Low-vaient ruthenium complexes of the non-innocent 2, 6-bis(imino)pyridine ligand

Article abstract of DOI:10.1021/om9009075

A series of low-valent ruthenium complexes bearing 2, 6-bis(imino)pyridyl ( [N₃] ) ligands has been synthesized and characterized. Reduction of [N₃]RuCl₂(C₂H₄) ([N₃^xyl] = 2,6-(XyINCMe)₂-C₅H ₃N, la; [N₃^mes] = 2, 6-(MesNCMe) ₂C₅H₂N, lb; [^tBu-N₃ ^mes] = 2, 6-(MeSNCMe)₂p-^tBuC₅H ₂N, 1c) with hydridosilanes in an arene solvent such as toluene yields new 18e^- η⁶-arene complexes [κ²-N₃]Ru(η⁶-MeC₆H ₅), 2a, b, c, in which the [N₃] ligand is bidentate and only one imine group is coordinated to the metal. The arene ligand can be displaced with dinitrogen in non-arene solvents to yield the binuclear, four-coordinate, formally Ru(0) complexes {[N₃]RU} ₂(μ-N₂), 3a,b,c. Pyrophoric complex 3c is a rare example of a structurally characterized Ru(0) dinitrogen complex. Treatment of low-valent complexes 2 or 3 with donor ligands generates five-coordinate complexes [N₃^xyl]RuL_1.2 (L_1.2 = C₂H₄, 4a; L_1.2 = PMe₃, 5a; L _1.2 = CO, 6a; L₁ = PMe₃, L₂ = CO, 7a). Complexes 2a, 3c, 5a, 6a, and 7a are diamagnetic and have been structurally characterized by single-crystal X-ray diffraction methods. New six-coordinate Ru(II) complexes [N₃^xyl]RuCl₂(L) (L = PMe ₃, CO) were also isolated and structurally characterized. The infrared data, observed geometrical parameters, and reactivity patterns of the formally Ru(0) centers suggest varying degrees of electron derealization to the non-innocent bis(imino)pyridyl, but probably not to the extent implied by the valence tautomeric [N₃]^2-/Ru(II) canonical form. Although the [N₃]^-/Ru(I) representation may portray the electron distribution more accurately than Ru(O) , the inherent odd electron counts on both ligand and metal-and requisite antiferromagnetic coupling-provides little in the way of useful distinctions or predictive value for the low-valent [N₃]Ru(L)₂ complexes with strong-field co-ligands such as CO and PMe₃, These five-coordinate adducts seem to be adequately described as Ru(0) complexes of the neutral [N₃] ligand. However, non-innocent valence tautomeric canonical forms such as [N₃]^-/Ru⁺ may be more applicable to the four-coordinate dinitrogen complexes ([N ₃]Ru)₂(μ-N₂).

Full text of DOI:10.1021/om9009075

Organometallics 2010, 29, 591–603 591
DOI: 10.1021/om9009075
Low-Valent Ruthenium Complexes of the Non-innocent
2,6-Bis(imino)pyridine Ligand
Michelle Gallagher, Noah L. Wieder, Vladimir K. Dioumaev, Patrick J. Carroll, and
Donald H. Berry*
Department of Chemistry, University of Pennsylvania, Philadelphia, Pennsylvania 19104-6323
Received October 14, 2009
A series of low-valent ruthenium complexes bearing 2,6-bis(imino)pyridyl (“[N3]”) ligands has
been synthesized and characterized. Reduction of [N3]RuCl2(C2H4) ([N3xyl] = 2,6-(XylNCMe)2-
C5H3N, 1a; [N3mes]=2,6-(MesNCMe)2C5H2N, 1b; [tBu-N3mes]=2,6-(MesNCMe)2p-tBuC5H2N, 1c)
with hydridosilanes in an arene solvent such as toluene yields new 18e- η6-arene complexes
[κ2-N3]Ru(η6-MeC6H5), 2a,b,c, in which the [N3] ligand is bidentate and only one imine group is
coordinated to the metal. The arene ligand can be displaced with dinitrogen in non-arene solvents to
yield the binuclear, four-coordinate, formally Ru(0) complexes {[N3]Ru}2(μ-N2), 3a,b,c. Pyrophoric
complex 3c is a rare example of a structurally characterized Ru(0) dinitrogen complex. Treatment of
low-valent complexes 2 or 3 with donor ligands generates five-coordinate complexes [N3xyl]RuL1,2
(L1,2=C2H4, 4a; L1,2=PMe3, 5a; L1,2=CO, 6a; L1=PMe3, L2=CO, 7a). Complexes 2a, 3c, 5a, 6a,
and 7a are diamagnetic and have been structurally characterized by single-crystal X-ray diffraction
methods. New six-coordinate Ru(II) complexes [N3xyl]RuCl2(L) (L=PMe3, CO) were also isolated
and structurally characterized. The infrared data, observed geometrical parameters, and reactivity
patterns of the formally Ru(0) centers suggest varying degrees of electron delocalization to the
“non-innocent” bis(imino)pyridyl, but probably not to the extent implied by the valence tautomeric
[N3]2-/Ru(II) canonical form. Although the [N3]-/Ru(I) representation may portray the electron
distribution more accurately than “Ru(0)”, the inherent odd electron counts on both ligand
and metal;and requisite antiferromagnetic coupling;provides little in the way of “useful”
distinctions or predictive value for the low-valent [N3]Ru(L)2 complexes with strong-field co-ligands
such as CO and PMe3. These five-coordinate adducts seem to be adequately described as Ru(0)
complexes of the neutral [N3] ligand. However, “non-innocent” valence tautomeric canonical
forms such as [N3]-/Ruþ may be more applicable to the four-coordinate dinitrogen complexes
{[N3]Ru}2(μ-N2).
Introduction
pincer-type ligands have gained attention due to their use
in catalysis and in other chemical transformations.1-5 For
example, bis(imino)pyridyl complexes of Ni, Pd, Fe, and Co
are effective catalysts for the polymerization of ethylene,1b-h
and [N3] complexes of ruthenium in particular catalyze a
variety of reactions, including cyclohexene epoxidation2 and
cyclopropanation of styrene.3 Chirik and co-workers have
also demonstrated that [N3]Fe complexes catalyze reactions
such as hydrogenation, hydrosilylation, and [2π þ 2π]
cycloaddition reactions.4 These ligands and other similar
pincer complexes have thus emerged as a convenient and
synthetically flexible alternative to the well-studied
2,20:60,200-terpyridine (terpy) ligand.5-7
Over the past decade, late transition metal complexes
bearing 2,6-bis(imino)pyridyl ligands, [N3], and other
*Corresponding author. E-mail: berry@sas.upenn.edu.
(1) (a) Gibson, V. C.; Redshaw, C.; Solan, G. A. Chem. Rev. 2007,
1745–1776. (b) Ittel, S. D.; Johnson, L. K.; Brookhart, M. Chem. Rev. 2000,
100, 1169–1203. (c) Paulino, I. S.; Schuchardt, U. Catal. Commun. 2004, 5,
5–7. (d) Chen, Y.; Chen, R.; Qian, C.; Dong, X.; Sun, J. Organometallics
2003, 22, 4312–4321. (e) Britovsek, G. J. P.; Bruce, M.; Gibson, V. C.;
Kimberley, B. S.; Maddox, P. J.; Mastroianni, S.; McTavish, S. J.; Redshaw,
€
C.; Solan, G. A.; Stromberg, S.; White, A. J. P.; Williams, D. J. J. Am. Chem.
Soc. 1999, 121, 8728–8740. (f) Small, B. L.; Brookhart, M. J. Am. Chem.
Soc. 1998, 120, 7143–7144. (g) Small, B. L.; Brookhart, M.; Bennett, A. M.
A. J. Am. Chem. Soc. 1998, 120, 4049–4050. (h) Bouwkamp, M. W.;
Lobkovsky, E.; Chirik, P. J. J. Am. Chem. Soc. 2005, 127, 9660–9661.
(2) Cetinkaya, B.; Cetinkaya, E.; Brookhart, M.; White, P. S. J. Mol.
Catal. A 1999, 142, 101–112.
Tridentate bis(imino)pyridyl ligands enforce the mer geo-
metry that has been found to promote catalytic activity in
(3) Bianchini, C.; Lee, H. M. Organometallics 2000, 19, 1833–1840.
(4) (a) Bart, S. C.; Lobkovsky, E.; Chirik, P. J. J. Am. Chem. Soc.
2004, 126, 13794–13807. (b) Bart, S. C.; Lobkovsky, E.; Bill, E.; Chirik, P. J.
J. Am. Chem. Soc. 2006, 128, 5302–5303. (c) Archer, A. M.; Bouwkamp,
M. W.; Cortez, M.-P.; Lobkovsky, E.; Chirik, P. J. Organometallics 2006, 25,
4269–4278. (d) Bouwkamp, M. W.; Bowman, A. C.; Lobkovsky, E.; Chirik,
P. J. J. Am. Chem. Soc. 2006, 128, 13340–13341.
(5) Van der Boom, M. E.; Milstein, D. Chem. Rev. 2003, 103, 1759–
1792.
´
(6) Abbenhuis, R. A. T. M.; del Rıo, I.; Bergshoef, M. M.; Boersma,
J.; Veldman, N.; Spek, A. L.; van Koten, G. Inorg. Chem. 1998, 37, 1749–
1758.
(7) Thompson, A. M. W. C. Coord. Chem. Rev. 1997, 160, 1–52.
r
2010 American Chemical Society
Published on Web 01/13/2010
pubs.acs.org/Organometallics

592 Organometallics, Vol. 29, No. 3, 2010
Gallagher et al.
certain ruthenium phosphine complexes.8 For example, we
have previously reported that the active catalytic species
for dehydrocoupling of silanes to carbosilanes are
16e- (PMe3)3Ru(SiR3)(R) complexes (R=H or alkyl) with
the meridional geometry.8,9 These complexes undergo facile,
productive intra- and intermolecular addition of H-H,
Si-H, C-H, and Si-C bonds, leading to C-H activation
and functionalization. Similarly, mer-(PMe3)3Ru(GeR3)2
(R = alkyl or aryl), generated by phosphine loss from the
tetrakis-phosphine complex, has been postulated to be the
active catalyst in the synthesis of polygermanes by demetha-
native coupling of hydridogermanes.10 One impediment to
rapid dehydrogenative catalytic turnover is the thermo-
dynamic and kinetic stability of fac-(Me3P)3Ru(H)3(SiR3)
complexes. The tridentate bis(imino)pyridine family of
ligands seemed appropriate to enforce the desired mer-
geometry and offered the advantage of convenient
ligand tuning.2 Furthermore, the known Ru(II) complex
[N3xyl]RuCl2(C2H4)11 could serve as a convenient entry into
the desired silyl and germyl complexes. Although a large
number of tridentate pyridine ligand complexes of
Ru(II), Ru(III), and Ru(IV) are known,2,3,5-7,11,12 there
are very few reports of comparable Ru(0) complexes,
and all of those involve strongly π-accepting co-ligands.13
Shiotsuki and co-workers isolated mononuclear ruthe-
nium complexes containing terpy, 2,6-bis(imino)pyridine,
and 2,6-bis(oxazolinyl)pyridine ligands as the bis(fuma-
rates).13 Milstein and co-workers proposed the intermediacy
of a [PNP]Ru(0) complex ([PNP]=bis(phosphinomethyl)-
pyridine) in the catalytic dehydrogenation of alcohols to
ketones. Although only Ru(II) complexes were isolated or
directly observed in the catalytic reaction,14a a related
[PNP]Ru(0) dicarbonyl derivative has been described very
recently.14b In this contribution we report the synthesis and
characterization of five new formally [N3]Ru(0) complexes
containing arene, dinitrogen, phosphine, and carbonyl li-
gands. The synthesis of silylene complexes from the reactions
of [N3]Ru(0) complexes with hydrido- and chlorosilanes was
Figure 1. ORTEP drawing of [N3xyl]RuCl2(PMe3) (30% ther-
mal ellipsoids). Selected bond lengths (A) and angles (deg):
Ru1-N1, 2.116(3); Ru1-N2, 1.971(3); Ru1-N3, 2.121(2);
Ru1-P1, 2.3567(10); Ru1-Cl1, 2.3902(10); Ru1-Cl2, 2.4016(10);
N2-Ru1-P1, 179.68(9); N1-Ru1-N3, 155.15(10); Cl1-
Ru1-Cl2, 177.54(4).
recently described in a preliminary communication from this
group.15
Results
Phosphine and Carbonyl Complexes of [N3]RuCl2. Treat-
ment of [N3xyl]RuCl2(C2H4) (1a) with PMe3 or carbon
monoxide at room temperature leads to ethylene loss and
formation of the corresponding phosphine or carbonyl
adducts trans-[N3xyl]RuCl2(L) (L=PMe3, CO), which can
be isolated in 80-90% yields (eq 1). The 31P{1H} NMR of
[N3xyl]RuCl2(PMe3) exhibits one singlet at δ -5.77 for the
PMe3 ligand. The CO ligand in [N3xyl]RuCl2(CO) is observed
as a singlet at δ 202.51 in the 13C{1H} NMR spectrum. In the
IR spectrum, the carbonyl complex exhibits νCO at 1963
cm-1, only slightly higher than the value of 1948 cm-1
reported for [(mer,trans-RuCl2(CO)(NN0N)] (NN0N = 2,
6-bis[(dimethylamino)methyl]pyridine), which features satu-
rated amines in the chelating ligand, rather the conjugated
imines in the [N3] system.6
(8) Dioumaev, V. K.; Procopio, L. J.; Carroll, P. J.; Berry, D. H.
J. Am. Chem. Soc. 2003, 125, 8043–8058.
(9) (a) Procopio, L. J.; Berry, D. H. J. Am. Chem. Soc. 1991, 113,
€
4039–4040. (b) Procopio, L. J.; Mayer, B.; Plossl, K.; Berry, D. H. Polym.
Prepr. Am. Chem. Soc., Div. Polym. Chem. 1992, 33, 1241–1242. (c)
Dioumaev, V. K.; Yoo, B. R.; Procopio, L. J.; Carroll, P. J.; Berry, D. H.
€
J. Am. Chem. Soc. 2003, 125, 8936–8948. (d) Dioumaev, V. K.; Plossl, K.;
Carroll, P. J.; Berry, D. H. Organometallics 2000, 19, 3374–3378. (e)
€
Dioumaev, V. K.; Plossl, K.; Carroll, P. J.; Berry, D. H. J. Am. Chem. Soc.
1999, 121, 8391–8392.
(10) (a) Reichl, J. A.; Popoff, C. M.; Gallagher, L. A.; Remsen, E. E.;
Berry, D. H. J. Am. Chem. Soc. 1996, 118, 9430–9431. (b) Katz, S. M.;
Reichl, J. A.; Berry, D. H. J. Am. Chem. Soc. 1998, 120, 9844–9849. (c)
Huo, Y.; Berry, D. H. Chem. Mater. 2004, 17, 157–163.
(11) Dias, E. L.; Brookhart, M.; White, P. S. Organometallics 2000,
19, 4995–5004.
(12) (a) del Rıo, I.; Gossage, R. A.; Hannu, M. S.; Lutz, M.;
´
Spek, A. L.; van Koten, G. Organometallics 1999, 18, 1097–1105. (b)
Vogler, L. M.; Brewer, K. J. Inorg. Chem. 1996, 35, 818–824. (c)
Chanda, N.; Sarkar, B.; Fiedler, J.; Kaim, W.; Lahiri, G. K. J. Chem.
Soc., Dalton Trans. 2003, 3550–3555. (d) Abbenhuis, R. A. T. M.;
Boersma, J.; van Koten, G. J. Org. Chem. 1998, 63, 4282–4290. (e) del
Río, I.; van Koten, G. Organometallics 2000, 19, 361–364. (f) Welch,
T. W.; Ciftan, S. A.; White, P. S.; Thorp, H. H. Inorg. Chem. 1997, 36,
4812–4821. (g) Meyer, T. J.; Huynh, M. H. V. Inorg. Chem. 2003, 42,
8140–8160.
(13) Shiotsuki, M.; Suzuki, T.; Iida, K.; Ura, Y.; Wada, K.; Kondo,
T.; Mitsudo, T. Organometallics 2003, 22, 1332–1339.
(14) (a) Zhang, J.; Gandelman, M.; Shimon, L. J.; Rozenberg, H.;
Milstein, D. Organometallics 2004, 23, 4026–4033. (b) Salem, H.; Shimon,
L. J. W.; Diskin-Posner, Y.; Leitus, G.; Ben-David, Y.; Milstein, D. Organo-
metallics 2009, 28, 4971–4806.
The solid-state structures of complexes trans-[N3xyl]-
RuCl2(L) (L=PMe3, CO) were determined by single-crystal
X-ray diffraction methods. Both complexes exhibit a
six-coordinate, pseudo-octahedral geometry with a trans
arrangement of the two chloride ligands (L = PMe3,
Figure 1; L=CO, Figure 2). The main deviation from the
ideal octahedral geometry is the trans Nimine-Ru-Nimine
bond angle (L = PMe3, 155.15(10)°; L = CO, 153.58(8)°),
(15) Yoo, H.; Carroll, P. J.; Berry, D. H. J. Am. Chem. Soc. 2006, 128,
6038–6039.

Article
Organometallics, Vol. 29, No. 3, 2010 593
(or Me3SiH) in an arene solvent (e.g., toluene or benzene)
leads to formation of intensely purple [κ2-N3]Ru(η6-arene)
complexes (2a,b,c) isolated in 75-85% yield (eq 2). GC and
1H NMR analysis indicates the concurrent production of
Et3SiCl, Et4Si, and ethane. An intermediate ruthenium
hydride complex can be observed by 1H NMR during
the reduction of 1a and is tentatively assigned as
[N3xyl]Ru(H)(Cl)(C2H4) on the basis of an upfield reso-
nance for the Ru-H characteristic of a trans Cl-Ru-H
arrangement (δ -21.91, 1H) and a resonance for coordi-
nated ethylene (δ 4.12, 4H) that is fluxional on the NMR
time scale. The PMe3 complex [N3xyl]RuCl2(PMe3) does
not react with silanes at room temperature, but leads to
[N3xyl]Ru(H)(Cl)(PMe3) after 3.5 h at 150 °C. The Ru-H
resonance for the phosphine complex is observed as a
1
doublet at δ -19.17 (2JP-H = 42.5 Hz) in the H NMR,
similar to the intermediate in the reduction of 1a. Further
heating of [N3xyl]RuCl2(PMe3) with silanes leads to exten-
sive decomposition.
Figure 2. ORTEP drawing of [N3xyl]RuCl2(CO) (30% thermal
ellipsoids). Selected bond lengths (A) and angles (deg):
Ru1-N1, 2.108(2); Ru1-N2, 2.130(2); Ru1-N6, 2.0056(19);
Ru1-C26, 1.870(3); Ru1-Cl1, 2.4045(7); Ru1-Cl2, 2.3930(7);
C26-O1, 1.149(3); N6-Ru1-C26, 177.11(10); N1-Ru1-N2,
154.68(8); Cl1-Ru1-Cl2, 173.15(2). Chlorobenzene solvent of
crystallization omitted.
Arene complexes 2 exhibit “two-legged piano-stool” geo-
metries, with a bidentate [N3] ligand necessary to maintain
the 18e- count at ruthenium and provide steric relief at the
metal center. This “arm-off” coordination mode leads to
chemically inequivalent xylyl (or mesityl) and imine methyl
groups in the NMR spectra. The coordinated arene ligand in
2 is labile and undergoes exchange with other arenes within
minutes at room temperature. The structure of 2a was
confirmed by crystallography (Figure 3). Bidentate (κ2)
coordination of potentially tridentate [N3] ligands has been
previously observed in low-valent complexes of several
metals.4c,17
which is typical for bis(imino)pyridyl complexes.2,11 The
xylyl groups are oriented approximately perpendicular to
the [N3] ligand plane. Although only the trans isomers of
[N3xyl]RuCl2(L) are observed, both cis and trans isomers
have been previously reported for the triphenylphosphine
and carbonyl adducts of related ruthenium complexes.2,6,16
Synthesis and Structure of Ru(0) μ-Dinitrogen Complexes.
Treatment of the [κ2-N3]Ru(η6-arene) complexes with
dinitrogen in non-arene solvents at room temperature
leads to a color change from dark purple to intense
turquoise and formation of equilibrium mixtures contain-
ing 2 and Ru(0) bridging dinitrogen complexes, 3 (eq 3). In
the case of xylyl-substituted 2a, the slightly higher solu-
bility of the starting complexes in hexanes can be
exploited, and 3a can be prepared in 80-90% yield by
stirring a slurry of 2a under N2 for 2 days. The bridging
dinitrogen ligand is labile, and 3 rapidly reverts to 2 in
arene solvents. It should be noted that the dinitrogen
complexes are pyrophoric solids and burst into flames
Reduction of [N3]RuCl2(C2H4) with R3SiH: Synthesis of
Ru(0) Arene Complexes. Treatment of the [N3]Ru(II) ethy-
lene adducts 1a,b,c ([N3xyl] = 2,6-(XylNCMe)2C5H3N, 1a;
[N3mes] = 2,6-(MesNCMe)2C5H2N, 1b; [tBu-N3mes] =
2,6-(MesNCMe)2p-tBuC5H2N, 1c) with excess Et3SiH
(17) (a) Shirin, Z.; Mukherjee, R.; Richardson, J. F.; Buchanan,
R. M. J. Chem. Soc., Dalton Trans. 1994, 465–469. (b) Lu, S.; Selbin, J.
Inorg. Chim. Acta 1987, 134, 229–232. (c) Albon, J. M.; Edwards, D. A.;
Moore, P. J. Inorg. Chim. Acta 1989, 159, 19–22.
(16) Ben-Hadda, T.; Mountassir, C.; Le Bozec, H. Polyhedron 1995,
14, 953–955.

594 Organometallics, Vol. 29, No. 3, 2010
Gallagher et al.
Figure 4. ORTEP drawing of ([tBu-N3mes]Ru)2(μ-N2), 3c (30%
thermal ellipsoids). Selected bond lengths (A) and angles (deg):
Ru1-N1,2.020(5);Ru1-N2, 1.946(7); Ru1-N3, 2.990(5); Ru1-N7,
1.931(8); N7-N8, 1.161(5); Ru2-N8, 1.899(8); Ru2-N4, 2.020(5);
Ru2-N5, 1.938(7); Ru2-N6, 2.026(4); N1-Ru1-N3, 155.29(18);
N7-Ru1-N2, 175.6(4); Ru1-N7-N8, 173.1(4); N4-Ru2-N6,
156.28(17); N8-Ru2-N5, 172.6(4); Ru2-N8-N7, 173.6(4).
ligand (Figure 4). Although μ-dinitrogen complexes of
ruthenium(II) are very well known,6,9e,14,18 complexes 3a-c
appear to be the first examples of formally Ru(0) dinitrogen-
bridged complexes. Interestingly, it was only very recently that
Field and co-workers described the first Ru(0) complex bearing
a terminal dinitrogen ligand.19 The two ruthenium centers of 3c
are connected by an essentially linear bridge (RuNN=173.1(4),
173.6(4) A), andthetwoRu[N3] planes are perpendicular, which
allows the bulky mesityl rings to interdigitate and provide steric
protection for the 16e- Ru(0) centers. This type of arrangement
has been observed in other μ-dinitrogen complexes containing
pincer ligands, such as [Ir{C6H3-2,6-(CH2PtBu2)2}]2(μ-N2)20
and [(mer,trans-RuCl2{NN0N})2(μ-N2)], where NN0N =
2,6-bis[(dimethylamino)methyl]pyridine.6 The N-N bond dis-
tance in 30 (1.161(5) A) is longer than that in Ru(II) μ-dinitrogen
complexes (1.104(8)-1.138(8) A)6,9e,14,18 and much longer than
in free dinitrogen (1.0976(15) A).21 Similarly, the average
Ru-N2 distance in 3c (1.915(8) A) is on the shorter end of the
range found in Ru(II) complexes (1.916(4)-2.073(4) A). The
average Ru-Npyr distance is 1.942(7) A, whereas the average
Ru-Nim distance is 2.014(5) A.
Figure 3. ORTEP drawing of [κ2-N3xyl]Ru(η6-MeC6H5), 2a (30%
thermal ellipsoids). Selected bond lengths (A) andangles(deg):Ru1-
N2, 2.0315(18); Ru1-N3, 1.9854(17); N2-Ru1-N3, 77.62(7).
upon exposure to air. Complexes 3 are, however, stable
under nitrogen as solids or in solution.
The IR spectrum of 3a (Nujol) exhibits a very weak signal
at 1856 cm-1, tentatively assigned as the N-N stretching
frequency. Although this band should not be IR-active in the
idealized D2d geometry, some intensity may be gained due to
slight nonlinearity of the Ru-N-N-Ru linkage and by
weak coupling to other normal modes in the molecule.18g
Attempts to measure the Raman spectrum of 3 were
unsuccessful due to the very intense absorption in the
UV-vis region and photolability of the complex.
The 1H NMR spectrum of 3a exhibits one singlet each for
xylyl methyl and imine methyl groups (2:1), indicative of two
planes of symmetry;one containing and one bisecting the
[N3] ligand;which would be consistent with either mono-
nuclear (C2v) or binuclear (D2d) structures in solution. How-
ever, conclusive evidence of the binuclear structure in
solution is found in the results of a crossover experiment.
Treatment of xylyl-substituted 3a with the mesityl analogue
The geometry of 3c in the solid state consists of two
square-planar ruthenium centers bridged by a dinitrogen
(18) (a) Shaver, M. P.; Fryzuk, M. D. Adv. Synth. Catal. 2003, 345,
1061–1076. (b) Fryzuk, M. D.; Johnson, S. A. Coord. Chem. Rev. 2000,
200-202, 379–409. (c) Hidai, M.; Mizobe, Y. Chem. Rev. 1995, 95, 1115–
ꢀ
1133. (d) Aneetha, H.; Jimenez-Tenorio, M.; Puerta, M. C.; Valerga, P.
Organometallics 2002, 21, 628–635. (e) Treitel, I. M.; Flood, M. T.; Marsh,
R. E.; Gray, H. B. J. Am. Chem. Soc. 1969, 91, 6512–6513. (f) Rodriguez,
V.; Atheaux, I.; Donnadieu, B.; Sabo-Etienne, S.; Chaudret, B. Organo-
metallics 2000, 19, 2916–2926. (g) Abdur-Rashid, K.; Gusev, D. G.; Lough,
A. J.; Morris, R. H. Organometallics 2000, 19, 1652–1660. (h) Gusev, D. G.;
Dolgushin, F. M.; Antipin, M. Y. Organometallics 2000, 19, 3429–3434.
(19) Field, L. D.; Guest, R. W.; Vuong, K. Q.; Daigarno, S. J.; Jensen,
P. Inorg. Chem. 2009, 48, 2246–2253.
(20) Lee, D. W.; Kaska, W. C.; Jensen, C. M. Organometallics 1998,
17, 1–3.
(21) Wilkinson, P. G.; Houk, N. B. J. Chem. Phys. 1956, 24, 528–534.

Article
Organometallics, Vol. 29, No. 3, 2010 595
3b at room temperature rapidly leads to an equilibrium
distribution of the starting homobinuclear complexes and a
new species, assigned as the heterobinuclear complex
[N3xyl]Ru{μ-N2}Ru[N3mes] (eq 4). Note that a new mixed
species would not be expected if 3a and 3b were mononuclear
in solution. This experiment also establishes that although
the binuclear form is predominant in solution, dissociation
occurs readily.
Although the 1H NMR spectrum of 3 is broadly consistent
with a closed-shell configuration and exhibits temperature-
independent resonances generally within the normal shift ranges,
the imine methyl resonance (δ -0.08) is found ca. 2 ppm upfield
from the free ligand (δ 2.61 in benzene-d6) and the other
[N3xyl]Ru complexes (e.g., δ 2.09 for [N3xyl]Ru(CO)2 (6a), vide
infra). Furthermore, the imine shift is essentially invariant from
300 to 230 K. Chirik and co-workers reported temperature-
independent 1H chemical shifts for the imine-methyl groups in a
series of closely related [N3Ar] iron complexes that ranged from
2.31 to -6.90 (Δδ ≈ 0-9 ppm) and attributed the unusual shifts
to temperature-independent paramagnetism (TIP) arising from
the admixture of low-lying states of higher multiplicity into the
diamagnetic ground states.22 Similarly shifted imine-methyl
resonances have also been reported by Budzelaar and co-work-
ers for a series of [N3Ar] cobalt complexes.23 In the present case,
the unusual;and temperature-independent;shift of the imine
methyl group resonance of 3 most likely indicates a similar
contribution of an S = 1 excited state to the singlet ground state.
Synthesis and Characterization of Ru(0) Complexes [N3xyl]-
RuL2. Treatment of both [N3xyl]Ru(0) complexes 2a and 3a with
donor or π-acceptor ligands leads to rapid formation of mono-
nuclear [N3xyl]RuL2 complexes (L2=(C2H4)2, 4a; (PMe3)2, 5a;
(CO)2, 6a) (eq 5). Compounds 5a and 6a have been isolated in
high yield (86% and 78%, respectively). Compound 4a has been
characterized in solution by multinuclear NMR, although the
lability of the ethylene ligands has prevented isolation.
Figure 5. ORTEP drawing of [N3xyl]Ru(PMe3)2, 5a (30% ther-
mal ellipsoids). Selected bond lengths (A) and angles (deg):
Ru1-N1, 2.0631(16); Ru1-N2, 1.9705(16); Ru1-N3, 2.0673(16);
Ru1-P1, 2.3121(5); Ru1-P2, 2.2489(6); N3-Ru1-N1, 146.02(6);
N2-Ru1-P2, 95.47(5); N2-Ru1-P1, 166.41(5); P2-Ru1-P1,
98.12(2).
nor reversion to 2a was observed after days at 150 °C in
toluene solution. Single-crystal X-ray analysis of complex 5a
reveals a slightly distorted square-pyramidal geometry at the
metal (Figure 5), with one phosphine trans to the pyridine
(N2-Ru1-P1=166.41(5)°) and the other cis (N2-Ru1-P2
= 95.47(5)°). The metal atom is displaced by 0.469(1) A
toward the apical phosphine, out of the plane formed by the
tridentate [N3] ligand. Other bond distances and angles are
within normal ranges.
1
The H NMR spectrum of 5a at -75 °C in toluene-d8
exhibits two doublets (δ 0.24 and 0.77), corresponding to the
two chemically inequivalent phosphines. Two doublets are
also observed in the 31P{1H} NMR spectrum at -75 °C
(δ -3.16 and 13.95, 2JPP=13 Hz). Compound 5a is fluxional,
however, and coalescence of the phosphine resonances is
observed at ca. -20 °C in the proton NMR and ca. 10 °C in
the 31P NMR spectrum (ΔGq ≈ 11.8 kcal/mol at 10 °C). At
room temperature, the 1H NMR spectrum exhibits only one
doublet for the phosphine ligands (δ 0.52), and a single broad
phosphine resonance is observed by 31P{1H} NMR (δ 4.83).
Treatment of 5a with excess PMe3-d9 did not lead to incorpo-
ration of labeled phosphine after 4 days at room temperature
(1H and 2H NMR), indicating that intermolecular phosphine
exchange is very slow and that fluxionality on the NMR
time scale is intramolecular (Scheme 1). Rotation of the
xylyl groups of the [N3] ligand is precluded by steric factors;
thus inequivalent pairs of xylyl methyl groups are also
observed in the low-temperature 1H NMR spectrum of
5a, whereas a single resonance for all four xylyl methyl
groups is observed in the high-temperature limit, when
phosphine ligand exchange is rapid. Note that interchange
of the xylyl methyls is concurrent with exchange of the
The bis(phosphine) complex (5a) is a red solid that is quite
stable under an insert atmosphere: neither decomposition
(22) (a) Bart, S. C.; Lobkovsky, E.; Bill, E.; Wieghardt, K.; Chirik,
P. J. Inorg. Chem. 2007, 46, 7055–7063. (b) 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.
(23) Knijnenburg, Q; Hetterscheid, D.; Kooistra, T.; Budzelaar, P.
Eur. J. Inorg. Chem. 2004, 1204–1211.

596 Organometallics, Vol. 29, No. 3, 2010
Gallagher et al.
140.76(11)°). In other words, the Ru(CO)2 fragment is
slightly (5°) canted toward the square-pyramidal geometry
exhibited by 5a. The ruthenium center lies slightly above the
plane (0.074(1) A) of the [N3] ligand, but much less so than in
the bis(phosphine) complex, 5a.
Reaction of 6a with excess PMe3 at 25 °C leads to the
formation of [N3xyl]Ru(PMe3)(CO), 7a, isolated as an orange
solid in 72% yield (eq 6). Similarly, treatment of 5a with CO
also generates 7a, but this path is less useful synthetically, as
the reaction proceeds readily to produce 6a. The 31P{1H}
NMR spectrum of 7a exhibits one singlet at δ 1.54 for the
PMe3 ligand, and the CO ligand is observed as a doublet at
δ 206.12 (2JPC =27.7 Hz) in the 13C{1H} NMR spectrum.
Complex 7a exhibits νCO at 1865 cm-1 in the IR spectrum,
consistent with a monocarbonyl Ru(0) species. In compa-
rison, the carbonyl stretch in the Ru(II) monocarbonyl
complex [N3xyl]RuCl2(CO) is found at much higher energy
(1963 cm-1, vide supra). It is interesting that the bis(phos-
phine) complex reacts readily with CO, but does not undergo
intermolecular exchange with PMe3-d9, suggesting substitu-
tion does not occur by initial dissociation of phosphine from
5a, but rather by a sterically sensitive associative path. The
most likely possibility would involve dissociation of an imine
arm of the chelate (cf. compound 2), although associative
attack of CO on the 18e- 5a cannot be excluded at this
time.
Figure 6. ORTEP drawing of [N3xyl]Ru(CO)2, 6a (30% thermal
ellipsoids). Selected bond lengths (A) and angles (deg):
Ru1-N1, 2.072(2); Ru1-N2, 1.947(2); Ru1-N3, 2.062(2);
Ru1-C26, 1.878(3); Ru1-C27, 1.884(3); C26-O1, 1.144(3);
C27-O2, 1.146(3); N3-Ru1-N1, 153.58(8); N2-Ru1-C26,
130.60(10); N2-Ru1-C27, 140.74(11); C26-Ru1-C27,
88.6512).
Complex 7a has also been characterized by X-ray diffrac-
tion methods (Figure 7). The geometry of 7a can best be
described as situated between square pyramidal (SP) and
trigonal bipyramidal (TBP), with the CO ligand lying in the
apical position (N2-Ru1-C26=117.59(15)°; C26-Ru1-P1
=85.24(13)°), pseudo-trans to the vacant coordination site.
The ruthenium center is displaced by 0.262(3) A toward the
carbonyl ligand from the plane formed by the [N3] ligand, or
about half the displacement in the bis(phosphine) 5a
(0.469(1) A).
Scheme 1. Intramolecular Exchange of Phosphines in Compound
5a
The vast majority of five-coordinate d6, Ru(II) complexes
exhibit a square-pyramidal geometry, except for sterically
hindered complexes that require the less congested TBP
environment.24 On the other hand, most five-coordinate
d8, Ru(0) complexes containing carbonyl ligands prefer a
TBP ground-state geometry.25,26 The preference for equa-
torial binding by π-acceptor ligands predicted by Hoffman
appears to be a contributing factor.25 Comparison of the
five-coordinate complexes 5-7 reveals a consistent progres-
sion from square pyramidal to trigonal pyramidal with the
phosphine ligand environments and, thus, does not require
rotation around the N-xylyl bonds. Intramolecular phos-
phine exchange presumably occurs via a symmetrical TBP
geometry, which is the observed ground state for the ana-
logous dicarbonyl complex (vide infra).
Dicarbonyl complex 6a was isolated in 78% yield as a
green solid following treatment of 2a with carbon monoxide
(room temperature, 1 atm). Strong bands for the terminal
carbonyl ligands (νCO=1925 and 1978 cm-1) are observed in
the IR spectrum of 6a. The structure of dicarbonyl complex
6a as determined in the solid state is shown in Figure 6.
Although the geometry is best described as trigonal bipyra-
midal with both CO ligands in the equatorial plane, the
OC-Ru-CO angle is quite small (88.66(12)°) and the two
Npyr-Ru-CO angles differ by ca. 10° (130.60(10)° and
(24) (a) Huang, D.; Streib, W. E.; Bollinger, J. C.; Caulton, K. G.;
Winter, R. F.; Scheiring, T. J. Am. Chem. Soc. 1999, 121, 8087–8097. (b)
Crochet, P.; Gimeno, J.; García-Granda, S.; Borge, J. Organometallics 2001,
20, 4369–4377. (c) MacFarlane, K. S.; Joshi, A. M.; Rettig, S. J.; James, B. R.
Inorg. Chem. 1996, 35, 7304–7310. (d) Heyn, R. H.; Huffman, J. C.;
Caulton, K. G. New J. Chem. 1993, 17, 797–803.
(25) (a) Rossi, A. R.; Hoffmann, R. Inorg. Chem. 1975, 14, 365–374.
(b) Elian, M.; Hoffmann, R. Inorg. Chem. 1975, 14, 1058–1076.
(26) Muetterties, E. L. Acc. Chem. Res. 1970, 3, 266–273.

Article
Organometallics, Vol. 29, No. 3, 2010 597
carbonyl derivatives, the bis(ethylene) complex 4a is labile
and is stable only in solution under an ethylene atmosphere.
Complex 4a has not been isolated as a pure solid, but
displays three singlets at δ 2.28 (6H), 1.76 (12H), and 1.63
(8H) in the 1H NMR spectrum, assigned as the chemically
equivalent imine methyl groups, xylyl methyl groups, and
1
two ethylene ligands respectively. The H NMR data are
consistent with a symmetrical trigonal-bipyramidal geo-
metry in which the ethylene protons are averaged by rapid
rotation or dissociation. A comparable geometry was pre-
viously observed for the bis(dimethyl fumarate) pyridine
complexes reported by Mitsudo and co-workers, although
the olefin ligands are not dynamic on the NMR time scale in
those complexes.13
Discussion
Stability and Scope of Reactivity of the Ruthenium(0)
Complexes. The chemical reduction of 1a provides conve-
nient access to 2-7, all of which are formally Ru(0), d8
complexes. Other than mononuclear and cluster carbonyl
complexes,29 there are relatively few examples of isolated
Ru(0) species.15,27,30 Triethylsilane is an excellent reducing
agent for 1, as it is sufficiently mild to avoid over-reduction
or deprotonation,4c,31 and also provides for the irreversible
removal of ethylene by hydrogenation or hydrosilylation.
Furthermore, only volatile byproducts are generated, which
simplifies workup.
Figure 7. ORTEP drawing of [N3xyl]Ru(PMe3)(CO), 7a (30%
thermal ellipsoids). Selected bond lengths (A) and angles (deg):
Ru1-N1, 2.094(3); Ru1-N2, 1.964(3); Ru1-N3, 2.088(3);
Ru1-C26, 1.835(4); Ru1-P1, 2.3446(11); C26-O1, 1.165(5);
N3-Ru1-N1, 150.47(12); N2-Ru1-C26, 117.59(16); N2-
Ru1-P1, 157.17(9); C26-Ru1-P1, 85.24(13).
The η6-arene complexes 2 are the easiest to prepare and
serve as versatile starting materials for subsequent reactions.
It is interesting to note that 1 is produced by displacing an
arene from [(p-cymene)RuCl2]2, but that reduction of 1 to 2
favors coordination of arene at the expense of coordination
of one of the chelate arms. This “arm on/arm off” pheno-
menon has been observed for bis(imino)pyridyl complexes of
Mo and W17b,c and for related tridentate ligands with
ruthenium.17a Recently, Chirik and co-workers reported an
intramolecular version of the equilibrium in eq 3 for an iron
complex, in which a coordinated imine nitrogen dissociates
to permit intramolecular η6 coordination of an arene sub-
stituent on the imine nitrogen.4c The facile arene exchange in
2 and equilibrium reaction to form the tridentate 3 highlights
the lability of the arene and versatility of the bis(imino)-
pyridyl ligand.
introduction of smaller, π-acceptor CO ligands. The degree
of displacement of the metal from the [N3] ligand plane also
changes with the nature of the co-ligands: 5a (0.469(1) A) >
7a (0.262(3) A) > 6a (0.074(1) A). Virtually identical trends
are seen in the structurally characterized Ru(0) complexes,
(dmpe)2Ru(L) (L = PMe3, CO), in which the phosphine
derivative adopts a SP geometry with the ruthenium above
the ligand square plane (0.53 A), and the carbonyl adduct
exhibits a TBP geometry.27d Caulton, Eisenstein, and co-
workers recently reported theoretical calculations of steric
and electronic effects in the model complex Ru(CO)2(PH3)3,
revealing a very small (<5 kcal/mol) electronic preference
for carbonyl ligands in the equatorial position, and thus
steric factors can easily dictate the geometry.28,37 Steric
factors may well determine the apical;not equatorial;
position of the CO ligand in the solid-state structure of 7a.
Treatment of either of the [N3xyl]Ru(0) complexes 2 or 3
with ethylene leads to rapid formation of the [N3xyl]RuL2
complex 4a (L1,2 = C2H4). Unlike the phosphine and
The bridging dinitrogen complex 3 is of interest for several
reasons. The first example of a bridging N2 ligand was the
diruthenium complex prepared by the Taube group in 196832
(29) (a) Whittlesey, M. K. In Comprensive Organometallic Chemistry
III; Mingos, D. P. M., Crabtree, R. H., Eds.; Elsevier: Amsterdam, 2007. (b)
Bruce, M. I.; Stone, F. G. A. Angew. Chem., Int. Ed. Engl. 1968, 7, 427–432.
(c) Sappa, E. J. Cluster Sci. 1994, 5, 211–263.
(30) (a) Suzuki, T.; Shiotsuki, M.; Wada, K.; Kondo, T.; Mitsudo, T.
Organometallics 1999, 18, 3671–3678. (b) Suzuki, T.; Shiotsuki, M.; Wada,
K.; Kondo, T.; Mitsudo, T. J. Chem. Soc., Dalton Trans. 1999, 4231–4237.
(c) Shiotsuki, M.; Suzuki, T.; Kondo, T.; Wada, K.; Mitsudo, T. Organo-
metallics 2000, 19, 5733–5743. (d) Shiotsuki, M.; Miyai, H.; Ura, Y.; Suzuki,
T.; Kondo, T.; Mitsudo, T. Organometallics 2002, 21, 4960–4964. (e) Ura,
Y.; Sato, Y.; Shiotsuki, M.; Suzuki, T.; Wada, K.; Kondo, T.; Mitsudo, T.
Organometallics 2003, 22, 77–82.
(31) (a) Sugiyama, H.; Ghazar, A.; Gambarotta, S.; Yap, G. P. A.;
Budzelaar, P. H. M. J. Am. Chem. Soc. 2002, 124, 12268–12274. (b)
Khorobkov, I.; Gambarotta, S.; Yap, G. P. A.; Budzelaar, P. H. M. Organo-
metallics 2002, 21, 3088–3090. (c) Enright, D.; Gambarotta, S.; Yap, G. P.
A.; Budzelaar, P. H. M. Angew. Chem., Int. Ed. 2002, 41, 3873–3876. (d)
Bouwkamp, M. W.; Lobkovsky, E.; Chirik, P. J. Inorg. Chem. 2006, 45, 2–4.
(32) Harrison, D. F.; Weissberger, E.; Taube, H. Science 1968, 159,
320–322.
(27) (a) McKinney, R. J.; Colton, M. C. Organometallics 1986, 5,
1080–1085. (b) Al-Ohaly, A.; Head, R. A.; Nixon, J. F. J. Organomet. Chem.
1981, 205, 99–110. (c) Pertici, P.; Vitulli, G.; Paci, M.; Porri, L. J. Chem.
Soc., Dalton Trans. 1980, 1961–1964. (d) Jones, W. D.; Libertini, E. Inorg.
Chem. 1986, 25, 1794–1800. (e) Fl€ugel, R.; Windm€uller, B.; Gevert, O.;
Werner, H. Chem. Ber. 1996, 129, 1007–1013. (f) Ogasawara, M.;
Macgregor, S. A.; Streib, W. E.; Folting, K.; Eisenstein, O.; Caulton, K. G.
J. Am. Chem. Soc. 1995, 117, 8869–8870. (g) Huang, D.; Streib, W. E.;
Eisenstein, O.; Caulton, K. G. Organometallics 2000, 19, 1967–1972.
(28) (a) Kaupp, M.; von Schnering, H. G. Angew. Chem., Int. Ed.
Engl. 1995, 34, 986. (b) Snyder, J. P. Angew. Chem., Int. Ed. Engl. 1995, 34,
986–987. (c) Ward, M. D.; McCleverty, J. A. J. Chem. Soc., Dalton Trans.
2002, 275–288. (d) Pierpont, C. G. Coord. Chem. Rev. 2001, 216, 99–125.

598 Organometallics, Vol. 29, No. 3, 2010
Gallagher et al.
and structurally characterized by Gray and co-workers in
1969.18e Although many similar complexes have been sub-
sequently isolated, all involve unambiguously d6, Ru(II)
centers and typically with octahedral geometries.6,9e,18 The
16e- Ru(0) dinitrogen complex 3 exhibits a square-planar
geometry in the solid state, mostly likely due to steric factors
that disfavor coordination of an additional terminal nitro-
gen ligand in the apical position, although reversible binding
of additional nitrogen ligands in solution has not been
excluded. The bulky aryl rings on each half of the complex
effectively encapsulate the metal centers and bridging nitro-
gen ligand, which also contributes to the stability of this
reduced species. Until quite recently, non-macrocyclic, four-
coordinate ruthenium complexes were relatively rare. Tetra-
hedral geometry is most typical,33 but Wilkinson reported a
square-planar bis(imido) ruthenium complex in 1992, and
Caulton, Werner, and others have isolated several square-
planar ruthenium complexes, including a 14e- (amido)-
(bisphosphine) complex that exhibits a high-spin ground
state.27e,g,34 The paucity of square-planar Ru(0) complexes
is somewhat surprising, as isoelectronic 16e-, d8 Rh(I) and
Pt(II) species are extremely common. However, such species
have been proposed as reaction intermediates or generated as
a short-lived transient species through flash photolysis and
matrix isolation.14,35
inorganic and organometallic chemistry.36,37,28,38 The ambi-
guity of formal oxidation states;particularly in complexes
containing ligands capable of accepting metal electron den-
sity via π-back-bonding;has long been recognized. Re-
cently, increased attention has been devoted to
understanding the “non-innocent” character of some ligand
classes in somewhat different terms. In certain cases,
[N3Ar]M complexes may be best described by forms repre-
senting full transfer of one or more electrons to relati-
vely remote orbitals on the ligand. Several groups have
shown that surprisingly complex electronic structures and
magnetic behavior can result in these formally low-valent
complexes.22,23,38a,39
In extreme cases, such as in some metal catecholate/semi-
quinonate complexes, equilibria can be observed between
separate valence tautomers in which an unpaired electron
can be localized on either the metal or the ligand.28c,d,36,40
For example, Kaim and co-workers investigated ruthenium
complexes of “non-innocent” ligands12c such as o-iminoqui-
none or o-iminothioquinone and documented “redox iso-
merism” (valence tautomerism)28d that effectively allows
variation of formal oxidation states from Ru(II) to Ru(IV).36
On the basis of structural and EPR studies it was concluded
that most members of that particular series are best described
as Ru(III).
It is worth noting the relative advantages and disadvan-
tages of 2 and 3 as starting materials for subsequent reac-
tions. The arene complexes are generally easier to synthesize
and purify, are more soluble in nonpolar solvents than the
dinitrogen complexes, and are stable in arene solvents
(although one must be mindful of the facile arene exchange).
However, an inherent disadvantage of the arene complexes is
that most reactions liberate the arene ligand (toluene,
benzene), and this byproduct can complicate the workup
of reactions that remain in equilibrium with the starting
complex. On the other hand, the pyrophoric dinitrogen
complex 3 is prepared from arene complex 2, which intro-
duces an additional synthetic step and precludes use of arene
solvents in subsequent reactions. Compound 3 is also less
stable than the arene complexes, decomposing above 60 °C in
solution, whereas 2 is stable for hours at 95 °C. However, 3 is
more reactive than 2, exhibits shorter reaction times, and
produces only gaseous nitrogen as a byproduct.
Similarly, the Ru(0)/Ru(II) continuum has been probed by
Caulton, Eisenstein, and co-workers as part of an extensive
series of papers on low-coordinate and low-valent ruthenium
complexes. These ruthenium phosphine nitrosyl complexes
can accommodate Ru(0) and Ru(II) oxidation states
through tautomerization of the NO ligand.27e,g,37
More commonly, π-back-bonding is not sufficient to
warrant a change in oxidation state formalism, but rather
involves the more subtle issue of relative “electron richness”
at the metal center. For example, the metal centers in both
W(CO)6 and W(PMe3)6 are both assigned an unambiguous
formal oxidation state of W(0), yet there would be little
disagreement that the carbonyl ligands delocalize substantial
electron density away from the metal center in the former
and that the tungsten in the latter is more “electron rich”.
More importantly, the latter behaves in chemical reactions as
a highly reduced species, whereas the reactivity of W(CO)6 is
more temperate.
Given the potential “non-innocent” character of bis-
(imino)pyridyl ligands, the question arises as to the best
electronic description of complexes 2 and 3. The bis-
(imino)pyridine in four-coordinate 3 would be classified
using common conventions as a neutral donor ligand, and
the metal as formally zerovalent, although even novice
practitioners would note the possibility for substantial elec-
tron delocalization into the ligand to reduce electron density
at Ru. Taking into account ligand non-innocence by expli-
citly specifying metal-to-ligand electron transfer results in
Ligand Non-innocence in [N3]Ru Complexes and Formal
Oxidation States. Electronic structure and the nature of the
bonding in metal/ligand complexes are fundamental issues in
ꢀ
(33) (a) Sanchez-Delgado, R. A.; Navarro, M.; Lazardi, K.; Atencio,
R.; Capparelli, M.; Vargas, F.; Urbina, J. A.; Bouillez, A.; Noels, A. F.;
Masi, D. Inorg. Chim. Acta 1998, 275-276, 528–540. (b) Hay-Motherwell,
R. S.; Wilkinson, G.; Hussain-Bates, B.; Hursthouse, M. B. J. Chem. Soc.,
Dalton Trans. 1992, 3477–3482. (c) Savage, P. D.; Wilkinson, G.; Motevalli,
M.; Hursthouse, M. B. J. Chem. Soc., Dalton Trans. 1988, 669–673.
(d) Gaughan, A. P., Jr.; Corden, B. J.; Eisenberg, R.; Ibers, J. A. Inorg.
Chem. 1974, 13, 786–791.
(34) (a) Danopoulos, A. A.; Wilkinson, G.; Hussain-Bates, B.;
Hursthouse, M. B. Polyhedron 1992, 11, 2961–2964. (b) Watson, L. A.;
Ozerov, O. V.; Pink, M.; Caulton, K. G. J. Am. Chem. Soc. 2003, 125, 8426–
8427. (c) Yamamoto, Y.; Satoh, R.; Tanase, T. J. Chem. Soc., Dalton Trans.
1995, 307–311.
(35) (a) Cole-Hamilton, D. J.; Wilkinson, G. J. Am. Chem. Soc. 1978,
883–884. (b) Chatt, J.; Davidson, J. M. J. Chem. Soc. 1965, 843–855. (c)
Mawby, R. J.; Perutz, R. N.; Whittlesey, M. K. Organometallics 1995, 14,
3268–3274. (d) Hall, C.; Jones, W. D.; Mawby, R. J.; Osman, R.; Perutz,
R. N.; Whittlesey, M. K. J. Am. Chem. Soc. 1992, 114, 7425–7435. (e)
Bogdan, P. L.; Weitz, E. J. Am. Chem. Soc. 1989, 111, 3163–3167.
(36) Patra, S.; Sarkar, B.; Mobin, S. M.; Kaim, W.; Lahiri, G. K.
Inorg. Chem. 2003, 42, 6469–6473.
(37) Ogasawara, M.; Huang, D.; Streib, W. E.; Huffman, J. C.;
Gallego-Planas, N.; Maseras, F.; Eisenstein, O.; Caulton, K. G.
J. Am. Chem. Soc. 1997, 119, 8642–8651.
(38) (a) Vidyaratne, I.; Gambarotta, S.; Korobkov, I.; Budzelaar,
P. H. M. Inorg. Chem. 2005, 44, 1187–1189. (b) Vidyaratne, I.; Scott, J.;
Gambarotta, S.; Korobkov, I.; Budzelaar, P. H. M. Inorg. Chem. 2007, 46,
7040–7049. (c) Knijnenburg, Q.; Gambarotta, S.; Budzelaar, P. H. M. Dalton
Trans. 2006, 5442–5448.
€
(39) Sugiyama, H.; Korobkov, I.; Gambarotta, S.; Moller, A.;
Budzelaar, P. H. M. Inorg. Chem. 2004, 43, 5771–5779.
(40) Da Silva, R. S.; Gorelsky, S. I.; Dodsworth, E. S.; Tfouni, E.;
Lever, A. B. P. J. Chem. Soc., Dalton Trans. 2000, 4078–4088.

Article
Organometallics, Vol. 29, No. 3, 2010 599
Scheme 2. Diamagnetic Electron Transfer Valence Tautomers
Involving Low-Spin Ruthenium
[N3iPr]Fe complexes,22 the authors comment that both cal-
culations and Mossbauer spectroscopy are consistent with
the neutral ligand/Fe(0) formulation for [N3iPr]Fe(CO)2.
Again, the IR data for both the iron and ruthenium com-
plexes are easily accommodated by the traditional M(0)
oxidation state, combined with moderate competition for
π-electron density between carbonyl and [N3] ligands.
A similar picture emerges for the monocarbonyl complex
7a, which exhibits a CO stretch (1865 cm-1) somewhat
higher in energy than that of (dmpe)2Ru(CO) (1845 cm-1
,
Ru(0)),27d,35d but significantly lower than Ru(II) mono-
carbonyl complexes devoid of other π-acceptors, such as
(PiPr3)2RuX2(CO) (X=Cl, H, I; 1910-1920 cm-1).43 The IR
spectrum of the formally Ru(II) complex [N3xyl]RuCl2(CO)
is also consistent with some delocalization of electron density
from Ru to the [N3] ligand, as the CO stretch (1963 cm-1) is
about 50 cm-1 higher energy than in (PiPr3)2RuX2(CO).
In the absence of isotopic labeling studies and Raman
data, the assignment of a weak band at 1856 cm-1 in the IR
spectrum of dinitrogen complex 3a as the N-N stretch must
be viewed as tentative, but does suggest a rather electron-rich
metal center. Although much lower than reported for
μ-dinitrogen complexes containing unambiguously Ru(II)
centers (2000-2150 cm-1),5,9e,18 this value for the N-N
stretch is not extraordinary compared with recent reports
of low-valent Fe-Fe systems such as (LFe)2(μ-N2) (L =
bulky β-diketiminate ligand),44 which exhibits νNN at 1778
cm-1 (Raman), an extremely long N-N triple-bond distance
(1.182(5) A), and short Fe-N2 distances (1.77-1.78(5) A).
Thus, the low N-N stretch (1856 cm-1) in 3a and the
elongated N-N bond length (1.161(5) A) in 3c are internally
consistent and indicate greater reduction of the N-N bond
order;and hence greater “low-valent” character;than in
typical Ru(II) nitrogen complexes. Somewhat in contrast,
Gambarotta and co-workers have made a strong case for the
dianionic formalism for the bis(imino)pyridine ligand in the
potential formalisms ranging from Ru(0) with a neutral
ligand, to Ru(I) with a radical anion ligand, to Ru(II) with
a dianionic ligand (Scheme 2). The matter of magnetic
properties is further complicated by the possibility of low-
or high-spin states on the metal, the small singlet/triplet
separation of the [N3]2- dianion,22 and the potential for spin
coupling between unpaired electrons on the metal and
ligand. For example, some of Chirik’s [N3]Fe(L) complexes
exhibit overall S = 1 and are conventionally depicted as
neutral [N3] ligand/Fe(0) (d8), but are more accurately
described as triplet [N3]2-/high-spin Fe(II) (d6), with two
unpaired electrons on the ligand and four unpaired electrons
of opposite spin on the iron!22 In the present case, ruthenium
nitrogen complex 3 is diamagnetic, and invoking the ana-
logous triplet ligand dianion coupled to an S=1, d6 second-
row metal center seems unnecessarily convoluted in the
absence of any experimental evidence. Thus only formalisms
involving low-spin ruthenium are considered in Scheme 2.
The question of whether an alternative representation
(e.g., [N3]2-/Ru(II)) is more “accurate” is nontrivial and
can require nonroutine computational studies such as
“broken symmetry” methods. However, the value of formal
oxidation states and d-electron counts lies in their utility, and
the question of which formal representation is most “useful”
or “descriptive” can be partially addressed by examining the
spectroscopic data and structural parameters of the “Ru(0)”
and “Ru(II)” complexes.
The dicarbonyl complex 6a exhibits two CO stretches at
1925 and 1978 cm-1. Ru(0) dicarbonyl complexes containing
donor co-ligands that are not significant π-acceptors such as
Ru(CO)2L3 (L = PMePh2, PEt3, and [P(2-furyl)3]) display
two CO stretching bands in the ranges 1827-1844 and
1883-1900 cm-1, about 90 cm-1 lower energy than in
6a.41 On the other hand, CO vibrations in Ru(II) complexes
such as L2RuCl2(CO)2 (L=PMe3 or PPh3) (1985-1996 and
2049-2058 cm-1) are about 60-70 cm-1 higher than in 6a.42
Thus this traditional measure suggests the electron density at
the metal is intermediate between that in less ambiguous
examples of the Ru(0) and Ru(II) formalisms.
case of
a
vanadium bridging dinitrogen complex,
{[N3iPr]V(THF)}2(μ-N2).38a The bridging dinitrogen ligand
in this early metal complex is apparently highly reduced
(D(NN)=1.259(6) A) and can be viewed as a dianion. On the
basis of bond lengths within the [N3] ligand and DFT
calculations, these workers interpret the paramagnetism of
the complex as resulting from two noninteracting high-spin
V(III) centers, each with a [N3] ligand dianion and bridged by
a N22- ligand. In this context, the ruthenium centers in 3b are
clearly less electron releasing toward the dinitrogen than the
[N3]V centers.
The degree of electron delocalization;whether best des-
cribed as π-back-bonding or complete electron transfer to
the [N3] ligand;is also reflected in the intraligand bond
distances. The CdNim and Cpyr-Cim distances in bis-
(imino)pyridine ligands have been described as particularly
diagnostic of the degree of electron transfer from the central
metal,38c with longer CdN and shorter Cpyr-Cim bonds
associated with the [N3]- and [N3]2- formulations. It must
be noted, however, that traditional “π-back-bonding” pre-
dicts exactly the same trends. Gambarotta has suggested
rough values corresponding to reduction of [N3] ligands by
The carbonyl stretching energies in 6a are quite similar to
those in an analogous Fe(0) dicarbonyl complex,
[N3iPr]Fe(CO)2 (1914 and 1974 cm-1), recently reported by
Chirik and co-workers.4a Interestingly, although broken
symmetry DFT calculations suggest the [N3]2-/Fe(II) form-
alism is most appropriate for a variety of other reduced
(41) Ogasawara, M.; Maseras, F.; Gallego-Planas, N.; Kawamura,
K.; Ito, K.; Toyota, K.; Streib, W. E.; Komiya, S.; Eisenstein, O.;
Caulton, K. G. Organometallics 1997, 16, 1979–1993.
(42) (a) Chung, M.; Ferguson, G.; Robertson, V.; Schlaf, M. Can. J.
Chem. 2001, 79, 949–957. (b) Olmstead, M. M.; Maisonnat, A.; Farr, J. P.;
Balch, A. L. Inorg. Chem. 1981, 20, 4060–4064.
(43) (a) Werner, H.; Tena, M. A.; Mahr, N.; Peters, K.; von Schnering,
H. Chem. Ber. 1995, 128, 41–47. (b) Esteruelas, M. A.; Werner, H.
J. Organomet. Chem. 1986, 303, 221–231.
(44) Smith, J. M.; Lachicotte, R. J.; Pittard, K. A.; Cundari, T. R.;
Lukat-Rodgers, G.; Rodgers, K. R.; Holland, P. L. J. Am. Chem. Soc.
2001, 123, 9222–9223.

600 Organometallics, Vol. 29, No. 3, 2010
Gallagher et al.
a
˚
Table 1. Selected [N3] Ligand Bond Distances (A) of [N3]RuCl2(PMe3), [N3]RuCl2(CO), 2a, 3c, 5a, 6a, and 7a
xyl
bond
[N3
]
[N3]RuCl2(PMe3)
[N3]RuCl2(CO)
2a
3c
5a
6a
7a
Ru-Nim
CdN
C-Cpyr
Ru-Npyr
2.119(3);
1.313(4);
1.466(5);
1.971(3);
2.119(2);
1.301(3);
1.478(4);
2.006(2);
1.985(2);
2.014(5);
1.359(7);
1.432(8);
1.942(7);
2.065(2);
1.352(3);
1.420(3);
1.971(2);
2.067(2);
1.334(3);
1.420(3);
1.947(2);
2.091(3);
1.348(5);
1.418(5);
1.964(3);
1.268(2)
1.495(2)
1.344(3), 1.282(3);
1.397(3), 1.491(3);
2.032(2);
a Numbers in parentheses are estimated standard deviations in the least significant digits. Chemically equivalent distances are averaged.
hand, the effect of charge delocalization will be manifest on
bond distances in only one imine, rather than spread over
two, and hence more may appear more dramatic.
Although all of the formally Ru(0) complexes exhibit
longer CdN and shorter Cpyr-Cim distances than the for-
mally Ru(II) species, no consistent trends are apparent,
despite the large variation in π-acidity of co-ligands (N2,
(PMe3)2, (PMe3)(CO), and (CO)2). The nitrogen complex 3c
does exhibit the longest CdN distances, and the average
of 1.359(7) A is nearly in the range suggested as for
[N3]2-/Ru(II).38c However, it should also be noted that the
CdN and C-C distances in dicarbonyl 6a (1.335(3) and
1.422(4) A) are extremely similar to those in Chirik’s iron
analogue, [iPrN3]Fe(CO)2 (1.333(2) and 1.424(2) A.) Given
the similar ligand metrical parameters and nearly identical
CO stretching frequencies (vide supra), it seems reasonable
that the same formalism should be used to describe both the
iron and ruthenium dicarbonyl complexes. On the one hand,
the CdN and C-C distances fall between the values sug-
gested for 1e- and 2e- reduction of the [N3] ligands. On the
Figure 8. Comparison of the bond distances in [N3]Ru com-
plexes and the free ligand, [N3xyl].45 Chemically equivalent
distances are averaged.
one electron (D(CdN) ≈ 1.32 A, D(C-C) ≈ 1.44 A) and two
electrons (1.36, 1.40 A), although again it is not clear if one
can use metrical parameters to practically distinguish
between the electron transfer implied by noninnocent
valence tautomers and more classical metal-imine (or
diimine) back-bonding. Values for the ruthenium complexes
are summarized in Table 1 and in Figure 8. For comparison,
the structurally characterized free [N3xyl] ligand has also been
included.45 The coordinated and free imine groups of the
bidentate [N3] complex 2a are listed individually.
One primary trend emerges: the CdN bonds lengthen, and
the Cpyr-Cim bonds contract on going from the free ligand to
the Ru(II) complexes to the formally Ru(0) complexes,
consistent with increasing electronic delocalization into the
[N3] ligand via π-back-bonding and/or contribution from a
reduced-ligand tautomer (Scheme 2). The bidentate Ru(0)
complex 2 shows the greatest diversity of distances. There is
reasonable agreement between the distances in the uncoor-
dinated imine in 2 and the free ligand (D(CdN)=1.282(3)
versus 1.268(2) A; D(C-C) = 1.491(3) versus 1.495(2) A).
The coordinated imine in 2 exhibits the shortest Cpyr-Cim
distance of all the complexes (1.397(3) A), and correspond-
ingly, the coordinated CdN distance in 2 is fairly long
(1.344(3) A), although not as exceptional within the series
as is the C-C distance. Bidentate 2 is an interesting case;
and not particularly representative of the other reduced
[N3]Ru complexes;in that only the coordinated imine is in
conjugation with the pyridine, with the free imine oriented
out of the pyridine plane (torsional angle=55°). As such, the
conjugated system available to accommodate electron den-
sity from Ru is substantially smaller than in a tridentate [N3]
ligand, and non-innocent valence tautomers (i.e., [N3]- and
[N3]2-) might be expected to be less favorable. On the other
€
other hand, the Mossbauer data and broken symmetry DFT
calculations are consistent with the neutral [N3]/Fe(0) for-
mulation of [N3]Fe(CO)2.22 We suggest the analogous;and
“innocent”;view of complex 6a as formally Ru(0) remains
appropriate and descriptive.
Conclusions
A new class of low-valent ruthenium complexes containing
2,6-bis(imino)pyridyl ligands has been prepared, and these
labile complexes serve as versatile reagents for the synthesis
of a variety of Ru(0) and Ru(II) complexes. The [N3] ligand
provides tunable steric protection of the metal center, but
also can serve as a hemilabile ligand able to reversibly
dissociate an imine group. The 16e- complex 3c is the first
structurally characterized Ru(0) dinitrogen complex and
also exhibits a square-planar geometry that is unusual for
Ru(0).
Overall, it can be concluded that electron delocalization
into the bis(imino)pyridyl ligand reduces electron density at
the metal in the formally Ru(0) complexes, but probably not
to the extent implied by the valence tautomeric [N3]2-/Ru(II)
canonical form. Although the [N3]-/Ru(I) representation
may portray the electron distribution more accurately than
“Ru(0)”, the suggested odd electron count on both ligand
and metal;with antiferromagnetic coupling to yield net
diamagnetism;is cumbersome and provides little predictive
value for [N3]Ru complexes with strong-field co-ligands such
as CO and PMe3. Application of Occam’s Razor, or perhaps
more appropriately the “Duck Test”,46 suggests the Ru(0)
formalism is appropriate for these five-coordinate com-
pounds. We note in closing, however, that the metalloradical
(46) (a) “If it looks like a duck, swims like a duck, and quacks like a
duck, then it probably is a duck.” (b) Hecht-Nielsen, R. Neural
Networks 2005, 18, 111–115.
(45) Huang, Y.; Ma, X.; Zheng, S.; Chen, J.; Wei, C. Acta Crystal-
logr., Sect. E: Struct. Rep. Online 2006, 62, o3044–o3045.

Article
Organometallics, Vol. 29, No. 3, 2010 601
character implied by the [N3]-/Ru(I) formalism may prove
helpful in understanding the observation that the four-
coordinate dinitrogen complex 3 reacts with hydrogen to
yield paramagnetic monohydride species, as will be detailed
in a forthcoming report.47 Regardless of the exact distri-
bution of electron density between the metal and ligands, it is
clear that bis(imino)pyridyl ligands allow access to reduced
complexes that maintain the reactivity implied by the “Ru-
(0)” formalism. The electronic and geometrical flexibility of
these bis(imino)pyridine ruthenium complexes should prove
useful in the development of new catalysts and bond-forming
processes. Reactions of 2 and 3 with hydrogen, silanes,
alkynes, and other substrates, along with initial DFT calcu-
lations on this system, will be described in subsequent
reports.
Xyl-Me), 17.61 (s, 2C, Im-Me), 15.22 (d, JPC =24.7 Hz, 3C,
1
PMe3). 31P{1H} NMR (toluene-d8): δ -5.77 (s, 1P, PMe3). Anal.
Calcd for C28H36N3PCl2Ru: C, 54.45; H, 5.88; N, 6.80. Found:
C, 54.56; H, 6.18; N, 6.61.
Synthesis of [N3xyl]RuCl2(CO). A CH2Cl2 solution (40 mL) of
[N3xyl]RuCl2(C2H4) (1a) (60 mg, 0.105 mmol) was stirred under
CO (1 atm) for 45 min at room temperature, during which time
the color of the reaction mixture changed from purple to red.
Volatiles were removed in vacuo, and the residue was washed
with C6H6 and filtered in air, yielding 50 mg of air-stable
1
burgundy solid (83% yield). H NMR (chloroform-d): δ 8.09
3
3
(t, JHH = 8.1 Hz, 1H, Py-Hp), 7.94 (d, JHH = 8.0 Hz, 2H,
Py-Hm), 7.04 (m, 6H, Xyl-Hm,p), 2.46 (s, 6H, Im-Me), 2.29 (s,
12H, Xyl-Me). 13C{1H} NMR (chloroform-d): δ 202.51 (s, 1C,
CO), 174.52 (s, 2C, CdN), 157.49 (s, 2C, Py-Co), 148.90 (s, 2C,
Xyl-C-N), 138.37, 130.39, 129.24, 127.00, and 124.72 (s, 13C,
aryl C), 20.91 (s, 4C, Xyl-Me), 18.98 (s, 2C, Im-Me). IR (Nujol):
ν(CO)=1963 cm-1. HRMS (ES): calcd 592.047 (M(102Ru) þ
Na)þ, found 592.0491; calcd 594.047 (M(104Ru) þ Na)þ, found
594.0478.
Experimental Section
All manipulations were performed in Schlenk-type glassware
on a dual-manifold Schlenk line or a nitrogen-filled Vacuum
Atmospheres glovebox.48 1H NMR spectra were obtained at
300, 360, and 500 MHz on Bruker DMX-300, AM-360, and
AMX-500 FT NMR spectrometers, respectively. 2H NMR
spectra were obtained at 76.8 MHz on a Bruker AM-500
spectrometer. 31P{1H} NMR spectra were recorded with broad-
band 1H decoupling at 121.5 MHz on a Bruker DMX-300
spectrometer. 13C{1H} NMR spectra were obtained at 125.8
MHz on an AMX-500 NMR spectrometer. All NMR spectra
were recorded at 303 K unless stated otherwise. Chemical shifts
are reported relative to tetramethylsilane for 1H and 13C spectra
and external 85% H3PO4 for 31P resonances. The temperature of
the NMR probe was calibrated against methanol (estimated
error 0.3 K). Infrared spectra were recorded on a Perkin-Elmer
model 1430 spectrometer. HRMS was acquired on an AutoSpec
(Micromass) with chemical ionization (CH4). Elemental
analyses were performed by Robertson Laboratory, Inc.
(Madison, NJ).
Synthesis of [K2-N3xyl]Ru(η6-MeC6H5), 2a. A toluene solution
(35 mL) of [N3xyl]RuCl2(C2H4) (1a) (1.00 g, 0.00176 mol) and
Et3SiH (1.7 mL, 1.22 g, 0.0105 mol) was stirred under nitrogen
for 16 h at room temperature. The reaction mixture was then
reduced in volume to approximately 5 mL in vacuo. The product
was recrystallized from 2:1 pentane/toluene at -78 °C, yielding
0.770 g of purple 2a (78% yield). 1H NMR (toluene-d8): δ 6.98
and 6.63 (m, 9H, aryl H), 5.10 (t, 3JHH=5.4 Hz, 1H, Tol-Hp),
4.56 (t, 3JHH=5.5 Hz, 2H, Tol-Hm), 4.29 (d, 3JHH=5.5 Hz, 2H,
Tol-Ho), 2.39 (s, 3H, Im-Me), 2.34 (s, 6H, Xyl-Me), 1.98 (s, 6H,
Xyl-Me), 1.66 (s, 3H, Im-Me), 1.60 (s, 3H, Tol-Me). 13C{1H}
NMR (toluene-d8): δ 170.05 (s, 1C, C=N), 165.50 (s, 1C, CdN),
162.85 (s, 1C, Py-Co), 155.02 (s, 1C, Py-Co), 148.84 (s, 1C, Xyl-
C-N), 146.02 (s, 1C, Xyl-C-N), 129.32, 129.30, 128.95, 128.51,
128.30, 128.13, 125.66, 125.25, 125.04, 123.51, 118.65, 116.98 (s,
19C, aryl C), 21.39 (s, 1C, Tol-Me), 19.66 (s, 1C, Im-Me), 18.83
(s, 2C, Xyl-Me), 18.25 (s, 2C, Xyl-Me), 14.24 (s, 1C, Im-Me).
Anal. Calcd for C32H35N3Ru: C, 68.30; H, 6.27; N, 7.46. Found:
C, 68.40; H, 6.15; N, 7.15.
Hydrocarbon solvents were dried over Na/K alloy-benzophe-
none. Benzene-d6, toluene-d8, cyclohexane-d12, and tetrahydro-
furan-d8 were dried over Na/K alloy. Chloroform-d was dried
over molecular sieves. CO and C2H4 (Airco) were used as
received. [Ru(p-cymene)Cl2]2,49 [N3xyl],2 [tBu-N3mes],50 [N3xyl]-
[K2-N3mes]Ru(η6-MeC6H5), 2b. A toluene solution (30 mL) of
[N3]RuCl2(C2H4) (1b), (1.08 g, 0.0018 mol), and Et3SiH (1.8 mL,
1.22 g, 0.0111 mol) was stirred under nitrogen for 16 h at room
temperature. The reaction mixture was then reduced in volume
to approximately 5 mL in vacuo. The product was recrystallized
from 2:1 pentane/toluene at -78 °C using a swivel frit under
51
RuCl2(C2H4),11 [tBu-N3mes]RuCl2(C2H4), and PMe3 were
synthesized according to the literature procedures. HSiMe3
was prepared by the reaction of Me3SiCl and LiAlH4 in nBu2O
and purified by trap-to-trap vacuum fractionation. Triethylsi-
lane (Aldrich) was dried over Na prior to use.
1
inert conditions, yielding 0.700 g of purple 2b (66% yield). H
NMR (toluene-d8): δ 7.06, 6.88, and 6.65 (m, 7H, aryl H), 5.15 (t,
3JHH=5.4 Hz, 1H, Tol-Hp), 4.59 (t, 3JHH=5.4 Hz, 2H, Tol-Hm),
4.35 (d, 3JHH=5.4 Hz, 2H, Tol-Ho), 2.43 (s, 3H, Im-Me), 2.34 (s,
6H, Mes- Meo), 2.29 and 2.26 (s, each 3H, Mes-Mep), 1.99 (s,
6H, Mes-Meo), 1.70 (s, 3H, Im-Me), 1.66 (s, 3H, Tol-Me).
[K2-tBu-N3mes]Ru(η6-MeC6H5), 2c. A toluene solution (7 mL)
of [tBu-N3mes]RuCl2(C2H4) (1c) (200 mg, 0.306 mmol) and
Et3SiH (0.3 mL, 0.216 g, 1.846 mmol) was stirred under nitrogen
for 20 h at room temperature. The reaction mixture was then
reduced in volume to approximately 2 mL in vacuo. The product
was recrystallized from 2:1 pentane/toluene at -78 °C, yielding
166.5 mg of purple 2c (84% yield). 1H NMR (toluene-d8): δ 7.26
(s, 2H, Py-Hm), 6.89 (s, 4H, Mes-Hm), 5.18 (t, 3JHH=5.2 Hz, 1H,
Tol-Hp), 4.61 (t, 3JHH=5.4 Hz, 2H, Tol-Hm), 4.37 (d, 3JHH=5.4
Hz, 2H, Tol-Ho), 2.45 (s, 3H, Im-Me), 2.39 (s, 6H, Mes-Meo),
2.30 (s, 3H, Mes-Mep), 2.27 (s, 3H, Mes-Mep), 2.16 (s, 3H, Im-
Me), 2.05 (s, 6H, Mes-Meo), 1.82 (s, 3H, Im-Me), 1.24 (s, 9H,
tBu).
Synthesis of [N3xyl]RuCl2(PMe3). PMe3 (328 mg, 4.31 mmol)
was vacuum transferred into a solution of toluene (5 mL) and
[N3xyl]RuCl2(C2H4) (1a) (114 mg, 0.200 mmol). The solution
was stirred for two days at room temperature. Volatiles were
removed in vacuo, and the residue was recrystallized from 2:1
pentane/toluene at 0 °C, yielding 107 mg of air-stable purple
[N3xyl]RuCl2(PMe3) (87% yield). 1H NMR (benzene-d6): δ 7.03
(d, 3JHH=8.0 Hz, 2H, Py-Hm), 6.88 (m, 7H, Xyl-Hm,p, Py-Hp),
2.40 (s, 12H, Xyl-Me), 1.93 (s, 6H, Im-Me), 1.08 (d, 2JPH=8.5
Hz, 9H, PMe3). 13C{1H} NMR (benzene-d6): δ 173.80 (s, 2C,
CdN), 159.91 (s, 2C, Py-Co), 151.68 (s, 2C, Xyl-C-N), 132.15,
129.58, 129.40, 126.40, and 123.15 (s, 13C, aryl C), 21.73 (s, 4C,
(47) Wieder, N. L.; Gallagher, M.; Carroll, P. J.; Berry, D. H. To be
submitted.
Observation of [N3xyl]Ru(H)(Cl)(C2H4). A NMR tube was
loaded with a toluene-d8 solution (0.3 mL) of [N3xyl]RuCl2-
(C2H4) (1a) (10 mg, 0.0176 mmol) and Et3SiH (25 μL, 0.155
mmol) under nitrogen. The reaction was monitored by 1H NMR
spectroscopy. After 30 min at room temperature, the solution
contained a mixture of [N3xyl]Ru(H)(Cl)(C2H4) (ca. 68%) and
(48) Experimental Organometallic Chemistry; Wayda, A. L., Darensbourg,
M. Y., Eds.; American Chemical Society: Washington, D.C., 1987.
(49) Bennett, M. A.; Huang, T. N.; Matheson, T. W.; Smith, A. K.
Inorg. Synth. 1982, 21, 74–78.
(50) Nuckel, S.; Burger, P. Organometallics 2001, 20, 4345–4359.
(51) Luetkens, M. L.; Sattelberger, A. P.; Murray, H. H.; Basil, J. D.;
Fackler, J. P. Inorg. Synth. 1989, 26, 7–12.
€

602 Organometallics, Vol. 29, No. 3, 2010
Gallagher et al.
2a (ca. 32%). Et3SiCl, Et4Si, and CH3CH3 were also observed.
1H NMR (toluene-d8): δ 6.90 (m, 9H, aryl H), 4.12 (s, 4H, C2H4),
2.51 (s, 6H, Im-Me), 1.89 (s, 6H, Xyl-Me), 1.73 (s, 6H, Xyl-Me),
-21.91 (s, 1H, Ru-H).
Observation of [N3xyl]Ru(C2H4)2 4a. A NMR tube sample of
[κ2-N3xyl]Ru(η6-MeC6H5) (2a) (4 mg, 0.00711 mmol) in cyclo-
hexane-d12 (0.3 mL) was flame-sealed under 1 atm of C2H4
at -196 °C. Upon warming to room temperature, the reac-
tion mixture changed from purple to burgundy in 5 min. The
resulting 1H NMR spectrum showed complete conversion of 2a
to 4a, with the concurrent release of toluene. The reaction
mixture decomposed during attempts to remove the volatiles
Synthesis of [N3xyl]Ru(H)(Cl)(PMe3). A toluene solution
(3 mL) of [N3xyl]RuCl2(PMe3) (1b) (85 mg, 0.138 mmol) and
Et3SiH (350 μL, 2.16 mmol) was added to a 100 mL thick-walled
pressure flask under nitrogen. The purple solution was heated to
150 °C for 3.5 h, during which time the color changed to red.
Volatiles were removed in vacuo, and the product was recrys-
tallized from 2:1 pentane/toluene at -78 °C, yielding 48 mg of
red powder (60% yield). 1H NMR (toluene-d8): δ 6.90 (m, 9H,
aryl H), 2.51 (s, 6H, Im-Me), 1.84 (s, 6H, Xyl-Me), 1.82 (s, 6H,
1
in vacuo or recrystallize the product. H NMR (cyclohexane-
d12): δ 7.99 (d, 3JHH=7.6 Hz, 2H, Py-Hm), 7.32 (t, 3JHH=7.4 Hz,
=
1H, Py-Hp), 6.93 (d, 3JHH=7.5 Hz, 4H, Xyl-Hm), 6.86 (t, 3JHH
7.5 Hz, 2H, Xyl-Hp), 2.28 (s, 6H, Im-Me), 1.76 (s, 12H, Xyl-Me),
1.63 (s, 8H, C2H4). 13C{1H} NMR (cyclohexane-d12): δ 155.82
(s, 2C, CdN), 150.38 (s, 2C, Py-Co), 146.90 (s, 2C, Xyl-C-N),
130.25 (s, 1C, Py-Cp), 128.74 (s, 2C, Xyl-Cp), 125.20 and 123.04
(s, 8C, Xyl-Co,m), 117.28 (s, 2C, Py-Cm), 59.00 (s, 4C, C2H4),
18.94 (s, 4C, Xyl-Me), 16.31 (s, 2C, Im-Me).
Xyl-Me), 0.83 (d, 2JPH=7.4 Hz, 9H, PMe3), -19.17 (d, 2JPH
=
42.5 Hz, 1H, RuH). 13C{1H} NMR (toluene-d8): δ 165.77 (s, 2C,
C dN), 157.60 (s, 2C, Py-Co), 153.10 (s, 2C, Xyl-C-N), 129.63,
129.12, 128.64, 125.93, 125.81, 119.30, and 119.19 (s, 13C, aryl
C), 18.08 (d, 1JPC=23.8 Hz, 3C, PMe3), 20.49 and 18.62 (s, 4C,
Xyl-Me), 16.33 (s, 2C, Im-C). 31P{1H} NMR (toluene-d8): -4.25
(s, 1P, PMe3).
Synthesis of [N3xyl]Ru(PMe3)2, 5a. PMe3 (73.20 mg, 0.962
mmol) was added by vacuum transfer to a toluene solution
(5 mL) of [κ2-N3xyl]Ru(|6-MeC6H5) (2a) (60 mg, 0.107 mmol) at
-196 °C. Upon warming to room temperature, a color change
from purple to red was observed within seconds. The reaction
mixture was stirred for 30 min, and volatiles were removed
in vacuo. The red product was recrystallized from pentane
at -78 °C, yielding 57 mg of 5a (86%). 1H NMR (toluene-d8):
δ 8.08 (d, 3JHH=7.6 Hz, 2H, Py-Hm), 7.25 (t, 3JHH=7.6 Hz, 1H,
Py-Hp), 6.94 (m, 6H, Xyl-Hm,p), 2.28 (s, 6H, Im-Me), 1.72 (br s,
Synthesis of {[N3xyl]Ru}2(μ-N2), 3a. [κ2-N3xyl]Ru(η6-MeC6-
H5) (2a) (215 mg, 0.382 mmol) was loaded into a 100 mL round-
bottom flask under N2 (1 atm) in cyclohexane (8 mL). The
purple slurry was stirred at room temperature for 2 days and
periodically exposed to fresh N2. The color gradually became
dark blue. The product was isolated by filtration under N2 and
residual 2a removed by washing with cyclohexane until the
filtrate became turquoise in color. The resulting turquoise solid
was dried in vacuo, yielding 165 mg of 3a (89% yield). Small
solid samples of 3a spontaneously enflame upon exposure to air.
2
12H, Xyl-Me), 0.52 (d, JPH = 7.5 Hz, 18H, PMe3). 13C{1H}
NMR (benzene-d6): δ 157.42 (s, 2C, CdN), 144.91 (s, 2C, Py-
Co), 143.90 (s, 2C, Xyl-C-N), 131.44 (s, 1C, Py-Cp), 129.43 (s,
2C, Xyl-Cp), 124.50 and 118.18 (s, 8C, Xyl-Co,m), 114.29 (s, 2C,
3
1H NMR (cyclohexane-d12): δ 8.59 (t, JHH=7.4 Hz, 2H, Py-
1
Hp), 8.12 (d, 3JHH=7.4 Hz, 4H, Py-Hm), 6.79 (d, 3JHH=7.4 Hz,
8H, Xyl-Hm), 6.57 (t, 3JHH=7.4 Hz, 4H, Xyl-Hp), 1.81 (s, 24H,
Xyl-Me), -0.08 (s, 12H, Im-Me). 13C{1H} NMR (tetrahydro-
furan-d8): δ 158.00 (s, 4C, CdN), 157.36 (s, 4C, Py-Co), 156.16
(s, 4C, Xyl-C-N), 130.16, 128.61, 124.94, 124.59, and 114.41 (s,
26C, aryl C), 20.84 (s, 8C, Xyl-Me), 19.86 (s, 4C, Im-Me). IR
(Nujol): 1856 cm-1 (w), tentatively assigned as ν(NN).
Py-Cm), 20.36 (s, 4C, Xyl-Me), 18.75 (d, JPC =21.9 Hz, 6C,
PMe3), 15.98 (s, 2C, Im-Me). 31P{1H} NMR (toluene-d8): δ 4.83
(br s, 2P, PMe3).
VT 1H NMR of [N3xyl]Ru(PMe3)2, 5a. 1H NMR (toluene-d8,
198 K): δ 8.07 (d, 3JHH=7.5 Hz, 2H, Py-Hm), 7.29 (t, 3JHH=7.6
Hz, 1H, Py-Hp), 6.93 (m, 6H, Xyl-Hm,p), 2.30 (s, 6H, Im-Me),
1.85 (s, 6H, Xyl-Me), 1.48 (s, 6H, Xyl-Me), 0.77 (d, 2JPH=4.2
Hz, 9H, PMe3), 0.24 (d, 2JPH=8.2 Hz, 9H, PMe3); (toluene-d8,
228 K): δ 8.07 (d, 3JHH=7.5 Hz, 2H, Py-Hm), 7.28 (t, 3JHH=7.6
Hz, 1H, Py-Hp), 6.95 (m, 6H, Xyl-Hm,p), 2.30 (s, 6H, Im-Me),
1.85 (s, 6H, Xyl-Me), 1.48 (s, 6H, Xyl-Me), 0.74 (d, 2JPH=4.2
Hz, 9H, PMe3), 0.29 (d, 2JPH=8.2 Hz, 9H, PMe3); (toluene-d8,
237 K): δ 8.07 (d, 3JHH=7.2 Hz, 2H, Py-Hm), 7.27 (t, 3JHH=7.6
Hz, 1H, Py-Hp), 6.95 (m, 6H, Xyl-Hm,p), 2.29 (s, 6H, Im-Me),
1.85 (s, 6H, Xyl-Me), 1.48 (s, 6H, Xyl-Me), 0.54 (br, 18H,
{[N3mes]Ru}2(μ-N2), 3b. A round-bottomed flask was charged
with [κ2-N3mes]Ru(η6-MeC6H5) (2b) (100 mg, 0.169 mmol) and
decane (25 mL). The resultant slurry was stirred under 1 atm of
N2 until the color of the solution changed from purple to
turquoise. The volume of the solution was then reduced in vacuo
by ca. 50% in an attempt to reduce the amount of free toluene,
and the slurry was again stirred under 1 atm N2 until the solution
changed to turquoise again, indicating displacement of coordi-
nated toluene by N2. The process was repeated twice, or until the
color of the solution remained turquoise upon stirring under
vacuum. The slurry was then filtered to collect the solid, and the
solid was washed with hexanes to afford 22 mg of 3b (25% yield).
Note: the [N3mes] ligand imparts greater solubility to 3b, redu-
cing the yield of solid collected in this manner. Additional crops
of 3b contaminated with varying amounts of 2b, but suitable for
preparative reactions, can also be recovered. 1H NMR
3
PMe3); (toluene-d8, 263 K): δ 8.06 (d, JHH=7.5 Hz, 2H, Py-
Hm), 7.27 (t, 3JHH=7.6 Hz, 1H, Py-Hp), 6.95 (m, 6H, Xyl-Hm,p),
2.28 (s, 6H, Im-Me), 1.71 (br s, 12H, Xyl-Me), 0.50 (br, 18H,
2
PMe3). 31P{1H} NMR (toluene-d8, 198 K): δ 13.95 (d, JPP
=
2
12.2 Hz, 1P, PMe3), -3.16 (d, JPP = 13.3 Hz, 1P, PMe3);
(toluene-d8, 228 K): δ 13.42 (br s, 1P, PMe3), -3.40 (br s, 1P,
PMe3); (toluene-d8, 237 K): δ 13.42 (br s, 1P, PMe3), -3.35 (br s,
1P, PMe3); (toluene-d8, 263 K): δ 13 (vbr s, 1P, PMe3), -2 (vbr s,
1P, PMe3).
3
(cyclohexane-d12): δ 8.58 (t, JHH =7.9 Hz, 2H, Py-Hp), 8.10
(d, JHH=7.9 Hz, 4H, Py-Hm), 6.60 (s, 8H, Mes-Hm), 1.97 (s,
3
Reaction of [N3xyl]Ru(PMe3)2 with PMe3-d9. A NMR tube
12H, Mes-Mep), 1.75 (s, 24H, Mes-Meo), -0.05 (s, 12H, Im-
Me). UV-vis: λmax=638 nm, ε=11 900.
was loaded with a benzene-d6 solution (0.3 mL) of
[N3xyl]Ru(PMe3)2 (5a) (5 mg, 0.00804 mmol), and the solution
was degassed in vacuo. PMe3-d9 (0.0243 mmol) was added by
vacuum transfer, and the tube sealed with a torch. The at-
tempted reaction was monitored periodically by 1H and 2H
NMR for 4 weeks at room temperature. No exchange of PMe3-
d9 with coordinated PMe3 was observed.
{[tBu-N3mes]Ru}2(μ-N2), 3c. [κ2-tBu-N3mes]Ru(η6-MeC6H5)
(2c) (166.5 mg, 0.258 mmol) was loaded into a 100 mL round-
bottom flask under N2 (1 atm) in cyclohexane (5 mL). The
purple slurry was stirred at room temperature for 2 days and
periodically exposed to fresh N2. The color gradually became
dark blue. The product was isolated by filtration under N2 and
residual 2c removed by washing with cyclohexane until the
filtrate became turquoise in color. The resulting turquoise solid
was dried in vacuo, yielding 87 mg of 3c (57% yield) 1H NMR
(cyclohexane-d12): δ 8.25 (s, 4H, Py-Hm), 6.59 (s, 8H, Mes-Hm),
1.97 (s, 12H, Mes-Mep), 1.75 (s, 24H, Mes-Meo), 1.29 (s, 18H,
tBu), -0.09 (s, 12H, Im-Me).
Synthesis of [N3xyl]Ru(CO)2, 6a. A toluene solution (10 mL) of
[κ2-N3xyl]Ru(η6-MeC6H5) (2a) (57 mg, 0.101 mmol) was placed
in a swivel frit and stirred under CO (1 atm) for 20 min at room
temperature, at which time the color changed from purple to
green. The solvent was removed in vacuo, and the product was
recrystallized from 3:1 pentane/toluene at -78 °C, yielding
1
42 mg of green 6a (78% yield). H NMR (benzene-d6): δ 7.69

Article
Organometallics, Vol. 29, No. 3, 2010 603
(d, 3JHH=7.7 Hz, 2H, Py-Hm), 6.98 (d, 3JHH=7.5 Hz, 4H, Xyl-
Hm), 6.94 (m, 3H Py-Hp, Xyl-Hp), 2.09 (s, 6H, Im-Me), 1.99 (s,
12H, Xyl-Me). 13C{1H} NMR (benzene-d6): δ 200.93 (s, 2C,
CO), 153.89 (s, 2C, CdN), 153.61 (s, 2C, Py-Co), 141.87 (s, 2C,
Xyl-C-N), 129.74 (s, 1C, Py-Cp), 129.09 (s, 2C, Xyl-Cp), 126.17
and 124.19 (s, 8C, Xyl-Co,m), 113.34 (s, 2C, Py-Cm), 18.57 (s, 4C,
Xyl-Me), 15.02 (s, 2C, Im-Me). IR (Nujol): ν(CO) = 1925,
centric structure resulting from a 1:1 disorder of 1- and 4-
positions on the ring (Cl3 and H27).
In the case of nitrogen complex 3c, a significant solvent-
accessible void was observed in the unit cell, although relatively
little electron density was observed in difference Fourier maps.
Analysis using SQUEEZE55 indicated the void is accommodat-
ing atoms containing only ca. 12 electrons. The small void size
and low electron density are inconsistent with cocrystallized
cyclohexane solvent, but the possibility of cocrystallized dini-
trogen gas (14 electrons) was considered. Inspection of the
packing diagram indicates that the void forms an infinite
channel parallel to the c-axis. Although refinement including a
full occupancy N2 molecule did lead to a decrease in R1 from
5.74% to ca. 4.9%, the atomic positions and thermal parameters
of the “N2” were not well-behaved and were omitted from the
final model. The residual electron density was accounted for in
the final structure factor calculation using the SQUEEZE
program. Nonbonded, cocrystallized dinitrogen is not common,
but has been observed in several other instances.56 Given that
the voids form continuous channels through the crystal, it is also
not surprising that the dinitrogen could be positionally disor-
dered, or at only partial occupancy in the lattice, and hence
difficult to model.
1978 cm-1
.
Synthesis of [N3xyl]Ru(PMe3)(CO), 7a. PMe3 (33 mg, 0.433
mmol) was vacuum transferred into a benzene solution (3 mL)
of [N3xyl]Ru(CO)2 (6a) (46 mg, 0.0874 mmol) at -196 °C. The
mixture was thawed and stirred for 15 h at room temperature,
during which time the color changed from green to orange. The
volatiles were removed in vacuo, and the resulting residue was
recrystallized from 2:1 pentane/toluene at -35 °C, yielding 36
mg of 7a (72% yield). 1H NMR (benzene-d6): δ 7.88 (d, 3JHH
=
7.5 Hz, 2H, Py-Hm), 6.95 (m, 7H, Xyl-Hm,p, Py-Hp), 2.25 (s, 6H,
Im-Me), 2.21 (s, 6H, Xyl-Me), 1.60 (s, 6H, Xyl-Me), 0.63 (d, 2JPH
=7.9 Hz, 9H, PMe3). 13C{1H} NMR (benzene-d6): δ 206.12 (d,
2JPC=27.7 Hz, 1C, CO), 155.44 (s, 2C, C=N), 150.44 (s, 2C, Py-
Co), 142.59 (s, 2C, Xyl-C-N), 131.74, 131.57, 129.48, 128.44,
125.43, 123.63, and 111.73 (s, 13C, aryl C), 19.09 (d, 1JPC=14.2
Hz, 3C, PMe3), 18.95 (s, 4C, Xyl-Me), 15.34 (s, 2C, Im-Me).
31P{1H} NMR (benzene-d6): δ 1.54 (s, 1P, PMe3). IR (benzene):
ν(CO)=1963 cm-1
.
Acknowledgment. We are grateful to the National
Science Foundation for support of this work under grant
CHE-9904798. We also thank Andrew Blum for his
assistance with the synthesis of the [tBu-N3mes] ligand.
Single-Crystal X-ray Diffraction Analyses. X-ray intensity
data were collected on a Rigaku Mercury CCD area detector
employing graphite-monochromated Mo KR radiation (λ =
0.71069 A) at a temperature of 143 K. Oscillation images were
processed using CrystalClear,52 producing a listing of unaver-
aged F2 and σ(F2) values, which were then passed to the Crystal
Structure53 program package for further processing. Refine-
ment was by full-matrix least-squares based on F2 using
SHELXL-97.54 All reflections were used during refinement
(F2’s that were experimentally negative were replaced by F2 =
0). Non-hydrogen atoms were refined anisotropically, and
hydrogen atoms were refined using a “riding” model and
structure refinement. In all cases, non-hydrogen atoms were
refined anisotropically and hydrogen atoms were refined using a
“riding” model.
Supporting Information Available:
X-ray crystallographic
data (in CIF format) and 1H NMR spectra of purified samples are
available free of charge via the Internet at http://pubs.acs.org.
(55) SQUEEZE: Sluis, P. v.d.; Spek, A. L. Acta Crystallogr. 1990,
A46, 194.
(56) (a) Phillips, A. E.; Goodwin, A. L.; Halder, G. J.; Southon, P. D.;
Kepert, C. J. Angew. Chem., Int. Ed. 2008, 47, 1369–1399. (b) Ma, X.;
You, Z. Transition Met. Chem. 2008, 33, 961–965. (c) Kachi-Terajima, C.;
Akatsuka, T.; Takamizawa, S. Chem. Asian J. 2007, 2, 40–50. (d) Makarov,
A. Y.; Tersango, K.; Nivesanond, K.; Blockhuys, C. V. A.; Kovalev, M. K.;
Bagryanskaya, I, Y.; Gatilov, Y. V.; Shakirov, M. M.; Zibarev, A. V. Inorg.
Chem. 2006, 45, 2221–2228. (e) Rowsell, J. L. C.; Spencer, E. C.; Eckert, J.;
Howard, J. A. K.; Yaghi, O. M. Science 2005, 309, 1350–1354. (f) Clarke, C.
S.; Haynes, D. A.; Rawson, J. M.; Bond, A. D. Chem. Commun. 2003, 2774–
2775. (g) Bryan, C. D.; Cordes, A. W.; Haddon, R. C.; Hicks, R. G.;
Kennepohl, D. K.; MacKinnon, C. D.; Oakley, R. T.; Palstra, T. M. M.;
Perel, A. S.; Scott, S. R.; Schneemeyer, L. F.; Waszczak, J. V. J. Am. Chem.
Soc. 1994, 116, 1205–1210. (h) Balch, A. L.; Olmstead, M. M.; Safari, N. J.
Inorg. Chem. 1993, 32, 291–296. (i) Wood, F. E.; Olmstead, M. M.; Balch, A.
L. J. Am. Chem. Soc. 1983, 105, 6332–6334. (j) Gies, H. Z. Kristallogr.
Kristallgem. Kristallphys. Kristallchem. 1983, 164, 247–257.
The dichloride carbonyl complex [N3xyl]RuCl2(CO) cocrys-
tallized with one-half chlorobenzene molecule per unit cell. The
chlorobenzene sits on an inversion center, with the pseudo-
(52) CrystalClear; Rigaku Corporation, 1999.
(53) Crystal Structure: Crystal Structure Analysis Package; Rigaku
Corporation, 2002.
(54) Sheldrick, G. M. SHELXL-97: Program for the Refinement of
€
Crystal Structures; University of Gottingen, Germany, 1997.