A Ru(I) metalloradical that catalyzes nitrene coupling to azoarenes from arylazides

Article abstract of DOI:10.1021/ja211603f

Unusual N-N coupling of aryl azides to yield azoarenes is demonstrated by the Ru(I) metalloradical, [SiP^iPr₃]Ru(N₂) (4) ([SiP^iPr₃] = (2-iPr₂PC₆H ₄)₃Si^-). The yield of the azoarene is dependent on the substituent on the aryl azide, and the reaction is catalytic for p-methoxy and p-ethoxy phenyl azides, while no azoarene is observed for p-trifluoromethylphenyl azide. Studies aimed at probing the viability of a bimolecular coupling mechanism of metal imide species, as shown in the related [SiP^iPr₃]Fe system, have led to the isolation of several structurally unusual complexes including the ruthenium(IV) imide, 7-OMe, as well as the Ru(II) azide adduct 8-OMe. One electron reduction of 7-OMe complex led to the isolation of the formally Ru(III) imide complex, [SiP^iPr ₃]Ru(NAr) (Ar = p-MeOC₆H₄, 5-OMe). EPR spectroscopy on 5-OMe suggests that the complex is electronically similar to the previously reported imide complex, [SiP^iPr₃]Ru(NAr) (Ar = p-CF₃C₆H_4,5-CF₃), and features radical character on the NAr moiety, but to a greater degree. The stability of 5-OMe establishes that bimolecular coupling of 5-OMe is kinetically inconsistent with the reaction. Further studies rule out mechanisms in which 5-OMe reacts directly with free aryl azide or a transient Ru(I) azide adduct. Together, these studies show that 5-OMe is likely uninvolved in the catalytic cycle and demonstrates the influence of the metal center on the mechanism of reaction. Instead, we favor a mechanism in which free aryl nitrene is released during the catalytic cycle and combines with itself or with free aryl azide to yield the azoarene.

Full text of DOI:10.1021/ja211603f

Article
pubs.acs.org/JACS
A Ru(I) Metalloradical That Catalyzes Nitrene Coupling to Azoarenes
from Arylazides
Ayumi Takaoka, Marc-Etienne Moret, and Jonas C. Peters*
California Institute of Technology, Division of Chemistry and Chemical Engineering, Pasadena, California 91125, United States
S
* Supporting Information
ABSTRACT: Unusual N−N coupling of aryl azides to yield azoarenes
is demonstrated by the Ru(I) metalloradical, [SiPiPr3]Ru(N2) (4)
([SiPiPr3] = (2-iPr2PC6H4)3Si−). The yield of the azoarene is dependent
on the substituent on the aryl azide, and the reaction is catalytic for p-
methoxy and p-ethoxy phenyl azides, while no azoarene is observed for
p-trifluoromethylphenyl azide. Studies aimed at probing the viability of
a bimolecular coupling mechanism of metal imide species, as shown in
the related [SiPiPr3]Fe system, have led to the isolation of several
structurally unusual complexes including the ruthenium(IV) imide, 7-
OMe, as well as the Ru(II) azide adduct 8-OMe. One electron reduction of 7-OMe complex led to the isolation of the formally
Ru(III) imide complex, [SiPiPr3]Ru(NAr) (Ar = p-MeOC6H4, 5-OMe). EPR spectroscopy on 5-OMe suggests that the complex
is electronically similar to the previously reported imide complex, [SiPiPr3]Ru(NAr) (Ar = p-CF3C6H4, 5-CF3), and features
radical character on the NAr moiety, but to a greater degree. The stability of 5-OMe establishes that bimolecular coupling of 5-
OMe is kinetically inconsistent with the reaction. Further studies rule out mechanisms in which 5-OMe reacts directly with free
aryl azide or a transient Ru(I) azide adduct. Together, these studies show that 5-OMe is likely uninvolved in the catalytic cycle
and demonstrates the influence of the metal center on the mechanism of reaction. Instead, we favor a mechanism in which free
aryl nitrene is released during the catalytic cycle and combines with itself or with free aryl azide to yield the azoarene.
the most commonly observed binding mode for structurally
characterized azide complexes.5,7,8,10 A more recent mechanistic
study by Hillhouse described an unusual η2 bound azide
1. INTRODUCTION
Organic azides are valuable sources of nitrenes (NR). They are
easily synthesized for a wide array of R substitutents and release
N2 as the only byproduct. These aspects make organic azides
preferable for nitrene transfer/insertion in organic synthesis
over hypervalent iodine compounds (PhINTs, etc.) and N-
halogenated sulfonamides (chloramine-T, bromamine-T),1−3
which have received more attention but are less flexible with
respect to substituent variability and/or release of undesirable
byproducts. As a result, recent efforts have increasingly focused
on developing metal catalyzed nitrene transfer reactions with
organic azides.4 Because the commonly invoked intermediate in
these reactions is the metal nitrene/imide species, its reactivity
and mechanism of formation from the precursor metal azide
adduct have been an important topic of study.
Well-defined metal azide complexes are relatively uncommon
species and their decay has been mechanistically examined in
limited cases. Bergman and Cummins were the first to report
such studies. Bergman’s Cp2Ta(CH3)(N3Ar) complexes were
found to decay cleanly in a unimolecular fashion to afford the
corresponding imide complexes, Cp2Ta(CH3)(NAr),5 akin to
N2 extrusion in phosphazides (ArN3PR3) that proceeds through
a four-membered transition state to yield iminophosphoranes.6
Cummins’ V(N3Mes)(I)(NRArF)2 (ArF = 2,5-C6H3FMe)
system, in contrast, followed bimolecular decay to the
corresponding imide complex, V(NMes)(I)(NRArF)2.7 In
both studies, the γ-N atom of the azide ligand is bound to
the metal center in the precursor complex and, indeed, this is
complex, (dtbpe)Ni(N3 R) (dtbpe
= bis(ditert-
butylphosphino)ethane, R = adamantyl), that decayed unim-
olecularly to the imide complex, (dtbpe)Ni(NR). In this study,
a large and negative entropy of activation was observed,
consistent with a highly ordered transition state.9
Our group has recently studied the interaction between aryl
azides and the Fe(I) complex, [SiPiPr3]Fe(N2) (1).10,11 This
reaction was found to initially form an Fe(I) azide adduct,
[SiPiPr3]Fe(N3Ar) (2), which subsequently exhibited clean
unimolecular decay as in the Bergman and Hillhouse systems.
Interestingly, the major product of this decay was shown to be
azoarene and 1, and use of excess aryl azide demonstrated
catalytic azoarene formation from 1. While several stoichio-
metric reactions had been known,12 this example was
noteworthy in that it represented a rare example of catalytic
N−N coupling to yield azoarene from organic azides, and the
first in which azoarene was the major product.13 Mechanistic
studies suggested the formation of a transient Fe(III) imide
complex, [SiPiPr3]Fe(NAr) (3-R), following decay of 2, which
subsequently underwent 4e− reductive N−N coupling to
produce azoarene (Scheme 1).14
Received: December 12, 2011
Published: March 1, 2012
© 2012 American Chemical Society
6695
dx.doi.org/10.1021/ja211603f | J. Am. Chem. Soc. 2012, 134, 6695−6706

Journal of the American Chemical Society
Article
Scheme 1. Catalytic Azobenzene Formation with 1
Complex 3-R is a reactive species that was only observable by
EPR spectroscopy in a frozen glass (for R = tol). While the EPR
features of 3-R were indicative of an S = 1/2 ground state, DFT
calculations predicted a small doublet-quartet gap of 2.8 kcal/
mol, perhaps suggesting that two-state reactivity15 may be
responsible for its rich reactivity; this included hydrogen atom
abstraction from 9,10-dihydroanthracene and carbodiimide
formation with t-butylisocyanide.10 In this regard, the recent
isolation of the Ru(I) metalloradical, [SiPiPr3]Ru(N2) (4), and
its interaction with p-CF3C6H4N3 to yield the formally Ru(III)
imide complex, [SiPiPr3]Ru(NC6H4CF3) (5-CF3), is note-
worthy (Scheme 2).16 This work highlighted the first reaction
Figure 1. Key compounds involved in this work. Lines between core
atoms only denote connectivity.
MeOC6H4), in 93(7) % yield as judged by 1H NMR
spectroscopy with ferrocene (Fc) as an internal standard
(Scheme 3). The major metal containing product was 4, with
Scheme 3. Azoarene Formation Catalyzed by 4
Scheme 2. Synthesis of 5-CF3 from 4
small amounts of previously reported16 [SiPiPr3]Ru(H)(N2) (6)
and a minor paramagnetic product, which will be shown below
to be the imide species, 5-OMe. In contrast to the Fe system,
which required heating at 70 °C, the reaction between p-
MeOC6H4N3 and 4 was complete within seconds at room
temperature. As 4 was the major metal containing product in
the stoichiometric reaction, catalysis could be expected and
addition of 10 equiv of p-MeOC6H4N3 led to catalytic
generation of roughly 50% yield of azoarene with catalyst
decomposition. Low temperature 1H NMR studies showed that
azoarene began to form at about −65 °C, and no obvious
buildup of intermediates was observed; the broadness of the
paramagnetic resonances prevented our ability to draw any
definitive conclusions about the metal containing species
present in solution at these temperatures. Performing the
reaction in the presence of excess elemental mercury (>500
equiv) had little effect, suggesting that colloidal metal is unlikely
to be responsible for the catalysis.
chemistry of an unusual mononuclear ruthenium(I) complex.
Although 5-CF3 did not yield azoarene upon decay, it was
stable enough for thorough characterization. We thus
envisioned that use of other substituted aryl azides might
yield similar metal imide species (5-R) that would retain some
stability for characterization, yet also exhibit N−N coupling
reactivity as observed in the Fe system. Further, because the
doublet-quartet gap in the heavier congeners of 1 is expected to
be much greater than 1, any azoarene formation in the Ru
system could also be used as indirect support for doublet state
involvement in 3-R,17 provided that the same mechanism is
operative in both systems.
We herein report the results of our studies on 4 with
substituted aryl azides. The key compounds involved are shown
in Figure 1 for reference. The tendency of the aryl azide to
degrade in the presence of 4 to azoarene product, either
stoichiometrically or catalytically, is dependent on the aryl-ring
substitution pattern. Terminally bonded [SiPiPr3]Ru(NAr)
complexes 5-R were observed in some stoichiometric reactions
and independent synthesis of 5-OMe allowed thorough
characterization and detailed mechanistic studies. Contrary to
our initial expectation, we establish that species 5-R is not
responsible for the observed reactivity and instead determine
that a dramatic change in mechanism occurs in moving from Fe
to Ru within the [SiPiPr3]M(N2) system.
Use of other para-substituted aryl azides showed that the
electronic influence of the substituent, R, on the phenyl group
greatly affected the yield of azoarene. Similar to R = OMe, use
of azide with R = OEt led to near quantitative yield of azoarene
(Table 1). Use of aryl azides where the aryl group is either a
Table 1. Yield (%) of ArNNAr from 4 and 1 equiv of
Substituted ArN3
R =
p-OEt
p-OMe
Mes
p-Me
p-CF3
ArNNAr
91(9)
93(7)
42(3)
29(1)
0
mesityl or a p-tolyl substituent, in contrast, led to diminished
yields of azoarene with greater amounts of 6. Finally, as
previously reported, R = CF3 did not yield any azoarene but
exclusively provided 5-CF3. Thus, electron donating groups
increase the yield of azoarene formation. Because the aryl azide
para-substituted by R = OMe gave the highest yield of
2. RESULTS AND DISCUSSION
2.1. Reaction between 4 and p-MeOC6H4N3 and Other
Aryl Azides. Addition of 1 equiv of p-MeOC6H4N3 to 4 in
Et2O yielded the substituted azoarene, ArNNAr (Ar = p-
6696
dx.doi.org/10.1021/ja211603f | J. Am. Chem. Soc. 2012, 134, 6695−6706

Journal of the American Chemical Society
Article
azoarene, this aryl azide provided the main focus of subsequent
mechanistic studies.
scale. The diamagnetic spectra are suggestive of a distorted 5-
coordinate structure that lies between a square pyramid (SP)
and trigonal bipyramid (TBP) (vide infra), as seen in other
2.2. A Strategy toward the Synthesis of 5-OMe. We
have previously concluded that Fe(III) imide complexes
bimolecularly couple to yield azoarenes in the [SiPiPr3]Fe
system.10 We thus conducted experiments to probe whether
such species were also responsible for azoarene formation in
this Ru system. Mechanistic analysis in this study, however, was
complicated by the presence of multiple paramagnetic species.
Since no buildup of intermediates was observed during low-
temperature NMR experiments, attempts to independently
synthesize 5-OMe were made to assess the viability of 5-OMe
as a chemically and kinetically competent species for the nitrene
coupling process. In this context, it is noteworthy that 5-CF3
features a reversible oxidation event at −1.0 V (vs Fc/Fc+, see
Supporting Information (SI)). Chemical oxidation of 5-CF3
with AgOTf cleanly led to the Ru(IV) imide complex,
{[SiPiPr3]Ru(NAr)}OTf (7-CF3). This compound is thermally
robust and represents an unusual example of a structurally
characterized (see SI) terminal imide of Ru(IV).18 The
increased thermal stability of 7-CF3 relative to 5-CF3 pointed
to increased stability of 7-OMe relative to 5-OMe. We thus
aimed at first synthesizing 7-OMe, which would enable
chemical reduction to 5-OMe. Since the stoichiometric
reactions in this Ru system were complete within seconds at
room temperature, complex 5-OMe would be expected to
rapidly yield azoanisole upon reduction from 7-OMe if the
same mechanism as in the Fe system were at play (Scheme 4).
structurally characterized d6 ruthenium complexes of [SiPR ] (R
3
= iPr, Ph).16,19 The IR spectrum of 8-OMe exhibits a strong
N−N stretch at 2106 cm−1. This value is close to the value of
free p-MeOC6H4N3 (2103 cm−1), and is in accord with an
unactivated azide ligand as expected for a cationic Ru(II)
complex. Upon photolysis of 8-OMe with a mercury lamp in
the presence of excess azide, a color change from red to green
took place and the 31P{1H} signal of 8-OMe at 72 ppm
decreased as a new peak at 106 ppm gained in intensity.
Monitoring the complete decay of the signal at 72 ppm and
following with workup yielded diamagnetic and green 7-OMe
in good yield. Consistent with N2 loss, the IR spectra of 7-OMe
showed no significant stretch near 2100 cm−1 after photolysis.
Interestingly, new signals at 86, 79, and 33 ppm in the
31P{1H} spectrum developed at the expense of the signal for 7-
OMe upon prolonged photolysis. The conversion was
complete after two days and X-ray diffraction studies (vide
infra) revealed nitrene insertion into the Ru−P bond of 7-OMe
and binding of the aryl ring in the nitrene moiety to the metal
1
center in an η6 mode (complex 10, Scheme 5). The H NMR
spectrum of the nitrene insertion product 10 exhibits three
resonances between 6.7 and 5.3 ppm that correspond to the
resonances of the η6 arene moiety. The 31P{1H} resonance at
33 ppm is attributed to the oxidized phosphine center. For ease
of purification, subsequent studies utilized the PF6− anion for 7-
OMe, 8-OMe, and 9. These were synthesized in the same
manner as the BArF − anion complexes (Scheme 5), except for
Scheme 4. Strategy for Probing the Decay of 5-OMe
4
9-PF6, which was synthesized by oxidation of 4 with AgPF6
(see Experimental Section).
2.3. Solid-State Structures of 8-OMe, 7-OMe, and 10.
The solid-state structures of 8-OMe, 7-OMe, and 10 have been
obtained and are shown in Figure 2. The geometry about the
metal center in 8-OMe is between that of a trigonal bipyramid
and a square pyramid with a τ value20 of 0.39. The N(1)−N(2)
and N(2)−N(3) distances of 1.107(5) and 1.254(6) Å are
short and similar to the values of crystallographically
characterized free organic azides.21 The N(1)−N(2) distance
in 8-OMe is significantly shorter than the bond distance in the
Fe(I) azide adduct,10 2-Ad (1.27(1) Å), suggesting little
activation of the azide ligand. The near linear N(1)−N(2)−
N(3) angle of 168.3(6) is also in contrast to 2-Ad (147(4)°)
but similar to the value of 173.1(3)° observed in a Cu(I) azide
complex, [HB(3,5-(CF3)2Pz)3]Cu(N3Ad) ([HB(3,5-
(CF3)2Pz)3]− = hydridotris(3,5-bis(trifluoromethyl)pyrazolyl)-
borate).8f Interestingly, 8-OMe appears to be the only
crystallographically characterized ruthenium complex with a
coordinated organic azide reported to date.
Addition of 1 equiv of p-MeOC6H4N3 to previously
reported16 {[SiPiPr3]Ru(N2)}BArF (9-BArF ) resulted in a
4
4
rapid color change from orange to red, yielding the diamagnetic
Ru(II) azide adduct, {[SiPiPr3]Ru(N3Ar)}BArF (8-OMe)
4
(Scheme 5). Both the 1H and 31P{1H} NMR spectra are
consistent with a 3-fold symmetric structure on the NMR time
Scheme 5. Synthesis of 5-OMe
The solid-state structure of 7-OMe features a metal center
with a τ value of 0.90, approximating a trigonal bipyramidal
geometry. The ruthenium is significantly displaced out of the
plane of the phosphines, however, and the three P−Ru−P
angles sum to 340.93(5)°. The Ru−Si bond length of 2.491(1)
Å is accordingly much longer than typically observed distances
in other [SiPiPr3]RuX complexes, where X is a neutral or
anionic ligand trans to the Si anchor in the [SiPiPr3] scaffold
(Table 2).16 The elongation of the Ru−Si bond reflects an
approach of the metal center toward a pseudotetrahedral
geometry, with a corresponding weakening of the Ru−Si bond.
This change in geometry has significant consequences toward
its electronic structure (see Section 2.9). The Ru−N distance of
6697
dx.doi.org/10.1021/ja211603f | J. Am. Chem. Soc. 2012, 134, 6695−6706

Journal of the American Chemical Society
Article
Figure 3. Cyclic voltammogram of 7-OMe. Conducted in 0.3 M
TBAPF6 in THF.
resulting from one electron reduction of 7-OMe is stable on the
electrochemical time scale. Chemical reduction was accom-
plished by addition of 1 equiv of CoCp2 to 7-OMe, which
caused a color change from green to red/brown. Removal of
[CoCp2]PF6 and extraction into pentane led to the isolation of
5-OMe in moderate yield. While its solid-state structure was
1
not be obtained, the H NMR of 5-OMe is reminiscent of the
spectrum of 5-CF3. The room temperature (RT) EPR
spectrum (vide infra) of 5-OMe is also similar to that of 5-
CF3, corroborating its assignment. Interestingly, 5-OMe is
relatively stable at room temperature, showing signs of
decomposition only after several hours in solution. Further,
5-OMe does not produce azoanisole upon decay. These
observations rule against a mechanism in common with the
related Fe system, in which two imide species undergo N−N
coupling (Scheme 6A).10
Figure 2. Solid-state structure of 8-OMe (top, left), 7-OMe (top,
right), and 10 (bottom). Anions, hydrogen atoms, and solvent
molecules are removed for clarity.
Scheme 6. Plausible Candidate Mechanisms for Liberation
of Azoarenes Involving 5-OMe
Table 2. Comparison of Ru−Si bond lengths (Å) and Sum of
P−Ru−P Angles (°) for Representative 5-Coordinate Ru(I)
and Ru(II) Complexes Relative to 7-OMe
a
a
[SiPiPr3]RuI
4
8-OMe
7-OMe
Ru−Si (Å)
∑(P−Ru−P)
2.284(1)
353.3(1)
2.319(1)
352.7(1)
2.305(1)
352.1(1)
2.491(1)
340.9(1)
a
From ref 16.
1.802(3) Å, while significantly shorter than in 5-CF3, is longer
than the two other structurally characterized Ru(IV) imide
complexes (1.716(3) and 1.785(6) Å).18 The solid-state
structure of 7-CF3 is very similar to 7-OMe (see SI).
Finally, the solid-state structure of 10 features a phosphini-
mide moiety resulting from NAr insertion of 7-OMe into one
of the M−P bonds. This complex is best described as a three-
legged piano stool complex. Dechelation of one P atom results
from its oxidation to a formally pentavalent phosphorus atom,
and an η6 interaction between the aryl ring and the metal center
is observed with Ru−C distances ranging between 2.23 and
2.40 Å. These distances are within, if not slightly longer, than
reported Ru−C(aryl) distances for ruthenium complexes with
an η6 coordinated anisole ligand.22
2.4. Synthesis of 5-OMe and Mechanistic Studies. The
cyclic voltammetry of 7-OMe is shown in Figure 3 and displays
a reversible reduction event at −1.24 V and an irreversible
reduction event at −2.17 V. The first wave is assigned to a
formal Ru(IV) to Ru(III) reduction, as this wave is close to the
formal Ru(IV)/Ru(III) couple of 7-CF3. The irreversible wave
is assigned to a Ru(III)/Ru(II) redox event. The reversible
nature of the first redox event indicates that the product
Other plausible mechanisms that involve 5-OMe in the
catalytic cycle are shown in Scheme 6B,C. In mechanism B, 5-
OMe reacts with free azide to yield azoanisole and 4. This
mechanism is related to one recently reported for a nickel
system,12b in which azide addition to a nickel imide results in
1,3-dipolar addition to yield a tetrazene type intermediate/
transition state that releases azoarene after N2 extrusion. In
mechanism C, 5-OMe reacts with a transient Ru(I) azide
adduct, [SiPiPr3]Ru(N3Ar) (11-OMe), to release azoanisole and
regenerate 4. To test the validity of mechanism B, p-
MeOC6H4N3 was added to 5-OMe. Neither decay of 5-OMe
nor azoanisole formation was observed, ruling it out. To test
mechanism C, a crossover experiment was designed with p-
EtOC6H4N3. First, a control experiment, in which a 1:1 mixture
of p-MeOC6H4N3 and p-EtOC6H4N3 was added to 4, was
conducted. This reaction produced a statistical mixture of 1:2:1
ArNNAr/ArNNAr′/Ar′NNAr′ (Ar = p-MeOC6H4, Ar′ = p-
EtOC6H4) (Scheme 7). Next, 1 equiv of p-EtOC6H4N3 was
added to 1 equiv each of 4 and 5-OMe. If mechanism C was
responsible for azoarene formation, the hetero azoarene,
6698
dx.doi.org/10.1021/ja211603f | J. Am. Chem. Soc. 2012, 134, 6695−6706

Journal of the American Chemical Society
Article
our hands, generation of singlet nitrene by room temperature
photolysis of tolylazide in neat diethylamine produced azepine
as the major product as judged by GC−MS and H NMR
Scheme 7. Reaction between 4 and a 1:1 Mixture of Two
Similar Aryl Azides
1
spectroscopy. A similar trapping experiment with p-
MeOC6H4N3 produced little, if any, azepine, possibly due to
the >100-fold greater rate of ISC for this nitrene relative to
singlet tolylnitrene.23b Thus, for the trapping experiment with
4, tolyl azide was used instead of p-MeOC6H4N3. Addition of
tolyl azide to 4 in neat diethylamine, however, produced no
azepine but only azotoluene. Thus, if singlet nitrene is released
by 11-Me, it must be undergoing ISC to the triplet state sooner
than it is being trapped by diethylamine. In this case, ISC would
have to be assumed to be accelerated by the influence of a
nearby ruthenium center, since efficient trapping was
established without the presence of ruthenium. Trapping of
triplet nitrene is more difficult since most of the conventional
triplet nitrene traps, such as nitrosobenzene,26 would react with
metalloradical 4. Direct observation of triplet nitrene during the
reaction by EPR spectroscopy was also considered. While
photolysis of both tolylazide and p-MeOC6H4N3 at 77 K in a
frozen glass produced the EPR signal characteristic of triplet
nitrene (Figure 4),27 thawing at −78 °C resulted in rapid loss of
ArNNAr′, would be expected to form to some extent. In
contrast, only the homocoupled azoarene, Ar′NNAr′ (Ar′ =
p-EtOC6H4) was observed (Scheme 8). This series of
Scheme 8. Crossover Experiment Discussed in the Text
experiments rules out all three mechanisms that are shown in
Scheme 6 and establish that the formally Ru(III) imide complex
5-OMe is not involved in the catalytic cycle.
2.5. Considering the Release of Free Aryl Nitrene. An
interesting alternative to the mechanistic scenarios shown in
Scheme 6 is the release of aryl nitrene during the catalytic cycle
from a ruthenium aryl azide precursor, 11-OMe. It is well
established that aryl nitrenes have triplet ground states23 that
react at near diffusion controlled rates with either themselves or
with free aryl azide to produce azoarenes.24 This reactivity is
unique to the triplet state; singlet aryl nitrene is not known to
form azoarene. In the present Ru system under consideration,
because 4 and 11-OMe both have doublet ground states (as
shown by EPR spectroscopy, see Figure 5), both singlet and
triplet nitrene could be released from 11-OMe (Scheme 9). If
Figure 4. X-Band EPR spectra of triplet p-tolyl nitrene (red) and
triplet p-methoxyphenyl nitrene (blue).
the signal within seconds. Thus, no buildup of triplent nitrene
is expected, since the ruthenium catalyzed azoarene reaction is
known only to take place at higher temperatures (vide supra).
2.6. EPR Spectroscopy on 5-R and 11-R. The room
temperature and 77 K EPR spectra of 5-OMe are shown in
Figure 5. The room temperature spectrum depicts a three line
pattern due to large coupling of 119 MHz to the nitrogen of the
NAr moiety and smaller coupling of 48 MHz to one
phosphorus atom. The N hyperfine coupling in 5-OMe is
larger relative to 5-CF3 while the P coupling is smaller, which
yields a combined effect of creating a three line pattern for 5-
OMe instead of a four line pattern as in 5-CF3.16 Also worthy
of note is the smaller Ru hyperfine coupling of 38 MHz,
observed as satellites due to spin active isotopes of Ru
(approximately 30%), in 5-OMe relative to 5-CF3 (48 MHz).
Taken together, the hyperfine coupling to N and Ru indicate a
greater spin density on the NAr moiety for 5-OMe relative to
5-CF3. By analogy to the assignment of 5-CF3 containing a
NAr•− radical moiety, we assign 5-OMe as also possessing a
NAr•− radical.4k,18b,28 The radical character on the NAr moiety
is also supported from the isotropic g-value, giso, of 2.002, which
shows little deviation from the value of the free electron,
2.0023. Further, the anisotropy in the g-values (Δg = 0.063) at
Scheme 9. Aryl Nitrene Release from 8-OMe
triplet nitrene is released, rapid recombination with itself or
with free azide would yield azoanisole. If singlet nitrene is
released, it could undergo intersystem crossing (ISC) to triplet
nitrene to then produce azoanisole. This mechanism would be
consistent with the absence of involvement of 5-OMe and also
with the results summarized in Scheme 7, where a statistical
mixture of the three azoarenes is produced upon addition of a
1:1 mixture of two similar aryl azides to 4.
Singlet aryl nitrenes have been trapped with nucleophiles
77 K is to be contrasted with the value for the Ru(I)
such as diethyamine to yield azepine (Scheme 9).25 Indeed, in
metalloradical 4 (0.135) or the Ru(III) complex, {[SiPiPr
]
3
-
6699
dx.doi.org/10.1021/ja211603f | J. Am. Chem. Soc. 2012, 134, 6695−6706

Journal of the American Chemical Society
Article
Figure 6. (Left) RT EPR spectrum of crude mixture from reaction
between 4 and tolylazide. (Right) Contribution of 4 (red) and 5-Me
(blue) to crude spectrum. The blue line results from subtracting the
contribution of the red line from the black experimental spectrum.
route to either 5-R or azoarene at low temperature. To this end,
a solution of 1 equiv of p-CF3C6H4N3 was layered over a frozen
solution of 4. The resulting layered frozen solution was rapidly
thawed in a dry ice/isopropyl alcohol bath, quickly mixed, and
frozen again for analysis by EPR spectroscopy. The EPR
spectrum showed new signals that were distinct from 4, but
conversion was not complete. Addition of excess azide (>50
equiv) resulted in complete conversion to a new signal that was
assigned as 11-CF3 (Figure 5). Notably, the anisotropy in the g-
values in this spectrum (Δg = 0.17) indicate significant
metalloradical character (Figure 5). Similarly, addition of 1
equiv of tolylazide was enough to see new features attributable
to 11-Me, although excess was required for full conversion. In
contrast, addition of 1 equiv of p-MeOC6H4N3 led to
unnoticeable changes in the EPR signatures of 4, indicating
the presence of an equilibrium between 4 and 11-OMe that
strongly favored 4. Indeed, addition of over 50 equiv of p-
MeOC6H4N3 and evacuation of N2 from the EPR tube was
necessary for the signals of 11-OMe to be observed. The
direction of the equilibrium in this case appears to be dictated
by the π-accepting properties of the aryl azide, since p-
MeOC6H4N3 is expected to be a better σ-donor than p-
CF3C6H4N3.
Figure 5. (A and B) EPR spectra of 5-CF3 (A, RT; B, 77 K); (C and
D) EPR spectra of 5-OMe (C, RT; D, 77 K); (E) 77 K EPR spectrum
of 11-OMe; (F) 77 K EPR spectrum of 11-CF3; (G) 77 K EPR
spectrum of 12. For simulation parameters, see SI.
Thawing a frozen solution of freshly prepared 11-CF3 at −76
°C and recording the EPR spectrum as a function of time led to
decay of 11-CF3 to 5-CF3 as shown in Figure 7. The decay
followed first-order kinetics with a half-life of 23 min at −76 °C
(Figure 8), and was independent of the concentration of p-
CF3C6H4N3. The kinetics are consistent with N2 extrusion from
a transient Ru(I) azide adduct and corroborate the assignment
RuCl}PF6 (12), which shows a Δg = 0.242. The latter complex
-
was synthesized by oxidation of previously reported [SiPiPr
]
3
RuCl16 with AgPF6 and represents a formally Ru(III)
metalloradical.
Measuring the RT EPR spectrum of the crude mixture from
the stoichiometric reaction between 4 and p-MeOC6H4N3
(Section 2.1) yielded a complicated spectrum. Deconvolution
of the spectrum was performed by subtracting out the
contribution from 4 and provided a spectrum of nearly pure
5-OMe. This result suggests that, although 5-OMe is not
responsible for the catalysis of azoanisole formation, it is
nevertheless formed under the reaction conditions. To probe
whether complexes 5-R (R = OEt, Me, Mes) were also formed
during stoichiometric reactions, the RT EPR spectra of crude
reaction mixtures were also measured (see Figure 6 for R =
Me). Again, a complicated pattern was observed (Figure 6, SI),
but similar deconvolution resulted in spectra similar to that of
5-OMe. The results again point to distinct pathways that give
rise to 5-R and azoarene, in which the azoarene formation step
is favored for more electron donating substituents.
As EPR proved to be a convenient tool in detecting minor
amounts of paramagnetic products for this system, attempts
were made to detect transient Ru(I) azide adducts, 11-R, en
Figure 7. Decay of 11-CF3 to 5-CF3. Black curve, 11-CF3; red curve,
5-CF3.
6700
dx.doi.org/10.1021/ja211603f | J. Am. Chem. Soc. 2012, 134, 6695−6706

Journal of the American Chemical Society
Article
structures (α and γ-bound) and two η2-structures (α,β and β,γ-
bound) were examined (Figure 10). Of these four isomers, the
Figure 10. Energies of DFT optimized structures of 11-OMe.
Figure 8. First-order decay plot of 11-CF3 (3.4 mM) in 2-MeTHF at
−76 °C. The value of [11-CF3]t/[11-CF3]0 was obtained by
subtracting the contribution of 5-CF3 from the experimental spectrum
at each time and comparing the intensity of [11-CF3]t to [11-CF3]0.
γ-bound azide adduct is found to be the most stable structure.
Interestingly, two electronically distinct energy minima were
found for the γ-bound structure. The lowest energy structure A
shows localization of spin on the metal center, with little spread
among the NAr moiety. In the optimized geometry, one P atom
contains a much greater amount of spin density relative to the
other two P atoms in the equatorial plane of the trigonal
bipyramid; this trend is not noticeable for 5-OMe, in which the
radical is located in the NAr moiety and little spin density is
distributed among the phosphines. This feature is consistent
with the Ru(I) metalloradicals, 4 and [SiPiPr3]Ru(PMe3),16 and
the Group 9 metalloradicals, {[SiPiPr3]M(PMe3)}+ (M = Co,
Rh, Ir).29 The unequal spread in spin is reflected in the 77 K
EPR spectrum, where only one P atom is found to be largely
responsible for the fine features (Figure 5).
of 11-CF3 as well as the assignment of 11-OMe, which has
similar EPR signatures.
2.7. DFT Calculations on 5-OMe and 11-OMe. DFT
calculations on 5-OMe and 11-OMe were performed to further
probe their electronic structures. The optimized structure of 5-
16
OMe is very similar to that of structurally characterized 5-CF3
and features a geometry that is between a TBP and SQP with τ
= 0.44 (Figure 9). One of the P−Ru−P angles is considerably
The second lowest energy structure (B) exhibits the same
connectivity as A but differs by its electronic structure: 80% of
its spin-density is located on the azide nitrogen atoms. The
ligand radical character of B is inconsistent with the EPR
spectrum, which exhibits large g-anisotropy suggestive of
metalloradical character. However, the rather low energy
difference (7.8 kcal/mol) indicates that such charge-transfer
structures may be accessible and could in principle contribute
to the observed reactivity. The α-bound azide adduct C was
calculated less stable than the γ-bound adduct A by 20.6 kcal/
mol (Figure 10), which must largely reflect the steric mismatch
between the [SiPiPr3] scaffold and the aryl substituent; the
unfavorable nature of the α-N binding mode is reflected in the
long Ru−N distance of 2.568 Å. Of the two η2-bound
structures, the β,γ-structure E is more stable and only 9.9
kcal/mol above the energy of the γ-structure. This binding
mode is rare but has precedent.9 The geometry about the metal
center is also rather close to that of the γ-structure, and the spin
density distribution is found to be very similar. As a result,
conclusive assignment of the ground state structure is difficult
to make, and an equilibrium may even exist. The α,β-structure
D, in turn, is 18.7 kcal/mol above the γ-structure, and exhibits
large spin density (76%) on the unbound γ-N atom, while only
3% is found on the metal center. This electronic structure,
similar to that of isomer B, is inconsistent with the EPR data.
Overall, DFT calculations support the formulation of adducts
11 as terminally bound azide complexes of type A (Figure 10)
with metalloradical character, in accord with the large g-
anisotropy observed in the EPR spectrum.
Figure 9. DFT optimized structure of 5-OMe (left) and spin density
plot (right).
larger (133.1°) than the other two and the NAr moiety is found
to slightly slant into the pocket created by this large angle,
giving rise to the observed distorted structure. The Ru−N
distance is 1.887 Å, close to the optimized value for 5-CF3
(1.872 Å). Previous DFT calculations on 5-CF3 supported the
EPR simulations that assigned a significant amount of spin
density on the NAr moiety. The calculation on 5-OMe also
points to a similar conclusion. In fact, larger delocalization of
spin density is seen on the NAr moiety (70%) for 5-OMe
relative to 5-CF3 (54 %).16 Conversely, the spin density on the
metal center is lower (26% for 5-OMe vs 40% for 5-CF3).
These values are qualitatively consistent with the Ru and N
hyperfine coupling constants found in the simulations of the
RT EPR spectrum for 5-OMe (Section 2.6).
2.8. Nitrene Release. Having established the η1-γ-bound
geometry A as the most likely ground state structure for azide
adducts 11, we located a transition state for the release of free
triplet nitrene from the truncated model [SiPMe3]Ru(N3Ph)
(Figure 11, see SI for details). An interesting feature of the
The calculations on 11-OMe, in contrast, support the EPR
simulations that point to a metalloradical species. As the azide
adduct may bind in several different modes, two η1-bound
6701
dx.doi.org/10.1021/ja211603f | J. Am. Chem. Soc. 2012, 134, 6695−6706

Journal of the American Chemical Society
Article
Figure 11. DFT-optimized transition state for the release of triplet
phenylnitrene from [SiPMe3]Ru(N3Ar) and the corresponding spin
densities.
calculated transition structure is the contraction of the Si−Ru−
N angle to 152.1° versus 165.1° in the azide adduct and 179.4°
in the final Ru−N2 complex. This presumably increases the
overlap between the Ru-centered SOMO and the π orbitals of
the azide ligand, conferring some charge-transfer character to
the transition state, as also suggested by inspection of the
corresponding spin density (Figure 11). A quantitative
interpretation of the free enthalpy of activation is difficult
because of the limitations of single-determinantal DFT
methods, but the calculated value of 12.6 kcal/mol is consistent
with nitrene release from structure A being an accessible
pathway. An alternative route involving the charge-transfer
structure B as an intermediate cannot be ruled out, but overall
Figure 12. (Left) Core atoms of 7-OMe with relevant bond lengths
(Å) and angles (°). (Right) MO diagram of 7-OMe obtained from a
single point calculation on X-ray coordinates. The energy levels are
drawn to scale. Calculated using the B3LYP functional with the
LANL2TZ(f) for Ru and 6-311G** for all other atoms.
[SiPiPr3] ligand. The metal center in this complex is more
pyramidal than 7-OMe (340.9°), with P−M−P angles
summing to 330.0°. That the two complexes exhibit similar
MO diagrams is perhaps expected as the [Fe−B] unit in
[TPB]Fe(NAr) is valence isoelectronic to the [Ru−Si]+ unit in
7-OMe (Figure 13). One difference between these units is the
3
the DFT study clearly substantiates the feasibility of ArN
release from Ru−N3Ar adducts 11 to concomitantly liberate the
Ru−N2 complex 1.
2.9. Electronic Structure of 7-OMe. Complex 7-OMe is a
trigonal bipyramidal, formally Ru(IV) complex with a d4
electronic configuration. The thermal stability of this
compound is striking given that a simplified molecular orbital
diagram for a trigonal bipyramid (TBP) would place 4 electrons
in π* dxz and dyz orbitals, yielding a formal Ru−N bond order
of 1. DFT calculations were thus performed to analyze the
frontier orbitals of 7-OMe. These calculations indicate that in
contrast to a typical TBP MO diagram, the π*orbitals, dxz and
Figure 13. A comparison of [TPB]Fe(NAr) with 7-OMe.
2
dyz, lie above the σ* orbitals dxy and dx2−y (Figure 12).
2
2
Accordingly, four electrons are placed in the dxy/dx −y instead
of the dxz/dyz orbitals, conserving a formal bond order of 3 for
the Ru−N bond. The stability of 7-OMe and the origin of the
reversal of orbital ordering are likely due to the pyramidaliza-
tion of the ruthenium center. The ruthenium is displaced out of
the plane of the three phosphines, which leads to decreased P−
Ru−P angles, an increased Ru−Si bond length (Table 2), and
an approach of the complex toward a pseudotetrahedral
geometry. The ample precedent for stabilization of multiply
bonded metal complexes in late transition metals under
pseudotetrahedral metal centers supports this argument.30
Interestingly, the orbital located directly below the π* orbitals
possesses significant Si pz and Ru dz2 character. The overall MO
diagram is very reminiscent of that of the recently reported
[TPB]Fe(NAr) complex ([TPB] = (2-iPr2C6H4)3B), Ar = p-
C6H4OMe),31 which features an iron center chelated by a
tris(phosphino)borane ligand that is topologically related to the
nature of the anchoring ligand; traditionally, silyl ligands are
thought of as anionic electron donating ligands while boranes
are considered neutral electron accepting ligands. While we
continue to denote 7-OMe as a formally Ru(IV) complex, the
thought of the silicon atom acting as an electron acceptor (Si+)
in 7-OMe is not unreasonable given the particularly low
Ru(IV/III) redox potential of −1.24 V.
3. MECHANISTIC CONSIDERATIONS
Catalytic formation of azoarenes from aryl azides has little
precedent. To our knowledge, the only example prior to our
recent work on the [SiPiPr3]Fe system was reported by Cenini
as a side reaction of catalytic C−H amination with Co
porphyrins.13 Recently, Heyduk32 has reported a similar
reaction. The work herein adds to the library of catalytic
group transfer of two nitrene moieties to form azoarenes, and
6702
dx.doi.org/10.1021/ja211603f | J. Am. Chem. Soc. 2012, 134, 6695−6706

Journal of the American Chemical Society
Article
showcases the first reaction chemistry mediated by the unusual
Ru(I) oxidation state. Importantly, the mechanistic analysis in
this work illustrates that the mechanism responsible for
azoarene formation is distinct from the [SiPiPr3]Fe system. In
Cenini’s work, the azoarene is believed to form through the
reaction between an imide complex of a cobalt porphyrin and a
free organic azide (Scheme 10A). This mechanism is akin to
Scheme 11. Favored Mechanism for the Azoarene Catalytic
Cycle
Scheme 10
capture. The observation that 5-CF3 is the sole product when
p-CF3C6H4N3 is used is intuitively consistent with rapid nitrene
capture, as nitrene capture formally involves a two electron
oxidation of the metal center, which is expected to be facilitated
by electron withdrawing groups on the aryl moiety.
the reactivity observed in the stoichiometric azoarene formation
mediated by a phosphine-supported Ni(II) center reported by
Hillhouse (Scheme 10B).12b In contrast, the [SiPiPr3]Fe system
invokes bimolecular coupling of two Fe(III) imide complexes;
Heyduk’s system also appears to operate through a related
mechanism (Scheme 10C,D).10,32 The experimental studies
described in the present work rule out either of these
mechanisms for the [SiPiPr3]Ru system. If a metal imide were
involved, the catalytic cycle would involve a formal Ru(I)−N2/
Ru(III) NAr redox cycle, as in the [SiPiPr3]Fe system. This
redox cycle is inconsistent with the mechanistic studies
performed in this work, since 5-OMe was detected in
stoichiometric reactions and found to be stable over hours in
solution while the catalytic azoarene formation in this system
proceeds within seconds at room temperature. Moreover,
independent synthesis of 5-OMe as described above demon-
strated that the decay product of 5-OMe did not contain
azoanisole. The Cenini type mechanism was also ruled out by
observing that p-MeOC6H4N3 does not yield any azoanisole on
addition of 5-OMe. Finally, crossover experiments described in
Section 2.4 also ruled out a mechanism that involves a transient
Ru(I) azide adduct, 11-OMe, reacting with 5-OMe. Thus, 5-
OMe does not appear to be responsible for azoarene formation at
all.
An alternative to consider is the release of free triplet nitrene
from 11-OMe (Scheme 11). After release of free triplet nitrene
from 11-OMe, it could react with itself or with free azide and
yield azoanisole while regenerating 4 (Scheme 11). Such a
mechanism would be consistent with the results in Scheme 7
and is additionally supported by the DFT optimization of a
low-lying transition state for nitrene release (Figure 11) . The
presence of 5-OMe in the reaction mixture can be reconciled
by the formation of a transient cage complex where the free
nitrene is in the vicinity of the ruthenium. If the rate of nitrene
capture by the metal fragment is rapid relative to nitrene escape
from the cage, then the imide species, 5-R, would be the major
product. If cage escape is rapid, however, nitrene will escape
and rapidly form azoarene. The different product distributions
(5-R and azoarene) from altering the R group, from CF3 to
OMe, could be explained by the differing rates of nitrene
One last mechanism to consider involves the reaction
between 11-OMe with itself or with free p-MeOC6H4N3. This
mechanism is conceptually related to the ‘third oxidant’
mechanism, as described by Goldberg in the oxygen atom
transfer from iodosylarene to substrate by manganese
corralazine complexes.33 In this study, the often invoked
Mn(V) oxo functionality was not found to be the source of the
oxygen atom transferred to substrate; instead, the oxygen atom
was transferred directly from a coordinated iodosylarene ligand.
While we cannot rule this mechanism out, we disfavor it since
the electronic polarization between the two reactants that exists
in the Mn system does not exist in our system. Specifically, in
the Mn corralazine study, the coordinated iodosylbenzene acts
as the electrophile and substrate (olefin, sulfide) acts as the
nucleophile, and thus, a distinct electronic polarization exists
between the two. In our system, the transient Ru(I) azide
adduct would need to act as both the nucleophile and
electrophile, where such polarization is expected to be minimal.
We thus favor the nitrene release mechanism.
4. CONCLUSIONS
In conclusion, we have demonstrated facile catalytic N−N
coupling of aryl azides to yield azoarenes mediated by the Ru(I)
metalloradical, [SiPiPr3]Ru(N2) (4). Studies aimed at probing
the viability of a bimolecular coupling mechanism of metal
imide species as found in the related iron system have led to the
isolation of several structurally unusual complexes, including
the ruthenium imides, 5-OMe and 7-OMe, as well as the azide
adduct 8-OMe. Mechanistic studies showed that 5-OMe is not
involved in the catalytic cycle and demonstrates the influence of
the metal center on the mechanism of reaction. Instead, we
favor a mechanism in which free aryl nitrene is released during
the catalytic cycle and combines with itself or with free aryl
azide to yield the azoarene.
5. EXPERIMENTAL SECTION
5.1. General Considerations. All manipulations were carried out
using standard Schlenk or glovebox techniques under an atmosphere
of dinitrogen. Unless otherwise noted, solvents were degassed and
6703
dx.doi.org/10.1021/ja211603f | J. Am. Chem. Soc. 2012, 134, 6695−6706

Journal of the American Chemical Society
Article
dried by thoroughly sparging with N2 gas followed by passage through
an activated alumina column. Hexamethyldisiloxane was dried over
CaH2 and distilled. Pentane, hexamethyldisiloxane, benzene, toluene,
tetrahydrofuran, and diethylether were tested with a standard purple
solution of sodium benzophenone ketyl in tetrahydrofuran. Unless
noted otherwise, all reagents were purchased from commercial vendors
and used without further purification. Celite (Celite 545) was dried at
dissolved in 4 mL of Et2O and cooled to −78 °C. p-MeOC6H4N3
(4.4 mg, 0.029 mmol) was dissolved in 1 mL of Et2O and also cooled
to −78 °C. The azide solution was added dropwise to the
{[SiPiPr3]Ru(N2)}BArF solution, which resulted in an immediate
4
color change from orange to red. The solution was stirred at room
temperature for 10 min and concentrated to yield red 8-OMe (44 mg,
89%). Crystals suitable for X-ray diffraction were grown from layering
1
150 °C overnight before use. Complexes 4, 5-CF3, 9-BArF ,
pentane over a concentrated ether solution of 8-OMe at −35 °C. H
4
[SiPiPr3]RuCl were previously reported.16 [(C6H6)RuCl2]2,34 tris(2-
NMR (d8-THF, δ): 8.34 (d, J = 7.2 Hz, 3H), 7.79 (s, 8H), 7.64 (br,
3H), 7.57 (s, 4H), 7.47 (t, J = 7.2 Hz, 3H), 7.38 (t, J = 7.5 Hz, 3H),
7.16 (d, J = 9.0 Hz, 2H), 7.00 (d, J = 9.0 Hz, 2H), 3.78 (s, 3H), 2.10
(br, 6H), 1.17 (br, 18H), 0.83 (br, 18H). 13C{1H} NMR (d8-THF, δ):
163.8 (m), 160.4, 154.4, 144.0, 136.5, 134.9, 131.6, 131.0, 130.8, 130.0,
129.6, 127.4, 123.1, 123.0, 119.1, 117.1, 67.2, 56.5, 20.8 (br), 16.5.
19F{1H} NMR (d8-THF, δ): −61.2. 31P{1H} NMR (d8-THF, δ): 72.9.
(diisopropylphosphino)phenyl)silane ([SiPiPr3]H),11 aryl azides,35 and
36
KC8 were synthesized according to literature procedures. Triethyl-
amine was dried over calcium hydride and distilled. Deuterated
solvents were purchased from Cambridge Isotope Laboratories, Inc.,
degassed, and stored over 3-Å molecular sieves prior to use. Elemental
analyses were performed by Midwest Microlabs. Varian Mercury-300
and Varian Inova-500 were used to collect 1H, 13C, 29Si, and 31P
spectra at room temperature unless otherwise noted. 1H and 13C
spectra were referenced to residual solvent resonances. 29Si spectra
were referenced to external tetramethylsilane (δ = 0 ppm), and 31P
spectra were referenced to external 85% phosphoric acid (δ = 0 ppm).
IR measurements were obtained on samples prepared as KBr pellets
using a Bio-Rad Excalibur FTS 3000 spectrometer. X-band EPR
spectra were obtained on a Bruker EMX spectrometer. Spectra were
simulated using Easyspin37 program.
5.4.3. Synthesis of {[SiPiPr3]Ru(NAr)}BArF (7-OMe). {[SiPiPr3]Ru-
4
(N3Ar)}BArF (8-OMe, Ar = C6H4OMe) (40 mg, 0.023 mmol) was
4
dissolved in 2 mL of THF and charged in a 4 mL quartz cuvette.
Excess p-MeOC6H4N3 (4 mg, 0.027 mmol) was added to the cuvette
and the red solution was photolyzed. The progress of the conversion
was monitored by 31P{1H} NMR spectroscopy. After the conversion
was complete (approximately 1 h), the green solution was
concentrated, and the oily material was triturated with pentane (5 ×
3 mL) to yield the green 7-OMe (32 mg, 83%). Crystals suitable for
X-ray diffraction were grown by layering pentane over a concentrated
ether solution of 7-OMe at −35 °C. 1H NMR (d8-THF, δ): 8.14 (d, J
= 7.5 Hz, 3H), 7.78 (s, 8H), 7.77 (m, 3H), 7.57 (s, 4H), 7.44−6.85
(m, 10H), 3.83 (s, 3H), 2.74 (br, 6H), 1.12 (m, 18H), 0.55 (m, 18H).
13C{1H} NMR (d8-THF, δ): 163.6 (m), 162.1, 157.2 (m), 142.6 (m),
5.2. Crystallographic Details. X-ray diffraction studies were
carried out at the Beckman Institute Crystallography Facility on a
Bruker KAPPA APEX II diffractometer and at the MIT Department of
̈
Chemistry X-Ray Diffraction Facility on a Bruker three-circle Platform
APEX II diffractometer solved using SHELX v. 6.14. The crystals were
mounted on a glass fiber with Paratone-N oil. Data was collected at
100 K using Mo Kα (λ = 0.710 73 Å) radiation and solved using
SHELXS38 and refined against F2 on all data by full-matrix least-
squares with SHELXL. X-ray quality crystals were grown as described
in the experimental procedures.
136.5, 134.9, 133.0, 131.0, 130.8, 130.5, 130.2, 129.6, 127.5, 125.3,
125.2, 123.1, 119.1, 117.6, 57.1, 33.3, 20.5, 20.3.). 19F{1H} NMR (d8-
THF, δ): −61.4. 31P{1H} NMR (d8-THF, δ): 106.4.
5.4.4. Synthesis of {[SiPiPr3]Ru(N2)}PF6 (9-PF6), {[SiPiPr3]Ru(N3Ar)}-
PF6 (8-OMe), and {[SiPiPr3]Ru(NAr)}PF6 (Ar = C6H4OMe) (7-OMe). 9-
PF6: [SiPiPr3]Ru(N2) (35 mg, 0.048 mmol) was dissolved in 5 mL of
THF. AgPF6 (12 mg, 0.048 mmol) was dissolved in 1 mL of THF and
both solution were cooled to −78 °C. The AgPF6 solution was added
dropwise to the solution of [SiPiPr3]Ru(N2), causing an immediate
darkening of the solution. The solution was stirred for 10 min, filtered
through Celite, and concentrated. The solid was washed with Et2O
5.3. Electrochemical Details. Electrochemical measurements
were carried out in a glovebox under a dinitrogen atmosphere in a
one-compartment cell using a CH Instruments 600B electrochemical
analyzer. A glassy carbon electrode was used as the working electrode
and platinum wire was used as the auxiliary electrode. The reference
electrode was Ag/AgNO3 in THF. The ferrocene couple Fc+/Fc was
used as an external reference. Solutions (THF) of electrolyte (0.3 M
tetra-n-butylammonium hexafluorophosphate) and analyte were also
prepared under an inert atmosphere.
1
and dried to yield 9-PF6 (31 mg, 74%). H NMR (d8-THF, δ): 8.25
(d, J = 6.9 Hz, 3H), 7.64 (d, J = 7.2 Hz, 3H), 7.42 (t, J = 6.9 Hz, 3H),
7.33 (t, J = 6.6 Hz, 3H), 2.44 (br, 6H), 1.22 (s, 18H), 0.86 (s, 18H).
%). 13C {1H} NMR (d8-THF, δ): 155.2, 144.7, 134.2, 131.8, 131.0,
129.4, 29.3, 21.3, 20.5. 19F{1H} NMR (d8-THF, δ): −72.9 (d, J = 715
Hz). 31P{1H} NMR (d8-THF, δ): 67.5, −142.1 (sep, J = 715 Hz). 8-
5.4. Synthetic Details. 5.4.1. Synthesis of {[SiPiPr3]Ru(NAr)}OTf
(Ar = p-C6H4CF3) (7-CF3). [SiPiPr3]Ru(N2) (100 mg, 0.14 mmol)
was dissolved in 10 mL of Et2O and cooled to −78 °C. p-
CF3C6H4N3 (26 mg, 0.14 mmol) was diluted with 2 mL of Et2O
and also cooled to −78 °C. The azide solution was added
dropwise to the solution of [SiPiPr3]Ru(N2), resulting in an
immediate color change from green to red/purple. The solution
was stirred for 10 min at −78 °C and for 10 min at room
temperature. The solution was cooled to −78 °C again, and
AgOTf (36 mg, 0.14 mmol) was added in one portion. The
solution gradually precipitated a green solid, along with black
Ag metal. The mixture was filtered through Celite, and the
green product was extracted into THF. The dark green solution
was concentrated, and the product was recrystallized from
layering pentane over a concentrated THF solution of green 7-
CF3 to yield crystals suitable for X-ray diffraction (82 mg, 59%).
1H NMR (d8-THF, δ): 8.31 (d, J = 8.0 Hz, 2H), 8.11 (d, J = 7.5
−
OMe, PF6 anion: {[SiPiPr3]Ru(N2)}PF6 (17 mg, 0.020 mmol) was
dissolved in 6 mL of THF and cooled to −78 °C. p-MeOC6H4N3 (2.9
mg, 0.020 mmol) was added to the solution in one portion, resulting
in an immediate color change to red. The solution was stirred for 10
1
min, and concentrated to yield red 8-OMe (18 mg, 94%). H NMR
(d8-THF, δ): 8.26 (d, J = 7.0 Hz, 3H), 7.61 (br, 3H), 7.41 (t, J = 8.0
Hz, 3H), 7.32 (t, J = 8.0 Hz, 3H), 7.07 (d, J = 9.0 Hz, 2H), 6.97 (t, J =
9.0 Hz, 3H), 3.74 (s, 3H), 2.15 (br, 6H), 1.13 (s, 18H), 0.78 (s, 18H).
13C{1H} NMR (d8-THF, δ): 159.8, 155.1, 144.5, 134.6, 132.8, 131.7,
131.3, 129.6, 122.6, 117.2, 56.6, 30.4, 21.2, 20.7. 19F{1H} NMR (d8-
THF, δ): −71.7 (d, J = 711 Hz). 31P{1H} NMR (d8-THF, δ): 72.6,
−
−142.9 (sep, J = 711 Hz). 7-OMe, PF6 anion: The synthesis of 7-
OMe was performed in a three step sequence, without isolation of
intermediate products, 9-PF6 and 8-OMe. [SiPiPr3]Ru(N2) (50 mg,
0.068 mmol) was dissolved in 8 mL of THF. AgPF6 (17 mg, 0.068
mmol) was dissolved in 2 mL of THF. Both were cooled to −78 °C
and the AgPF6 solution was added dropwise to the solution of
[SiPiPr3]Ru(N2), leading to an immediate color change from green to
dark brown. The mixture was stirred at −78 °C for 5 min, and stirred
at room temperature for 10 min. The mixture was filtered through
Celite, and the filtrate was cooled to −78 °C. A THF solution of p-
MeOC6H4N3 (20 mg, 0.14 mmol) was added dropwise to the filtrate,
resulting in a color change to red. The red solution of 8-OMe was
stirred for 10 min, and charged into a 100 mL quartz flask. The
Hz, 3H), 7.77 (m, 3H), 7.59 (d, J = 8.5 Hz, 2H), 7.52 (t, J = 8.0
Hz, 3H), 7.41 (t, J = 7.5 Hz, 3H), 2.76 (br, 6H), 1.11 (dd, J =
14.5, 6.5 Hz, 18H), 0.52 (m, 18H). 13C{1H} NMR (d8-THF,
δ): 156.1, 141.2, 133.7, 132.0, 130.4, 129.2, 128.3, 120.8, 32.4,
19.5, 19.2. 19F{1H} NMR (d8-THF, δ): −62.3, −77.3. 31P{1H}
NMR ( d 8 - T H F , δ ) : 1 0 9 . 8 . A n a l . C a l c d f o r
C44H58NO3F6SiP3SRu: C, 51.96; H, 5.75; N. 1.38. Found: C,
51.59; H, 5.76; N, 1.25.
5.4.2. Synthesis of {[SiPiPr3]Ru(N3Ar)}BArF (8-OMe, Ar =
4
C6H4OMe). {[SiPiPr3]Ru(N2)}BArF (46 mg, 0.029 mmol) was
4
6704
dx.doi.org/10.1021/ja211603f | J. Am. Chem. Soc. 2012, 134, 6695−6706

Journal of the American Chemical Society
Article
solution was photolyzed and the progress of the reaction was
monitored by 31P{1H} NMR spectroscopy. After the conversion was
complete (approximately 1 h), the solution was concentrated, and the
residues were washed with Et2O to yield green 7-OMe (57 mg, 86%
overall). Recrystallization by layering pentane over a THF solution of
5.4.8. Preparation of EPR Samples for Detecting Ru(I) Azide
Adducts, 11-R. A solution of [SiPiPr3]Ru(N2) (0.7 mg in 0.1 mL 2-
MeTHF) was added to an EPR tube and the solution was frozen inside
the glovebox using the glovebox cold well. A solution of azide (>30
equiv in 0.1 mL of 2-MeTHF) was layered above the frozen
[SiPiPr3]Ru(N2) solution and also frozen. The tube was quickly taken
out of the glovebox and immersed in liquid nitrogen. The tube was
immersed in a dry ice/isopropyl alcohol bath quickly to thaw the
solution, and the tube was rapidly shaken to homogenize the solution.
The tube was reimmersed in liquid nitrogen and frozen, and was
placed inside the EPR cavity for measurement. For detecting 11-OMe,
several freeze−pump−thaw cycles were additionally applied to remove
N2 to favor formation of 11-OMe.
1
7-OMe yielded crystals suitable for X-ray diffraction. H NMR (d8-
THF, δ): 8.18 (d, J = 8.4 Hz, 2H), 8.12 (d, J = 7.2 Hz, 3H), 7.77 (m,
3H), 7.52 (t, J = 7.2 Hz, 3H), 7.41 (t, J = 6.9 Hz, 3H), 6.93 (d, J = 9.0
Hz, 2H), 3.85 (s, 3H), 2.76 (m, 6H), 1.11 (m, 18H), 0.55 (m, 18H).
13C NMR (d8-THF, δ): 162.3 (m), 157.3(m), 142.7 (m), 134.8 (m),
132.8, 131.2, 130.1, 125.5, 117.9, 57.3, 33.4 (m), 20.6, 20.3. 31P NMR
(d8-THF, δ): 106.5, −143.6 (sep, J = 738 Hz). Anal. Calcd for
C43H61NOF6SiP4Ru: C, 52.97; H, 6.31; N, 1.44. Found: C, 52.23; H,
6.18; N, 1.32.
5.4.5. Synthesis of [SiPiPr2P(NAr)Ru]PF6 (10). [SiPiPr3]Ru(N2)
(17 mg, 0.023 mmol) was dissolved in 5 mL of THF. AgPF6 (5.8 mg,
0.023 mmol) was dissolved in 1 mL of THF. Both solutions were
cooled to −78 °C and the AgPF6 solution was added dropwise to the
[SiPiPr3]Ru(N2) solution, resulting in an immediate darkening of the
solution. The mixture was stirred for 2 min at low temperature, and
then stirred for 2 min at room temperature. The mixture was filtered
through Celite, and the filtrate was cooled to −78 °C again. p-
MeOC6H4N3 (6.9 mg, 0.046 mmol) in 1 mL of THF was added to the
solution dropwise, leading to a color change to red. The solution was
stirred for 5 min, and charged into a 100 mL quartz flask. The solution
was photolyzed for two days. The solution was concentrated, and
washed with Et2O to yield 10 (14 mg, 62%). Crystals suitable for X-ray
diffraction were obtained by vapor diffusion of pentane into a
ASSOCIATED CONTENT
* Supporting Information
■
S
NMR spectra, EPR spectra with simulation parameters, DFT
calculations, details of crystallographic data, and details of
calculation methods and results. This material is available free
of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
jpeters@caltech.edu
■
Notes
The authors declare no competing financial interest.
1
concentrated solution of 10. H NMR (d8-THF, δ): 7.92−7.83 (m,
ACKNOWLEDGMENTS
2H), 7.66 (dd, J = 7.5, 4.2 Hz, 2H), 7.49 (d, J = 7.2 Hz, 2H), 7.43−
7.23 (m, 6H), 6.66 (d, J = 6.0 Hz, 1H), 5.47 (d, J = 6.0 Hz, 1H), 5.31
(t, J = 6.9 Hz, 2H), 3.88 (s, 3H), 3.19 (sep, J = 6.6 Hz, 1H), 3.07 (sep,
J = 5.7 Hz, 1H), 2.51−2.11 (m, 3H), 1.70−0.81 (m, 30H), 0.68 (dd, J
= 14.7, 7.2 Hz, 3H), −0.33 (dd, J = 12.9, 7.8 Hz, 3H). 13C{1H} NMR
(d8-THF, δ): 155.2, 149.4 (d, J = 45.1 Hz), 148.1 (d, J = 15.7 Hz),
144.8 (d, J = 43.9 Hz), 142.4 (d, J = 13.2 Hz), 140.8, 137.3 (d, J = 17.3
Hz), 136.2 (d, J = 18.6 Hz), 134.2 (m), 132.7, 132.1 (d, J = 17.1 Hz),
131.1, 130.8, 130.3, 129.8, 129.1, 126.2, 116.0 (br), 90.9, 84.9, 81.8,
81.3, 81.2, 57.6, 56.6 (br), 35.9 (d, J = 29.2 Hz), 33.5 (d, J = 12.4 Hz),
32.3 (J = 18.9 Hz), 31.0, 30.4, 28.3, 27.9, 27.2, 24.4, 23.8, 23.4, 23.0,
21.4, 20.8, 20.0, 17.8, 17.7, 16.3, 2.2. 31P{1H} NMR (d8-THF, δ): 85.4
(d, J = 28.3 Hz), 73.9 (d, J = 28.3 Hz), 32.4.
■
We acknowledge the generosity of NSF (Grant CHE-0750234)
and the Gordon and Betty Moore Foundation for support of
this work. M.-E.M. acknowledges a Fellowship for Advanced
Researchers from the Swiss National Science Foundation. We
also thank Dr. Hill Harman for insightful discussions. Dr. Peter
Muller, Larry Henling, and Charlene Tsay provided crystallo-
̈
graphic assistance.
REFERENCES
■
(1) Selected reviews: (a) Muller, P.; Fruit, C. Chem. Rev. 2003, 103,
2905. (b) Che, C.-M.; Lo, V. K.-Y.; Zhou, C.-Y.; Huang, J.-S. Chem.
̈
5.4.6. Synthesis of [SiPiPr3]Ru(NAr) (Ar = C6H4OMe, 5-OMe).
{[SiPiPr3]Ru(NAr)}PF6 (Ar = C6H4OMe, 7-OMe) (11 mg, 0.011
mmol) was dissolved in 4 mL of THF and cooled to −78 °C. CoCp2
(2.0 mg, 0.011 mmol) was added in one portion and the resulting
solution was stirred at −78 °C for 15 min. The solution was warmed
to room temperature and stirred for an additional 15 min. The
solution was concentrated, and the product was extracted into
pentane. The pentane solution was filtered through Celite. The
extraction process was repeated once more to yield red/brown 5-OMe
Rev. Soc. 2011, 40, 1950.
(2) Selected Examples: (a) Breslow, R.; Gellman, S. H. J. Chem. Soc.,
Chem. Commun. 1982, 1400. (b) Breslow, R.; Gellman, S. H. J. Am.
Chem. Soc. 1983, 105, 6728. (c) Mansuy, D.; Mahy, J. −P.; Dureault,
A.; Bedi, G.; Battioni, P. J. Chem. Soc., Chem. Commun. 1984, 1161.
(d) Mahy, J. P.; Battioni, P.; Mansuy, D. Tetrahedron Lett. 1988, 29,
1927. (e) Evans, D. A.; Faul, M. M.; Bilodeau, M. T. J. Org. Chem.
1991, 56, 6744. (f) Muller, P.; Baud, C.; Jacquier, Y. Tetrahedron 1996,
̈
52, 1543. (g) Albone, D. P.; Aujla, P. S.; Taylor, P. C.; Challenger, S.;
Derrick, A. M. J. Org. Chem. 1998, 63, 9569. (h) Zhou, X.-G.; Yu, X.-
Q.; Huang, J.-S.; Che, C.-M. Chem. Commun. 1999, 2377. (i) Au, S.-
M.; Huang, J.-S.; Yu, W.-Y.; Fung, W.-H.; Che, C.-M. J. Am. Chem. Soc.
1999, 121, 9120. (j) Kohmura, Y.; Katsuki, T. Tetrahedron Lett. 2001,
42, 3339. (k) Simkhovich, L.; Gross, Z. Tetrahedron Lett. 2001, 42,
8089. (l) Espino, C. G; Du Bois, J. Angew. Chem., Int. Ed. 2001, 40,
598. (m) Mar Díaz-Requejo, M.; Belderraín, T. R.; Carmen Nicasio,
1
(4.6 mg, 52%). H NMR (C6D6, δ): 9.0, 7.8, 6.8, 5.9, 4.0 (extending
from 8 to 0 ppm). μeff (Evan’s Method, C6D6/C6H6) = 1.5 μB. UV−vis
(in THF): (nm, ε [mol−1 cm−1]), 473 (4200), 737 (1700).
5.4.7. Synthesis of {[SiPiPr3]RuCl}PF6 (12). [SiPiPr3]RuCl (30 mg,
0.040 mmol) was dissolved in 6 mL of THF. AgPF6 (10 mg, 0.040
mmol) was dissolved in 1 mL of THF. Both were cooled to −78 °C
and the AgPF6 solution was added to the [SiPiPr3]RuCl solution. A
gradual color change from red to brown took place. The solution was
stirred at −78 °C for 15 min., and was subsequently stirred at room
temperature for 30 min. The mixture was filtered through Celite, and
the filtrate was concentrated. The residues were washed with ether and
benzene, and the product was extracted into THF and filtered through
Celite. Concentration of the purple filtrate yielded 12 (22 mg, 61%).
1H NMR (CD2Cl2, δ): 12 (extending from 16 to 8 ppm), 9.7, 6.7, 6.4.
μeff (Evan’s Method, CD2Cl2) = 1.73 μB. UV−vis (in CH2Cl2): (nm, ε
[mol−1 cm−1]), 454 (290), 494 (1300). Anal. Calcd for
C36H54F6SiP4RuCl: C, 48.62; H, 6.12; N, 0. Found: C, 47.95; H,
6.05; N, 0.00.
́
M.; Trofimenko, S.; Perez, P. J. J. Am. Chem. Soc. 2003, 125, 12078.
(n) Cui, Y.; He, C. Angew. Chem., Int. Ed. 2004, 43, 4210. (o) Zdilla,
M. J.; Abu-Omar, M. M. J. Am. Chem. Soc. 2006, 128, 16971.
(p) Milczek, E.; Boudet, N.; Blakey, S. Angew. Chem., Int. Ed. 2008, 47,
6825.
(3) A recent report describes the use of a hypervalent bromine
reagent for the amination of unactivated C−H bonds. See; Ochiai, M.;
Miyamoto, K.; Kaneaki, T.; Hayashi, S.; Nakanish, W. Science 2011,
332, 448.
(4) Selected examples: (a) Kwart, H.; Khan, A. A. J. Am. Chem. Soc.
1965, 89, 1951. (b) Cenini, S.; Tollari, S.; Penoni, A.; Cerada, C. J.
6705
dx.doi.org/10.1021/ja211603f | J. Am. Chem. Soc. 2012, 134, 6695−6706

Journal of the American Chemical Society
Article
Mol. Catal. A: Chem. 1999, 137, 135. (c) Cenini, S.; Gallo, E.; Penoni,
A.; Ragaini, F.; Tollari, S. Chem. Commun. 2000, 2265. (d) Omura, K.;
Murakami, M.; Uchida, T.; Irie, R.; Katsuki, T. Chem. Lett. 2003, 32,
354. (e) Omura, K.; Uchida, T.; Irie, R.; Katsuki, T. Chem. Commun.
2004, 2060. (f) Li, Z.; Quan, R. W.; Jacobsen, E. N. J. Am. Chem. Soc.
1995, 117, 5889. (g) Liu, Y.; Che, C.-M. Chem.−Eur. J. 2010, 16,
Sadaoka, K.; Iwasa, T.; Wada, K.; Mitsudo, T. Organometallics 2005,
24, 905.
(23) (a) Schuster, G. B.; Platz, M. S. Photochemistry of Phenyl
Azide. In Advances in Photochemistry; Volman, D., Hammond, G.,
Neckers, D., Eds.; John Wiley & Sons, Inc.; New York, 1992; Vol. 17,
p 69. (b) Platz, M. S. Nitrenes. In Reaction Intermediate Chemistry;
Moss, R. A., Platz, M. S., Jones, M., Jr., Eds.; John Wiley & Sons, Inc.;
New York, 2004; p 501.
10494. (h) Bach, T.; Korber, C. Eur. J. Org. Chem. 1999, 1033.
̈
(i) Bach, T.; Korber, C. Tetrahedron Lett. 1998, 39, 5015. (j) Goa, G.-
̈
(24) Reiser, A.; Willets, F. W.; Terry, G. C.; Williams, V.; Marley, R.
Trans. Faraday Soc. 1968, 64, 3265.
Y.; Jones, J. E.; Vyas, R.; Harden, J. D.; Zhang, X. P. J. Org. Chem.
2006, 71, 6655. (k) Badiei, Y. M.; Dinescu, A.; dai, X.; Palomino, R.
M.; Heinemann, F. W.; Cundari, T. R.; Warren, T. H. Angew. Chem.,
Int. Ed. 2008, 47, 9961. (l) Stokes, B. J.; Dong, H.; Leslie, B. E.;
Pumphrey, A. L.; Driver, T. G. J. Am. Chem. Soc. 2007, 129, 7500.
(5) (a) Proulx, G.; Bergman, R. G. J. Am. Chem. Soc. 1995, 117, 6382.
(b) Proulx, G.; Bergman, R. G. Organometallics 1996, 15, 684.
(6) Gololobox, Y. G.; Zhmurova, I. N.; Kasukhin, L. F. Tetrahedron
1981, 37, 437.
(25) Borden, W. T.; Gritsan, N. P.; Hadad, C. M.; Karney, W. L.;
Kemnitz, C. R.; Platz, M. S. Acc. Chem. Res. 2000, 33, 765.
(26) Liang, T.-Y.; Schuster, G. B. J. Am. Chem. Soc. 1987, 109, 7803.
(27) (a) Smolinsky, G.; Wasserman, E.; Yager, W. A. J. Am. Chem.
Soc. 1962, 84, 3220. (b) Herbert Hall, J..; Fargher, J. A.; Gisler, M. R. J.
Am. Chem. Soc. 1978, 100, 2029.
(28) Other complexes assigned as containing an radical on the NAr
ligand: (a) Kogut, E.; Wiencko, H. L.; Zhang, L.; Cordeau, D. E.;
Warren, T. H. J. Am. Chem. Soc. 2005, 127, 11248. (b) Lu, C. C.;
(7) Fickes, M. G.; Davis, W. M.; Cummins, C. C. J. Am. Chem. Soc.
1995, 117, 6384.
George, S. D.; Weyhermuller, Bill, E.; Bothe, E.; Wieghardt, K. Angew.
̈
(8) γ-N bound azide complexes: (a) Guillemot, G.; Solari, E.;
Floriani, C. Organometallics 2001, 20, 607. (b) Evans, W. J.; Mueller,
T. J.; Ziller, J. W. J. Am. Chem. Soc. 2009, 131, 2678. (c) Hanna, T. A.;
Baranger, A. M.; Bergman, R. G. Angew. Chem., Int. Ed. Engl. 1996, 35,
653. α-N bound azide complex: (d) Barz, M.; Herdtweck, E.; Thiel,
W. R. Angew. Chem., Int. Ed. Engl. 1998, 37, 2262. (e) Albertin, G.;
Chem., Int. Ed. 2008, 47, 6384. (c) King, E. R.; Hennessy, E. T.; Betley,
T. A. J. Am. Chem. Soc. 2011, 133, 4917. (d) Lyaskovskyy, V.; Olivos
Suarez, A. I. O.; Lu, H.; Jiang, H.; Zhang, X. P.; de Bruin, B. J. Am.
Chem. Soc. 2011, 133, 12264.
(29) Takaoka, A.; Peters, J. C. Inorg. Chem. 2012, 51, 16.
(30) Sauouma, C.; Peters, J. C. Coord. Chem. Rev. 2011, 255, 920.
(31) Moret, M.-E.; Peters, J. C. Angew. Chem., Int. Ed. 2011, 50, 2063.
(32) Heyduk, A. F.; Zarkesh, R. A.; Nguyen, A. I. Inorg. Chem. 2011,
50, 9849.
(33) Wang, S. H.; Mandimutsira, B. D.; Todd, R.; Ramdhanie, B.;
Fox, J. P.; Goldberg, D. P. J. Am. Chem. Soc. 2004, 126, 18.
(34) Palmer, M. J.; Kenny, J. A.; Walsgrove, T.; Kawamoto, A. M.;
Wills, M. J. J. Chem. Soc., Perkin Trans. 1 2002, 416.
(35) Hu, M.; Li, J.; Yao, S. Q. Org. Lett. 2008, 10, 5529.
(36) Weitz, I. S.; Rabinovitz, M. J. Chem. Soc., Perkin Trans. 1 1993,
117.
́
Antonuitti, S.; Baldan, D.; Castro, J.; García-Fontan, S. Inorg. Chem.
2008, 47, 742. For a Cu complex that binds 1-adamantyl azide at the
γ-N and a Ag complex that binds it at the α-N using the same ancillary
ligand see, (f) Rasika Dias, H. V.; Polach, S. A.; Goh, S.-K.; Archibong,
E. F.; Marynick, D. S. Inorg. Chem. 2000, 39, 3894.
(9) Waterman, R.; Hillhouse, G. L. J. Am. Chem. Soc. 2008, 130,
12628.
(10) Mankad, N. P.; Muller, P.; Peters, J. C. J. Am. Chem. Soc. 2010,
̈
132, 4083.
(11) (a) Mankad, N. P.; Whited, M. T.; Peters, J. C. Angew. Chem.,
Int. Ed. 2007, 46, 5768. (b) Whited, M. T.; Mankad, N. P.; Lee, Y.;
Oblad, P. F.; Peters, J. C. Inorg. Chem. 2009, 48, 2507. (c) Lee, Y.;
Mankad, N. P.; Peters, J. C. Nat. Chem. 2010, 2, 558.
(12) (a) Hansert, B.; Vahrenkamp, H. J. Organomet. Chem. 1993, 459,
265. (b) Harrold, N. D.; Waterman, R.; Hillhouse, G. L.; Cundari, T.
R. J. Am. Chem. Soc. 2009, 131, 12872.
(37) Stoll, S.; Schweiger, A. J. Magn. Reson. 2006, 178, 42.
(38) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112.
(13) The only prior report was by Cenini, in which azoarene was a
minor product of catalysis. See; Ragaini, F.; Penoni, A.; Gallo, E.;
Tollari, S.; Gotti, C. L.; Lapadula, M.; Mangioni, E.; Cenini, S. Chem.−
Eur. J. 2003, 9, 249.
(14) N−N coupling to yield azobenzenes from metal imide
complexes: Zarkesh, R. A.; Ziller, J. W.; Heyduk, A. F. Angew.
Chem., Int. Ed. 2008, 47, 4715.
(15) (a) Schroder, D.; Shaik, S.; Schwarz, H. Acc. Chem. Res. 2000,
̈
33, 139. (b) Shaik, S.; Hirao, J.; Kumar, D. Acc. Chem. Res. 2007, 40,
532.
(16) Takaoka, A.; Gerber, L. C. H.; Peters, J. C. Angew. Chem., Int. Ed.
2010, 49, 4088.
(17) Dhuri, S. N.; Seo, M. S; Lee, Y.-M.; Hirao, H.; Wang, Y.; Nam,
W.; Shaik, S. Angew. Chem., Int. Ed. 2008, 47, 3356.
(18) (a) Danopoulos, A. A.; Wilkinson, G.; Hussain-Bates, B.;
Hursthouse, M. B. Polyhedron 1992, 11, 2961. (b) Walstron, A. N.;
Fullmer, B. C.; Fan, H.; Pink, M.; Buschhorn, D. T.; Caulton, K. G.
Inorg. Chem. 2008, 47, 9002.
(19) Takaoka, A.; Mendiratta, A.; Peters, J. C. Organometallics 2009,
28, 3744.
(20) Addison, A. W.; Rao, T. N.; Van Rijn, J. J.; Verschoor, G. C. J.
Chem. Soc., Dalton Trans. 1984, 1349.
(21) Select examples: (a) Klapotke, T. M.; Krumm, B.; Piotrowski,
̈
H.; Polborn, K.; Holl, G. Chem.−Eur. J. 2003, 9, 687. (b) Kawano, M.;
Takayama, T.; Uekusa, H.; Ohashi, Y.; Ozawa, Y.; Matsubara, K.;
Imabayashi, H.; Mitsumi, M.; Toriumi, K. Chem. Lett. 2003, 32, 922.
(22) Selected examples: (a) Hayashida, T.; Nagashima, H. Organo-
metallics 2002, 21, 3884. (b) Kondo, T.; Tsunawaki, F.; Ura, Y.;
6706
dx.doi.org/10.1021/ja211603f | J. Am. Chem. Soc. 2012, 134, 6695−6706