. Chem. 1999, 38, 4135–4138.
(21) Derived from the enthalpy of {2-mod}2 minus twice the enthalpy of 2-mod,
each corrected to 298.15 K and for zero-point energies, but not for
counterpoise (BSSE).
(22) (a) Cui, W.; Wayland, B. B. J. Am. Chem. Soc. 2004, 126, 8266–8274. (b)
Delrossi, K. J.; Zhang, X. X.; Wayland, B. B. J. Organomet. Chem. 1995,
504, 47–56. (c) Zhang, X. X.; Wayland, B. B. J. Am. Chem. Soc. 1994,
116, 7897–7898. (d) Wayland, B. B.; Ba, S. J.; Sherry, A. E. Inorg. Chem.
1992, 31, 148–150. (e) Wayland, B. B.; Ba, S.; Sherry, A. E. J. Am. Chem.
Soc. 1991, 113, 5305–5311. (f) Sherry, A. E.; Wayland, B. B. J. Am. Chem.
Soc. 1990, 112, 1259–1261.
Acknowledgment. We are grateful to the National Science
Foundation for support of this work under Grant CHE-9904798.
We also acknowledge the Property Measurement Shared Equip-
ment Facility at the University of Pennsylvania for magnetization
measurements, supported in part by NSF MRSEC DMR05-
20020.
(23) Puschmann, F. F.; Gru¨tzmacher, H.; de Bruin, B. J. Am. Chem. Soc. 2010,
132, 73–75.
JA100894H
9
J. AM. CHEM. SOC. VOL. 132, NO. 12, 2010 4109