Copper(II) anilides in sp3 C-H amination

Article abstract of DOI:10.1021/ja5026547

We report a series of novel β-diketiminato copper(II) anilides [Cl₂NN]Cu-NHAr that participate in C-H amination. Reaction of H ₂NAr (Ar = 2,4,6-Cl₃C₆H₂ (Ar ^Cl3), 3,5-(CF₃)₂C₆H₃ (Ar^F6), or 2-py) with the copper(II) t-butoxide complex [Cl ₂NN]Cu-^tOBu yields the corresponding copper(II) anilides [Cl₂NN]Cu-NHAr. X-ray diffraction of these species reveal three different bonding modes for the anilido moiety: κ¹-N in the trigonal [Cl₂NN]Cu-NHAr^Cl3 to dinuclear bridging in {[Cl₂NN]Cu}₂(μ-NHAr^F6)₂ and κ²-N,N in the square planar [Cl₂NN] Cu(κ²-NH-2-py). Magnetic data reveal a weak antiferromagnetic interaction through a π-stacking arrangement of [Cl₂NN]Cu- NHAr^Cl3; solution EPR data are consistent with monomeric species. Reaction of [Cl₂NN]Cu-NHAr with hydrocarbons R-H (R-H = ethylbenzene and cyclohexane) reveals inefficient stoichiometric C-H amination with these copper(II) anilides. More rapid C-H amination takes place, however, when ^tBuOO^tBu is used, which allows for HAA of R-H to occur from the ^tBuO? radical generated by reaction of [Cl ₂NN]Cu and ^tBuOO^tBu. The principal role of these copper(II) anilides [Cl₂NN]Cu-NHAr is to capture the radical R? generated from HAA by ^tBuO? to give functionalized aniline R-NHAr, resulting in a novel amino variant of the Kharasch-Sosnovsky reaction.

Full text of DOI:10.1021/ja5026547

Article
pubs.acs.org/JACS
Copper(II) Anilides in sp3 C‑H Amination
Eun Sil Jang,†,‡ Claire L. McMullin,§,# Martina Kaß,∥ Karsten Meyer,∥ Thomas R. Cundari,*,§
̈
and Timothy H. Warren*,†,‡
†Department of Chemistry, Georgetown University, Box 571227-1227, Washington, D.C. 20057, United States
‡Department of Chemistry, Purdue University, 560 Oval Drive, West Lafayette, Indiana 47907, United States
§Department of Chemistry, Center for Advanced Scientific Computing and Modeling (CASCaM), University of North Texas,
Denton, Texas 76203, United States
∥Department of Chemistry and Pharmacy, Friedrich-Alexander-University Erlangen-Nuremberg, Egerlandstrasse 1, 91058 Erlangen,
Germany
S
* Supporting Information
ABSTRACT: We report a series of novel β-diketiminato copper-
(II) anilides [Cl2NN]Cu-NHAr that participate in C-H amination.
Reaction of H2NAr (Ar = 2,4,6-Cl3C6H2 (ArCl3), 3,5-(CF3)2C6H3
(ArF6), or 2-py) with the copper(II) t-butoxide complex [Cl2NN]-
Cu-tOBu yields the corresponding copper(II) anilides [Cl2NN]-
Cu-NHAr. X-ray diffraction of these species reveal three different
bonding modes for the anilido moiety: κ1-N in the trigonal
[Cl2NN]Cu-NHArCl3 to dinuclear bridging in {[Cl2NN]Cu}2(μ-
NHArF6)2 and κ2-N,N in the square planar [Cl2NN]Cu(κ2-NH-2-
py). Magnetic data reveal a weak antiferromagnetic interaction
through a π-stacking arrangement of [Cl2NN]Cu-NHArCl3; solution EPR data are consistent with monomeric species. Reaction
of [Cl2NN]Cu-NHAr with hydrocarbons R-H (R-H = ethylbenzene and cyclohexane) reveals inefficient stoichiometric C-H
amination with these copper(II) anilides. More rapid C-H amination takes place, however, when tBuOOtBu is used, which allows
for HAA of R-H to occur from the tBuO• radical generated by reaction of [Cl2NN]Cu and tBuOOtBu. The principal role of these
copper(II) anilides [Cl2NN]Cu-NHAr is to capture the radical R• generated from HAA by tBuO• to give functionalized aniline
R-NHAr, resulting in a novel amino variant of the Kharasch−Sosnovsky reaction.
common, catalyst systems based on Earth abundant metals such
as Mn,9 Fe,10,11 Co,12 and Cu6,7,13 may also be employed.
General non-nitrene routes involving metal-amide ([M]-
NRR′) intermediates are potentially more versatile and would
allow for the construction of both 2° and 3° amines from C-H
substrates R-H when 1° (H2NR′) and 2° (HNR1R2) amine
reagents are employed. For instance, White and Liu illustrated
the formation of linear amines from terminal alkenes using a
Pd-catalyzed protocol with N-tosylcarbamates HN(Ts)C(O)-
OR.14 C-H activation provides a Pd(II) π-allyl intermediate
subject to intermolecular attack by a mildly basic amide anion
such as [N(SO2Ph)C(O)OMe]−. Other mildly basic 2° amines
such as HNPh2 may be similarly employed for allylic C-H
amination.15 Other Pd-catalyzed variations take advantage of
directed C-H activation to give amide-aryl intermediates subject
to C−N reductive elimination such as in the synthesis of
carbazoles.16 Additionally, peroxides have been used as oxidants
for sp3 C-H amination with electron-poor nitrogen sources
such as organic amides.17,18 In some cases, peroxycarbamates
ROOC(O)NHR could be directly employed with metal amides
[CuII]-NHR suggested as intermediates.19
INTRODUCTION
■
C-H amination allows the construction of C-N bonds directly
from C-H bonds and leads to rapid incorporation of N atoms
into molecules.1,2 This reaction proceeds in fewer reaction steps
and with lower potential environmental impact as compared to
traditional functional group manipulations. Perhaps the best
studied sp3 C-H amination systems involve metal-nitrenes [M]
= NR′ (Scheme 1).3 These reactions produce secondary amines
(amides) R-NHR′ employing iminoiodinanes PhINR′4 or in
some cases organoazides N3R′.5−7 Alternatively, Lebel and co-
workers have used N-tosyloxycarbamates that provide nitrene-
like reactivity.8 While dirhodium or ruthenium catalysts are
Scheme 1. C-H Amination: Nitrene and Amide
Intermediates
Received: March 26, 2014
Published: June 18, 2014
© 2014 American Chemical Society
10930
dx.doi.org/10.1021/ja5026547 | J. Am. Chem. Soc. 2014, 136, 10930−10940

Journal of the American Chemical Society
Article
A common limitation to many C-H amination systems is the
requirement for electron-deficient N-substituents in both the
nitrene and non-nitrene protocols. While C-H amination has
been used in complex molecule synthesis,2,20 often these
electron-withdrawing N-based activating groups (e.g., sulfonyl,
carbamoyl) require deprotection and refunctionalization for the
synthesis of new alkyl- and aryl-amines via C-H functionaliza-
tion. This detracts from the potential atom economy that C-H
amination can offer. Notable exceptions are reactions that make
use of organoazides and proceed via Fe10 and Cu7,21 alkyl- or
aryl-nitrene intermediates as well as net C-H amination of
carbonyl α-C-H bonds with 2° amines via α-C-Br carbonyl
intermediates formed upon bromination of carbonyl substrates
with CuBr2.22
[Cl2NN]Cu-NHAd (3) led to the development of a catalytic
cycle that features copper(II) amides [CuII]-NR1R2 as key
intermediates (Scheme 2).23 The copper(I) β-diketiminate
t
[Cl2NN]Cu (1) reacts with BuOOtBu to give the copper(II)
alkoxide [Cl2NN]Cu-OtBu (2), which undergoes reversible
exchange with H2NAd (Keq = 4(1) in benzene) to give the
copper(II) amide [Cl2NN]Cu-NHAd (3). This three coor-
dinate amido complex stoichiometrically aminates hydro-
carbons such as ethylbenzene and indane. Kinetic studies
demonstrated that the rate limiting step is H atom abstraction
of R-H by [CuII]-NHAd to give a radical R• that may rapidly
combine with an additional equivalent of [Cl2NN]Cu-NHAd to
give the functionalized amine R-NHAd. The HAA step
proceeds at modest rates with benzylic substrates. Activation
parameters ΔH⧧ = 11.4(4) kcal/mol, ΔS⧧ = −38.7(14) eu, and
ΔG⧧(298K) = 22.9(8) kcal/mol were obtained in the
stoichiometric reaction of 3 with indane. DFT analysis predicts
a modest N-H bond strength of 68 kcal/mol for [Cl2NN]Cu-
(NH2Ad) corresponding to an uphill HAA step that contributes
to the large KIE in the C-H amination of ethylbenzene (kH/kD
= 70(9)).23
C-H amination with anilines revealed that oxidation of the
aniline H2NAr to diazenes ArNNAr competes with
successful C-H amination (Scheme 3b).24 Nonetheless, C-H
amination may compete successfully with diazene formation,
particularly for electron-poor anilines, to give high yields of C-
H amination products with substrates such as ethylbenzene or
cyclohexene.
To expand the range of amine coupling partners in C-H
functionalization reactions, we recently reported a novel Cu-
based catalyst system that employs 1° and 2° alkyl- and
23,24
t
arylamines with the mild oxidant BuOOtBu.
Strong C-H
bonds such as those in cyclohexane may be efficiently
functionalized with amines such as H2NAd and H2NArCl3
(ArCl3 = 2,4,6-trichloroaniline, Ad = 1-adamantyl) while
secondary amines such as morpholine and N-methylaniline
may also be used (Schemes 2 and 3a).
Scheme 2. Catalytic Cycle for Amination with Alkylamines
To provide a more concrete mechanistic foundation for C-H
amination with anilines employing the [Cl2NN]Cu catalyst, we
targeted the synthesis of the anticipated copper(II) anilide
intermediates [Cl2NN]Cu-NHAr. Based on their disinclination
to diazene formation, we synthetically target copper(II) anilide
intermediates based on electron-poor anilines for our studies.
Herein, we discuss their synthesis, structures, and reactivity in
the context of catalytic C-H amination. Calculations at the
hybrid ONIOM(BP86/6-311+G(d):UFF) level of theory have
also been completed to examine the reactivity of copper(II)
anilides and elucidate experimental observations.
RESULTS AND DISCUSSION
■
Scheme 3. C-H Amination with Aromatic Amines
Synthesis and Structure of Copper(II) Anilides. The
reaction of [Cl2NN]Cu-OtBu (2) with electron-deficient
anilines H2NAr at RT in ether provides copper(II) anilides
that may be obtained from ether/pentane as crystals in
moderate yields (Scheme 4). For instance, blue crystals of
[Cl2NN]Cu-NHArCl3 (4) are obtained from ether/pentane at
−35 °C in 62% yield. The X-ray structure of 4 reveals two
independent molecules featuring three coordinate copper
centers (Figure 1). The Cu-Nanilide distances Cu1−N3 and
Cu2−N6 of 1.847(3) and 1.841(3) Å are similar to those found
in the copper(II) amide [Cl2NN]Cu-NHAd (3) (1.839(9) Å)23
and the diphenylamide [Me2NN]Cu-NPh2 (1.841(6) Å).25
These Cu−Nanilide distances may be compared to the somewhat
longer Cu−Nβ‑dik distances [Cu1−N1 1.895(3), Cu1−N2
1.926(3), Cu2−N4 1.893(3), Cu2−N5 1.919(3)]. The Cu−
N−Cipso angles are each 130.0(2)°, considerably wider than the
Cu−Namide−C angle in [Cl2NN]Cu-NHAd (117.69(11)°).
Perhaps the most striking feature of the X-ray structure of 4
is the interaction of the two crystallographically independent
molecules of 4 through a π-stacking interaction (Figure S32 in
the Supporting Information). The slightly offset carbon atoms
of each N-aryl ring are found separated from their nearest
Catalytic reactivity studies of [Cl2NN]Cu (1) with H2NAd
t
and BuOOtBu informed by the isolation and reactivity of the
copper(II) alkoxide [Cl2NN]Cu-OtBu (2) and amide
10931
dx.doi.org/10.1021/ja5026547 | J. Am. Chem. Soc. 2014, 136, 10930−10940

Journal of the American Chemical Society
Article
copper anilides in the solid state. Figure 2 shows plots of μeff vs
T for compounds 4−6. Based on the dimeric X-ray structures
of 4 and 5, molecular weights in the calculations of the
magnetic data were based on the dimeric compounds. For
compound 4, μeff ≈ 2.5 μB, which is essentially the spin-only
value of 2.45 μB expected for two unpaired electrons per
dimeric unit. A sharp drop in μeff at temperatures lower than 50
K is indicative of antiferromagnetic coupling. Modeling these
Scheme 4. Synthesis of Copper(II) Anilides 4−6
̂
magnetic data with the Hamiltonian H = −2JS1S2 and the g
value of 2.09, taken from EPR data (see below), gives a good fit
from which J12 = −10.3 cm−1 is derived. Thus, at low
temperature the two copper centers couple antiferromagneti-
cally, likely mediated by the π-stacked arene rings in 4. The
solid state magnetic data for compounds 5-dimer and 6 are less
remarkable. The magnetic moment at RT for 5-dimer is 2.55
μB and exhibits ferromagnetic coupling of 56 cm−1, whereas
mononuclear 6 behaves like a simple Curie paramagnet with a
nearly temperature independent magnetic moment of μeff
=
carbon atom in the opposing N-aryl ring within a narrow range
(3.443−3.596 Å) with a centroid−centroid distance of 3.470 Å.
The somewhat less sterically hindered, electron-deficient
aniline H2NArF6 reacts to give the corresponding copper(II)
anilide that may be isolated as navy blue crystals in 53% yield
from pentane. X-ray studies reveal that it exists as the dimer
{[Cl2NN]Cu}2(μ-NHArF6)2 (5-dimer) in the solid state with
the two copper centers separated by 3.102(1) Å (Figure 1).
This distorted square planar structure has a twist angle of 27.6°
between the Nβ‑dik−Cu−Nβ‑dik and Nanilide−Cu−Nanilide planes,
possessing a structure reminiscent of the hydroxide bridged
dimer {[Me2NN]Cu}2(μ-OH)2 (Cu···Cu = 3.05 Å).26 The
bridging Cu−Nanilide distances are ca. 0.2 Å longer than those in
monomeric 4 with distances that span 2.035(16)−2.0448(18)
Å. The Nanilide atoms are separated by 2.650 Å.
Reaction of 2-aminopyridine with 2 in ether leads to a green
solution from which green needles of [Cl2NN]Cu(κ2-NH-2-py)
(6) may be isolated in 25% yield. Though monomeric, it
exhibits a square planar structure due to the κ2-amidopyridine
interaction27 with Cu−Nanilide and Cu−Npy distances of
1.975(2) and 2.102(2) Å, respectively (Figure 1). The Cu−
Nβ‑dik distances are 1.944(2) and 1.948(2) Å.
1.75 μB, close to the expected spin-only value of 1.73 μB for a
CuII, S = 1/2 system.
Solution Studies: EPR and UV−Vis Spectroscopy. EPR
spectra of 4−6 at room temperature in toluene solution are
consistent with monomeric [Cl2NN]Cu-NHAr species. The
EPR spectra of both 4 and 5 at RT each give a broad resonance
at giso = 2.09 and 2.083, respectively, with little resolvable fine
structure (Figures S22 and S24 in the Supporting Information).
The lack of distinguishable Cu hyperfine structure (I = 3/2 for
63/65Cu) indicates rapid relaxation. This is likely due to facile
spin−spin exchange via transient dimeric species {[Cl2NN]Cu-
NHAr}2 with interactions related to those seen in the solid
state structures for 4 and 5. Alternatively, each may possess an
intrinsically broad line width at room temperature. In contrast,
Cu hyperfine and N superhyperfine coupling is apparent in the
RT solution EPR spectra of [Cl2NN]Cu(κ2-NH-2-py), though
significant broadening is also apparent (Figure S26 in the
Supporting Information).
Frozen glass spectra of 4−6 at 80 K (Figure 3) are each
pseudoaxial with g1 (e.g., g(par)) ranging from 2.18 to 2.21,
similar to that found for [Cl2NN]Cu-NHAd and [Me2NN]Cu-
NPh2 (Table 1). The similarity of EPR spectra of 4 and 5
suggests that [Cl2NN]Cu-NHArF6 exists in appreciable
amounts as its monomer in solution. Nonetheless, the square
planar nature of [Cl2NN]Cu(κ2-NH-2-py) leads to a much
Magnetic Behavior of the Copper(II) Anilides.
Prompted by the interesting π-stacking interaction between
the two anilido rings in the X-ray structure of [Cl2NN]Cu-
NHArCl3 (4), we undertook a magnetic investigation of these
Figure 1. X-ray structures of [Cl2NN]Cu-NHArCl3 (4), {[Cl2NN]Cu}2(μ-NHArF6) (5-dimer), and [Cl2NN]Cu(κ2-NH-2-py) (6-κ2).
10932
dx.doi.org/10.1021/ja5026547 | J. Am. Chem. Soc. 2014, 136, 10930−10940

Journal of the American Chemical Society
Article
Figure 2. Variable temperature SQUID magnetization data for copper(II) anilides 4−6 from 2 to 300 K with an applied field of 1 T (three
independent samples each). Data are corrected for underlying diamagnetism, and simulation gives the following parameters: 4, μeff (RT) = 2.42 μB,
J12 = −10.3 cm−1; 5, μeff (RT) = 2.55 μB, J12 = 56 cm−1, tip = −150 × 10−6; 6, μeff (RT) = 1.75 μB, tip = −54.4 × 10−6 (tip = temperature
independent paramagnetism).
Figure 3. Frozen toluene glass EPR spectra of 4−6 at 85 K. Full simulation parameters may be found in Figures S15, S17, and S19 in the Supporting
Information.
complexes is noticeably less than that for the alkylamide
[Cl2NN]Cu-NHAd or the alkoxide [Cl2NN]Cu-OtBu.23 This is
likely a result of the ability of the unpaired electron density in
[Cl2NN]Cu-N(R)Ar species to delocalize over the N-aryl ring
(Figure 4).
The UV−vis spectra of copper(II) anilides 4 and 5 are very
similar, suggesting that in solution 5 is essentially dissociated to
monomeric [Cl2NN]Cu-NHArF6. Each complex exhibits two
low energy optical bands between 700 and 900 nm with high
molar absorptivities (3000−5000 M−1 cm−1) in aromatic
solvents such as ethylbenzene (Table 2). The solution spectra
of 4 and 5 are similar to that of [Me2NN]Cu-NPh2, which has
two low energy bands at 774 nm (1200 M−1 cm−1) and 908 nm
(1200 M−1 cm−1) that were attributed to Cu−Namide σ−π* and
π−π* transitions, respectively.25 Furthermore, cryoscopic
Table 1. EPR Parameters for Compounds
compound
giso
g1
A1(Cu) (MHz)
[Cl2NN]Cu-OtBu (3)
[Cl2NN]Cu-NHAd (1)
[Cl2NN]Cu-NHArCl3 (4)
[Cl2NN]Cu-NHArF6 (5)
[Cl2NN]Cu-NH-2-py (6)
[Me2NN]Cu-NPh2
2.11(1)
2.243(5)
2.133(5)
2.208(5)
2.180(5)
2.201(2)
2.146(3)
2.175(2)
353(10)
365(10)
288(5)
314(5)
535(5)
298(5)
410(5)
2.067(2)
2.090(2)
2.083(2)
2.101(2)
2.071(3)
2.104(2)
[Me2NN]Cu−I
larger A1(Cu) = 590 MHz that distinguishes it from the trigonal
species with A1(Cu) in the range 280−370 MHz. Such large
A1(Cu) are typical for square planar Cu(II) complexes.28
Closer inspection reveals that A1(Cu) for the trigonal anilido
Figure 4. Electronic structure of mononuclear copper(II) anilides with spin density plots for 4−6 (ADF2007.1/ZORA/BP/TZ2P(+); excess spin α
(blue), spin β (red)).
10933
dx.doi.org/10.1021/ja5026547 | J. Am. Chem. Soc. 2014, 136, 10930−10940

Journal of the American Chemical Society
Article
Table 2. Optical Spectral Bands for Copper(II) Anilides
(Ethylbenzene, RT)
Table 3. Spin Densities (e−) in Copper(II) Amido
Complexes (ADF2007.1/ZORA/BP/TZ2P(+))
compound
λ1 (nm) (M−1 cm−1
)
λ2 (nm) (M−1 cm−1
)
compound
Cu
Namide
0.23
Nβ‑dik
[Cl2NN]Cu-NHArCl3 (4)
[Cl2NN]Cu-NHArF6 (5)
[Cl2NN]Cu(κ2-NH-2-py) (6)
[Me2NN]Cu-NPh2
741 (4500)
726 (3700)
663 (230)
774 (1200)
883 (4200)
863 (3000)
766 (190)
908 (1200)
4
0.36
0.36
0.45
0.07, 0.13
0.09, 0.13
0.12, 0.14
5-mono
6-κ2
0.25
0.15 (NH)
0.08 (py)
0.25
6-κ1
0.38
0.31
0.09, 0.13
0.08, 0.10
3
0.49
molecular weight measurements of 5 indicate a monomeric
solution structure in cyclohexane; the UV−vis spectrum of 5 in
cyclohexane is closely related to spectra taken in ethylbenzene
(Figures S3 and S4 in the Supporting Information). In contrast,
the optical spectrum of square planar [Cl2NN]Cu-NH-2-py (6)
possesses two bands at higher energy (660 and 760 nm) but of
much lower intensity (ε = 230 and ∼200 M−1 cm−1,
respectively).
roles that copper(II) anilides [Cl2NN]Cu-NHAr play in C-H
amination. Each of the corresponding anilines, H2NAr,
employed in this study serves as an efficient coupling partner
in C-H amination with ethylbenzene and cyclohexane under
t
catalytic conditions using BuOOtBu as a substrate with 1 mol
% [Cl2NN]Cu in neat hydrocarbon solvent (Scheme 5). While
Solutions of the copper(II) anilides may be prepared in
essentially quantitative yield by reaction of [Cl2NN]Cu-OtBu
(3) with H2NAr. Monitoring the course of reaction by UV−vis
in benzene solutions with 0.5 mM concentrations of 3, the
band of the alkoxide at λmax = 471 nm converts cleanly to the
corresponding bands of the respective copper(II) anilides 4−6
(Scheme 4; Figures S6−S8 in the Supporting Information).
Electronic Structure. DFT (ONIOM(BP86/6-311+G-
(d):UFF) methods were used to optimize the geometries and
analyze the electronic structures of the monomeric trigonal
CuII-anilide complexes. Both [Cl2NN]Cu-NHArCl3 (4) and
[Cl2NN]Cu-NHArF6 (5-mono) possess an electronic structure
similar to that discussed for the similar alkylamide [Cl2NN]Cu-
NHAd (3)23 and anilide [Me2NN]Cu-NPh2.25 Prominent in
these trigonal copper(II)-amides is the 2-center, 3-electron π-
interaction between the formally d9 CuII center and the lone
pair of an anionic sp2-hybridized Nanilide donor (Figure 4). The
high covalency of this interaction leads to significant
delocalization of the unpaired electron formally ascribed to
the d9 CuII center onto the anilido N atom and indeed into the
anilido aromatic ring. The planar Nanilide atom along with the
orthogonal orientation of the anilido aromatic ring orthogonal
to the β-diketiminato backbone allows for the anilido π-system
to efficiently participate in this interaction. TD-DFT (ADF
2007.1 ZORA/BP/TZ2P(+)) calculations suggest that the two
low energy optical transitions observed each for mononuclear,
trigonal 4 and 5 correspond to Cu−Namide σ−π* and π−π*
transitions (Tables S6−S9 and Figures S41, S43, S45, and S47
in the Supporting Information). In particular, the lowest energy
optical absorptions correspond to clean π−π* transitions
(Figure 4) as found for the related diphenylamide complex
[Me2NN]Cu-NPh2.25
Scheme 5. Catalytic C-H Amination with Electron-Poor
Anilines
the C-H amination of these two C-H substrates was previously
reported with H2NArCl3 24 the two other anilines H2NArF6 and
,
H2N-2-py also participate in C-H amination with these two
substrates. Higher yields are observed with neat ethylbenzene
(85−99%) than cyclohexane (46−97%) attributed to the lower
C-H bond strength of ethylbenzene (87 kcal/mol)29 vs
cyclohexane (97 kcal/mol).29
[Cl2NN]Cu-NHArCl3 (4) is clearly present in catalytic C-H
amination of ethylbenzene with H2NArCl3. Following a catalytic
reaction in neat ethylbenzene by UV−vis spectroscopy in a
cuvette thermostated at 80 °C clearly reveals the strong optical
bands of 4 at λ = 741 and 883 nm. Based on the amount of
catalyst used, we estimate that 70% of the [Cl2NN]Cu species
present exists as [Cl2NN]Cu-NHArCl3 initially, which declines
as the reaction nears completion over 5 h (Figure S9 in the
Supporting Information).
The direct observation of [Cl2NN]Cu-NHAr species 4−6 in
catalytic C-H amination reactions employing the corresponding
anilines H2NAr begs the question of what role these [CuII]-
NHAr species may play in the reaction. Aiming at answering
The square planar complex [Cl2NN]Cu(κ2-NH-2-py) (6-κ2)
possesses an electronic structure that is more typical of square
planar Cu(II) species. Consistent with its large A1(Cu) =
535(5) MHz, there is considerably more unpaired electron
density at Cu in 6-κ2 (0.45 e−) than in trigonal 3−5 (0.31−0.36
e−) and the putative κ1-N isomer of 6 (6-κ1) (0.38 e−).
Moreover, the remaining unpaired electron density in 6-κ2 is
spread more evenly among the 4 N donor atoms whereas in
trigonal 3−5 and 6-κ1 the amido N atom possesses
considerably more spin density than the β-diketiminato N-
donors (Table 3).
this question, we first focused on [Cl2NN]Cu-NHArCl3
.
Surprisingly, this copper(II) anilide is rather stable in neat
ethylbenzene at 80 °C. Preliminary kinetic analysis suggests
slow, second-order decay of [CuII]-NHArCl3 in neat ethyl-
benzene with a half-life of ca. 11 h (initial concentration = 0.40
mM; Figure S10 in the Supporting Information). After the
sample has completely bleached (40 h), analysis of the mixture
by GC−MS revealed that only 20% of the C-H amination
product was present; the remainder is H2NArCl3. Based on the
stoichiometry observed in the C-H amination of ethylbenzene
by isolated [Cl2NN]Cu-NHAd (1), which proceeds via smooth
pseudo first-order decay of 1 in excess ethylbenzene at 25 °C,23
we anticipate a maximum yield of 50% conversion of
[Cl2NN]Cu-NHArCl3 (4) to the C-H amination product
Catalytic C-H Amination of Ethylbenzene and Cyclo-
hexane with Electron-Poor Anilines. The isolation of the
copper(II) anilides 4−6 offers an opportunity to examine the
10934
dx.doi.org/10.1021/ja5026547 | J. Am. Chem. Soc. 2014, 136, 10930−10940

Journal of the American Chemical Society
Article
PhCH(NHArCl3)Me along with an equivalent amount of
H2NArCl3. While we observe some C-H amination product
from the isolated copper(II) anilides [Cl2NN]Cu-NHAr (4−6)
(Scheme 6), it is nonetheless clear that the rate and efficiency
Scheme 6. Stoichiometric C-H Amination with Cu-anilides
4-6
of stoichiometric C-H amination is not commensurate with that
observed in the catalytic variant employing H2NAr with 1.2
t
equiv of BuOOtBu in the presence of 1 mol % [Cu] (Scheme
5).
Figure 5. Zero-order decay of [Cl2NN]Cu-NHArCl3 (4) and zero-
order growth of [Cl2NN]Cu-OtBu (1) in ethylbenzene at 80 °C in the
Since it appears that [Cl2NN]Cu-NHAr species are not
particularly efficient at functionalizing neat ethylbenzene by
themselves, we considered an alternative reagent for HAA: the
tBuO• radical (Scheme 7). We recently demonstrated that the
t
presence of BuOOtBu.
the zero-order rate for [CuII]-NHArCl3 decay (Figures S13−16
in the Supporting Information), signaling a first-order depend-
ence on [tBuOOtBu].
Scheme 7. C-H Amination via HAA by BuO• radical
t
The zero-order dependence in [CuII]-NHAr (4) signals that
another species is responsible for initiating H atom abstraction
of the ethylbenzene C-H substrate. Based on the rapid
generation of the tBuO• radical from [Cl2NN]Cu and
tBuOOtBu (Scheme 7a), we suspect that a small amount of
solvent-free copper(I) species [Cl2NN]Cu generated by slow
decay of [Cl2NN]Cu-NHAr in ethylbenzene is responsible for
reaction of [Cl2NN]Cu with BuOOtBu in an inert solvent
t
t
the formation of BuO• and [Cl2NN]Cu-OtBu (2) in the
(fluorobenzene) proceeds readily via overall second order
kinetics (rate = k[CuI][tBuOOtBu]) with a rate constant of
5.9(2) M−1 s−1 at 20 °C consistent with generation of the
tBuO• radical (Scheme 7a).30 In addition, use of dicumylper-
oxide (PhMe2COOCMe2Ph) led to increased amounts of
acetone formation as the [CuI] concentration decreased,
providing firm evidence for the production of the cumyloxy
radical PhMe2CO• that rapidly decays by β-scission to give
t
reaction of [Cl2NN]Cu-NHArCl3 and BuOOtBu. Reaction of
tBuO• with ethylbenzene (R-H) to generate the corresponding
ethylbenzene radical (R•) swiftly proceeds (Scheme 7b)
followed by capture of the copper anilide [Cl2NN]Cu-NHAr
(4) by R• to give the C-H amination product R-NHArCl3
(Scheme 7c).
Direct Observation of C−N Bond Formation−Capture
of Trityl Radical by [Cu]-NHAr. We took advantage of the
ready generation of the trityl radical Ph3C• by equilibrium
dissociation of Gomberg’s dimer {Ph3C}2 to probe the
reactivity of [CuII]-NHAr species with a C-centered radical
under mild conditions. Addition of 1/2 equiv of {Ph3C}2 to
[CuII]-NH-py (6) in benzene-d6 results in the formation of the
corresponding N-trityl substituted anilines Ph3C-NH-2-py and
an equivalent of [Cl2NN]Cu(benzene) each in 78% yield
(Scheme 8). While reaction of Gomberg’s dimer was not as
acetophenone and Me•. Furthermore, BuOOtBu reacts with
t
[CuI] in cyclohexane to give Cy-OtBu, strongly suggesting that
the BuO• radical formed can undergo HAA reactions with
t
strong C-H bonds.31
Addition of 1/2 equiv of BuOOtBu to 1 mM solutions of
t
[CuII]-NHAr in ethylbenzene followed by heating at 80 °C for
24 h led to dramatic increases in the amount of functionalized
amine PhCH(NHAr)Me observed (Scheme 5). Following the
reaction of [CuII]-NHArCl3 (4) with tBuOOtBu in ethylbenzene
at 80 °C by UV−vis reveals zero-order decay of [CuII]-NHArCl3
(4) coupled to zero-order growth of [CuII]-OtBu (3) with
essentially identical zero-order rate constants (Figure 5). This
zero-order decay of 4 coupled with growth of 1 also occurs at
25 °C, albeit more slowly and with an induction period (Figure
S11 in the Supporting Information). Following the decay of
Scheme 8. Capture of Ph3C• by [CuII]-NHAr
[CuII]-NHArCl3 in the presence of 3−10 equiv of BuOOtBu
and 1 equiv of added [Cl2NN]Cu results in a linear increase of
t
10935
dx.doi.org/10.1021/ja5026547 | J. Am. Chem. Soc. 2014, 136, 10930−10940

Journal of the American Chemical Society
Article
clean with the Cu(II) anilides 4 and 5, this experiment reveals
that attack of [CuII]-NHAr species by radicals R• can deliver
the corresponding functionalized amine R-NHAr.
22.9(8) kcal/mol; ΔH‡ = 11.5 kcal/mol, ΔS‡ = −46.8 e.u. with
ΔG‡ = 25.2 kcal/mol). On the other hand, the transition states
for HAA of ethylbenzene by 4−6 are calculated to be much
higher in free energy, largely due to increased enthalpies of
activation (ΔH‡ = 21.3, 17.4, and 16.8 kcal/mol for 4, 5, and 6-
κ1, respectively). No transition state was identified from the
square planar 6-κ2, due to the bidentate coordination mode of
the amido-2-pyridine substrate. The significant increase in
enthalpy of activation does not originate from the C-H
substrate since the benzylic C-H bonds of ethylbenzene and
indane are of similar strength (calc. BDE = 83.1 (ethylbenzene)
vs 83.5 kcal/mol (indane)). Instead, the HAA enthalpies of
activation reflect the weaker N−H bonds calculated in the
copper(I) aniline adducts [CuI](NH2Ar) (48.1−52.7 kcal/mol)
vs the copper(I) amine adduct [CuI](NH2Ad) (68.0 kcal/mol).
Theory−Formation and Reactivity of Copper(II)
Anilides. At the ONIOM(BP86/6-311+G(d):UFF) level of
theory, the acid−base exchange reactions between [Cl2NN]Cu-
OtBu and H2NAr are computed to be slightly exergonic for
differing aryls (Ar); ΔG in kcal/mol = −2.6 (Ar = Ph), −0.4
(Ar = py), −2.4 (Ar = ArCl3) and −2.8 (Ar = ArF6) (Scheme 4).
DFT predicts the observed κ2-NH-2-py coordination observed
to be 2.4 kcal/mol lower in free energy than κ1-NH-2-py
coordination to give a trigonal [CuII]-NHAr species.
The arylamine N−H bond strength in each copper(I) adduct
[Cl2NN]Cu(NH2Ar) reflects the driving force for HAA by the
corresponding copper(II) anilide [Cl2NN]Cu-NHAr. These
N−H bond strengths are calculated to be 48.3, 48.1, and 52.7
kcal/mol for the copper(I) aniline adducts derived from 4, 5,
and 6, respectively (Scheme 9). These values for the N−H
t
On the other hand, HAA of ethylbenzene by the BuO•
radical is calculated to be highly thermodynamically favorable
(ΔG = −15.4 kcal/mol) driven by the strong O−H bond in
tBuO-H (calc: 98.3 and expt: 105 kcal/mol29) that leads to
kinetically competent HAA of ethylbenzene (ΔH‡ = −1.8 kcal/
mol, ΔS‡ = 36.8 e.u. and ΔG‡(298 K) = 9.2 kcal/mol). These
values compare well to experimental values for HAA reactions
Scheme 9. Calculated Thermodynamics for HAA of
Ethylbenzene by 4, 5 and 6-κ1
t
of BuO• with related hydrocarbons such as toluene and
cyclohexane (ΔG‡(298 K) = 10.2(5) and 9.4(7) kcal/mol,
respectively).31 Thus, generation of the tBuO• radical would be
expected to swiftly yield the 2° benzylic ethylbenzene radical
(Scheme 7b).
t
The radical R• generated via HAA from R-H by BuO•
conceptually may be captured in two conceptually different
ways by the [CuII]-NHAr intermediates to ultimately give the
C-H amination product R-NHAr (Schemes 7 and 10). Capture
Scheme 10. HAA of R-H by tBuO• and Radical Capture of R•
by Copper(II) Anilides 4−6
bond enthalpy are approximately ca. 40 kcal/mol lower than
that found in the free anilines (exp. 90−95 kcal/mol)29
calculated to be 89.4, 91.3, and 91.9 kcal/mol for H2NArCl3
,
H2NArF6, and H2N-2-py, respectively. The modest N−H bond
strengths in the copper(I) aniline adducts [Cl2NN]Cu(NH2Ar)
reflect the relative stabilization of the aminyl radical as
copper(II) anilides [Cl2NN]Cu-NHAr that possess highly
delocalized unpaired electron-density (Figure 4, Table 3).
Furthermore, the more acidic N−H bond in arylamine adducts
[Cl2NN]Cu(NH2Ar) as compared to the corresponding
alkylamine species [Cl2NN]Cu(NH2Ad) may also contribute
to lower N−H bond strengths in [Cl2NN]Cu(NH2Ar)
complexes.32 These relatively low values of N−H bond strength
result in highly endothermic HAA of ethylbenzene by [CuII]-
NHAr species.
directly at the copper center would result in the copper(III)
organometallic species [CuIII](R)(NHAr), which possess
square planar coordination at Cu owing to their d8 electronic
configuration. Direct capture at copper of the 2° benzylic
radical R• derived from R-H = ethylbenzene by [CuIII]-NHAr
species 4−6 is calculated to be nearly thermoneutral (ΔH =
+3.7, −1.5, and −4.2 kcal/mol, respectively), but significantly
entropically disfavored (ΔS = −57.4, −53.3, −58.7 e.u. at 298
K) rendering capture at Cu endergonic with ΔG values of 20.8,
14.4, and 13.3 kcal/mol at 298 K. Due to the high oxidation
state at Cu as well as the crowded square planar environment,
reductive elimination to give the corresponding [CuI](NH(R)-
Ar) adducts containing the coordinated product amine R-
The transition states for HAA of ethylbenzene by [CuII]-
NHAr (Scheme 9) are similar in structure to that found for
HAA of indane by [Cl2NN]Cu-NHAd (3), yet importantly are
appreciably higher in energy for the anilides 4−6. As a point of
comparison, the experimental activation parameters for HAA of
‡
indane by 3 closely match those found by theory. (ΔHexp
=
=
‡
‡
11.4(4) kcal/mol and ΔSexp = −38.7(14) e.u. with ΔGexp
10936
dx.doi.org/10.1021/ja5026547 | J. Am. Chem. Soc. 2014, 136, 10930−10940

Journal of the American Chemical Society
Article
NHAr is favored (Scheme 10). On the other hand, direct
reaction of R• at the N atom to give the product amine R-
NHAr coordinated to the copper center is much more
thermodynamically favorable. Moreover, η2-bound amine
adducts (via the N-aryl group) in [CuI](η2-ArNHR) species
are more thermodynamically stable than the corresponding N-
bound adducts [CuI](κ1-NH(R)Ar). Considering the [CuI](η2-
ArNHR) complexes, addition of a radical R• to copper(II)
anilides 4−6 is calculated to be exothermic (ΔH = −21.4,
−27.6, and −32.3 kcal/mol, respectively), but significantly
entropically disfavored (ΔS = −58.6, −57.3, −57.3 e.u. at
298.15 K). Nonetheless, radical capture at N to give the
[CuI](η2-ArNHR) adducts is still favorable with ΔG values of
−4.0, −10.5, and −15.2 kcal/mol at 298.15 K. While these
calculated energetics indicate that radical capture at either Cu
or N are each favorable in terms of enthalpy (ΔH), free energy
(ΔG) favors radical capture at N due to unfavorable entropy
(TΔS) terms. While direct capture of R• at N is clearly
thermodynamically favored (Scheme 10, Figure S36 in the
Supporting Information), radical addition reactions to either
Cu or N are likely to have low barriers.
subsequent HAA of ethylbenzene by tBuO• successfully
competes with slower HAA by [Cl2NN]Cu-NHAd (3).
While theoretical studies demonstrate that these three-
coordinate copper(II) anilides may participate in HAA
reactions with ethylbenzene, they possess rather high barriers
with ΔG‡(298 K) = 30−35 kcal/mol depending on the anilide
substituent. These barriers are 5−10 kcal/mol higher than
established for the copper(II) alkylamide [Cl2NN]Cu-NHAd
(3).23 These differences reflect the lower N−H bond strength
in the HAA product [Cl2NN]Cu(NH2Ar) as compared to
[Cl2NN]Cu(NH2Ad) and track differences in the N−H bond
strength in the corresponding free primary amines H2NR. The
N−H bond in alkylamines (100 kcal/mol) is generally 5−10
kcal/mol stronger than in anilines (90−95 kcal/mol); electron-
deficient aromatic substituents generally enhance the N−H
bond strength in anilines.29 As shown by experiment, capture of
the radical R• by [CuII]-NHAr is facile. Capture at either Cu or
N is enthalpically favorable, though the former is endergonic
while direct addition to N is favored. Similar trends were found
in the thermodynamics of addition of the cyclohexyl radical to
[CuII]-OtBu that strongly favors formation of the correspond-
ing copper(I) ether adduct [CuI](Cy(OtBu)).30
Taken together, these studies suggest a catalytic cycle for C-
H amination with anilines that is a variation of the Kharasch-
Sosnovsky reaction. First reported in 1958, Kharasch and
Sosnovsky found that peracetates such as PhC(O)OOtBu
functionalize allylic C-H bonds in substrates R-H to directly
form benzoates R−OCH(O)Ph (Scheme 11).33,34 A variety of
DISCUSSION AND CONCLUSIONS
■
The synthesis of copper(II) anilides 4−6 allowed a detailed
investigation into the mechanism of efficient C-H amination
with the corresponding aromatic amines. Copper(II) anilides
engage in a diversity of bonding modes including terminal,
bridging dinuclear and κ2-N,N for the anilide derived from 2-
aminopyridine. Based on earlier catalytic C-H functionalization
studies, these copper(II) anilides derived from electron-poor
anilines appear to be quite stable toward N−N bond formation
that could lead to diazenes ArNNAr, perhaps via coupling of
two [CuII]-NHAr species. While the X-ray structure of
[Cl2NN]Cu-NHArF6 clearly illustrates that this species can
dimerize, it does so via Cu−N−Cu bridging interactions. The
crowded nature of this dimer results in facile formation of
monomeric [Cl2NN]Cu-NHArF6 as evidenced by solution EPR
and UV−vis studies.
Scheme 11. Kharasch-Sosnovsky Reaction
simple copper(I) salts such as CuCl serve as catalysts for this
reaction at higher temperatures (60−90 °C). Importantly,
enantioselective variants have been developed based on cationic
copper(I) complexes of bis(oxazolines) and other chiral
scaffolds that deliver ee’s as high as 96%.34
Unlike the alkylamide complex [Cl2NN]Cu-NHAd (3), the
corresponding anilide complexes 4−6 do not engage in efficient
HAA reactions with ethylbenzene under mild conditions.
Higher temperatures (e.g., 80 °C) are required, and yields of
the C-H functionalized amine PhCH(NHAr)Me are low to
negligible. In contrast when a [CuI]/tBuOOtBu mixture is
added to an ethylbenzene solution of 4, facile conversion takes
place, completely consuming the anilide and delivering the
functionalized amine R-NHAr in high yield. These studies
The copper(I) species [Cl2NN]Cu serves two essential roles
in C-H functionalization catalysis. It swiftly generates the very
t
reactive BuO• radical capable of HAA of R-H (Scheme 12b)
t
upon reaction with BuOOtBu that also provides [CuII]-OtBu
(2) (Scheme 12a). Facile acid/base chemistry between amines
H2NR′ (especially electron-poor amines such as anilines) and
[CuII]-OtBu to generate [CuII]-NHR represents the second key
role for the copper complex (Scheme 12c). Coordination of the
t
suggest that BuO• is the predominant HAA agent in the
efficient C-H amination with anilines, generating the R• radical
from R-H which may be captured by [CuII]-NHAr to give R-
NHAr.
Scheme 12. General C-H Amination via Copper(II) Amides
This mechanism also extends to C-H amination by
alkylamines. Whereas [Cl2NN]Cu-NHAd (3) undergoes
clean, first-order decay in excess ethylbenzene at 25 °C (rate
= k[[CuII]-NHAd][ethylbenzene], k = 1.51(3) × 10−5 M−1 s−1;
Figure S21 in the Supporting Information) addition of
tBuOOtBu greatly accelerates the reaction and leads to zero-
order decay of 3 (Figure S20 in the Supporting Information).
At 0.50 mM initial concentrations of 3 in ethylbenzene the half-
life for decay of 3 is 5590 s whereas in the presence of 1.2 equiv
of tBuOOtBu the initial (zero-order) half-life shortens
dramatically to 555 s. Thus, generation of tBuO• and
10937
dx.doi.org/10.1021/ja5026547 | J. Am. Chem. Soc. 2014, 136, 10930−10940

Journal of the American Chemical Society
Article
bis(trifluoromethyl)aniline (0.0342 mL, 0.22 mmol). The suspension
turned into an intense navy solution and was stirred for 5 min at RT,
and then the solvent was removed in vacuo. The remaining solid was
extracted with 10 mL of n-pentane and filtered through Celite. It was
then concentrated to 5 mL and allowed to stand overnight at −35 °C
to yield navy crystals (0.078 g) in 53% yield. λmax (ethylbenzene) =
731 nm (ε = 3740 M−1 cm−1). Anal. Calcd for C25H17Cl4CuF6N3: C,
44.24; H, 2.52; N, 6.19. Found: C, 44.59; H, 2.21; N, 6.02.
amide to the copper center renders the NHR′ moiety
particularly reactive toward reaction with C-based radicals R•
(Scheme 12d).
These mechanistic studies also open the possibility of
employing other functional groups (H-FG) that could undergo
facile acid/base exchange with [CuII]-OtBu with loss of HOtBu
t
to give [CuII]-FG intermediates. Generation of BuO• radical
(through reaction of [CuI] and tBuOOtBu)30 can lead to facile
HAA reactions with a wide range of sp3-hybridized C-H bonds
in substrates R-H to give R• under mild conditions with
bimolecular rate constants in the range of 105 to 108 M−1 s−1.31
Reaction of R• with [CuII]-FG could conceivably deliver new
C-H functionalized species R-FG with concomitant reduction
of the Cu center to [CuI].35 In fact, Hartwig and co-workers
recently found that amidation of unactivated sp3 C-H bonds
[Cl2NN]Cu(κ2-NH-2-py) (6). To a suspension of [Cl2NN]Cu-OtBu
(0.114 g, 0.22 mmol) in ca. 8 mL of n-pentane was added 2-
aminopyridine (0.021 g, 0.22 mmol). The suspension turned into a
green solution and was stirred for 5 min at RT, and then the solvent
was removed in vacuo. The remaining solid was extracted with 5 mL of
Et2O, filtered through Celite, then concentrated to 2 mL, layered with
1 mL of pentane, and then allowed to stand overnight at −35 °C to
yield green crystals (0.030 g) in 25% yield. λmax (ethylbenzene) = 664
nm (ε = 230 M−1 cm−1). Anal. Calcd for C22H18Cl4CuN4: C, 48.59; H,
3.34; N, 10.30. Found: C, 48.44; H, 3.17; N, 9.97.
t
can occur with organic amides, BuOOtBu and phenanthroline
copper(I) complexes by a related variation of the Kharasch-
Catalytic Amination Ethylbenzene and Cyclohexane. A
solution was prepared under inert atmosphere employing 1.0 mmol
of H2NArCl3, H2NArF6, or NH2-2-py (1.0 equiv) along with 1.20 mmol
of DTBP (1.2 equiv) in 20 mL of C-H substrate (ethylbenzene or
cyclohexane). To this solution was added 1 mol % [Cl2NN]Cu from a
stock solution of the catalyst (0.100 mL = 0.01 mmol [Cl2NN]Cu).
The reactions were stirred inside a sealed pressure vessel at 90 °C for
24 h. Afterward the reaction mixture was allowed to cool, filtered
t
Sosnovsky reaction in which the BuO• radical was also
identified as the H atom abstracting reagent.18 These concepts
underscore the possibility of a family of new, mechanistically
related C-H functionalization reactions that are under active
development in our laboratory.
1
EXPERIMENTAL SECTION
through Celite, and analyzed by GC−MS and H NMR after adding
■
1.0 mmol of 1,2,4,5-tetrachlorobenzene as a standard. Product
characterization these substrates is found in the Supporting
Information.
General Procedures. All experiments were carried out in a dry
nitrogen atmosphere using an MBraun glovebox and/or standard
Schlenk techniques, unless otherwise stated. 4A molecular sieves were
activated in vacuo at 180 °C for 24 h. Dry benzene and
dichloromethane were purchased from Aldrich and were stored over
activated 4A molecular sieves under nitrogen. Diethyl ether and
tetrahydrofuran (THF) were first sparged with nitrogen and then
dried by passage through activated alumina columns. Pentane was first
washed with conc. HNO3/H2SO4 to remove olefins, stored over CaCl2
and then passed through activated alumina columns. All deuterated
solvents were sparged with nitrogen, dried over activated 4A molecular
sieves and stored under nitrogen. 1H and 13C NMR spectra were
recorded on Varian 400 MHz spectrometer (400 and 100.4 MHz,
respectively). All NMR spectra were recorded at room temperature
unless otherwise noted and were indirectly referenced to TMS using
residual solvent signals as internal standards. GC−MS spectra were
recorded on a Varian Saturn 2100T, elemental analyses were
performed on a PerkinElmer PE2400 microanalyzer at Georgetown,
and UV−vis spectra were recorded on a Cary 50 spectrophotometer.
All EPR, SQUID magnetization, and X-ray characterization details may
be found in the Supporting Information.
Stoichiometric Reaction of [Cl2NN]Cu-NHAr with Ethyl-
benzene. Stock solutions of the copper(II) anilides were prepared
by dissolving each of [Cl2NN]Cu-NHArCl3 (32 mg, 0.050 mmol),
[Cl2NN]Cu-NHArF6 (34 mg, 0.050 mmol), and [Cl2NN]Cu(κ2-NH-
2-py) (27 mg, 0.050 mmol) in 10.0 mL of ethylbenzene (5.00 mM).
t
Stock solutions of [Cl2NN]Cu-OtBu, BuOOtBu, and the 1,2,4,5-
tetrachlorobenzene standard were prepared by dissolving [Cl2NN]Cu-
OtBu (114.6 mg, 0.219 mmol) in 10.0 mL of ethylbenzene (0.0216
t
M), by mixing BuOOtBu (11 mL, 0.060 mmol) in a total volume of
1.00 mL of ethylbenzene (0.100 M), or by dissolving 1,2,4,5-
tetrachlorobenzene (216 mg, 1.00 mmol) in 10.0 mL of ethylbenzene
(0.100 M). Three sets of reactions employing each species
[Cl2NN]Cu-NHAr were carried out in sealed pressure vessels at 80
°C for 24 h each in either (a) neat ethylbenzene, (b) neat
ethylbenzene with 1.0 equiv of [Cl2NN]Cu-OtBu, or (c) neat
t
ethylbenzene with 0.6 equiv of BuOOtBu. These reaction mixtures
were prepared as follows: (a) 2.00 mL of 5.00 mM [Cl2NN]Cu-NHAr
and 0.100 mL of 0.100 M 1,2,4,5-tetrachlorobenzene diluted to total
volume of 10.00 mL with ethylbenzene; (b) 2.00 mL of 5.00 mM
[Cl2NN]Cu-NHAr, 0.457 mL of 0.0216 M [Cl2NN]Cu-OtBu, and
0.100 mL of 0.100 M 1,2,4,5-tetrachlorobenzene diluted to total
volume of 10.00 mL with ethylbenzene; and (c) 2.00 mL of 5.00 mM
[Cl2NN]Cu-NHAr, 0.100 mL of 0.100 M DTBP, and 0.100 mL of
0.100 M 1,2,4,5-tetrachlorobenzene diluted to total volume of 10.00
mL with ethylbenzene. See Tables S1 and S2 in the Supporting
Information for product characterization details.
All reagents were obtained commercially unless otherwise noted
and typically stored over activated 4A molecular sieves. Ethylbenzene
was obtained from Acros and purified by passing each through
activated alumina. The neutral β-diketimine H[Cl2NN]36 and
corresponding copper catalyst {[Cl2NN]Cu}2(μ-benzene)7 were
prepared by literature methods or may be obtained from Strem
Chemicals. [Cl2NN]Cu-OtBu23 and Gomberg’s dimer [Ph3C]2 were
prepared by a slight variation of literature procedures37(see Supporting
Information).
Kinetic Analysis of the Reaction of [Cl2NN]Cu-NHArCl3 with
t
Ethylbenzene in the Presence of BuOOtBu. A stock solution of
Preparation of Compounds. [Cl2NN]Cu-NHArCl3 (4). To a
suspension of [Cl2NN]Cu-OtBu (0.114 g, 0.22 mmol) in ca. 8 mL
of n-pentane was added 2,4,6-trichloroaniline (0.043 g, 0.22 mmol).
The suspension turned intense blue solution and was stirred for 5 min
at RT then the solvent was removed in vacuo. The remaining solid was
extracted with 10 mL of n-pentane and filtered through Celite. The
solution was then concentrated to 5 mL and allowed to stand
overnight at −35 °C to yield blue crystals (0.087 g) in 62% yield. λmax
(ethylbenzene) = 741 nm (ε = 4480 M−1 cm−1). Anal. Calcd for
C23H16Cl7CuN3: C, 42.76; H, 2.50; N, 6.50. Found: C, 43.12; H, 2.35;
N, 6.32.
[Cl2NN]Cu-NHArCl3 in ethylbenzene was prepared by dissolving
[Cl2NN]Cu-NHArCl3 (41 mg, 0.063 mmol) in ethylbenzene and
diluting to 10.00 mL. 0.250 mL of this solution was diluted to 3.00 mL
of ethylbenzene in a cuvette to achieve 0.49 mM solution. The
solution was thermostated in a UV−vis spectrometer to either 25.0 or
80.0 °C. After being held at the appropriate temperature for 5 min, 0.1
mL of a stock solution of tBuOOtBu (3.5 μL in 1.00 mL of
ethylbenzene) was added to the cuvette. Kinetic analysis monitored
the decay of [Cl2NN]Cu-NHArCl3 at λmax = 741 nm and growth of
[Cl2NN]Cu-OtBu at λmax = 471 nm. Both decay of 4 and growth of 3
followed zero-order kinetics with rate constants of 1.2(2) × 10−8 s−1
and 1.1(2) × 10−8 s−1 at 25 °C, respectively (Figure S10 in the
{[Cl2NN]Cu}2(μ-NHArF6)2 (5). To a suspension of [Cl2NN]Cu-OtBu
(0.115 g, 0.22 mmol) in ca. 8 mL of n-pentane was added 3,5-
10938
dx.doi.org/10.1021/ja5026547 | J. Am. Chem. Soc. 2014, 136, 10930−10940

Journal of the American Chemical Society
Article
Supporting Information). At 80 °C, the zero-order rate constants for
decay of 3 and growth of 4 are 6.5(3) × 10−7 s−1 and 7.0(3) × 10−7
s−1, respectively (Figure S11 in the Supporting Information).
Present Address
#Institute of Chemical Sciences, Heriot-Watt University,
Edinburgh EH14 4AS, U.K.
Reaction of Gomberg’s Dimer with [Cl2NN]Cu-NH-2-py (6). A
stock solution of Gomberg’s dimer with an internal standard was
prepared by dissolving (Ph3C)2·pentane (36 mg, 0.0644 mmol) and
1,2,4,5-tetrachlorobenzene (13.9 mg, 0.0644 mmol) in 3.00 mL of
C6D6 (0.0215 M). A vial containing 0.500 mL of this stock solution
(0.0107 mmol (Ph3C)2·pentane and 0.0107 mmol standard) was
concentrated to dryness in vacuo to remove excess pentane. 2.0 equiv
of [Cl2NN]Cu(κ2-NH-2-py) (11.7 mg, 0.0215 mmol) in 0.500 mL of
C6D6 was added to the vial containing 1.0 equiv of (Ph3C)2 and then
transferred to an NMR tube for measurement. An immediate color
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We are grateful to NSF (CHE-1012523 and CHE-1300774,
T.H.W.; CHE-1057785, T.R.C.) for support of this work. K.M.
acknowledges financial support from the Friedrich-Alexander-
University (FAU) Erlangen-Nurnberg.
̈
1
change was observed resulting in a yellow solution, and the H NMR
spectrum of this solution was taken after 20 min identifying [Cl2NN]
Cu and Ph3C-NH-2-py in 78% yield each (Figure S12 in the
Supporting Information). The formation of Ph3C-NH-2-py was
confirmed by its independent synthesis according to a literature
REFERENCES
■
(1) (a) Dauban, P.; Dodd, R. H. In Amino Group Chemistry; Ricci, A.,
Ed.; Wiley-VCH: Weinheim, 2008; pp 55−92. (b) Collet, F.; Lescot,
C.; Liang, C.; Dauban, P. Dalton Trans. 2010, 39, 10401−10413.
(c) Zalatan, D. N.; Du Bois, J. Top. Curr. Chem. 2010, 292, 347−378.
(d) Zhang, X. P.; Lu, H. Chem. Soc. Rev. 2011, 40, 1899−1909.
(2) Collet, F.; Dodd, R. H.; Dauban, P. Chem. Commun. 2009, 5061−
5074.
(3) Berry, J. F. Comments Inorg. Chem. 2009, 30, 28−66.
(4) (a) Fiori, K. W.; Du Bois, J. J. Am. Chem. Soc. 2007, 129, 562−
568. (b) Liang, C.; Collet, F.; Robert-Peillard, F.; Muller, P.; Dodd, R.
H.; Dauban, P. J. Am. Chem. Soc. 2008, 130, 343−350. (c) Du Bois, J.
Org. Process Res. Dev. 2011, 15, 758−762. (d) Au, S.-M.; Huang, J.-S.;
Yu, W.-Y.; Fung, W.-H.; Che, C.-M. J. Am. Chem. Soc. 1999, 121,
9120−9132. (e) Leung, S. K.-Y.; Tsui, W.-M.; Huang, J.-S.; Che, C.-
M.; Liang, J.-L.; Zhu, N. J. Am. Chem. Soc. 2005, 127, 16629−16640.
(f) Au, S.-M.; Huang, J.-S.; Che, C.-M.; Yu, W.-Y. J. Org. Chem. 2000,
65, 7858−7864. (g) Harvey, M. E.; Musaev, D. G.; Du Bois, J. J. Am.
Chem. Soc. 2011, 133, 17207−17216. (h) Liang, J.-L.; Yuan, S.-X.;
Huang, J.-S.; Yu, W.-Y.; Che, C.-M. Angew. Chem., Int. Ed. 2002, 41,
3465−3468. (i) Milczek, E.; Boudet, N.; Blakey, S. Angew. Chem., Int.
Ed. 2008, 47, 6825−6828. (j) Yu, X.-Q.; Huang, J.-S.; Zhou, X.-G.;
Che, C.-M. Org. Lett. 2000, 2, 2233−2236.
(5) (a) Kwart, H.; Khan, A. A. J. Am. Chem. Soc. 1967, 89, 1951−
1953. (b) Badiei, Y. M.; Krishnaswamy, A.; Melzer, M. M.; Warren, T.
H. J. Am. Chem. Soc. 2006, 128, 15056−15057.
(6) Gephart, R. T., III; Warren, T. H. Organometallics 2012, 31,
7728−7752.
(7) 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,
9961−9964.
(8) (a) Lebel, H.; Huard, K.; Lectard, S. J. Am. Chem. Soc. 2005, 127,
14198−14199. (b) Huard, K.; Lebel, H. Chem.Eur. J. 2008, 14,
6222−6230.
(9) Zhang, J.; Chan, P. W. H.; Che, C.-M. Tetrahedron Lett. 2005, 46,
5403−5408.
(10) (a) King, E. R.; Hennessy, E. T.; Betley, T. A. J. Am. Chem. Soc.
2011, 133, 4917−4923. (b) Hennessy, E. T.; Betley, T. A. Science
2013, 340, 591−595.
(11) (a) Liu, Y.; Che, C.-M. Chem.Eur. J. 2010, 16, 10494−10501.
(b) Wang, Z.; Zhang, Y.; Fu, H.; Jiang, Y.; Zhao, Y. Org. Lett. 2008, 10,
1863−1866.
method38 and observing its resulting H NMR spectrum in C6D6 in
1
the presence of 1 equiv of [Cl2NN]Cu (Figure S13 in the Supporting
Information).
Computational Methods. All computations employed the
Gaussian 09 program.39 Hybrid quantum mechanics/molecular
mechanics (QM/MM) calculations on full experimental chemical
models utilize the BP8640 density functional for the QM region in
conjunction with the 6-311+G(d) Pople basis set, which is a triple-ζ
basis set augmented by diffuse and polarization functions on main
group elements, and an f-polarization function on copper. The classical
MM region was modeled by the Universal Force Field41 (UFF) and
contained the 2,6-dichloropylphenyl and methyl substituents of the β-
diketiminate ligand. The remainder of the complex and ligands, as well
as substrate molecules, were modeled at the aforementioned BP86/6-
311+G(d) level of theory. The ONIOM methodology of Svensson et
al. was used for all QM/MM simulations.42
Open and closed shell species were modeled within the unrestricted
and restricted Kohn−Sham formalisms, respectively. All geometry
optimizations were conducted without symmetry constraint using
gradient methods. The energy Hessian was evaluated at all stationary
points to designate them as either minima or transition states at the
pertinent levels of theory. Free energies collected in Table S4 in the
Supporting Information are reported at 298.15 K and 1 atm and are
calculated with unscaled vibrational frequencies.
Spin density calculations and plots for [Cl2NN]Cu-NHAr (4−6)
collected in Table 3 and illustrated in Figure 4 were generated with
ADF43 (ADF 2007.1 BP/ZORA/TZ2P(+)) by optimizing coordinates
obtained above. Slater-type orbital (STO) basis sets employed for H,
C, and N atoms were of triple-ζ quality augmented with two
polarization functions (ZORA/TZ2P) while an improved triple-ζ basis
set with two polarization functions (ZORA/TZ2P+) was employed for
the Cu atom. Scalar relativistic effects were included by virtue of the
zero order regular approximation (ZORA). The 1s electrons of C and
N as well as the 1s−2p electrons of Cu were treated as frozen core.
ASSOCIATED CONTENT
■
S
* Supporting Information
Complete synthetic and kinetics details, additional character-
ization data for compounds 4−6 including EPR spectra,
computational methods, X-ray details with fully labeled thermal
ellipsoid plots for 4−6. The full citation for Gaussian 09.
Crystallographic information in CIF format for compounds 4−
6. This material is available free of charge via the Internet at
http://pubs.acs.org.
(12) (a) Lyaskovskyy, V.; Olivos Suarez, A. I.; Lu, H.; Jiang, H.;
Zhang, X. P.; de Bruin, B. J. Am. Chem. Soc. 2011, 133, 12264−12273.
(b) Caselli, A.; Gallo, E.; Fantauzzi, S.; Morlacchi, S.; Ragaini, F.;
Cenini, S. Eur. J. Inorg. Chem. 2008, 3009−3019. (c) Cenini, S.; Gallo,
E.; Penoni, A.; Ragaini, F.; Tollari, S. Chem. Commun. 2000, 2265−
2266. (d) Lu, H.; Subbarayan, V.; Tao, J.; Zhang, X. P. Organometallics
2010, 29, 389−393. (e) Lu, H.; Tao, J.; Jones, J. E.; Wojitas, L.; Zhang,
X. P. Org. Lett. 2010, 12, 1248−1251. (f) Ruppel, J. V.; Kamble, R. M.;
Zhang, X. P. Org. Lett. 2007, 9, 4889−4892.
AUTHOR INFORMATION
■
(13) (a) Díaz-Requejo, M. M.; Belderraín, T. R.; Nicasio, M. C.;
Corresponding Authors
*thw@georgetown.edu
*t@unt.edu
Trofimenko, S.; Per
12079. (b) Fructos, M. R.; Trofimenko, S.; Días-Requejo, M. M.;
́
ez, P. J. J. Am. Chem. Soc. 2003, 125, 12078−
Per
́
ez, P. J. J. Am. Chem. Soc. 2006, 128, 11784−11791. (c) Hamilton,
10939
dx.doi.org/10.1021/ja5026547 | J. Am. Chem. Soc. 2014, 136, 10930−10940

Journal of the American Chemical Society
Article
C. W.; Laitar, D. S.; Sadighi, J. P. Chem. Commun. 2004, 1628−1629.
(d) Liu, X.; Zhang, Y.; Wang, L.; Fu, H.; Jiang, Y.; Zhao, Y. J. Org.
Chem. 2008, 73, 6207−6212. (e) Vedernikov, A. N.; Caulton, K. G.
Chem. Commun. 2004, 162−163.
(14) (a) Reed, S. A.; Mazzotti, A. R.; White, M. C. J. Am. Chem. Soc.
2009, 131, 11701−11706. (b) Liu, G.; Yin, G.; Wu, L. Angew. Chem.
2008, 120, 4811−4814. (c) Liu, G.; Yin, G.; Wu, L. Angew. Chem., Int.
Ed. 2008, 47, 4733−4736.
(15) Shimizu, Y.; Obora, Y.; Ishii, Y. Org. Lett. 2010, 12, 1372−1374.
(16) Jordan-Hore, J. A.; Johansson, C. C. C.; Gulias, M.; Beck, E. M.;
Gaunt, M. J. J. Am. Chem. Soc. 2008, 130, 16184−16186.
(17) (a) Smith, K.; Hupp, C. D.; Allen, K. L.; Slough, G. A.
Organometallics 2005, 24, 1747−1755. (b) Pelletier, G.; Powell, D. A.
Org. Lett. 2006, 8, 6031−6034. (c) Powell, D. A.; Fan, H. J. Org. Chem.
2010, 75, 2726−2729. (d) Shuai, Q.; Deng, G.; Chua, Z.; Bohle, D. S.;
Li, C.-J. Adv. Synth. Catal. 2010, 352, 632−636.
(39) Frish, M. J.; et al. Gaussian 09, Revision A.1; Gaussian, Inc:
Wallingford, CT, 2009 (full citation in the Supporting Information).
(40) (a) Becke, A. D. Phys. Rev. A 1988, 38, 3098−3100. (b) Perdew,
J. P. Phys. Rev. B 1986, 33, 8822−8824.
́
(41) Rappe, A. K.; Casewit, C. J.; Colwell, K. S.; Goddard, W. A. I.;
Skiff, W. M. J. Am. Chem. Soc. 1992, 114, 10024−10035.
(42) Svensson, M.; Humbel, S.; Froese, R. D. J.; Matsubara, T.;
Sieber, S.; Morokuma, K. J. Phys. Chem. 1996, 100, 19357−19363.
(43) (a) te Velde, G.; Bickelhaupt, F. M.; Baerends, E. J.; Fonseca
Geurra, C.; Snijders, J. G.; Ziegler, T. J. Comput. Chem. 2001, 22, 931−
967. (b) Fonseca Geurra, C.; Snijders, J. G.; te Velde, G.; Baerends, E.
J. J. Theor. Chem. Acc. 1998, 99, 391−403.
(18) Tran, B. L.; Li, B.; Driess, M.; Hartwig, J. F. J. Am. Chem. Soc.
2014, 136, 2555−2563.
(19) (a) Ramirez, T. A.; Zhao, B.; Shi, Y. Chem. Soc. Rev. 2012, 41,
931−942. (b) Clark, J. S.; Roche, C. Chem. Commun. 2005, 5175−
5177.
(20) (a) Davies, H. M. L.; Manning, J. R. Nature 2008, 451, 417−
424. (b) Godula, K.; Sames, D. Science 2006, 312, 67−72. (c) Wehn, P.
M.; Du Bois, J. Angew. Chem., Int. Ed. 2009, 48, 3802−3805.
(21) Aguila, M. J. B.; Badiei, Y. M.; Warren, T. H. J. Am. Chem. Soc.
2013, 135, 9399−9406.
(22) Evans, R. E.; Zbieg, J. R.; Zhu, S.; Li, W.; Macmillan, D. W. C. J.
Am. Chem. Soc. 2012, 135, 16074−16077.
(23) Wiese, S.; Badiei, Y. M.; Gephart, R. T.; Mossin, S.; Varonka, M.
S.; Melzer, M. M.; Meyer, K.; Cundari, T. R.; Warren, T. H. Angew.
Chem., Int. Ed. 2010, 49, 8850−8855.
(24) Gephart, R. T.; Huang, D. L.; Aguila, M. J. B.; Schmidt, G.;
Shahu, A.; Warren, T. H. Angew. Chem., Int. Ed. 2012, 51, 6488−6492.
(25) Melzer, M. M.; Mossin, S.; Dai, X.; Bartell, A. M.; Kapoor, P.;
Meyer, K.; Warren, T. H. Angew. Chem., Int. Ed. 2010, 122, 904−907.
(26) Dai, X.; Warren, T. H. Chem. Commun. 2001, 1998−1999.
(27) (a) Kempe, R. Eur. J. Inorg. Chem. 2003, 791−803. (b) Deeken,
S.; Motz, G.; Kempe, R. Z. Anorg. Allg. Chem. 2007, 633, 320−325.
(28) (a) van Albada, G. A.; Smeets, W. J. J.; Veldman, N.; Spek, A. L.;
Reedijk, J. Inorg. Chim. Acta 1999, 290, 105−112. (b) Comba, P.;
Curtis, N. F.; Lawrance, G. A.; Sargeson, A. M.; Skelton, B. W.; White,
A. H. Inorg. Chem. 1986, 25, 4260−4267.
(29) Luo, Y.-R. Handbook of Bond Dissociation Energies in Organic
Compounds; CRC Press: Boca Raton, FL, 2002.
(30) Gephart, R. T.; McMullin, C. L.; Sapiezynski, N. G.; Jang, E. S.;
Aguila, M. J. B.; Cundari, T. R.; Warren, T. H. J. Am. Chem. Soc. 2012,
134, 17350−17353.
(31) Finn, M.; Friedline, R.; Suleman, N. K.; Wohl, C. J.; Tanko, J.
M. J. Am. Chem. Soc. 2004, 126, 7578−7584.
(32) Warren, J. J.; Tronic, T. A.; Mayer, J. M. Chem. Rev. 2010, 110,
6961−7001.
(33) (a) Kharasch, M. S.; Sosnovsky, G. J. Am. Chem. Soc. 1958, 80,
756−756. (b) Kharasch, M. S.; Sosnovsky, G.; Yang, N. C. J. Am.
Chem. Soc. 1959, 81, 5819−5824. (c) Rawlingson, D. J.; Sosnovsky, G.
Synthesis 1972, 1−28. (d) Andrus, M. B.; Lashley, J. C. Tetrahedron
2002, 58, 845−866.
(34) Eames, J.; Watkinson, M. Angew. Chem., Int. Ed. 2001, 40,
3567−3571.
(35) (a) Kochi, J. K. Science 1967, 155, 415−424. (b) Jenkins, C. L.;
Kochi, J. K. J. Am. Chem. Soc. 1972, 94, 856−865. (c) Kochi, J. K.;
Jenkins, C. L. J. Org. Chem. 1971, 36, 3095−3102. (d) Kochi, J. K.;
Subramanian, R. V. J. Am. Chem. Soc. 1965, 87, 1508−1514.
(36) Budzelaar, P. H. M.; Moonen, N. N. P.; de Gelder, R.; Smits, J.
M. M.; Gal, A. W. Eur. J. Inorg. Chem. 2000, 753−769.
(37) Gomberg, M. J. Am. Chem. Soc. 1900, 22, 757−771.
(38) Zunszain, P. A.; Shah, M. M.; Miscony, Z.; Javadzadeh-
Tabatabaie, M.; Haylett, D. G.; Ganellin, C. R. Arch. Pharm. 2002, 335,
159−166.
10940
dx.doi.org/10.1021/ja5026547 | J. Am. Chem. Soc. 2014, 136, 10930−10940