Reaction of 6-methyl-2,2′-bipyridine with 1,2-Os3(CO) 10(MeCN)2: Syntheses, reductive elimination/ligand displacement kinetics, and X-ray diffraction structures of the isomeric clusters HOs3(CO)9(μ2-N2C 11H9)

Article abstract of DOI:10.1021/om800020x

Treatment of Os₃(CO)₁₀(MeCN)₂ (1) with the heterocyclic ligand 6-methyl-2,2′-bipyridine (6-Me-2,2′-bpy) at room temperature leads to the formation of the isomeric hydride-bridged clusters HOs₃(CO)₉(μ₂-CH₂N ₂C₁₀H₇) (2) and HOs₃(CO) ₉(μ₂-N₂C₁₁H₉) (3). The cyclometalation of the ancillary 6-Me group in 2 and the ortho metalation of the nonsubstituted pyridyl ring in 3 have been confirmed by spectroscopic and crystallographic methods. Thermolysis of 2 leads to the formation of 3 and the dihydride cluster H₂Os₃(CO)₈(μ₃- N₂C₁₁H₈) (4); the latter cluster, whose structure has been crystallographically determined, derives from a formal loss of CO and C-H bond activation of the methylene moiety in 2. Heating 2 in the presence of ligand-trapping agents proceeds with the release of the 6-Me-2,2′-bpy ligand and formation of Os₃(CO)₉L ₃ [where L = CO, P(OMe)₃]. The kinetics for the reaction between 2 and added ligand have been investigated by UV-vis and NMR spectroscopies and found to be first-order in starting cluster and independent of the incoming ligand. Parallel kinetic experiments employing the deuterated cluster DOS₃(CO)₉(μ₂-CD₂N ₂C₁₀H₇) (2-d₃), which was prepared from cluster 1 and 6-Me-d₃-2,2′-bpy, confirm the existence of a primary kinetic isotope effect (KIE) of 1.78 at 323 K. The KIE data and the calculated activation parameters [ΔS? = 21.7(4) kcal/mol; ΔS? = -13(1) eu] are strongly suggestive of a reaction scheme involving a rate-limiting reductive coupling of the bridging hydride ligand and cyclometalated alkyl moiety in 2 to furnish a putative sigma complex containing an intact methyl group bound to the Os₃ cluster, prior to the generation of the unsaturated cluster Os₃(CO)₉(μ-N ₂C₁₁H₁₀). Thermolysis of 3 in the presence of added P(OMe)₃ does not furnish free 6-Me-2,2′-bpy but proceeds by a ligand-induced displacement of the methyl-substituted pyridyl ring and formation of the cluster compound HOs₃(CO)₉-[P(OMe) ₃](μ₂-N₂C₁₁H₉) (5). The kinetics for the reaction between 3 and P(OMe)₃ have been studied over the temperature range 333-356 K, and on the basis of the observed activation parameters [ΔH? = 13.0(3) kcal/mol; ΔS? = -30(1) eu] and the first-order dependence on the cluster and ligand, an associative process that involves P(OMe)₃ ligand attack on the cluster and release of the methyl-substituted pyridyl ring in the rate-limiting step is proposed.

Full text of DOI:10.1021/om800020x

3018
Organometallics 2008, 27, 3018–3028
Reaction of 6-Methyl-2,2′-bipyridine with 1,2-Os3(CO)10(MeCN)2:
Syntheses, Reductive Elimination/Ligand Displacement Kinetics, and
X-ray Diffraction Structures of the Isomeric Clusters
HOs3(CO)9(µ2-N2C11H9) and H2Os3(CO)8(µ3-N2C11H8)
Bhaskar Poola,† Carl J. Carrano,‡ and Michael G. Richmond*,†
Department of Chemistry, UniVersity of North Texas, Denton, Texas 76203-5070, and Department of
Chemistry and Biochemistry, San Diego State UniVersity, San Diego, California 82182-1030
ReceiVed January 8, 2008
Treatment of Os3(CO)10(MeCN)2 (1) with the heterocyclic ligand 6-methyl-2,2′-bipyridine (6-Me-2,2′-
bpy) at room temperature leads to the formation of the isomeric hydride-bridged clusters HOs3(CO)9(µ2-
CH2N2C10H7) (2) and HOs3(CO)9(µ2-N2C11H9) (3). The cyclometalation of the ancillary 6-Me group in
2 and the ortho metalation of the nonsubstituted pyridyl ring in 3 have been confirmed by spectroscopic
and crystallographic methods. Thermolysis of 2 leads to the formation of 3 and the dihydride cluster
H2Os3(CO)8(µ3-N2C11H8) (4); the latter cluster, whose structure has been crystallographically determined,
derives from a formal loss of CO and C-H bond activation of the methylene moiety in 2. Heating 2 in
the presence of ligand-trapping agents proceeds with the release of the 6-Me-2,2′-bpy ligand and formation
of Os3(CO)9L3 [where L ) CO, P(OMe)3]. The kinetics for the reaction between 2 and added ligand
have been investigated by UV-vis and NMR spectroscopies and found to be first-order in starting cluster
and independent of the incoming ligand. Parallel kinetic experiments employing the deuterated cluster
DOs3(CO)9(µ2-CD2N2C10H7) (2-d3), which was prepared from cluster 1 and 6-Me-d3-2,2′-bpy, confirm
the existence of a primary kinetic isotope effect (KIE) of 1.78 at 323 K. The KIE data and the calculated
activation parameters [∆Hq ) 21.7(4) kcal/mol; ∆Sq ) -13(1) eu] are strongly suggestive of a reaction
scheme involving a rate-limiting reductive coupling of the bridging hydride ligand and cyclometalated
alkyl moiety in 2 to furnish a putative sigma complex containing an intact methyl group bound to the
Os3 cluster, prior to the generation of the unsaturated cluster Os3(CO)9(µ-N2C11H10). Thermolysis of 3 in
the presence of added P(OMe)3 does not furnish free 6-Me-2,2′-bpy but proceeds by a ligand-induced
displacement of the methyl-substituted pyridyl ring and formation of the cluster compound HOs3(CO)9-
[P(OMe)3](µ2-N2C11H9) (5). The kinetics for the reaction between 3 and P(OMe)3 have been studied
over the temperature range 333-356 K, and on the basis of the observed activation parameters [∆Hq )
13.0(3) kcal/mol; ∆Sq ) -30(1) eu] and the first-order dependence on the cluster and ligand, an associative
process that involves P(OMe)3 ligand attack on the cluster and release of the methyl-substituted pyridyl
ring in the rate-limiting step is proposed.
Recent reports from Cabeza and co-workers have demon-
strated the facile activation of the C-H bonds in an ancillary
methyl group in 6,6′-Me2-2,2′-bipyridine and 2,9-Me2-1,10-
phenanthroline by Ru3(CO)12 to furnish the corresponding
cyclometalated clusters H2Ru3(CO)8(µ3-HCN-NMe) (where N-N
) 2,2′-bpy, 1,10-phen).3 Thermolysis of these products in the
presence of added Ru3(CO)12 affords the ortho-metalated carbide
clusters HRu5(µ5-C)(CO)13(µ-N-NMe) through a chelate-assisted
bond activation sequence, with eq 4 illustrating these ligand
Introduction
The catalytic activation of C-H bonds in nitrogen-containing
heterocyles continues to captivate the attention and research
efforts of many academic and industrial groups.1 Here the
efficient production of commodity pharmaceutical chemicals via
alternative methodologies provides the continuing allure and
incentive for practitioners wishing to demonstrate selective
protocols for the functionalization of C-H bonds in heterocyclic
substrates. Some of these transformations that have achieved
distinction include the coupling of N-heterocycles with alkenes
and arenes, site-specific carbonylations, and nucleophilic addi-
tion reactions to metal-activated heterocyclic platforms.2 Rep-
resentative paradigms of these reactions are shown in eqs 1-3.
(2) (a) Bergman, B.; Holmquist, R.; Smith, R.; Rosenberg, E.; Ciurash,
J.; Hardcastle, K.; Visi, M. J. Am. Chem. Soc. 1998, 120, 12818. (b) Smith,
R.; Rosenberg, E.; Hardcastle, K. I.; Vazquez, V.; Roh, J. Organometallics
1999, 18, 3519. (c) Chatani, N.; Asaumi, T.; Ikeda, T.; Yorimitsu, S.; Ishii,
Y.; Kakiuchi, F.; Murai, S. J. Am. Chem. Soc. 2000, 122, 12882. (d) Chatani,
N.; Asumi, T.; Yorimitsu, S.; Ikeda, T.; Kakiuchi, F.; Murai, S. J. Am.
Chem. Soc. 2001, 123, 10935. (e) Rosenberg, E.; Kabir, S. E.; Abedin,
M. J.; Hardcastle, K. I. Organometallics 2004, 23, 3982. (f) Wiedemann,
S. H.; Lewis, J. C.; Ellman, J. A.; Bergman, R. G. J. Am. Chem. Soc. 2006,
128, 2452. (g) Wiedemann, S. H.; Ellman, J. A.; Bergman, R. G. J. Org.
Chem. 2006, 71, 1969. (h) Lewis, J. C.; Wu, J. Y.; Bergman, R. G.; Ellman,
J. A. Angew. Chem., Int. Ed. 2006, 45, 1589. (i) Lewis, J. C.; Bergman,
R. G.; Ellman, J. A. J. Am. Chem. Soc. 2007, 129, 5332. (j) Musaev, D. G.;
Nowroozi-Isfahani, T.; Morokuma, K.; Abedin, J.; Rosenberg, E.; Hard-
castle, K. I. Organometallics 2006, 25, 203.
* Corresponding author. E-mail: cobalt@unt.edu.
† University of North Texas.
‡ San Diego State University.
(1) (a) Ritleng, V.; Sirlin, C.; Pfeffer, M. Chem. ReV. 2002, 102, 1731.
(b) Pastine, S. J.; Youn, S. W.; Sames, D. Tetrahedron 2003, 59, 8859. (c)
Zificsak, C. A.; Hlasta, D. J. Tetrahedron 2004, 60, 8991. (d) Tobisu, M.;
Chatani, N. Angew. Chem., Int. Ed. 2006, 45, 1683. (e) Godula, K.; Sames,
D. Science 2006, 312, 67. (f) Campos, K. R. Chem. Soc. ReV. 2007, 36,
1069.
10.1021/om800020x CCC: $40.75
2008 American Chemical Society
Publication on Web 06/07/2008

Coordination and Rupture of Methyl C-H Bonds
Organometallics, Vol. 27, No. 13, 2008 3019
methyl groups in the 6,6′-Me2-2,2′-bipyridine and 2,9-Me2-1,10-
phenanthroline ligands and readily accounts for the absence of
the putative cluster species Ru3(CO)10(6,6′-Me2-2,2′-bpy) and
Ru3(CO)10(2,9-Me2-1,10-phen) in the Cabeza report. The extant
steric crowding between the cluster-bound CO groups and the
alkyl appendages of the heterocyclic auxiliary in these decac-
arbonyl clusters forms the basis for the chelate-directed C-H
bond activation observed by Cabeza.
Given the steric-induced activation of the methyl groups in
the ortho-disubstituted 2,2′-bpy ligand in the reactions depicted
in eq 4, it seemed reasonable that reactions employing the
slightly less encumbered heterocycle 6-Me-2,2′-bpy might
facilitate the observation and isolation of the putative clusters
M3(CO)10(6-Me-2,2′-bpy).6 The conventional thermolysis condi-
tions that are required to effect ligand substitution in the parent
clusters M3(CO)12 (where M ) Ru, Os) guarantee difficulty in
the successful preparation of the target decacarbonyl clusters
due to low-energy manifolds involving additional CO loss and
metal skeletal reorganizations. The use of an activated cluster
platform, one that would allow the substitution of the 6-Me-
2,2′-bpy to occur at or near room temperature, could, in theory,
allow the initial coordination chemistry to be established.
Accordingly, we have selected Os3(CO)10(MeCN)2 as our
starting cluster to probe the early stages of the substitution
reaction with the 6-Me-2,2′-bpy ligand. This particular cluster
has a rich chemical history because the two MeCN ligands are
lightly bound and readily replaced by donor ligands, making it
an ideal precursor for studies dealing with the early stages of
substrate binding and ligand activation at an intact polynuclear
entity.7,8
activation sequences using the 6,6′-Me2-2,2′-bpy ligand. The
ortho-metalated carbide clusters react with terminal alkynes to
givethealkenyl-substitutedclustersRu5(µ5-C)(CO)13(µ-RCHCHN-
NMe), establishing a viable cluster-based catalytic cycle for an
alkyl-to-alkenyl functionalization of alkylated heterocyclic
substrates.
Herein we report our data on the room-temperature reaction of
6-Me-2,2′-bpy with Os3(CO)10(MeCN)2, which affords the isomeric
hydride clusters HOs3(CO)9(µ2-CH2N2C10H7) (2) and HOs3(CO)9(µ2-
N2C11H9) (3). Thermolysis of cluster 2 leads to the formation of 3
(major) and the dihydride cluster H2Os3(CO)8(µ3-N2C11H8) (4)
(minor). The new clusters 2-4 have been isolated and fully
characterized in solution and by X-ray crystallography. The
dynamics associated with the reductive coupling of the cyclom-
etalated methyl group and the bridging hydride in 2 have been
kinetically probed in the presence of trapping ligands.
Experimental Section
General Methods. Os3(CO)12 was prepared by high-pressure
carbonylation of OsO4 using a 450 mL Parr Series 4560 benchtop-
mini reactor,9 with Os3(CO)10(MeCN)2 synthesized from Os3(CO)12,
The coordination of the parent R-diimine ligands 2,2′-bpy
and 1,10-phen at the trimetallic clusters Ru3(CO)12 and
Os3(CO)12 gives the chelated clusters Ru3(CO)10(N-N) and
Os3(CO)10(N-N),4,5 whose ligand coordination sphere about the
metallic polyhedron has been noted to be unduly crowded. This
situation is further exacerbated by the presence of the two ortho
(6) Whereas the reaction chemistry of mononuclear metal compounds
with ortho-substituted 2,2′-bpy and 1,10-phen derivatives has been exten-
sively investigated, the reactivity of metal clusters with the same ligands
remains, to our knowledge, relatively unexplored. For some representative
examples involving mononuclear systems, see:(a) Newkome, G. R.; Evans,
D. W.; Kiefer, G. E.; Theriot, K. J. Organometalllics 1988, 7, 2537. (b)
Constable, E. C.; Henney, R. P. G.; Leese, T. A.; Tocher, T. A. J. Chem.
Soc., Dalton Trans. 1990, 443. (c) Zucca, A.; Stoccoro, S.; Cinellu, M. A.;
Minghetti, G.; Manassero, M. J. Chem. Soc., Dalton Trans. 1999, 3431.
(d) Romeo, R.; Plutino, M. R.; Scolaro, L. M.; Stoccoro, S.; Minghetti, G.
Inorg. Chem. 2000, 39, 4749.
(3) (a) Cabeza, J. A.; da Silva, I.; del R´ıo, I.; Mart´ınez-Me´ndez, L.;
Miguel, D.; Riera, V. I. Angew. Chem., Int. Ed. 2004, 43, 3464. (b) Cabeza,
J. A.; del R´ıo, I.; Mart´ınez-Me´ndez, L.; Miguel, D. Chem.-Eur. J. 2006,
12, 1529.
(7) (a) Burgess, K.; Johnson, B. F. G.; Lewis, J. J. Organomet. Chem.
1982, 233, C55. (b) Dahlinger, K.; Poe¨, A. J.; Sayal, P. K.; Sekhar, V. C.
J. Chem. Soc., Dalton Trans. 1986, 2145. (c) Zoet, R.; van Koten, G.; Vrieze,
K.; Duisenberg, A. J. M.; Spek, A. L. Inorg. Chim. Acta 1988, 148, 71. (d)
Akther, J.; Azam, K. A.; Das, A. R.; Hursthouse, M. B.; Kabir, S. E.; Abdul
Malik, K. M.; Rosenberg, E.; Tesmer, M.; Vahrenkamp, H. J. Organomet.
Chem. 1999, 588, 211. (e) Watson, W. H.; Poola, B.; Richmond, M. G. J.
Organomet. Chem. 2006, 691, 4676.
(4) (a) Venalainen, T.; Pursiainen, J.; Pakkanen, T. A. J. Chem. Soc.,
Chem. Commun. 1985, 1348. (b) Bruce, M. I.; Humphrey, M. G.; Snow,
M. R.; Tiekink, E. R. T.; Wallis, R. C. J. Organomet. Chem. 1986, 314,
311. (c) Leadbeater, N. E.; Raithby, P. R.; Ward, G. N. J. Chem. Soc.,
Dalton Trans. 1997, 2511. (d) Poola, B.; Wang, X.; Richmond, M. G.
J. Chem. Crystallogr. 2007, 37, 641.
(5) For a report on the ligand-activated cluster HOs3(CO)9(µ-C10H7N2),
which has been isolated from the extended thermolysis of Os3(CO)10(2,2′-
bpy), see:Deeming, A. J.; Peters, R.; Hursthouse, M. B.; Backer-Dirks,
J. D. J. J. Chem. Soc., Dalton Trans. 1982, 787.
(8) For a computational study dealing with the lability and isomerization
of the MeCN ligands in Os3(CO)10(MeCN)2, see:Musaev, D. G.; Nowroozi-
Isfahani, T.; Morokuma, K.; Rosenberg, E. Organometallics 2005, 24, 5973.
(9) Drake, S. R.; Loveday, P. A. Inorg. Synth. 1990, 28, 230.

3020 Organometallics, Vol. 27, No. 13, 2008
Poola et al.
3
3
MeCN, and Me3NO.10 The heterocyclic ligand 6-Me-2,2′-bpy was
prepared from 2,2′-bpy and MeLi,11 while the 6-Me-d3-2,2′-bpy
ligand was prepared from 2,2′-bpy and Me-d3-Li.12 The deuterated
ligand P(OCD3)3 was prepared according to the published procedure
starting from CD3OD (99.8 atom % D) and PCl3.13 The chemicals
2,2′-bpy, MeLi (1.6 M in Et2O), Me-d3-Li (99 atom % D; 0.5 M in
Et2O), and Me3NO · 2H2O were purchased from Aldrich Chemical
Co. and used as received, except for the latter chemical, which
was rendered anhydrous by azeotropic distillation from benzene.
The alkyllithium reagents were titrated against diphenylacetic acid
prior to use.14 All reaction solvents were distilled from an
appropriate drying agent under argon using Schlenk techniques and
stored in Schlenk storage vessels equipped with high-vacuum Teflon
stopcocks.15 The solvents used in the collection of routine IR and
NMR spectra were reagent grade and were typically degassed by
three pump-thaw-degas cycles prior to their use. The solvents
employed in all kinetic experiments were rigorously purified by
bulb-to-bulb distillation from an appropriate drying agent prior to
use. The photochemical studies were conducted with GE blacklight
bulbs that had a maximum output at 366 ( 20 nm and a photon
flux of ca. 1 × 10-6 einstein/min. The reported combustion analyses
were performed by Atlantic Microlab, Norcross, GA.
The infrared spectra were recorded on a Nicolet 20 SXB FT-IR
spectrometer in a 0.1 mm NaCl cell, using PC control and OMNIC
software. The reported 1H and 31P NMR spectra were recorded at
200 MHz on a Varian Gemini-200 spectrometer and 121 MHz on
a Varian 300-VXR spectrometer, respectively. The reported 31P
chemical shifts, which were recorded in the proton-decoupled mode,
are referenced to external H3PO4 (85%), taken to have δ ) 0. The
reported FAB and ESI mass spectral data were collected at the mass
spectrometry facility at the University of California at San Diego.
The ESI mass spectrum was recorded in the positive ion mode using
MeOH and a trace amount of AcOH as the sample matrix, while
the FAB mass spectra were recorded using 3-nitrobenzyl alcohol
as the sample matrix. The GC-MS data on the ligands 6-Me-2,2′-
bpy and 6-Me-d3-2,2′-bpy were recorded on a Finnigan/Thermo-
Electron GC-MS system at UNT.
1J ) 16 Hz, J ) 3 Hz, J ) 1 Hz), -20.25 (s, hydride). Anal.
Calc (Found) for C20H10N2O9Os3 · 1/4C6H6: C, 25.48 (25.06); H,
1.14 (1.22).
Preparation of HOs3(CO)9(µ2-N2C11H9) (3) and H2Os3(CO)8(µ3-
N2C11H8) (4) from the Thermolysis of HOs3(CO)9(µ2-CH2N2-
C10H7) (2). To 0.10 g (0.10 mmol) of cluster 2 in a Schlenk tube
under argon flush was added 40 mL of toluene via syringe, after
which the vessel was capped and placed in a thermostated bath at
75 °C. The solution was heated for 24 h, with periodic monitoring
by TLC analysis. The cooled solution revealed the presence of two
closely moving spots by TLC corresponding to cluster 3 (Rf ) 0.50)
and the dihydride cluster 4 (Rf ) 0.47) using a 1:1 mixture of
CH2Cl2/hexane as the eluent. The toluene was removed under
vacuum and the residue subjected to column chromatography using
the aforementioned mobile phase. Both product clusters 3 and 4
were subsequently recrystallized from benzene/hexane to furnish
3 as a red-orange solid (70 mg; 70% yield) and 4 as a yellow-
orange solid (21 mg; 22% yield). Cluster 3: IR (CH2Cl2): ν(CO)
2083 (s), 2041 (vs), 2003 (vs), 1987 (s), 1955 (s), 1933 (m) cm-1
.
1H NMR (CD2Cl2): δ 8.10 (d, 1H, J ) 8 Hz), 7.71 (t, 1H, J ) 8
Hz), 7.45-7.10 (m, 4H), 2.80 (s, 3H, Me), -21.48 (s, hydride).
Anal. Calc (Found) for C20H10N2O9Os3: C, 24.19 (24.38); H, 1.02
(1.49). Cluster 4: IR (CH2Cl2): ν(CO) 2089 (s), 2049 (vs), 2004
(vs), 1985 (s), 1970 (s), 1950 (m), 1923 (m) cm-1 1H NMR
.
(CD2Cl2): δ 9.09 (d, 1H, J ) 5 Hz), 7.90 (m, 1H), 7.50 (t, 1H, J
) 8 Hz), 7.38-7.28 (m, 2H), 7.18 (d, 1H, J ) 8 Hz), 6.97 (d, 1H,
J ) 8 Hz), 5.38 (s, 1H, carbyne CH), -10.81 (d, 1H, J ) 2 Hz),
-14.82 (d, 1H, J ) 2 Hz). Anal. Calc (Found) for C19H10N2O8Os3:
C, 23.63 (23.91); H, 1.04 (1.45).
Photochemical Preparation of Cluster 4 from Cluster 2. Here
the reaction was conducted in a NMR tube that was equipped with
a J-Young valve in order to release the liberated CO that
accompanies the formation of cluster 4. The NMR tube was charged
with 50 mg (0.050 mmol) of cluster 2 and 0.7 mL of benzene-d6,
after which the tube was subjected to three freeze-pump-thaw
degas cycles. The photolysis was carried out using two GE
Preparation of the Hydride-Bridged Clusters HOs3(CO)9(µ2-
CH2N2C10H7) (2) and HOs3(CO)9(µ2-N2C11H9) (3) from
Os3(CO)10(MeCN)2 (1) and 6-Me-2,2′-bpy. To an argon-filled
Schlenk flask containing 0.30 g (0.32 mmol) of cluster 1 was added
100 mL of CH2Cl2 via cannula, followed by 54 mg (0.32 mmol)
of 6-Me-2,2′-bpy. The reaction was stirred for 2 h at room
temperature and then examined by TLC analysis, which confirmed
the consumption of the starting cluster and the presence of clusters
2 and 3 as two close running spots [Rf ) 0.50 for 3 and Rf ) 0.45
for 2 in CH2Cl2/hexane (1:1)]. A considerable amount of decom-
posed material also accompanied the reaction based on the dark
brown spot that remained at the origin of the TLC plate. The solvent
was next evaporated, and the desired clusters 2 and 3 were isolated
by column chromatography over silica gel. Cluster 3 was obtained
by using CH2Cl2/hexane (1:3) as the eluent, with cluster 2 eluting
when the mobile phase was changed to a 3:7 mixture of CH2Cl2/
hexane. Crude yield of cluster 3: 9.4% (30 mg). Cluster 2 was
recrystallized from benzene/CH2Cl2 to furnish 2 as a red-orange
solid in 38% yield (0.12 g). Data for 2: IR (CH2Cl2): ν(CO) 2075
1
blacklights, and the progress of the reaction was assessed by H
NMR spectroscopy. Quantitative conversion of 2 to 4 was realized
after several days of continuous photolysis. The time for the reaction
was not optimized, as it is dependent on the removal of the liberated
CO, which in turn depends on the number of additional freeze-
pump-thaw degas cycles. Typically, the NMR tube was subjected
to one freeze-pump-thaw degas cycle per day.
Preparation of 6-Me-d3-2,2′-bpy from 2,2′-bpy and Me-d3-
Li. To 5.0 g (0.032 mol) of 2,2′-bpy in 250 mL of Et2O at 0 °C
was added 64 mL of a 0.5 M ether solution (0.032 mol) of Me-
d3-Li dropwise via a pressure-equalizing addition funnel. The
temperature was maintained at 0 °C during the addition phase of
Me-d3-Li to the reaction, after which the reaction solution was
allowed to warm to room temperature with continuous stirring. At
this point the resulting maroon solution was refluxed for ca. 3 h,
followed by cooling and quenching with 30 mL of water. The
organic layer was separated via cannula and the aqueous layer
extracted with 3 × 20 mL portions of Et2O. The combined organic
layers were dried over sodium sulfate and filtered, and the solvent
was removed under vacuum to afford the corresponding dihydro
adduct 6-methyl-d3-1,6-dihydro-2,2′-bipyridine, which was then
oxidized by using 300 mL of a saturated KMnO4 solution in
acetone. The accompanying MnO2 was filtered away and the
acetone removed under vacuum to furnish a dark red liquid, which
was passed across a column of neutral alumina using Et2O/hexane
(1:4) as the eluent. The desired ligand 6-Me-d3-2,2′-bpy was
1
(s), 2030 (vs), 2008 (vs), 1987 (s), 1967 (s), 1946 (s) cm-1. H
NMR (CD2Cl2): δ 9.27 (d, 1H, J ) 5 Hz), 8.08 (d, 1H, J ) 8 Hz),
7.95-7.80 (m, 2H), 7.53-7.24 (m, 3H), 2.87 (multiplet, CH2 group,
(10) Nicholls, J. N.; Vargas, M. D. Inorg. Synth. 1989, 26, 289.
(11) Schmalzl, K. J.; Summers, L. A. Aust. J. Chem. 1977, 30, 657.
(12) Garber, T.; Van Wallendael, S.; Rillema, D. P.; Kirk, M.; Hatfield,
W. E.; Welch, J. H.; Singh, P. Inorg. Chem. 1990, 29, 2863.
(13) Ferrari, A.; Polo, E.; Ru¨egger, H.; Sostero, S.; Venanzi, L. M. Inorg.
Chem. 1996, 35, 1602.
(14) Kofron, W. G.; Baclawski, L. M. J. Org. Chem. 1976, 41, 1879.
(15) Shriver, D. F. The Manipulation of Air-SensitiVe Compounds;
McGraw-Hill: New York, 1969.
1
obtained as a colorless oily liquid in 70% yield (3.9 g). H NMR
(CD2Cl2): δ 8.67 (d, 1H, J ) 5 Hz), 8.42 (dt, 1H, J ) 8 Hz, J )
2 Hz), 8.18 (d, 1H, J ) 8 Hz), 7.81 (dt, 1H, J ) 8 Hz, J ) 2 Hz),

Coordination and Rupture of Methyl C-H Bonds
Organometallics, Vol. 27, No. 13, 2008 3021
7.71 (t, 1H, J ) 8 Hz), 7.30 (ddd, 1H, J ) 8 Hz, J ) 5 Hz, J )
1 Hz), 7.18 (dd, 1H, J ) 8 Hz, J ) 1 Hz). GC-MS: m/z 173.0
[M]+.
band as a function of time for at least 4-6 half-lives. The kinetics
for the reaction between cluster 3 with added ligand were followed
similarly, using the 540 nm absorbance band of 3 to monitor the
progress of the reaction. The UV-vis-derived rate constants quoted
in Tables 3–5 were determined by nonlinear regression analysis
Preparation of the Deuteride-Bridged Cluster DOs3(CO)9(µ2-
CD2N2C10H7)(2-d3)andtheHydride-BridgedClusterHOs3(CO)9(µ2-
N2C11H6D3) (3-d3) from Os3(CO)10(MeCN)2 (1) and 6-Me-d3-
2,2′-bpy. The procedure employed for the preparation of the
d3-isotopomers parallels that reported above for clusters 2 and 3
using 6-Me-2,2′-bpy. Here an argon-filled Schlenk flask containing
0.30 g (0.32 mmol) of cluster 1 dissolved in 100 mL of CH2Cl2
was treated with 55 mg (0.32 mmol) of 6-Me-d3-2,2′-bpy. The
reaction was stirred for ca. 2 h at room temperature, after which
time TLC analysis confirmed the presence of the desired clusters
2-d3 and 3-d3. The solvent was evaporated, and both product clusters
were isolated by column chromatography over silica gel. Cluster
3-d3 was obtained by using CH2Cl2/hexane (1:3) as the eluent,
followed by cluster 2-d3 when the mobile phase was changed to a
3:7 mixture of CH2Cl2/hexane. Both products were recrystallized
from benzene/CH2Cl2 to furnish cluster 3-d3 in 18% yield (58 mg)
and cluster 2-d3 in 29% yield (93 mg). Data for 2-d3: IR (CH2Cl2):
ν(CO) 2075 (s), 2028 (vs), 2004 (vs), 1986 (s), 1970 (s), 1947 (s)
using the single-exponential function16 A(t) ) A∞ + ∆Ae(-kt)
.
The 1H NMR kinetic experiments were conducted in 5 mm NMR
tubes that possessed a J. Young value in CD2Cl2 solvent or in NMR
tubes equipped with a right-angle high-vacuum Teflon stopcock.
These latter NMR tubes were manipulated on the vacuum line and
then flamed sealed after multiple freeze-pump-thaw degassing
cycles. The NMR reactions were followed for a period of 2-3 half-
lives, and the extent of these reactions was determined by calibration
against a measured amount of the internal standard ferrocene. The
rate constants quoted from the NMR experiments were determined
by traditional graphical analysis of ln[cluster] versus time plots.
The activation parameters for the consumption of clusters 2 and
3 in the ligand reactivity studies were calculated from a plot of
ln(k/T) versus T-1 17
, with the error limits representing the deviation
of the data points about the least-squares line of the Eyring plot.
X-ray Structural Analyses. Single crystals of clusters 2-4
suitable for X-ray crystallography were grown from CH2Cl2 and
hexane. The X-ray data were collected on a Bruker X8 APEX CCD
diffractometer using Mo radiation and a graphite monochromator.
The frames were integrated with the available SAINT software
package using a narrow-frame algorithm,18 with each structure
solved and refined using the SHELXTL program package.19 The
molecular structures were checked by using PLATON,20 and all
non-hydrogen atoms were refined anisotropically, with the carbon-
bound hydrogens assigned to calculated positions and allowed to
ride on the attached carbon atom. The bridging hydrides in the three
structures were not located during refinement. Tables 1 and 2
summarize the pertinent X-ray data for these clusters.
1
cm-1. H NMR (CD2Cl2): δ 9.27 (d, 1H, J ) 5 Hz), 8.08 (d, 1H,
J ) 8 Hz), 7.92 (d, 1H, J ) 8 Hz), 7.80 (d, 1H, J ) 6 Hz),
7.53-7.24 (m, 3H). FAB-MS: m/z 997.2 (calcd for [M + H]+:
C20H8D3N2O9Os3 997.0), along with ions for the sequential loss of
seven CO groups. Data for Cluster 3-d3: IR (CH2Cl2): ν(CO) 2084
1
(s), 2040 (vs), 2005 (vs), 1985 (s), 1955 (s), 1932 (m) cm-1. H
NMR (CD2Cl2): δ 8.10 (d, 1H, J ) 8 Hz), 7.71 (t, 1H, J ) 8 Hz),
7.45-7.10 (m, 4H), -21.48 (s, hydride). FAB-MS: m/z 997.2 (calcd
for [M + H]+: C20H8D3N2O9Os3 997.0), along with ions for the
sequential loss of five CO groups.
Synthesis of HOs3(CO)9[P(OMe)3](µ2-N2C11H9) (5) from
Cluster 3 and P(OMe)3. To a large Schlenk tube containing 0.10 g
of 3 (0.10 mmol) and 50 mL of toluene was added 0.12 mL (1.0
mmol) of P(OMe)3 via syringe. The vessel was then heated at ca.
60 °C and monitored by TLC analysis until the starting cluster was
consumed, after which time the solvent and the excess P(OMe)3
were removed under vacuum. The yellow residue was then passed
across a short column of silica gel using CH2Cl2 as the eluent and
then recrystallized from benzene/CH2Cl2 to afford 84 mg (75%
yield) of yellow 5. IR (CH2Cl2): ν(CO) 2086 (s), 2043 (vs), 2019
Results and Discussion
I. Reaction of 1,2-Os3(CO)10(MeCN)2 with 6-Me-2,2′-bpy.
Treatment of 1,2-Os3(CO)10(MeCN)2 with an equimolar amount
of 6-Me-2,2′-bpy in CH2Cl2 at room temperature furnishes the
isomeric hydride-bridged clusters HOs3(CO)9(µ2-CH2N2C10H7)
(2) and HOs3(CO)9(µ2-N2C11H9) (3). The expected precursor
to 2 and 3, namely, the ligand-chelated cluster Os3(CO)10(µ-6-
Me-2,2′-bpy), was not observed at any time when our reactions
(vs), 1997 (vs), 1979 (sh), 1963 (s), 1920 (w) cm-1 1H NMR
.
(CD2Cl2): δ 7.55 (t, 1H, J ) 7 Hz), 7.18-6.90 (m, 4H), 6.50 (d,
1H, J 7 Hz), 3.59 and 3.39 (d, methoxy groups, J ) 12 Hz, 30%
and 70%, respectively), 2.54 and 2.43 (s, 6-Me groups, 30% and
70%, respectively), -14.62 and -14.80 (d, hydrides, JP-H ) 11
Hz for both, 30% and 70%, respectively). 31P NMR (CD2Cl2): δ
114.08 and 116.25 (s, 30% and 70%, respectively). ESI-MS: m/z
1118.57 (calcd for [M + H]+: C23H20N2O12Os3P 1118.07), along
with ions for the loss of one and three CO groups.
Kinetics Studies. The UV-vis studies employed a cluster
concentration of ca. 10-4 M using 1.0 cm quartz UV-visible cells
that were equipped with a high-vacuum Teflon stopcock to facilitate
handling on the vacuum line. Solutions of cluster 2 were prepared
under either argon or CO (1 atm) and used immediately before
each kinetic measurement. The experiments conducted with cluster
2 at high CO pressure were carried out in a carbon-steel 300 mL
autoclave, whose internal CO pressure was kept constant with a
Tescom pressure regulator. The autoclave was equipped with a tip
tube that enabled sample removal for UV-vis analysis while
maintaining constant CO pressure. The Hewlett-Packard 8452A
diode array spectrometer employed in our studies was configured
with a variable-temperature cell holder and was connected to a
VWR constant-temperature circulator, allowing for the quoted
temperatures to be maintained within (0.5 K. The UV-vis kinetics
for the reaction of cluster 2 in the presence of ligand-trapping agents
were monitored by following the decrease of the 540 nm absorbance
1
were monitored by TLC and IR and H NMR spectroscopies
(Vide infra) at room temperature. This indicates that once
formed, the logical intermediate must undergo rapid cyclom-
etalation and ortho metalation to afford clusters 2 and 3,
respectively. A control experiment involving the reaction of 1,2-
Os3(CO)10(MeCN)2 with 6-Me-2,2′-bpy in CDCl3 was per-
formed at -20 °C, in an effort to trap the putative cluster
Os3(CO)10(µ-6-Me-2,2′-bpy). No reaction was observed over
the course of 2 days at this temperature when monitored by 1H
NMR spectroscopy. Raising the temperature to 0 °C led to the
release of MeCN (δ 2.08) and formation of a new species
exhibiting 1H resonances at δ 9.48 (1H, d, J ) 6 Hz), 8.20-7.40
(6H, overlapping bpy hydrogens), and 2.90 (3H, s, Me), whose
(16) The rate calculations were performed by using the commercially
available program Origin6.0. Here the initial (Ao) and final (A∞) absorbances
and the rate constant were floated to give the quoted least-squares value
for the first-order rate constant k.
(17) Carpenter, B. K. Determination of Organic Reaction Mechanisms;
Wiley-Interscience: New York, 1984.
(18) SAINT, Version 6.02; Bruker Analytical X-ray Systems, Inc.:
Madison, WI, 1997-1999.
(19) SHELXTL, Version 5.1; Bruker Analytical X-ray Systems, Inc.:
Madison, WI, 1998.
(20) Spek, A. L. PLATON-A Multipurpose Crystallographic Tool;
Utrecht University: Utrecht, The Netherlands, 2001.

3022 Organometallics, Vol. 27, No. 13, 2008
Poola et al.
Table 1. X-ray Crystallographic Data and Processing Parameters
Table 2. Selected Bond Distances (Å) and Angles (deg) for the
for the Triosmium Clusters 2-4
Triosmium Clusters 2-4
2
3
4
Cluster 2
CCDC entry no.
T, K
cryst syst
space group
a, Å
626670
626671
298 (2)
626669
Bond Distances
2.9100(5) Os(1)-Os(3)
2.9033(4) Os(1)-C(7)
240 (2)
triclinic
220 (2)
triclinic
Os(1)-Os(2)
Os(2)-Os(3)
Os(1)-C(8)
Os(1)-C(30)
Os(2)-N(2)
Os(2)-C(2)
Os(3)-C(4)
Os(3)-C(6)
2.9063(4)
1.889(6)
1.904(6)
2.138(5)
1.866(7)
1.892(7)
1.907(6)
1.466(9)
triclinic
j
j
j
P1
P1
P1
1.943(7)
2.245(6)
2.092(5)
1.898(6)
1.962(7)
1.928(6)
Os(1)-C(9)
Os(2)-N(1)
Os(2)-C(1)
Os(3)-C(3)
Os(3)-C(5)
C(11)-C(30)
9.840(1)
9.957(1)
8.5944(4)
8.8308(4)
15.3940(7)
82.301(2)
86.900(2)
79.624(2)
1138.33(9)
C20H10 N2Os3O9 C19H10N2Os3O8
992.89
2
8.4706(7)
10.3476(8)
12.451(1)
89.440(4)
80.517(4)
89.108(4)
1076.2(2)
b, Å
c, Å
12.684(2)
82.793(6)
82.191(6)
67.556(6)
1134.3(2)
R, deg
ꢀ, deg
γ, deg
V, Å3
Bond Angles
mol formula
fw
C
20H10N2Os3O9
Os(2)-Os(1)-C(8)
Os(2)-Os(1)-C(30)
Os(1)-Os(2)-C(2)
101.43(2)
77.6(2)
113.3(2)
Os(3)-Os(1)-C(8)
Os(2)-Os(1)-C(7)
Os(1)-Os(2)-N(1)
N(1)-Os(2)-N(2)
83.4(2)
117.6(2)
84.9(1)
77.1(2)
992.89
964.87
2
2.968
0.71073
17.714
0.0361
SADABS,
formula units per cell (Z) 2
Dcalcd (Mg/m3)
λ(Mo KR), Å
absorp coeff (mm-1
Rmerge
2.904
0.71073
16.816
0.0573
SADABS,
2.894
Os(1)-Os(2)-N(2) 148.2(2)
Os(2)-N(1)-C(11) 123.8(4)
Os(2)-N(2)-C(16) 117.2(4)
Os(3)-Os(1)-C(30) 102.3(2)
0.71073
16.756
0.0504
SADABS,
Os(2)-N(1)-C(15) 115.5(4)
)
Os(2)-N(2)-C(20) 124.9(4)
Os(3)-Os(2)-N(1)
Cluster 3
85.6(1)
abs corr, max./min.
0.2841/0.0812 0.0815/0.0355 0.4712/0.0761
33 023
11 343
total no. of reflns
no. of indep reflns
no. of data/restr/params 11 343/0/307
26 223
10 435
10 35/0/307
0.0583
0.1363
1.003
35 753
11 183
11 183/0/284
0.0240
0.0516
Bond Distances
2.8910(5) Os(1)-Os(3)
2.8933(5) Os(1)-C(7)
Os(1)-Os(2)
Os(2)-Os(3)
Os(1)-C(8)
Os(1)-C(20)
Os(2)-N(2)
Os(2)-C(2)
Os(3)-C(4)
Os(3)-C(6)
2.8915(5)
1.89(1)
1.90(1)
2.179(9)
1.89(1)
1.90(1)
1.92(1)
R
0.0441
0.1099
0.976
Rw
1.94(1)
2.13(1)
2.057(9)
1.86(1)
1.94(1)
1.94(1)
Os(1)-C(9)
Os(2)-N(1)
Os(1)-C(1)
Os(3)-C(3)
Os(3)-C(5)
GOF on F2
1.008
∆F(max.), ∆F(min.) (e3) 3.237, -2.815 5.150, -5.446 1.479, -1.832
integral ratios are consistent with the proposed Os3 cluster
containing an ancillary 6-Me-2,2′-bpy ligand. Os3(CO)10(µ-6-
Me-2,2′-bpy) is not stable at this temperature and undergoes
competitive conversion to the hydride clusters 2 and 3. The latter
products, whose structures are shown below, were subsequently
isolated by careful column chromatography as red-orange,
relatively air-stable solids.
Bond Angles
Os(2)-Os(1)-C(8)
Os(2)-Os(1)-C(20)
Os(1)-Os(2)-C(2)
Os(1)-Os(2)-N(2)
104.8(4)
Os(3)-Os(1)-C(8)
Os(2)-Os(1)-C(7)
89.5(4)
116.0(4)
68.8(3)
109.0(3)
68.4(2)
Os(1)-Os(2)-N(1) 137.3(2)
N(1)-Os(2)-N(2) 75.1(3)
Os(2)-N(2)-C(20) 114.2(7)
Os(2)-N(2)-C(16) 121.4(7)
Os(2)-N(1)-C(11) 128.1(8)
Os(3)-Os(2)-N(2)
Os(2)-N(1)-C(15) 111.1(6)
Os(3)-Os(1)-C(20)
89.2(3)
84.9(2)
Cluster 5
Bond Distances
Os(1)-Os(2)
Os(2)-Os(3)
Os(1)-C(8)
Os(1)-C(10)
Os(2)-N(2)
Os(2)-C(2)
Os(3)-C(4)
Os(3)-C(10)
2.9556(3) Os(1)-Os(3)
2.7745(3) Os(1)-C(7)
2.7984(3)
1.892(4)
1.904(3)
2.102(2)
1.861(3)
1.893(3)
1.914(4)
1.940(4)
2.185(3)
2.160(2)
1.888(3)
1.926(3)
2.165(3)
Os(1)-C(9)
Os(2)-N(1)
Os(2)-C(1)
Os(3)-C(3)
Os(3)-C(5)
1
The H NMR spectrum for 2 exhibits its seven aromatic
Bond Angles
Os(2)-Os(1)-C(8)
Os(3)-Os(1)-C(8)
C(7)-Os(1)-C(10)
97.8(1)
117.5(1)
93.5(1)
Os(2)-Os(1)-C(10)
Os(3)-Os(1)-C(9)
C(8)-Os(1)-C(9)
C(9)-Os(1)-C(10)
Os(1)-Os(3)-C(4)
Os(2)-Os(3)-C(4)
C(3)-Os(3)-C(10)
72.75(8)
97.5(1)
94.2(2)
hydrogens as multiple resonances over the region δ 9.27-7.24,
of which the downfield doublet at δ 9.27 is readily assigned to
the ortho hydrogen or 6′ hydrogen on the unsubstituted bipyridyl
ring.21 Cyclometalation of the original 6-Me group affords a
methylene moiety centered at δ 2.87 and a slightly broadened
high-field hydride resonance at δ -20.25. The observed
broadening in the high-field hydride is ascribed to small but
finite couplings with the diastereotopic methylene hydrogens,
as verified by spectral simulation of these three hydrogens within
an ABX spin system. The thermal ellipsoid plot of 2 shown in
Figure 1 confirms the chelation of the bpy moiety and the
cyclometalation of the methyl carbon C(30) to the Os(2) and
Os(1) centers, respectively. The Os-Os bond distances range
from 2.9100(4) Å [Os(1)-Os(2)] to 2.9033(4) Å [Os(2)-Os(3)],
with an average distance of 2.9065 Å. While the hydride ligand
was not located during refinement, it is confidently assigned to
the Os(1)-Os(2) vector on the basis of structural trends found
C(8)-Os(1)-C(10) 166.7(1)
91.4(2)
Os(2)-Os(1)-C(7)
Os(1)-Os(3)-C(3)
Os(2)-Os(3)-C(10)
Os(1)-Os(2)-N(1)
Os(3)-Os(2)-N(1)
N(1)-Os(2)-N(2)
N(1)-Os(2)-C(2)
Os(1)-C(10)-Os(3)
109.1(1)
94.0(1)
118.81(8)
91.05(9)
90.4(1)
76.98(8)
80.89(6)
86.46(6)
75.04(9)
100.0(1)
80.07(8)
Os(1)-Os(2)-N(2) 122.57(6)
Os(3)-Os(2)-N(2) 160.59(6)
N(1)-Os(2)-C(1)
N(2)-Os(2)-C(1)
169.4(1)
104.9(1)
N(1)-C(11)-C(10) 117.8(2)
in related osmium clusters.22 The Os(1)-C(30) bond length of
2.245(6) Å is in good agreement with those distances reported
in analogous cyclometalated Os3 derivatives.23
Cluster 3 was fully characterized in solution by IR and 1H NMR
spectroscopies, and the molecular structure determined by X-ray
(21) (a) Anderson, S.; Constable, E. C.; Seddon, K. R.; Turp, J. E.;
Baggott, J. E.; Pilling, M. J. J. Chem. Soc., Dalton Trans. 1985, 2247. (b)
Prasad, R.; Kumar, A.; Kumar, R. J. Mol. Struct. 2006, 786, 68.
(22) (a) Watson, W. H.; Poola, B.; Richmond, M. G. Polyhedron 2007,
26, 3585. (b) Liu, Y.-C.; Yeh, W.-Y.; Lee, G.-H.; Peng, S.-M. J. Organomet.
Chem. 2005, 690, 163.

Coordination and Rupture of Methyl C-H Bonds
Organometallics, Vol. 27, No. 13, 2008 3023
Figure 1. Thermal ellipsoid plots of the triosmium clusters 2 (left) and 3 (right) showing the thermal ellipsoids at the 40% probability level.
crystallography. The ortho metalation attendant in 3 was initially
confirmed by 1H NMR spectroscopy, where three sets of aromatic
hydrogens that integrated for a total of six hydrogens were observed
from δ 8.10 to 7.10, along with methyl and hydride singlets
recorded at δ 2.80 and -21.48, respectively. The thermal ellipsoid
plot of 3 shown in Figure 1 establishes the oxidative coupling of
the 6′ hydrogen on the unsubstituted bipyridyl ring, and with it
the formation of the four-membered metallocyclic ring composed
by the atoms Os(1)-Os(2)-N(2)-C(20). The Os-Os bond
distances in 3 range from 2.8909(5) Å [Os(1)-Os(2)] to 2.8933(5)
Å [Os(2)-Os(3)], exhibiting a mean length of 2.8919 Å. The
bridging hydride ligand was not located but is assumed to bridge
the Os(1)-Os(2) vector as the hydride ligand in those triosmium
clusters possessing related µ2,3-bridging nitrogen ligands.5,24 The
N(1) and N(2) atoms are chelated to the Os(2) atom in what may
be formally viewed as equatorial and axial sites, respectively. The
Os(1)-C(20) bond distance of 2.13(1) Å is in concert with those
distances found in the ortho-metalated bpy ligand in HOs3(CO)9(µ-
Figure 2. Thermal ellipsoid plot of the triosmium cluster 4 showing
C10H7N2) and other Os3 clusters with metalated nitrogenous
the thermal ellipsoids at the 40% probability level.
ligands.5,25
II. Thermolysis and Photochemical Reactivity of Clusters
2 and 3. The possible interconversion between the isomeric
clusters 2 and 3 was next examined by independent thermolysis
and photolysis experiments. Here reductive coupling of the
hydride ligand with the methylene moiety (in 2) or the pyridyl
ring (in 3) could, in theory, furnish the common intermediate,
namely, the unsaturated 46e- cluster Os3(CO)9(6-Me-2,2′-bpy),
and allow us to probe the kinetic and thermodynamic relation-
ship between the two isomeric clusters. Thermolysis of 2 in
toluene at 75 °C led to the complete consumption of 2 and the
formation of two products, with TLC analysis confirming the
presence of cluster 3 as the major product of the mixture. Both
products were subsequently isolated, and the minor product
(cluster 4) was characterized by spectroscopic methods and
X-ray crystallography. The two doublets found at δ -10.81 and
-14.82 in the 1H NMR spectrum of 4 suggested the formation
of dihydride species through an additional C-H bond activation
process. That these hydrogens originated from the initial 6-Me
group was corroborated by the observation of seven bipyridyl
hydrogens over the region δ 9.09-6.97 in the 1H NMR
spectrum, of which the downfield doublet centered at δ 9.09
may be confidently ascribed to the 6′ hydrogen of the unsub-
stituted bipyridyl ring. The assignment of the remaining methine
resonance at δ 5.38 is straightforward and matches those NMR
data reported for the methine moieties in the structurally related
dihydrides H2Ru3(CO)8(µ3-HCN-NMe) (where N-N ) 2,2′-bpy,
1,10-phen).3 Photolysis of 2 at room temperature using near-
UV light also gives cluster 4 slowly as the sole observable
product, as shown in eq 5. Optical excitation in 2 promotes CO
loss, followed by activation of the methylene group of the
cyclometalated ring. Figure 2 shows the thermal ellipsoid plot
of cluster 4 and confirms the formal loss of CO and the C-H
(23) (a) Zoet, R.; van Koten, G.; Vrieze, K.; Jansen, J.; Goubitz, K.;
Stam, C. H. Organometallics 1988, 7, 1565. (b) Yeh, W.-Y.; Yang, C.-C.;
Peng, S.-M.; Lee, G.-H. J. Chem. Soc., Dalton Trans. 2000, 1649.
(24) (a) Day, M.; Espitia, D.; Hardcastle, K. I.; Kabir, S. E.; Rosenberg,
E.; Gobetto, R.; Milone, L.; Osella, D. Organometallics 1991, 10, 3550.
(b) Kabir, S. E.; Rosenberg, E.; Day, M.; Hardcastle, K.; Wolf, E.;
McPhillips, T. Organometallics 1995, 14, 721. (c) Azam, K. A.; Das, A. R.;
Hursthouse, M. B.; Kabir, S. E.; Malik, K. M. A. J. Chem. Crystallogr.
1998, 28, 283. (d) Wong, W.-Y.; Cheung, S.-H.; Huang, X.; Lin, Z. J.
Organomet. Chem. 2002, 655, 39.
(25) (a) Au, Y.-K.; Cheung, K.-K.; Wong, W. T. Inorg. Chim. Acta
1995, 238, 193. (b) Deeming, A. J.; Stchedroff, M. J.; Whittaker, C.; Arce,
A. J.; De Sanctis, Y.; Steed, J. W. J. Chem. Soc., Dalton Trans. 1999, 3289.
(c) Wong, W.-Y.; Cheung, S.-H.; Lee, S.-M.; Leung, S.-Y. J. Organomet.
Chem. 2000, 596, 36. (d) Clarke, L. P.; Cole, J. M.; Davies, J. E.; French,
A.; Koentjoro, O. F.; Raithby, P. R.; Shields, G. P. New J. Chem. 2005,
29, 145. (e) Machado, R. A.; Goite, M. C.; Arce, A. J.; De Sanctis, Y.;
Deeming, A. J.; D’Ornelas, L.; Oliveros, D. A. J. Organomet. Chem. 2005,
690, 622.

3024 Organometallics, Vol. 27, No. 13, 2008
Poola et al.
bond activation in cluster 2. Cluster 4 contains 48-valence
electrons and is electron precise. The capping of one of the
cluster faces by the six-electron 6-CH-2,2′-bpy ligand via both
nitrogen atoms and a µ2-carbene moiety and the presence of
eight terminal CO groups are confirmed. The three Os-Os bond
distances range from 2.7745(2) Å [Os(2)-Os(3)] to 2.9556(3)
Å [Os(1)-Os(2)], with the hydrides presumed to bridge the
Os(1)-Os(2) and Os(1)-Os(3) vectors in keeping with related
Os3 cluster analogues.2a,3,23a,25d
Figure 3. UV-vis spectral changes for cluster 2 in the presence
of P(OMe)3 (10 equiv) recorded at 333 K in toluene, with the inset
showing the absorbance versus time curve for the experimental data
of the first-order rate constant k.
concentration of added ligand supports a rate-limiting step that
involves only cluster 2. Although not examined in detail, the
effect of solvent on the reaction of 2 with added ligand is
assumed to be negligible, at least between toluene and CH2Cl2
given the similarity of the rate constants in these solvents. The
zero-order dependence on ligand and the Eyring activation
parameters [∆Hq ) 21.7(4) kcal/mol; ∆Sq ) -13(1) eu] clearly
negate rate-limiting steps based on a dissociative ligand loss
and an associative ligand attack on cluster 2.
The thermal and photochemical reactivity of cluster 3 exhibits
very different behavior in comparison to cluster 2. Thermolysis
of 3 in toluene at 75 °C revealed no perceptible changes when
monitored by IR and TLC analyses over the course of 24 h,
establishing cluster 3 as the thermodynamically more stable
cluster in the isomeric pair of monohydrides. The photochemical
sensitivity of 3 was also probed, and only a slow bleaching of
the cluster was observed by UV-vis spectroscopy over the
course of several days.
III. Kinetic Investigation on C-H Bond Reductive
Coupling in Cluster 2 and Release of the 6-Me-2,2′-bpy
Ligand. Wishing to probe the intermediacy of the cluster 1,1-
Os3(CO)10(6-Me-2,2′-bpy), the anticipated cluster species from
the reaction of 1 and 6-Me-2,2′-bpy, and its role in the C-H
bond activation that accompanies the formation of the kinetic
hydride 2, we performed a mechanistic study on the thermolysis
of 2 in the presence of trapping ligands. Here the successful
trapping of the formal product of C-H bond reductive coupling,
the unsaturated cluster 1,1-Os3(CO)9(6-Me-2,2′-bpy), would be
expected to afford 1,1-Os3(CO)9L(6-Me-2,2′-bpy) (where L )
trapping ligand), which could then provide insight into the C-H
bond oxidative coupling, as dictated by the principle of
microscopic reversibility.
The fate of 2 upon reaction with excess CO and P(OMe)3
1
was ascertained by a combination of IR and H NMR spec-
troscopies and independent syntheses of the suspected end
products. Heating 2 in CH2Cl2 or toluene solvent at 333 K under
6.8 atm CO led to the formation of parent cluster Os3(CO)12,
as determined by IR monitoring of the reaction solution.26
Parallel preparative experiments also produced Os3(CO)12 in
near quantitative yields (>95%) after chromatographic purifica-
tion. An analogous reaction between 2 and CO (6.8 atm) was
1
studied by H NMR spectroscopy in CDCl3 in the presence of
the internal standard ferrocene. Here samples that were analyzed
as a function of time verified the release of free 6-Me-2,2′-bpy
with a mass balance of 97%. The reaction of 2 with P(OMe)3
was also examined and spectroscopic evidence (IR and 1H and
31P NMR) was obtained for the liberation of 6-Me-2,2′-bpy and
formation of the cluster 1,2,3-Os3(CO)9[P(OMe)3]3, whose
synthesis and spectral properties have been reported earlier by
Pomeroy.27 Samples of 1,2,3-Os3(CO)9[P(OMe)3]3 that were
isolated from preparative reactions of 2 with P(OMe)3 matched
independently prepared samples of the Pomeroy cluster. The
outcome of the CO and P(OMe)3 trapping studies described here
is illustrated by eq 6.28,29
Kinetic studies for the reaction of 2 with excess P(OCD3)3
and PPh3 were also investigated by 1H NMR spectroscopy. Since
the concentration requirements for 2 in NMR kinetics are greater
by a factor of 102 compared to the UV-vis studies, we
employed CD2Cl2 rather than toluene-d8 as the NMR solvent
Our initial studies on the reactivity of 2 with various 2e-
donors were investigated by UV-vis spectroscopy over the
temperature range 318-338 K, with the progress of each
reaction monitored by following the decrease in the absorbance
of the 540 nm band belonging to 2. Figure 3 shows UV-vis
spectral changes for the reaction of 2 in toluene at 333 K in the
presence of a 10-fold excess of P(OMe)3, while the absorbance
versus time plot depicted in the inset reveals excellent agreement
between the least-squares fit for the first-order decay of 2 and
the absorbance data. The first-order rate constants for these
reactions are summarized in Table 3; the collective insensitivity
of the rate constant to different ligands and variation in the
(26) These carbonylation reactions were conducted at constant CO
pressure using a Fisher-Porter tube equipped with a tip tube for periodic
sampling.
(27) Alex, R. F.; Pomeroy, R. K. Organometallics 1987, 6, 2437.

Coordination and Rupture of Methyl C-H Bonds
Organometallics, Vol. 27, No. 13, 2008 3025
Table 3. Experimental Rate Constants for the Thermolysis of
Cluster 2 in the Presence of Ligand-Trapping Agentsa
entry
temp (K)
solvent
ligand trap
CO, 1 atm
CO, 1 atm
PPh3, 10 equiv
PPh3, 10 equiv
PPh3, 25 equiv
PPh3, 50 equiv
CO, 1 atm
CO, 1 atm
CO, 3.4 atm
CO, 6.8 atm
CO, 34 atm
PPh3, 10 equiv
PCy3, 10 equiv
P(OMe)3, 10 equiv
P(OMe)3, 10 equiv
P(OCD3)3,10 equiv
PMe3, 10 equiv
P(OCD3)3, 10 equiv
PPh3, 10 equiv
CO, 1 atm
105k (s-1
)
1
2
3
4
5
6
7
8
318
323
323
323
323
323
328
333
333
333
333
333
333
333
333
333
333
333
333
338
toluene
toluene
toluene
CH2Cl2
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
CH2Cl2
CH2Cl2
CH2Cl2
CD2Cl2
CD2Cl2
toluene
1.05 ( 0.02
1.72 ( 0.01
1.49 ( 0.04
2.70 ( 0.09
1.74 ( 0.01
2.31 ( 0.04
3.16 ( 0.03
5.02 ( 0.05
4.3 ( 0.2b
9
10
11
12
13
14
15
16
17
18
19
20
4.8 ( 0.1b
5.8 ( 0.4b
5.90 ( 0.03
4.65 ( 0.02
4.86 ( 0.04
5.72 ( 0.03
6.63 ( 0.05
8.0 ( 0.1
due to the higher solubility of cluster 2 in the former solvent.
Our choice of P(OCD3)3 as a trapping ligand was necessary in
order to eliminate the intense methoxy resonance from P(OMe)3
that overlapped and saturated the region containing the bipyridyl
methylene and methyl groups. A sealed NMR tube containing
2 and P(OCD3)3 was heated at 333 K, and the extent of the
reaction was analyzed by measuring the intensity changes of
the methylene hydrogens at δ 2.87. The reaction was kinetically
well-behaved, and a plot of ln[2] versus time afforded the rate
constant quoted in Table 3 (entry 18). A comparable rate
constant was also obtained for the consumption of 2 when the
high-field doublet of the 6′ hydrogen at δ 9.27 was employed
as the probe resonance and its intensity monitored relative to
the ferrocene reference. The release of free 6-Me-2,2′-bpy in
97% yield was verified by 1H NMR spectroscopy. The reaction
between 2 and PPh3 proceeded similarly (entry 19, Table 3)
with the liberation of 6-Me-2,2′-bpy, whose presence was
ascertained by the growth of the methyl singlet at δ 2.48. Under
no circumstances did we observe the buildup of any species
that could be reconciled with the formation of Os3(CO)9L(6-
Me-2,2′-bpy), the ligand-trapped product from the unsaturated
cluster Os3(CO)9(6-Me-2,2′-bpy).
The role played by reductive coupling of the hydride and
cyclometalated alkyl moiety in the rate-limiting step in the
conversion of 2 to Os3(CO)12 and 1,2,3-Os3(CO)9[P(OMe)3]3
was next explored through the synthesis of the isotopically
labeled heterocycle 6-Me-d3-2,2′-bpy. Treatment of 2,2′-bpy
with Me-d3-Li (>99% deuterium) in Et2O, followed by oxidative
workup of the dihydro adduct 6-methyl-d3-1,6-dihydro-2,2′-
bipyridine, gave the d3-isotopomer 6-Me-d3-2,2′-bpy as a
colorless oil. The 1H NMR spectrum of 6-Me-d3-2,2′-bpy
confirmed the absence of the high-field methyl resonance and
the presence of seven bipyridyl hydrogens, whose integrated
intensities and chemical shifts matched those of the d0 isoto-
pomer. The extent of deuteration was further corroborated by
GC-MS and the observation of a strong parent ion having a
m/z value of 173.0 (cf. the m/z value of 170.0 found for the
6-Me-d0-2,2′-bpy ligand recorded under identical conditions).
10.0 ( 0.4c
6.6 ( 0.2c
8.36 ( 0.04
a The UV-vis kinetic data were collected in the specified solvent
using a ca. 10-4 M solution of cluster 2 by following the decrease in
the absorbance of the 540 nm band. b Reaction carried out in an
autoclave at constant CO pressure using a ca. 10-4 M solution of cluster
2. c Rate constant determined by 1H NMR using a ca. 10-2 M solution
of 2 in CD2Cl2.
Table 4. Experimental Rate Constants for the Thermolysis of
Cluster 2-d3 in the Presence of Ligand-Trapping Agentsa
entry
temp (K)
solvent
ligand trap
CO, 1 atm
PPh3, 10 equiv
PPh3, 25 equiv
CO, 1 atm
105k (s-1
)
1
2
3
4
5
6
7
8
323
323
323
333
333
333
333
333
toluene
toluene
toluene
toluene
toluene
CH2Cl2
CH2Cl2
CD2Cl2
1.07 ( 0.01
1.12 ( 0.01
1.18 ( 0.03
2.55 ( 0.01
2.33 ( 0.09b
2.50 ( 0.02
3.24 ( 0.03
3.8 ( 0.2c
CO, 6.8 atm
P(OMe)3, 10 equiv
P(OCD3)3, 10 equiv
P(OCD3)3, 10 equiv
a The UV-vis kinetic data were collected in the specified solvent
using a ca. 10-4 M solution of cluster 2-d3 by following the decrease in
the absorbance of the 540 nm band. b Reaction carried out in an
autoclave at constant CO pressure using a ca. 10-4 M solution of cluster
2-d3. c Rate constant determined by 1H NMR using a ca. 10-2
solution of 2-d3 in CD2Cl2.
M
The desired cluster DOs3(CO)9(µ2-CD2N2C10H7) (2-d3) and
its isomeric isotopomer HOs3(CO)9(µ2-N2C11H6D3) (3-d3) were
prepared, isolated, and fully characterized in solution by IR and
NMR spectroscopies and by FAB mass spectrometry, as outlined
in the Experimental Section. Kinetic studies for the reaction of
2-d3 with CO, PPh3, and P(OMe)3-d0,3 were performed, and the
isotope effect on the reaction was evaluated at 323 and 333 K.
UV-vis analysis, as previously described, provided the rate
1
constants quoted in Table 4. A parallel H NMR experiment
using 2-d3 was carried out in CD2Cl2 at 333 K under pseudo-
first-order conditions with P(OCD3)3. The decay of the downfield
6′ hydrogen at δ 9.27 allowed the rate of the reaction to be
easily followed, and the rate constant (entry 8, Table 4) was
subsequently determined by graphical analysis. Using an average
value for the rate constants reported in Tables 3 and 4 leads to
a kinetic isotope effect (KIE) of 1.78 (323 K) and 2.09
(333 K).
Collectively taken, the kinetic data and the normal KIE for
the reaction of 2 with donor ligands strongly suggest a rate-
limiting step involving the reductive coupling of the hydride
ligand in 2 and the cyclometalated alkyl moiety. Scheme 1
outlines the likely sequence of events using CO as the trapping
ligand. It is proposed that C-H bond formation gives rise to a
(28) The kinetics for the reaction of cluster 2 and PPh3 (10 equiv) were
also monitored by 1H NMR spectroscopy (see entry 19, Table 3), and while
the experimentally measured rate constant matched the other rate data
reported in Table 3, the tris-substituted cluster, if formed, was not the major
osmium-containing material. Tentatively, we believe that two PPh3 ligands
add to the cluster, followed by an ortho metalation of one of the phosphine
ligands. This premise is based on the observation of two, overlapping high-
field doublets at ca. δ-15.80 in the 1H NMR spectrum. See ref 29 for
related triosmium clusters containing ortho-metalated PPh3 ligands from
the thermolysis of Os3(CO)12 and PPh3.
(29) (a) Gainsford, G. J.; Guss, J. M.; Ireland, P. R.; Mason, R.; Bradford,
C. W.; Nyholm, R. S. J. Organomet. Chem. 1972, 40, C70. (b) Bradford,
C. W.; Nyholm, R. S. J. Chem. Soc., Dalton Trans. 1973, 529.

3026 Organometallics, Vol. 27, No. 13, 2008
Poola et al.
Scheme 1
transient agostic complex where the methyl group functions as
a weakly bound donor ligand.30,31 Dissociation of the methyl
group from the cluster follows the rate-limiting step and
generates the unsaturated cluster Os3(CO)9(6-Me-2,2′-bpy),
which rapidly captures CO to afford the saturated decacarbonyl
cluster Os3(CO)10(6-Me-2,2′-bpy).32 The steric congestion in the
latter cluster provides the driving force for the rapid dissociation
of the 6-Me-2,2′-bpy ligand and uptake of two additional CO
groups.2a,33
III. Ligand Substitution Kinetics in Cluster 3 and
Formation of HOs3(CO)9[P(OMe)3](µ2-N2C11H9) Due to
Partial Dissociation of the Chelating Bipyridyl Ligand. The
reactivity of cluster 3 with CO and P(OMe)3 was investigated
in order to assess the ease of reductive coupling of the hydride
and the ortho-metalated pyridyl carbon. Preliminary experiments
quickly confirmed that the reaction between 3 and CO and
P(OMe)3 exhibited disparate rates, indicative of a ligand-
dependent process.34 The ligand dependence was unequivocally
established by using P(OMe)3 since the ligand concentrations
could be more easily adjusted and measured with greater
accuracy than CO. The reactions were explored by UV-vis
spectroscopy over the temperature range 333-356 K in toluene
and CH2Cl2 solvents, with reactions conducted in the former
solvent being faster by a factor of ca. 2×. The progress of each
reaction was followed by monitoring the decrease in the 540
nm absorbance band for 3, as illustrated in Figure 4 for the
absorbance changes for 3 in the presence of P(OMe)3 (26 equiv)
at 356 K in toluene. The pseudo-first-order rate constants quoted
in Table 5 were obtained from nonlinear regression analysis of
the absorbance versus time curve, as exemplified by the inset
accompanying Figure 4. Plots of the pseudo-first-order rate
constants (kobsd) as a function of P(OMe)3 were found to be
linear with intercepts of zero within experimental error, as
illustrated by Figure 5. The slopes of these plots afforded the
second rate constants shown in Table 6, which in turn yielded
the activations parameters ∆Hq ) 13.0(3) kcal/mol and ∆Sq )
-30(1) eu, which strongly support an associative mechanism
involving the attack of a ligand on cluster 3 in the rate-limiting
step.
(30) The proposed rate-limiting step involving cluster 2 and the
corresponding cluster containing an agostically bound methyl group (i.e., a
σ-bound methyl moiety), prior to the formal generation of a vacant
coordination site through dissociation of the methyl group, is consistent
with the current body of knowledge for alkane activation at mononuclear
compounds (see ref 31). An alternative rate-limiting step proceeding by a
direct reductive coupling of the C-H bond and formation of an unsaturated
cluster in a single step cannot be eliminated from consideration on the basis
of the data at hand. Differentiation of these two scenarios is thwarted by
our inability to isolate the carbonylation product Os3(CO)10(6-Me-2,2′-bpy)
in the reductive coupling step and in the initial reaction of 6-Me-2,2′-bpy
with cluster 1. Here the preparation of the isotopically substituted derivatives
Os3(CO)10(6-Me-d1,2-2,2′-bpy) would, in theory, help clarify the details
associated with the activation of the methyl group in the reductive coupling
and oxidative coupling steps.
(31) For reports and reviews on C-H bond activation in alkanes by
mononuclear compounds and the isotope techniques employed in the
mechanistic study of these reactions, see:(a) Buchanan, J. M.; Stryker, J. M.;
Bergman, R. G. J. Am. Chem. Soc. 1986, 108, 1537. (b) Parkin, G.; Bercaw,
J. Organometallics 1989, 8, 1172. (c) Bullock, R. M.; Headford, C. E. L.;
Hennessy, K. M.; Kegley, S. E.; Norton, J. R. J. Am. Chem. Soc. 1989,
111, 3897. (d) Jones, W. D. Acc. Chem. Res. 2003, 36, 140. (e) Churchill,
D. G.; Janak, K. E.; Wittenberg, J. S.; Parkin, G. J. Am. Chem. Soc. 2003,
125, 1403. (f) Lersch, M.; Tilset, M. Chem. ReV. 2005, 105, 2471. (g)
Northcutt, T. O.; Wick, D. D.; Vetter, A. J.; Jones, W. D. J. Am. Chem.
Soc. 2001, 123, 7257.
(32) In the absence of suitable trapping ligands the thermodynamically
more stable hydride-bridged cluster 3 and the dihydride-bridged cluster 4
are formed. The latter product is presumed to originate from cluster 2
through a slower CO loss pathway at elevated temperatures.
(33) Whether the displacement of the 6-Me-2,2′-bpy ligand from the
saturated Os3(CO)10(6-Me-2,2′-bpy) cluster by CO occurs by a dissociative
or associative process cannot be determined on the basis of the available
data.
(34) Using the published solubility data for CO in toluene, we estimate
the molar ratio of cluster 3 to CO (1 atm) at 333 K as 1:75. See: Basickos,
L.; Bunn, A. G.; Wayland, B. B. Can. J. Chem. 2001, 79, 854.

Coordination and Rupture of Methyl C-H Bonds
Organometallics, Vol. 27, No. 13, 2008 3027
Table 5. Experimental Rate Constants for the Thermolysis of
Cluster 3 in the Presence of Ligand-Trapping Agentsa
entry
temp (K)
solvent
ligand trap
CO, 1 atm
105kobsd (s-1
)
1
2
3
4
5
6
7
8
333
333
333
333
333
333
333
338
344
344
344
344
351
356
356
356
356
toluene
toluene
CH2Cl2
toluene
CH2Cl2
CD2Cl2
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
toluene
1.22 ( 0.04
2.92 ( 0.01
1.51 ( 0.01
6.83 ( 0.02
2.95 ( 0.01
18.8 ( 0.1b
12.13 ( 0.03
1.58 ( 0.01
2.50 ( 0.01
5.71 ( 0.04
12.88 ( 0.08
23.5 ( 0.2
P(OMe)3, 11 equiv
P(OMe)3, 11 equiv
P(OMe)3, 26 equiv
P(OMe)3, 26 equiv
P(OCD3)3, 11 equiv
P(OMe)3, 53 equiv
CO, 1 atm
9
CO, 1 atm
10
11
12
13
14
15
16
17
P(OMe)3, 11 equiv
P(OMe)3, 26 equiv
P(OMe)3, 53 equiv
CO, 1 atm
4.29 ( 0.02
6.22 ( 0.04
11.44 ( 0.05
26.5 ( 0.7
CO, 1 atm
P(OMe)3, 11 equiv
P(OMe)3, 26 equiv
P(OMe)3, 53 equiv
45.3 ( 0.6
a The UV-vis kinetic data were collected in the specified solvent
using a ca. 10-4 M solution of cluster 3 by following the decrease in
the absorbance of the 540 nm band. b Rate constant determined by 1H
NMR using a ca. 10-2 M solution of 3 in CD2Cl2.
Figure 4. UV-vis spectral changes for cluster 3 and P(OMe)3 (26
equiv) recorded at 356 K in toluene, with the inset showing the
absorbance versus time curve for the experimental data of the
Table 6. Ligand-Dependent Rate Constants for the Thermolysis of
Cluster 3 with P(OMe)3
psuedo-first-order rate constant kobsd
.
temp (K)
103k2 (M-1 s-1
)
333
344
356
7.08 ( 0.08
13.70 ( 0.02
26.0 ( 0.9
was also reflected in the 1H NMR spectrum, where two methoxy
doublets were observed δ 3.59 (30%) and 3.39 (70%). The other
1H resonances nicely matched those data recorded in the kinetics
experiment. The ESI mass spectrum revealed a parent ion at
m/e 1118.57 that is consistent with a cluster having the
composition HOs3(CO)9[P(OMe)3](µ2-N2C11H9). Taken col-
lectively, the kinetics, NMR, and mass spectral data support
the attack of a P(OMe)3 ligand on cluster 3 and release of the
methyl-substituted pyridine ring from the cluster to give the
phosphite-substituted cluster HOs3(CO)9[P(OMe)3](µ2-N2C11H9)
(5), which exists as a pair of isomers having a pendant or
dangling pyridyl group. This reaction is depicted in eq 7.
Figure 5. Plots of kobsd versus the concentration of P(OMe)3 in
toluene solution as a function of temperature for the reaction of
cluster 3 with ligand.
The kinetics for the reaction between 3 and P(OCD3)3 (11
equiv) were also studied by 1H NMR at 333 K. Monitoring the
intensity changes of the methyl singlet in 3 at δ 2.80 yielded
the pseudo-first-order constant of 18.8(1) × 10-5 s-1 that is
quoted in Table 5 (entry 6). The diminution of the methyl
resonance, as well as the hydride ligand at δ -21.48 in 3, was
accompanied by the growth of two new bipyridyl methyl singlets
at δ 2.54 (30%) and 2.43 (70%) and a pair of high-field hydrides
at δ -14.62 (30%) and -14.80 (70%). New aromatic resonances
at δ 7.55, 7.18-6.90, and 6.50 having an integral ratio of 1:4:
1, respectively, replaced the bipyridyl hydrogens of 3 as the
reaction proceeded to completion. A total of six aromatic
hydrogens in the new product indicates that the bipyridyl moiety
is bound to the cluster polyhedron by an ortho-metalated
bipyridyl moiety. The nature of the product formed from the
reaction of 3 with P(OMe)3 was subsequently established by
spectroscopic methods and mass spectrometry. The single
product isolated from a preparative reaction between 3 and
P(OMe)3 exhibited two 31P resonances at δ 114.08 (30%) and
116.25 (70%), confirming the presence of a P(OMe)3 ligand in
the products, whose relative ratios parallel the methyl and
hydride ratios found in the kinetics experiment. That two
P(OMe)3-containing clusters were present in the product mixture
Our formulation for cluster 5 is strengthened by an earlier
report from Lewis and co-workers, where the reaction of the

3028 Organometallics, Vol. 27, No. 13, 2008
Poola et al.
pyridyl-metalated cluster HOs3(CO)10(µ2-NC5H4) with P(OMe)3
was found to furnish HOs3(CO)9[P(OMe)3](µ2-NC5H4), as
depicted in eq 8.35,36 The close match between the IR and NMR
data for HOs3(CO)9[P(OMe)3](µ2-NC5H4) and cluster 5 supports
the existence of structurally similar pyridyl-bridged triosmium
clusters.
Conclusions
The reaction between the activated cluster Os3(CO)10(MeCN)2
and the ligand 6-Me-2,2′-bpy proceeds readily at room temperature
to give the isomeric hydride-bridged clusters HOs3(CO)9(µ2-
CH2N2C10H7) (2) and HOs3(CO)9(µ2-N2C11H9) (3). The kinetically
formed cluster 2 transforms into cluster 3 upon heating in an effort
to relieve the steric congestion that accompanies the chelation of
the R-diimine ligand and cyclometalation of the methyl group. The
driving force for the isomerization of 2 f 3 is currently under
computational investigation, and these data will provide the basis
of a future report. The kinetics for C-H bond reductive coupling
in 2 have been studied and data presented on the lability of the
putative cluster 1,1-Os3(CO)9L(6-Me-2,2′-bpy) [where L ) CO,
P(OMe)3], which readily releases the R-diimine ligand in the
presence of additional trapping ligand. The demonstration of
cluster-mediated activation of the R-methyl substituent in 6-Me-
2,2′-bpy, followed by release of the R-diimine ligand upon capture
of external trapping ligands by the cluster, provides the basis for
a potential synthetic protocol for side-chain modification of
heterocyclic ligands. The generality of the kinetic and thermody-
namic activation of different alkyl and aryl moieties in R-substituted
R-diimine ligands is currently under investigation by our groups.
Acknowledgment. Financial support from the NSF (C.J.C.)
and Robert A. Welch Foundation (Grant B-1093, M.G.R.)
is greatly appreciated. The NSF-MRI program grant CHE-
0320848 is gratefully acknowledged for support of the X-ray
diffraction facilities at San Diego State University.
(35) Ditzel, E. J.; Johnson, B. F. G.; Lewis, J. J. Chem. Soc., Dalton
Trans. 1987, 1293.
(36) The substitution site for the ancillary P(OMe)3 ligand in HOs3(CO)9-
[P(OMe)3](µ2-NC5H4) has been assigned by using NMR spectroscopy,
making use of the 2JP-H value and spectral comparisons to related clusters.
Here the specific locus of the phosphite ligand may be confidently assigned
to a site on the cluster that is cis to the bridging hydride ligand, with eq 8
showing one of four possible stereoisomeric products.
OM800020X