Acceptorless alkane dehydrogenation catalyzed by iridium CCC-pincer complexes

Article abstract of DOI:10.1021/om4006577

Iridium complexes of CCC-pincer bis-N-heterocyclic carbenes, including a newly synthesized trifluoromethyl-substituted complex, were examined as catalysts for the acceptorless dehydrogenation of cyclooctane and n-undecane. Up to 103 turnovers were observed for the dehydrogenation of cyclooctane, and up to 97 turnovers were observed for the dehydrogenation of n-undecane. The catalysts showed high initial turnover frequencies, followed by a gradual loss of activity over 24 h. Experiments indicate that this loss of activity is due to catalyst decomposition rather than product inhibition. Stoichiometric reactivity was investigated for the precatalysts, focusing on the synthesis of dihydride and trihydride complexes as well as the dissociation and addition of neutral ligands.

Full text of DOI:10.1021/om4006577

Article
pubs.acs.org/Organometallics
Acceptorless Alkane Dehydrogenation Catalyzed by Iridium CCC-
Pincer Complexes
Anthony R. Chianese,* Myles J. Drance, Kelsey H. Jensen, Samuel P. McCollom, Nevin Yusufova,
Sarah E. Shaner, Dimitar Y. Shopov, and Jennifer A. Tendler
Department of Chemistry, Colgate University, 13 Oak Drive, Hamilton, New York 13346, United States
S
* Supporting Information
ABSTRACT: Iridium complexes of CCC-pincer bis-N-heterocyclic carbenes, including a newly synthesized trifluoromethyl-
substituted complex, were examined as catalysts for the acceptorless dehydrogenation of cyclooctane and n-undecane. Up to 103
turnovers were observed for the dehydrogenation of cyclooctane, and up to 97 turnovers were observed for the dehydrogenation
of n-undecane. The catalysts showed high initial turnover frequencies, followed by a gradual loss of activity over 24 h.
Experiments indicate that this loss of activity is due to catalyst decomposition rather than product inhibition. Stoichiometric
reactivity was investigated for the precatalysts, focusing on the synthesis of dihydride and trihydride complexes as well as the
dissociation and addition of neutral ligands.
transfer dehydrogenation (up to 12 turnovers),8b but more
INTRODUCTION
■
active (up to 68 turnovers) for the acceptorless dehydrogen-
The selective, catalytic dehydrogenation of alkanes to give
ation of cyclooctane.8a A substantial steric effect was observed:
alkenes is a longstanding challenge in synthetic chemistry.1
mesityl- and 3,5-di-tert-butylphenyl-substituted variants were
Since the initial discoveries by Crabtree2 and Felkin3 more than
active, while 3,5-xylyl and 2,6-diisopropylphenyl-substituted
30 years ago, significant progress has been made. In 1996,
variants were inactive. Herein, we report more detailed studies
Kaska and Jensen reported that an iridium complex of a bulky
of the acceptorless dehydrogenation of alkanes, catalyzed by
PCP-pincer ligand was especially active for the transfer of
CCC-Ir complexes. Time-course studies show fast initial
hydrogen from cyclooctane to tert-butylethylene.4 Through
catalysis, which slows substantially over 24 h; the falloff arises
steric and electronic modification of the PCP ligand, turnover
numbers in the thousands have been observed for the iridium-
from catalyst decomposition rather than product inhibition. We
describe the stoichiometric reactivity of the CCC-Ir precata-
catalyzed transfer dehydrogenation of alkanes employing a
lysts, focusing on neutral ligand dissociation/addition and
sacrificial alkene acceptor.5 The acceptorless dehydrogenation
of alkanes, where strongly endothermic hydrogen removal is
reaction with the hydride donor LiHBEt3. We also report the
synthesis and characterization of an electronically modified,
CF3-substituted catalyst.
driven by continuously purging a refluxing solution with inert
gas, has also been achieved with high turnover numbers.5b,d,6
Though the PCP-iridium systems are by far the most active for
alkane dehydrogenation, it is noteworthy that a strongly
RESULTS AND DISCUSSION
■
Synthesis and Characterization of a CF3-Substituted
Precatalyst. Substitution of the para position of the aryl
fragment in PCP-pincer ligands has previously been shown to
have a substantial effect on the rate of iridium-catalyzed alkane
electron-withdrawing PCP-Ru catalyst has shown promise,
giving nearly 200 turnovers for the transfer dehydrogenation of
cyclooctane.7
Recently, we8 and others9 have begun developing the
chemistry of iridium complexes containing monoanionic
CCC-pincer bis-N-heterocyclic carbene ligands. We observed
that our CCC-Ir complexes were weakly active for alkane
Received: July 3, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/om4006577 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 1. Synthesis of CF3-Substituted Precatalyst10
Figure 1. ORTEP diagram of 1-CF3Mes, showing 50% probability ellipsoids. Hydrogen atoms are omitted for clarity; the iridium-bound hydride was
not located. Selected metric data (bond lengths in Å and angles in deg): Ir(1)−C(16), 1.958(3); Ir(1)−C(6), 2.026(3); Ir(1)−C(19), 2.030(3);
Ir(1)−Cl(2), 2.5012(7); Ir(1)−N(3), 2.101(2); C(19)−Ir(1)−C(16), 78.36(11); C(6)−Ir(1)−C(16), 78.70(11); C(6)−Ir(1)−C(19), 156.75(11).
dehydrogenation.5a By analogy with our previously described
synthetic route,8a we successfully synthesized an electronically
modified precatalyst as shown in Scheme 1. Palladium-catalyzed
Buchwald−Hartwig coupling followed by cyclization in acidic
triethylorthoformate gave the bis-benzimidazolium salt in good
yield. Metalation was accomplished by heating the bis-
benzimidazolium salt, [Ir(cod)Cl]2, and excess triethylamine
in acetonitrile. The precatalyst 1-CF3Mes was purified by flash
chromatography on silica gel followed by recrystallization.
compound.8a The two mesityl-substituted benzimidazol-2-
ylidene fragments are chemically equivalent. The nonequiva-
lence of the two ortho-methyl groups and meta-hydrogens on
each mesityl ring indicates that rotation about the C−N bond is
restricted. The iridium-bound hydride appears at −22.7 ppm,
and the presence of an acetonitrile molecule in the coordination
1
sphere is apparent in both the H and 13C NMR spectra.
NOESY indicates that the acetonitrile ligand is trans to the aryl
fragment of the pincer ligand, while the hydride is cis to the aryl
fragment.
1
By H and 13C NMR spectroscopy, 1-CF3Mes appears to
have Cs symmetry, and its structure is analogous to that
The structure of 1-CF3Mes was confirmed by X-ray
previously observed for the non-CF3-substituted parent
crystallography. Diffraction-quality crystals were obtained by
B
dx.doi.org/10.1021/om4006577 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
recrystallization from CH2Cl2/pentane. The ORTEP diagram is
presented in Figure 1. Overall, the structural parameters are
nearly identical to the non-CF3-substituted parent compound.8a
The coordination geometry at iridium is octahedral, distorted
by the small bite angles in the CCC-pincer ligand. The hydride
ligand was not located.
ation of cyclooctane. Control experiments indicated that
NaOtBu was needed to produce active catalysts; presumably
its function is to promote the reductive elimination of HCl,13
generating a reactive Ir(I) intermediate capable of entering the
catalytic cycle14 via C−H oxidative addition. 1-Mes and 1-
CF3Mes were similarly active, giving 103 and 84 turnovers after
22 h, respectively. In contrast, 1-dtbp was less active, producing
only 35 turnovers in 22 h. Due to the challenges inherent in
controlling the vigorousness of the reflux and the argon flow
rate, acceptorless dehydrogenation experiments are challenging
to reproduce precisely. Each cyclooctane dehydrogenation
experiment was repeated twice, and the average relative
standard deviation in the final TON across the three iridium
complexes was 29%. This implies that the activities of 1-Mes
and 1-CF3Mes are indistinguishable, but that 1-dtbp is clearly
less active.
Acceptorless Dehydrogenation of Cyclic Alkanes. The
precatalysts employed in this study are shown in Figure 2. In
Cyclodecane (bp 201 °C) is employed more frequently in
studies of acceptorless dehydrogenation. Due to its higher
boiling point, significantly greater rates and turnover numbers
(up to 3050)15 have been observed for the acceptorless
dehydrogenation of cyclodecane catalyzed by PCP-Ir catalysts.
Despite this precedent, we found 1-Mes to be similarly active
for the acceptorless dehydrogenation of cyclooctane and
cyclodecane. After 22 h at reflux, 109 turnovers of cyclodecenes
were produced, with a nearly constant cis:trans ratio of 1:5
(Figure 4).
Figure 2. Precatalysts examined for acceptorless dehydrogenation of
alkanes.
addition to the newly synthesized 1-CF3Mes, five previously
synthesized CCC-Ir complexes8 were tested. In all of our trials,
we found 1-dipp, 2-tBu, and 2-Ad to be inactive for alkane
dehydrogenation. We found that 1-Mes, 1-CF3Mes, and 1-
dtpb, when activated with NaOtBu, formed active catalysts for
the acceptorless dehydrogenation of a variety of alkanes.
Cyclooctane (bp 149 °C) is a common model substrate for
alkane dehydrogenation catalysis, due to its unusually low
enthalpy of dehydrogenation (+24.3 kcal/mol)11 and the
formation of only one product, cis-cyclooctene. Although
TONs in the thousands have been recorded for transfer
dehydrogenation of cyclooctane,5a,c,d,6b to our knowledge the
highest TON for its acceptorless dehydrogenation is 190 after
120 h, catalyzed by a tert-butyl-substituted PCP-Ir complex.12
Figure 3 shows plots of catalytic turnover for the dehydrogen-
Figure 4. Time course plot for acceptorless dehydrogenation of
cyclodecane. Reaction conditions: cyclodecane (3.0 mL, 19 mmol), 1-
Mes (1.5 μmol), NaOtBu (7.5 μmol), refluxing under a stream of Ar.
Test for Product Inhibition. The approximately 100
turnovers observed for cyclooctane dehydrogenation represent
a conversion of only 1.3%. In order to determine whether the
falloff in activity arises from catalyst decomposition or
inhibition by the buildup of product alkene, we compared the
dehydrogenation of cyclooctane catalyzed by 1-Mes/NaOtBu
in the presence or absence of 0.40 mmol of added cis-
cyclooctene, equivalent to 100 catalytic turnovers. Nearly
identical reactivity was observed, indicating that product
inhibition is not the major source of the loss of activity.
Cyclooctane Dehydrogenation under Different At-
mospheres. Alkane dehydrogenation catalyzed by di-tert-
butylphosphino-substituted PCP-Ir complexes has been
reported to be strongly inhibited by dinitrogen, due at least
in part to strong binding of N2 to the PCP-Ir fragment.16 In
contrast, a bis-trifluoromethylphosphino-substituted PCP-Ir
complex showed no such inhibition, giving the same reactivity
under an Ar or N2 atmosphere.6b We compared the activity of
Figure 3. Time course plots for acceptorless cyclooctane dehydrogen-
ation. Reaction conditions: cyclooctane (8.0 mL, 30 mmol), iridium
complex (4.0 μmol), NaOtBu (20 μmol), refluxing under a stream of
Ar.
C
dx.doi.org/10.1021/om4006577 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
1-Mes for cyclooctane dehydrogenation under a flow of argon,
dinitrogen, and air. Catalyst activity was indistinguishable under
nitrogen and argon. Under air, almost all catalyst activity was
destroyed: small amounts of cyclooctene (up to 5 turnovers)
were observed along with other unidentified products, but the
time course was not reproducible. On the suggestion of a
reviewer, we tested the reversibility of the catalyst deactivation
under air by running the reaction under air for 1 h, then
switching the atmosphere to argon for the remaining 19 h. In
this experiment, catalyst activity was not restored, indicating
that oxygen irreversibly deactivates the catalyst.
Acceptorless Dehydrogenation of n-Undecane. The
dehydrogenation of linear alkanes is particularly challenging,
especially with respect to product selectivity. PCP-Ir catalysts
have shown kinetic selectivity for terminal alkenes, which
erodes with longer reaction times due to competing alkene
isomerization catalysis.5d,15,17 In trials of PCP-Ir-catalyzed
acceptorless dehydrogenation of linear alkanes, up to 108
turnovers were observed after 24 h for n-undecane dehydrogen-
ation (bp 196 °C),15 and up to 71 turnovers were observed
after 72 h for n-dodecane (bp 216 °C).18 Our CCC-Ir catalysts
show comparable activity for the acceptorless dehydrogenation
of n-undecane, as depicted in Figure 5. Notably, 1-dtpb, which
fast alkene isomerization catalysis we have documented for
these systems.8b
Reaction with Hydride Donors. Although in situ
activation of Ir(III) hydridochloride complexes with NaOtBu
has been successfully employed to generate active catalysts for
alkane dehydrogenation,5a one-component catalysts consisting
of dihydride or tetrahydride complexes have been more
commonly employed. Notably, selective formation of 1-alkenes
via dehydrogenation of n-alkanes has been observed only using
one-component catalysts.5d,15,17 PCP-Ir-HCl complexes have
been converted to dihydride, trihydride, and tetrahydride
complexes using LiEt3BH/H2,5d,17b,19 NaH/H2,20 and NaOt-
Bu/H2.20b
By addition of LiEt3BH or NaEt3BH to 1-Mes or 2-Ad, we
observed the formation of compounds tentatively assigned by
1H NMR spectroscopy as the dihydride complexes 3-Mes and
3-Ad and the anionic trihydride complexes 4-Mes and 4-Ad
(Scheme 2). Treatment of a dilute solution of 1-Mes in
benzene-d6 with 1.2 equiv of LiEt3BH gave nearly complete
conversion to 3-Mes, while addition of 5 equiv of LiEt3BH
primarily gave 4-Mes. Addition of a large excess of NaEt3BH to
a THF-d8 solution of 2-Ad gave clean formation of 4-Ad.
Attempts to add 1 equiv of LiEt3BH to 2-Ad in THF-d8 gave
only mixtures of the reactant 2-Ad and the trihydride 4-Ad,
presumably due to the low solubility of 2-Ad. However,
addition of an excess of LiEt3BH to a suspension of 2-Ad in
benzene gave nearly complete conversion to the dihydride 3-
Ad; presumably its low solubility prevents further reaction to
form 4-Ad in benzene. Attempts to isolate samples of the
dihydrides 3-Mes and 3-Ad for use as one-component catalysts
were unsuccessful, apparently due to decomposition under
vacuum.
1
By H NMR, the trihydride species 4-Mes and 4-Ad occupy
the expected octahedral geometry, showing C2v symmetry in
solution. The hydride ligands appear in a 1:2 ratio. In both
cases, the NHC’s N-substituents act as stereochemical
reporters: equivalent mesityl ortho-methyl groups (4-Mes) or
equivalent adamantyl methylene hydrogens (4-Ad) indicate the
existence of a mirror plane coincident with the plane of the
CCC-pincer backbone, consistent with the structures depicted
in Scheme 2.
Figure 5. Time course plots for acceptorless n-undecane dehydrogen-
ation. Reaction conditions: n-undecane (8.0 mL, 38 mmol), iridium
complex (4.0 μmol), NaOtBu (20 μmol), refluxing under a stream of
Ar.
The dihydride species show an interesting structural
dichotomy, the origin of which is not obvious. Compound 3-
Mes possesses C2v symmetry in solution: only one hydride
resonance is observed, and the four ortho-methyl groups on the
mesityl N-substituents are chemically equivalent. This is
consistent with the trans arrangement of hydride ligands
shown in Scheme 2. On the suggestion of a reviewer, we
showed the lowest activity for cyclooctane dehydrogenation, is
the most active for n-undecane, giving 97 turnovers after 22 h.
Each n-undecane dehydrogenation experiment was repeated
twice, and the average relative standard deviation in the final
TON across the three iridium complexes was 9%. Therefore,
the greater activity observed for 1-dtbp as compared to 1-Mes
and 1-CF3Mes can be expressed with a moderate amount of
confidence. It is noteworthy that 1-dtbp remains active for the
entire course of the reaction here, whereas it loses activity more
rapidly in the dehydrogenation of cyclooctane, which occurs at
a lower reflux temperature; we cannot rationalize this
observation with confidence. In contrast to PCP-Ir catalysts,
which produced a mixture of undecenes containing 44−56% 1-
undecene after 8 h via acceptorless dehydrogenation of n-
undecane,5d our catalysts produce only internal undecene
isomers at all time-points. This result is unsurprising given the
1
recorded the H NMR spectrum at temperatures between 22
and −75 °C to examine the possibility that the symmetric
spectrum observed at room temperature for 3-Mes might result
from a rapidly exchanging complex of lower symmetry. The
spectrum was unchanged over this temperature range,
indicating that the most likely structure is a trans-dihydride.
In contrast, two distinct hydride resonances are observed at
room temperature for 3-Ad. The observation of diastereotopic
adamantyl methylene hydrogens beta to nitrogen supports the
less symmetrical geometry shown in Scheme 2. It is not clear
why this geometry would be favored, but one possibility is that
THF binds reversibly to the site trans to hydride in the
observed isomer of 3-Ad, but is too bulky to bind in the site
trans to Caryl, as would be required in a trans-dihydride isomer.
D
dx.doi.org/10.1021/om4006577 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Scheme 2. Reaction of CCC-Ir Hydridochloride Complexes with LiEt3BH or NaEt3BH
Ligand Dissociation/Addition Reactions. As we have
previously described,8b CCC-Ir hydridochloride complexes
have been isolated either as six-coordinate, octahedral,
acetonitrile-solvated species or as five-coordinate, square-
pyramidal species, depending on the shape and steric bulk of
the N-substituents (see Figure 2). Braunstein9b,c and Hollis9a
have also observed that CCC-pincer complexes with small (n-
butyl) N-substituents tend to favor the formation of six-
coordinate, octahedral iridium complexes. Because catalytic
competence likely depends on the accessibility of the
coordination site trans to the aryl fragment of the pincer
ligand, we decided to examine the reactivity of 1-Mes and 2-Ad
for ligand dissociation and ligand addition, respectively.
Upon treatment with up to 500 equiv of acetonitrile or
1
pyridine in CD2Cl2, the H NMR spectrum of 2-Ad shows no
change, indicating that even weak binding of those ligands is
not detectable.
To assess whether a five-coordinate structure is accessible for
1-Mes, a sample was heated to 200 °C under vacuum
overnight. The resulting brown solid was completely insoluble
in noncoordinating solvents, but dissolved readily in CH2Cl2 to
give yellow 5-Mes, where the acetonitrile ligand has been
replaced by dichloromethane (Scheme 3). The dichloro-
methane adduct 5-Mes has been characterized spectroscopically
and crystallographically (Figure 6). The formation of 5-Mes
provides strong evidence that acetonitrile was removed by
heating in a vacuum. Although the composition of the
intermediate brown solid has not been determined, its
Figure 6. ORTEP diagram of 5-Mes, showing 30% probability
ellipsoids. Hydrogen atoms are omitted for clarity; the iridium-bound
hydride was not located. Selected metric data (bond lengths in Å and
angles in deg): Ir(1)−C(6), 2.028(6); Ir(1)−C(9), 1.954(6); Ir(1)−
C(12), 2.037(6); Ir(1)−Cl(2), 2.4842(15); Ir(1)−Cl(3), 2.5328(16);
Cl(3)−C(4), 1.787(7); C(4)−Cl(5), 1.768(7); C(12)−Ir(1)−C(9),
79.2(2); C(6)−Ir(1)−C(9), 78.6(2); Ir(1)−Cl(3)−C(4), 109.3(2);
Cl(3)−C(4)−Cl(5), 109.8(4).
Scheme 3. Synthesis of the Dichloromethane Adduct 5-Mes
insolubility in noncoordinating solvents likely indicates a
polymeric structure.
Solution characterization of 5-Mes by NMR spectroscopy
was performed in CD2Cl2, in which 5-Mes is highly soluble. In
CD2Cl2, no separate resonance for iridium-bound CH2Cl2 was
observed, consistent with fast exchange with the solvent.
Addition of 1 drop of acetonitrile to a CD2Cl2 solution of 5-
Mes results in rapid, complete conversion to 1-Mes. When 5-
Mes is dissolved in benzene-d6, a sharp 2H singlet at 4.35 ppm
is observed, corresponding to bound dichloromethane (free
E
dx.doi.org/10.1021/om4006577 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
dichloromethane appears at 4.27 ppm in benzene-d6).21 When
2 equiv of dichloromethane is added to this solution, one sharp
signal for dichloromethane is observed at 4.32 ppm, the signals
for the iridium complex shift slightly, and the hydride resonance
is not observed. The hydride signal is likely broadened into the
baseline; heating to 80 °C did not allow its observation. This
indicates that free and bound dichloromethane exchange
rapidly on the NMR time scale at room temperature and that
5-Mes is likely in rapid exchange with another, unidentified
complex under these conditions, potentially an adduct of
benzene or adventitious water.
We are aware of only two published crystal structures
containing neutral dichloromethane bound to iridium; both are
Ir(III) complexes. The Ir−Cl distance of 2.5328 Å in 5-Mes can
be compared with Ir−Cl distances of 2.462 Å in Cp*Ir(PMe3)-
(CH3)(CH2Cl2)22 and 2.550 Å in IrH2(PR3)2(OTf)-
(CH2Cl2).23 In both the previously known examples, the
bound dichloromethane is significantly distorted. The C−Cl
bonds proximal to iridium are 1.820 and 1.830 Å, respectively,
while the C−Cl bonds distal to iridium are 1.730 and 1.695 Å,
respectively. In 5-Mes, this distortion is much less pronounced:
the proximal C−Cl bond is 1.787 Å, while the distal C−Cl
bond is 1.768 Å. As this distortion is likely induced by π-
donation from a filled d-orbital on iridium into the C−Cl σ*-
orbital, it appears that the iridium atom in 5-Mes is less π-basic
than in the previously described complexes.
Iridium Complex 1-CF3Mes. The bis-benzimidazolium salt shown
in Scheme 1 (1.77 mmol, 1.22 g), [Ir(cod)Cl]2 (0.872 mmol, 596 mg),
and triethylamine (53 mmol, 7.4 mL) were added to an oven-dried,
medium-walled pressure vessel in the glovebox. The mixture was
stirred at 100 °C for 24 h behind a blast shield and then allowed to
cool to room temperature. The flask was opened to air, the solvent was
evaporated, and the residue was purified by chromatography on silica
gel, eluting with a gradient of 0−20% ethyl acetate in dichloromethane.
After removing the solvent under vacuum, the complex was analytically
pure. Yield: 1.08 g, 61%. Recrystallization for catalytic trials was
performed at room temperature in an argon-filled glovebox by layering
1
a solution in dichloromethane with pentane. H NMR (CD2Cl2): δ
3
3
8.20 (d, 2H, JHH = 8.1 Hz); 7.89 Hz (s, 2H); 7.49 (t, 2H, JHH = 7.6
3
Hz), 7.32 (t, 2H, JHH = 7.7 Hz), 7.07 (s, 2H); 7.02 (s, 2H); 6.98 (d,
3
2H, JHH = 7.9 Hz); 2.35 (s, 6H); 2.22 (s, 6H); 1.96 (s, 6H); 1.45 (s,
3H); −22.7 (s, 1H). 13C NMR: δ 186.52; 152.87; 146.82; 139.18;
138.66; 137.53; 135.98; 133.31; 132.64; 129.45; 128.86; 126.08 (q,
1JCF = 271.4 Hz); 124.08; 123.51; 123.44 (q, 2JCF = 31.6 Hz); 116.56;
3
111.39; 111.06; 105.17 (q, JCF = 4.1 Hz); 21.27; 18.38; 18.08; 2.91.
19F NMR: δ −60.3. Anal. Calcd for C41H36ClF3IrN5: C, 55.74; H, 4.11;
N, 7.93. Found: C, 55.53; H, 4.00; N, 7.88.
Acceptorless Dehydrogenation of Cyclooctane. An oven-
dried 25 mL round-bottom flask was charged with an iridium complex
(4.0 μmol), sodium tert-butoxide (20 μmol, 1.9 mg), and a stir-bar. A
reflux condenser was attached, and the system was evacuated and
backfilled with argon three times. Degassed cyclooctane (59 mmol, 8.0
mL) was added via syringe. The solution was refluxed while stirring
and purging continuously with argon for 22 h. Aliquots were removed
1
for analysis via cannula and analyzed by H NMR in CDCl3 (100 μL
reaction mixture in 700 μL of CDCl3), comparing the integration of
cyclooctene product with that of the cyclooctane reactant. The results
in Figure 3 represent an average of two runs for each iridium complex.
Acceptorless Dehydrogenation of Cyclodecane. The proce-
dure was analogous to that for cyclooctane, using 1-Mes (1.22 mg, 1.5
μmol), sodium tert-butoxide (7.5 μmol, 0.7 mg), and degassed
cyclodecane (19 mmol, 3.0 mL).
Acceptorless Dehydrogenation of n-Undecane. The proce-
dure was analogous to that for cyclooctane, using an iridium complex
(4.0 μmol), sodium tert-butoxide (20 μmol, 1.9 mg), and degassed n-
undecane (38 mmol, 8.0 mL).
SUMMARY
■
We have synthesized a new CF3-substituted CCC-pincer ligand
and its acetonitrile-solvated iridium hydridochloride complex,
1-CF3Mes. We have studied the acceptorless dehydrogenation
of alkanes catalyzed by 1-CF3Mes and previously reported
CCC-Ir complexes. Three of these complexes, 1-CF3Mes, 1-
Mes, and 1-dtbp, are active for acceptorless alkane dehydrogen-
ation when activated by NaOtBu, giving up to 103 catalytic
turnovers. Rapid initial catalysis followed by a falloff in rate over
24 h is due to catalyst decomposition. Potentially useful CCC-
Ir dihydride complexes 3-Mes and 3-Ad were synthesized by
reduction of hydridochloride precursors, but instability under
vacuum prevented their isolation for catalytic trials. A study of
neutral ligand substitution on 1-Mes and ligand addition to 2-
Ad underscores the substantial effect of the varying steric bulk
of the N-substituents. For the six-coordinate 1-Mes, removal of
the acetonitrile ligand was accomplished by heating under
vacuum to give a material that is likely polymeric, but cleanly
gives the solvent adduct 5-Mes on dissolution in dichloro-
methane. In contrast, the five-coordinate 2-Ad is resistant to
addition of a neutral ligand trans to Caryl, due to the increased
steric bulk of the adamantyl N-substituent as opposed to
mesityl.
Iridium Complex 3-Mes. Compound 1-Mes (1.0 mg, 1.2 μmol)
was dissolved in 1 mL of benzene-d6 in the glovebox, and LiEt3BH
(1.0 M in THF, 1.5 μL, 1.5 μmol) was added. The solution was
transferred to a J-Young NMR tube. 1H NMR (C6D6): δ 8.04 (d, 2H,
3
3
3JHH = 8.1 Hz); 7.82 (d, 2H, JHH = 7.8 Hz); 7.52 (t, 1H, JHH = 7.8
Hz); 7.21 (t, 3JHH = 7.3 Hz); 7.07 (t, 2H, 3JHH = 7.6 Hz); 6.86 (d, 2H,
3JHH = 7.9 Hz); 6.75 (s, 4H); 2.26 (s, 12H); 2.11 (s, 6H); 0.94 (s, 3H);
−9.21 (s, 2H).
Iridium Complex 3-Ad. Compound 2-Ad (2.5 mg, 3.1 μmol) was
suspended in 1 mL of benzene-d6 in the glovebox, and LiEt3BH (1.0 M
in THF, 25 μL, 25 μmol) was added. The yellow suspension was
stirred at room temperature overnight, during which time the yellow
undissolved solid became white. The mixture was filtered, and the solid
residue was washed with a small amount of benzene-d6, dissolved in
THF-d8, and transferred to a J-Young NMR tube. 1H NMR (THF-d8):
δ 8.32 (d, 2H, 3JHH = 8.0 Hz); 7.81 (d, 2H, 3JHH = 7.9 Hz); 7.28−7.33
3
2
EXPERIMENTAL SECTION
(m, 3H); 7.19 (t, 2H, JHH = 7.3 Hz); 2.45 (d, 6H, JHH = 12.5 Hz);
■
2
2
2.27 (d, 6H, JHH = 12.5 Hz); 1.72 (s, 6H); 1.59 (d, 6H, JHH = 11.3
Hz); −8.15 (s, 1H); −16.41 (s, 1H). The 6H resonance at ∼1.80 ppm
for the adamantyl CH2 group delta to nitrogen was overlapped with
the THF signal.
General Methods. [Ir(cod)Cl]2 was prepared as previously
described.24 All other materials were commercially available and
were used as received, unless otherwise noted. Solvents were purified
by sparging with argon and passing through columns of activated
alumina, using an MBraun solvent purification system. Flash
chromatography using solvent gradients was performed using a
Combiflash RF system. NMR spectra were recorded at room
temperature on a Bruker spectrometer operating at 400 MHz (1H
NMR), 100 MHz (13C NMR), and 376 MHz (19F NMR). Elemental
analyses were performed by Robertson Microlit (Madison, NJ, USA).
Images of NMR spectra, along with procedures for ligand synthesis
and characterization, are given in the Supporting Information.
Iridium Complex 4-Mes. Compound 1-Mes (5.0 mg, 6.2 μmol)
was suspended in 1 mL of benzene-d6 in the glovebox, and LiEt3BH
(1.0 M in THF, 31 μL, 31 μmol) was added. The solution was
transferred to a J-Young NMR tube. 1H NMR (C6D6): δ 8.06 (d, 3JHH
3
3
= 8.1 Hz); 7.90 (d, 2H, JHH = 7.8 Hz); 7.58 (t, 1H, JHH = 7.8 Hz);
7.22 (t, 2H, 3JHH = 8.1 Hz); 7.04 (t, 2H, 3JHH = 7.9 Hz); 6.94 (s, 4H);
6.80 (d, 2H, JHH = 8.1 Hz); 2.32 (s, 6H); 2.17 (s, 12H); −12.15 (t,
1H, JHH = 5.8 Hz); −13.17 (d, 2H, JHH = 5.8 Hz).
3
2
2
F
dx.doi.org/10.1021/om4006577 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
Iridium Complex 4-Ad. Compound 2-Ad (7.0 mg, 8.7 μmol) was
suspended in 0.75 mL of THF-d8 in the glovebox, and NaEt3BH (23
mg, 170 μmol) was added. The mixture was stirred at 25 °C until clear
AUTHOR INFORMATION
Corresponding Author
*E-mail: achianese@colgate.edu.
Notes
■
1
and then transferred to a J-Young NMR tube. H NMR (THF-d8): δ
8.28 (d, 2H, 3JHH = 8.2 Hz); 8.10 (d, 2H, 3JHH = 8.6 Hz); 7.83 (d, 2H,
3
3
3JHH = 7.8 Hz); 7.18 (t, 2H, JHH = 7.8 Hz); 7.14 (t, 1H, JHH = 7.8
Hz); 7.04 (t, 2H, 3JHH = 7.8 Hz); 3.57 (s, 12H); 2.29 (s, 6H); 2.22 (d,
6H, 2JHH = 11.9 Hz); 1.76 (d, 6H, 2JHH = 11.9 Hz); −7.80 (t, 1H, 2JHH
= 3.2 Hz); −12.28 (d, 2H, 2JHH = 3.2 Hz). The resonance at 3.57 ppm
for the adamantyl CH2 groups beta to nitrogen was overlapped with
the residual THF-d8 signal.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank the National Science Foundation (Grant No. CHE-
1057792) for financial support.
Iridium Complex 5-Mes. Compound 1-Mes (25 mg, 31 μmol)
was heated under vacuum (10 mTorr) at 200 °C for 16 h, resulting in
the formation of a brown solid. This solid was recrystallized under
argon by layering a dichloromethane solution with pentane. Yield: 16
REFERENCES
■
(1) Choi, J.; MacArthur, A. H. R.; Brookhart, M.; Goldman, A. S.
Chem. Rev. 2011, 111, 1761−1779.
(2) (a) Crabtree, R. H.; Mihelcic, J. M.; Quirk, J. M. J. Am. Chem. Soc.
1979, 101, 7738−7740. (b) Burk, M. J.; Crabtree, R. H.; Parnell, C. P.;
Uriarte, R. J. Organometallics 1984, 3, 816−817.
1
3
mg, 60%. H NMR (C6D6): δ 7.74 (d, 2H, JHH = 8.4 Hz); 7.45 (d,
3
3
3
2H, JHH = 8.0 Hz); 7.26 (t, 1H, JHH = 8.0 Hz); 7.07 (t, 2H, JHH
=
=
7.9 Hz); 6.92 (t, 2H, 3JHH = 7.9 Hz); 6.82 (s, 2H); 6.71 (d, 2H, 3JHH
8.0 Hz); 6.50 (s, 2H); 4.35 (s, 2H); 2.31 (s, 6H); 2.04 (s, 6H); 1.88 (s,
6H). The iridium-bound hydride was not observed in C6D6. 1H NMR
(CD2Cl2): δ 8.18 (d, 2H, 3JHH = 8.4 Hz); 7.68 (d, 2H, 3JHH = 8.0 Hz);
(3) (a) Baudry, D.; Ephritikhine, M.; Felkin, H. J. Chem. Soc., Chem.
Commun. 1980, 1243−1244. (b) Baudry, D.; Ephritikhine, M.; Felkin,
H. J. Chem. Soc., Chem. Commun. 1982, 606−607. (c) Baudry, D.;
Ephritikhine, M.; Felkin, H.; Holmes-Smith, R. J. Chem. Soc., Chem.
Commun. 1983, 788−789.
3
3
7.47 (t, 2H, JHH = 7.6 Hz); 7.34 (t, 1H, JHH = 8.0 Hz); 7.31 (t, 2H,
3JHH = 7.7 Hz); 7.07 (s, 2H); 7.05 (s, 2H); 6.98 (d, 2H, JHH = 8.0
3
Hz); 2.39 (s, 6H); 2.18 (s, 6H); 1.97 (s, 6H); −22.68 (s, 1H). 13C
NMR (CD2Cl2): δ 188.76; 145.80; 140.29; 137.62; 137.39; 136.16;
132.80; 132.62; 129.73; 129.52; 124.05; 123.37; 121.94; 111.47;
111.16; 108.42; 21.34; 18.17; 18.04. The iridium-bound Caryl was not
observed, possibly due to broadening caused by exchange of the
CD2Cl2 ligand. Anal. Calcd for C39H36Cl3IrN4: C, 54.51; H, 4.22; N,
6.52. Found: C, 56.90; H, 4.52; N, 6.95. Repeated analyses gave similar
results, which may result from partial loss of dichloromethane upon
storage under vacuum.
(4) Gupta, M.; Hagen, C.; Flesher, R. J.; Kaska, W. C.; Jensen, C. M.
Chem. Commun. 1996, 2083−2084.
(5) (a) Gottker-Schnetmann, I.; White, P.; Brookhart, M. J. Am.
̈
Chem. Soc. 2004, 126, 1804−1811. (b) Morales-Morales, D.; Redon,
R.; Yung, C.; Jensen, C. M. Inorg. Chim. Acta 2004, 357, 2953−2956.
(c) Kuklin, S. A.; Sheloumov, A. M.; Dolgushin, F. M.; Ezernitskaya,
M. G.; Peregudov, A. S.; Petrovskii, P. V.; Koridze, A. A.
Organometallics 2006, 25, 5466−5476. (d) Kundu, S.; Choliy, Y.;
Zhuo, G.; Ahuja, R.; Emge, T. J.; Warmuth, R.; Brookhart, M.; Krogh-
Jespersen, K.; Goldman, A. S. Organometallics 2009, 28, 5432−5444.
(6) (a) Liu, F. C.; Goldman, A. S. Chem. Commun. 1999, 655−656.
(b) Adams, J. J.; Arulsamy, N.; Roddick, D. M. Organometallics 2012,
31, 1439−1447.
X-ray Crystallography, General Methods. Structure determi-
nations were performed on an Oxford Diffraction Gemini-R
diffractometer, using Mo Kα radiation (1-CF3Mes) or Cu Kα
radiation (5-Mes). Single crystals were mounted on Hampton
Research Cryoloops using Paratone-N oil. Unit cell determination,
data collection and reduction, and empirical absorption correction
were performed using the CrysAlisPro software package.25 Direct
methods structure solution was accomplished using SIR92,26 and full-
matrix least-squares refinement was carried out using CRYSTALS.27
All non-hydrogen atoms were refined anisotropically. Unless otherwise
noted, hydrogen atoms were placed in calculated positions, and their
positions were initially refined using distance and angle restraints. All
hydrogen positions were fixed in place for the final refinement cycles.
X-ray Structure Determination of 1-CF3Mes. X-ray quality
crystals of 1-CF3Mes were grown by layering a dichloromethane
solution with pentane. Minor conformational disorder was present for
the CF3 group, but accounting for this disorder did not improve the
agreement. Highly disordered solvent was present; correction for this
residual density was performed using the option SQUEEZE in the
program package PLATON.28 A total of 83 electrons per unit cell were
removed, from a total potentially solvent-accessible void of 568.0 Å3.
The iridium-bound hydride was not located in the difference map.
X-ray Structure Determination of 5-Mes. X-ray quality crystals
of 5-Mes were grown by layering a dichloromethane solution with
pentane. Highly disordered solvent was present; correction for this
residual density was performed using the option SQUEEZE in the
program package PLATON.28 A total of 104 electrons per unit cell
were removed, from a total potentially solvent-accessible void of 357.3
Å3.
(7) Gruver, B. C.; Adams, J. J.; Warner, S. J.; Arulsamy, N.; Roddick,
D. M. Organometallics 2011, 30, 5133−5140.
(8) (a) Chianese, A. R.; Mo, A.; Lampland, N. L.; Swartz, R. L.;
Bremer, P. T. Organometallics 2010, 29, 3019−3026. (b) Chianese, A.
R.; Shaner, S. E.; Tendler, J. A.; Pudalov, D. M.; Shopov, D. Y.; Kim,
D.; Rogers, S. L.; Mo, A. Organometallics 2012, 31, 7359−7367.
(9) (a) Bauer, E. B.; Andavan, G. T. S.; Hollis, T. K.; Rubio, R. J.;
Cho, J.; Kuchenbeiser, G. R.; Helgert, T. R.; Letko, C. S.; Tham, F. S.
Org. Lett. 2008, 10, 1175−1178. (b) Raynal, M.; Cazin, C. S. J.; Vallee,
C.; Olivier-Bourbigou, H.; Braunstein, P. Chem. Commun. 2008,
3983−3985. (c) Raynal, M.; Pattacini, R.; Cazin, C. S. J.; Vallee, C.;
Olivier-Bourbigou, H.; Braunstein, P. Organometallics 2009, 28, 4028−
4047. (d) Schultz, K. M.; Goldberg, K. I.; Gusev, D. G.; Heinekey, D.
M. Organometallics 2011, 30, 1429−1437. (e) Zuo, W. W.; Braunstein,
P. Organometallics 2012, 31, 2606−2615. (f) Zuo, W. W.; Braunstein,
P. Dalton Trans. 2012, 41, 636−643.
(10) Abbreviations: cod = 1,5-cyclooctadiene, DPEPhos = bis(2-
diphenylphosphinophenyl)ether; dba = dibenzylideneacetone.
(11) Afeefy, H. Y.; Liebman, J. F.; Stein, S. E. Neutral
Thermochemical Data. In NIST Chemistry WebBook, NIST Standard
Reference Database Number 69; Linstrom, P. J., Mallard, W. G., Eds.;
National Institute of Standards and Technology: Gaithersburg, MD,
http://webbook.nist.gov (retrieved June 7, 2013).
(12) Xu, W. W.; Rosini, G. P.; Gupta, M.; Jensen, C. M.; Kaska, W.
C.; Krogh-Jespersen, K.; Goldman, A. S. Chem. Commun. 1997, 2273−
2274.
ASSOCIATED CONTENT
* Supporting Information
■
(13) Gottker-Schnetmann, I.; Brookhart, M. J. Am. Chem. Soc. 2004,
̈
S
126, 9330−9338.
CIF file giving crystallographic data for complexes 1-CF3Mes
and 5-Mes. Complete experimental procedures for ligand
synthesis and characterization and images of NMR spectra for
all new compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
(14) Krogh-Jespersen, K.; Czerw, M.; Summa, N.; Renkema, K. B.;
Achord, P. D.; Goldman, A. S. J. Am. Chem. Soc. 2002, 124, 11404−
11416.
(15) Zhu, K. M.; Achord, P. D.; Zhang, X. W.; Krogh-Jespersen, K.;
Goldman, A. S. J. Am. Chem. Soc. 2004, 126, 13044−13053.
G
dx.doi.org/10.1021/om4006577 | Organometallics XXXX, XXX, XXX−XXX

Organometallics
Article
(16) Lee, D. W.; Kaska, W. C.; Jensen, C. M. Organometallics 1998,
17, 1−3.
(17) (a) Liu, F. C.; Pak, E. B.; Singh, B.; Jensen, C. M.; Goldman, A.
S. J. Am. Chem. Soc. 1999, 121, 4086−4087. (b) Huang, Z.; Brookhart,
M.; Goldman, A. S.; Kundu, S.; Ray, A.; Scott, S. L.; Vicente, B. C. Adv.
Synth. Catal. 2009, 351, 188−206.
(18) Punji, B.; Emge, T. J.; Goldman, A. S. Organometallics 2010, 29,
2702−2709.
(19) Gupta, M.; Hagen, C.; Kaska, W. C.; Cramer, R. E.; Jensen, C.
M. J. Am. Chem. Soc. 1997, 119, 840−841.
(20) (a) Haenel, M. W.; Oevers, S.; Angermund, K.; Kaska, W. C.;
Fan, H. J.; Hall, M. B. Angew. Chem., Int. Ed. 2001, 40, 3596−3600.
(b) Gottker-Schnetmann, I.; White, P. S.; Brookhart, M. Organo-
̈
metallics 2004, 23, 1766−1776.
(21) Fulmer, G. R.; Miller, A. J. M.; Sherden, N. H.; Gottlieb, H. E.;
Nudelman, A.; Stoltz, B. M.; Bercaw, J. E.; Goldberg, K. I.
Organometallics 2010, 29, 2176−2179.
(22) Arndtsen, B. A.; Bergman, R. G. Science 1995, 270, 1970−1973.
(23) Piras, E.; Lang, F.; Ruegger, H.; Stein, D.; Worle, M.;
Grutzmacher, H. Chem.Eur. J. 2006, 12, 5849−5858.
(24) Lin, Y.; Nomiya, K.; Finke, R. G. Inorg. Chem. 1993, 32, 6040−
6045.
(25) Xcalibur CCD system, CrysAlisPro software system, Version
1.171.32; Oxford Diffraction Ltd., 2007.
(26) Altomare, A.; Cascarano, M.; Giacovazzo, C.; Guagliardi, A.;
Burla, M. C.; Polidori, G.; Camalli, M. J. Appl. Crystallogr. 1994, 27,
435−436.
(27) Betteridge, P. W.; Carruthers, J. R.; Cooper, R. I.; Prout, K.;
Watkin, D. J. J. Appl. Crystallogr. 2003, 36, 1487.
(28) Spek, A. L. J. Appl. Crystallogr. 2003, 36, 7−13.
H
dx.doi.org/10.1021/om4006577 | Organometallics XXXX, XXX, XXX−XXX