Reactivity of bio-inspired Cu(II) (N2/Py2) complexes with peroxide at room temperature

Article abstract of DOI:10.1016/j.jinorgbio.2019.03.014

Developing coordination complexes of earth abundant metals that can perform substrate oxidations under benign conditions is an ongoing challenge. Herein, the reactivity of two mononuclear Cu-complexes toward the oxidant H₂O₂ is reported. Both complexes displayed ligand oxidation upon reaction with the oxidant. Analysis of spectroscopic data established that the respective product complexes contained mononuclear Cu(II) centers. Moreover, treatment of these Cu-complexes with oxidant in the presence of substrate resulted in the interception of ligand oxidation with preferential oxidation of the substrate. Computational studies identified plausible mechanistic pathways, suggesting a copper-oxyl intermediate as the likely reactive intermediate responsible for substrate and ligand oxidation. To our knowledge, this is the first Cu-mediated system that showed ligand oxidation, oxo-transfer capability, and external hydrocarbon oxidation under stoichiometric conditions.

Full text of DOI:10.1016/j.jinorgbio.2019.03.014

Accepted Manuscript
Reactivity of bio-inspired Cu(II) (N2/Py2) complexes with
peroxide at room temperature
Nirupama Singh, Niharika Botcha, Thomas M. Jones, Mehmed Z.
Ertem, Jens Niklas, Erik R. Farquhar, Oleg G. Poluektov, Anusree
Mukherjee
PII:
S0162-0134(18)30613-5
DOI:
Reference:
https://doi.org/10.1016/j.jinorgbio.2019.03.014
JIB 10674
To appear in:
Journal of Inorganic Biochemistry
Received date:
Revised date:
Accepted date:
16 October 2018
5 March 2019
18 March 2019
Please cite this article as: N. Singh, N. Botcha, T.M. Jones, et al., Reactivity of bio-inspired
Cu(II) (N2/Py2) complexes with peroxide at room temperature, Journal of Inorganic
Biochemistry, https://doi.org/10.1016/j.jinorgbio.2019.03.014
This is a PDF file of an unedited manuscript that has been accepted for publication. As
a service to our customers we are providing this early version of the manuscript. The
manuscript will undergo copyediting, typesetting, and review of the resulting proof before
it is published in its final form. Please note that during the production process errors may
be discovered which could affect the content, and all legal disclaimers that apply to the
journal pertain.

ACCEPTED MANUSCRIPT
Reactivity of Bio-inspired Cu(II) (N2/Py2) Complexes with Peroxide at
Room Temperature
Nirupama Singh, † Niharika Botcha, † Thomas M. Jones, † Mehmed Z. Ertem, ‡
1
Jens Niklas, § Erik R. Farquhar, ┴ Oleg G. Poluektov, § Anusree Mukherjee †*
† Department of Chemistry, University of Alabama in Huntsville, 301 Sparkman
Drive,
Huntsville,
AL
35899
‡
Chemistry Division, Brookhaven National Laboratory, Upton, New York,
11973-5000
§ Chemical Sciences and Engineering Division, Argonne National Laboratory,
9700 S. Cass Avenue, Lemont, IL 60439
┴
Case Western Reserve University, Center for Synchrotron Biosciences,
National Synchrotron Light Source II, Brookhaven National Laboratory, Upton,
New York, 11973-5000
Abstract
Developing coordination complexes of earth abundant metals that can
perform substrate oxidations under benign conditions is an ongoing challenge.
Herein, the reactivity of two mononuclear Cu-complexes toward the oxidant
H2O2 is reported. Both complexes displayed ligand oxidation upon reaction with
the oxidant. Analysis of spectroscopic data established that the respective
product complexes contained mononuclear Cu(II) centers. Moreover, treatment
*1Corresponding author: anusree.mukherjee@uah.edu (email) +1 2568243782 (phone)
1

ACCEPTED MANUSCRIPT
of these Cu-complexes with oxidant in the presence of substrate resulted in the
interception of ligand oxidation with preferential oxidation of the substrate.
Computational studies identified plausible mechanistic pathways, suggesting a
copper-oxyl intermediate as the likely reactive intermediate responsible for
substrate and ligand oxidation. To our knowledge, this is the first Cu-mediated
system that showed ligand oxidation, oxo-transfer capability, and external
hydrocarbon oxidation under stoichiometric conditions.
Keywords
Cu-reactivity with peroxide, hydrocarbon oxidation
1. Introduction
The functionalization of hydrocarbons in an environmentally benign manner is
one of the greatest challenges of chemistry in the twenty-first century. Oxidative
transformations are useful for hydrocarbon functionalization and are heavily
involved in the conversion of petroleum feedstocks to building blocks for the
synthesis of many useful chemicals and pharmaceutical compounds. Metabolic
processes incorporating C-H bond oxidation can serve as inspiration for
developing eco-friendly pathways for obtaining synthetic chemicals. In recent
decades, there has been enormous progress in the development and
incorporation of bio-inspired strategies to achieve regio- and stereo- selective
C-H bond oxidations.1
In the literature, reactivity studies of Cu-complexes with hydrogen peroxide
(H2O2) have primarily been performed in the interest of understanding the
2

ACCEPTED MANUSCRIPT
mechanism of the enzymes dopamine beta-monooxygenase
and
peptidylglycine alpha-hydroxylation monooxygenase.2 These biomimetic
studies have provided important insights that allowed key copper-oxygen
adducts to be trapped and characterized.3 Cu-complexes supported by the
bispicen family of ligands with amine and pyridine functionalities can also serve
as suitable functional models for the histidine-rich coordination environment of
the copper active sites of lytic polysaccharide monooxygenase and particulate
methane monooxygenase.4 While the reactivity of mononuclear copper
complexes toward dioxygen activation has been extensively investigated,
reports of the reactivity of mononuclear copper complexes toward hydrogen
peroxide remain limited.4-5 This biomimicry provides valuable insight into the
nature of copper-oxygen intermediate species and the factors that influence their
formation and reactivity.3 Previous studies have concluded that the reactivity of
mononuclear Cu-complexes with H2O2 is mostly governed by the coordination
geometry around the metal center.6 Specifically, mononuclear complexes that
generated a Cu(II)-hydroperoxo (Cu-OOH) intermediate with square pyramidal
geometry exhibited substrate oxidation, while those with trigonal bipyramidal
geometry displayed no reactivity.7 Additionally, another study found that a Cu-
OOH species with square pyramidal geometry showed oxidation of a ligand
arm.8 In all cases, the Cu-OOH species was implied as a reactive intermediate
but the involvement of other intermediates could not be ruled out.8-9
Mononuclear copper complexes have shown great promise for C-H bond
activation using O2 as the oxidant.4, 10 However, reports of mononuclear Cu-
complexes that oxidize C-H bonds using hydrogen peroxide as an oxidant
remain limited.
3

ACCEPTED MANUSCRIPT
In the present work, we have explored the reactivity of two mononuclear
complexes in square pyramidal geometry with H2O2 as the oxidant. BPMEN
(N,N′-dimethyl-N,N′-bis-(pyridine-2-ylmethyl)-1,2-diaminoethane), a ligand
from the bispicen family, has already been reported to support mononuclear Cu-
complexes in a square pyramidal geometry.11 Here, through a combined
experimental and computational approach, we examine the reactivity of two Cu-
complexes, [Cu(BPMEN)(ClO4)2] (1) and its quinoline derivative (BQDMEN:
N,N′-dimethyl-N,N′-bis-(quinoline-2-ylmethyl)-1,2-diaminoethane)
[Cu(BQDMEN)(ClO4)2] (2) (Figure 1), with H2O2, and show that the oxidant
generated from both complexes performs ligand oxidation that can be
intercepted by the addition of triphenylphosphine and external hydrocarbons.
Figure 1. Copper complexes studied in the present work
2. Experimental
2.1 Materials and methods: All reagents and solvents were purchased from
Sigma-Aldrich and Fisher Scientific and directly used for synthesis without
4

ACCEPTED MANUSCRIPT
further purification unless otherwise mentioned. Copper complexes and ligands
were synthesized following reported procedures.11 9,10-dihydroanthracene
(DHA) was purchased from Sigma-Aldrich and recrystallized from ethanol
prior to use.12 All electronic absorption spectra were measured using an Agilent
Cary 8454 UV-Vis (Ultraviolet-Visible) Spectrophotometer equipped with a
Unisoku low-temperature cryostat. Samples for optical spectroscopy were
prepared in quartz cells with an optical path length of 1 cm. The electrospray
ionization mass spectrometry (ESI-MS) experiments were performed with a
Thermo Scientific LTQ Orbitrap XL Hybrid Fourier Transform Mass
Spectrometer (FT-MS) in positive ionization mode. Graphical representation
and data analysis were performed in Origin. Nuclear magnetic resonance
(NMR) spectroscopy samples were prepared by redissolving the demetallated
product solution in deuterated methanol (CD3OD). All NMR experiments were
carried out using a Varian 500 MHz spectrometer. Heteronuclear multiple bond
correlation (HMBC) experiments were realized using the standard crisis2
13
heteronuclear multiple bond correlation (gc2hmbc) pulse sequence with a C
spectral width of 0-225 ppm, 256 t1 increments, and 64 scans per t1 increment.
Band-selective HMBC experiments were realized using the band-selective
gradient heteronuclear multiple bond correlation (bsgHMBC) pulse sequence
with a 13C spectral width of 153.1-202 ppm, 512 t1 increments, and 32 scans per
t1 increment at a sampling density of 75%. Data processing was done using
MestReNova (version 12.0.1).
Electron paramagnetic resonance (EPR) samples were prepared by
removing the solvent from the product solution mixture and then re-dissolving
the sample in an acetonitrile/dichloromethane (MeCN/DCM) 1:3 solvent
5

ACCEPTED MANUSCRIPT
mixture. All EPR spectroscopy experiments were carried out with a Bruker
ELEXSYS II E500 EPR spectrometer (Bruker Biospin, Rheinstetten,
Germany), equipped with a TE102 rectangular EPR resonator (Bruker ER
4102ST). A helium gas-flow cryostat (ICE Oxford, UK) and a temperature
controller (ITC503; Oxford Instruments, UK) were used for measurements at
cryogenic temperatures (T = 50 K). Data processing was done using Xepr
(Bruker BioSpin) and Matlab 7.11.2 (The MathWorks, Inc., Natick) software
environments. Simulations were performed using the EasySpin software
package (version 5.1.8).13
X-ray absorption spectroscopy (XAS) spectra were obtained at the
Stanford Synchrotron Radiation Lightsource (SLAC National Accelerator
Laboratory, Menlo Park, CA) on beamlines 2-2 (1, 2, and 2-Product) and 9-3
(1-Product). SPEAR3 operated at 3.0 GeV and 500 mA in top-off mode. On
beamline 2-2, a water-cooled Si(111) double-crystal monochromator was used
for energy selection, detuned by 30-40% to reject higher harmonics. On
beamline 9-3, a cryogenically-cooled Si(220) double-crystal monochromator
was used for energy selection, with a Rh-coated harmonic rejection mirror
before the monochromator rejecting higher harmonics (13 keV cutoff
configuration) and a Rh-coated toroidal focusing mirror after the
monochromator used for tuning beam focus. During data collection, samples
were maintained at 20K on beamline 2-2 with a He Displex cryostat and 10K
on beamline 9-3 with an Oxford liquid helium cryostat. A copper metal foil was
collected simultaneously using a photodiode for internal energy calibration,
with the first inflection point of the reference foil set to 8979.0 eV. XAS spectra
were collected in fluorescence using 13 element (beamline 2-2) and monolithic
6

ACCEPTED MANUSCRIPT
100 pixel (beamline 9-3) solid state germanium detectors (Canberra). On
beamline 9-3, use of a Z-1 filter (6µ Ni) and Soller slits was required to maintain
detector linearity. Spectra were monitored for evidence of radiation damage
during data collection, as indicated by progressive red-shifts in the absorption
edge energy, and new spots on the sample exposed as necessary. XAS samples
were prepared for 1, 2, 1-product and 2-product by freezing the starting
materials and the final product solutions, respectively.
All XAS data were processed using Athena, while Artemis was used for
extended X-ray absorption fine structure (EXAFS) analysis.14 Theoretical phase
and amplitude parameters were calculated for representative models of 1 and 2
using the FEFF6L program integrated into Artemis and the output inspected to
identify significant paths. For a given shell, the coordination number n was
2
fixed, while r and σ2 were allowed to float. The amplitude reduction factor S0
was fixed at 0.9, while the edge shift parameter ΔE0 was allowed to float at a
single common value for all shells. The fit was evaluated in k3-weighted R-
space, and fit quality was judged by the reported R-factor and reduced χ2.
Significant fits are tabulated in Tables S1-S4.
2.2 Reaction with H2O2: The reactions of 1 and 2 with H2O2 were performed
in a UV-Vis cell under anaerobic conditions. Solutions of 1 and 2 were prepared
in anhydrous acetonitrile (2 mM) and kept at room temperature. To this solution,
triethylamine (NEt3, 10 equivalents) and H2O2 (10 equivalents) were added
using a gas tight syringe.
7

ACCEPTED MANUSCRIPT
2.3 Demetallation studies: Demetallation studies were carried out as
previously reported.15 The reaction mixture was allowed to stir for 24h at room
temperature and was followed by the addition of ammonium hydroxide (2 mL,
28.0-30.0%). The resulting solution was passed through a mini-alumina
column. The alumina column was washed with methanol (10 mL) followed by
dichloromethane (20 mL). The combined filtrate was further washed with water
(3 x 10 mL). The organic layer was collected and dried over sodium sulfate, and
the solvent was removed in vacuo. Mass recovery was ca. 2 mg (ca. 60%,
[BPMEN+O-2H]+). Mass spectrometry was used to identify the product. The
1H-NMR spectrum of the demetallated solution indicated that while the degree
of ligand oxidation is not high, oxidized product clearly has been formed.
Diffusion-ordered spectroscopy (DOSY) indicated that the mixture was likely
more complicated than originally anticipated. This was corroborated via HMBC
experiments. The presence of two 13C resonances at 169.4 ppm and 169.9 ppm
in Figure S1 were further resolved through a band-selective HMBC experiment
(Figure S1 inset). The data showed two distinct carbonyl signals that fall within
the typical chemical shift range for amides.16 The 13C resonance 169.9 ppm was
assigned to the amide product resulting from oxidation of the benzylic position,
and the resonance at 169.4 ppm was assigned to the oxidation of the ligand’s
methylene backbone (one of the NCH2).17 The data suggest that the ligand is
oxidized at two different positions, both of which correspond to the peak at m/z
= 285, providing spectroscopic evidence that the mixture contains both
oxidation products. Because of the similarity of the functional groups, the
8

ACCEPTED MANUSCRIPT
different products were not distinguished through other spectroscopic
measurements.
A similar procedure was performed to demetallate complex 2 for which mass
recovery was ca. 3 mg (ca. 27%, [BQDMEN+O-2H]+). Product was similarly
identified by ESI-MS spectroscopy. As the yield of ligand oxidation is lower
with 2, NMR analysis focused on complex 1 only.
2.4 Gas chromatography analysis of external substrates oxidation:
Oxidation of external substrates was performed with complexes 1 and 2. All
solutions were kept under anaerobic conditions, and all experiments were
performed, at minimum, in triplicate.
2.5 General procedure for substrate oxidation: In a sample solution of Cu-
complexes, 40 equivalents of substrate were added, followed by the addition of
H2O2 and NEt3 solutions in acetonitrile (each 10 equivalents with respect to Cu-
complex). Reaction progress was monitored with UV-Vis absorption
spectroscopy, and products were identified after removal of the Cu-complex
using a mini-alumina column. Quantification of the oxidized products was
carried out with gas chromatography (GC) equipped with a flame ionization
detector (FID) using naphthalene as an internal standard.
2.6 Computational Methods
All geometries were fully optimized at the M06-L level of density
functional theory18 using the Stuttgart [8s7p6d2f | 6s5p3d2f] ECP10MWB
contracted pseudopotential basis set on Cu19 and the 6-31G(d) basis set20 on all
9

ACCEPTED MANUSCRIPT
other atoms. The grid=ultrafine option (in Gaussian 0921) was chosen for
integral evaluation, and an automatically generated density-fitting basis set was
used within the resolution-of-the-identity approximation for the evaluation of
Coulomb integrals. The nature of all stationary points was verified by analytic
computation of vibrational frequencies, which were also used for the
computation of zero-point vibrational energies, molecular partition functions,
and for determining the reactants and products associated with each transition-
state structure (by following the normal modes associated with imaginary
frequencies).
Partition functions were used in the computation of 298 K thermal
contributions to free energy employing the usual ideal-gas, rigid-rotator,
harmonic oscillator approximation.22 Solvation effects associated with
acetonitrile as solvent were accounted for by using the SMD continuum
solvation model.23 A 1 M standard state was used for all species in solution thus,
for all molecules, the free energy in solution is computed as the 1 atm gas-phase
free energy, plus an adjustment for the 1 atm to 1 M standard-state concentration
change of RT ln (24.5), or 1.9 kcal/mol, plus the 1 M to 1 M transfer (solvation)
free energy computed from the SMD model. Free energy contributions were
added to single-point M06-L electronic energies computed with the SDD basis
set on Cu and the 6-311+G(2df,p) basis set on all other atoms to arrive at final,
composite free energies.
Time-dependent density functional theory (TDDFT) calculations were
performed to predict the UV/visible electronic excitations of relevant structures.
The B3LYP density functional,24 Stuttgart [8s7p6d2f | 6s5p3d2f] ECP10MWB
contracted pseudopotential basis set on Cu19 and the 6-311+G(d,p) basis set20
10

ACCEPTED MANUSCRIPT
on all other atoms were used for the TDDFT calculations. Non-equilibrium
solvation effects were included via the linear response approximations25 in
combination with the SMD aqueous continuum solvation model.23
Standard reduction potentials were calculated for various possible redox
couples to assess the energetic accessibility of different intermediates at various
oxidation states. For a redox reaction of the form
O
(soln) + ne(-g) ® R
(1)
(soln)
where O and R denote the oxidized and reduced states of the redox couple,
respectively, and n is the number of electrons involved in redox reaction, the
o
O R
E
reduction potential
relative to SCE was computed as
DGOo R
EOo R = -
-DEroef
(2)
nF
o
where O R is the free energy change associated with eq. (2) (using Boltzmann
G
DEroef
Eo
statistics for the electron) and
is taken as -4.422 V, which is required for
Eo
the conversion of calculated O Rrelative to the vacuum level to O Rversus the
saturated calomel electrode (SCE) in acetonitrile.26
3. Results and Discussion
The syntheses of 1 and 2 and their characterization by different
spectroscopic techniques have been reported.11 Complexes 1 and 2 exhibit broad
d-d transition bands at approximately 624 and 626 nm with a low-energy tail.
Addition of H2O2 (10 equiv.) to a blue solution of 1 in acetonitrile (MeCN) in
11

ACCEPTED MANUSCRIPT
the presence of triethylamine (NEt3 ,10 equiv.) at room temperature resulted in
a rapid color change to green. This color change was accompanied by the decay
of the 624 nm peak and generation of a new peak at 655 nm that subsequently
disappeared to form a new product species with an absorption feature at 484 nm
and a shoulder at 641 nm (Figure 2). When the final product solution of complex
1 was analyzed by ESI-MS, a new peak at m/z = 347.17 was observed, along
with the parent ion peak at m/z = 333.25. Both peaks exhibit similar isotopic
distribution patterns (Figure 2). The peak with m/z = 333.25 is consistent with a
[(BPMEN)Cu]+ formulation, suggesting that the addition of 14 mass units for
m/z = 347.17 can be interpreted as addition of an oxygen atom to
[(BPMEN)Cu]+ with concomitant loss of two hydrogen atoms, resulting in the
formation of [(BPMEN)Cu+O-2H]+.
Figure 2. Left Panel: UV-Visible spectral changes of complex 1 (black) upon
addition of H2O2/NEt3. The initial solution (2 mM, black solid line) immediately
generates an intermediate solution (blue dash-dotted line) and slowly converts
12

ACCEPTED MANUSCRIPT
to a final solution (green dashed line) in acetonitrile at room temperature. Grey
lines represent spectral changes over the time. Right Panel: ESI-MS spectrum
of the final product solution of 1 (black, top) and its simulation (red, bottom).
In order to determine whether the oxygen atom was inserted in the
ligand, the product solution was demetallated with aqueous NH4OH and
extracted with dichloromethane. ESI-MS analysis of the extracted organic layer
showed two peaks, one of which is readily assigned to protonated BPMEN
ligand (m/z = 271), while the second at m/z = 285.25 corresponds to an
oxygenated ligand [BPMEN+O-2H]+ (Figure 3), with yield of 60%.
Figure 3. ESI-MS spectra of demetallated solution from complex 1 (final green
chromophore). (In red: Software simulation of expected mass peaks).
Treatment of complex 2 in MeCN with H2O2/NEt3 (10 equiv.) elicited
similar changes. Specifically, the bluish-green solution of the starting material
associated with the peak at 626 nm decreased in intensity and generated a final
13

ACCEPTED MANUSCRIPT
solution with dark green color with a peak at 642 nm along with a shoulder at
~504 nm (Figure 4). Mass spectrometric analysis of the final solution showed
two prominent peaks at m/z 433.3 ([(BQDMEN)Cu]+) and 447.3
([(BQDMEN)Cu+O-2H]+). The ligand oxidation product was isolated in 27%
yield by demetallation and confirmed by mass spectrometry that showed a peak
at m/z = 385 (Figure 5).
Figure 4. Left Panel: UV-Vis absorption spectral changes of complex 2 (Black,
2 mM) upon addition of NEt3 and H2O2. The initially formed intermediate
solution (blue dash-dotted line) slowly converted to final solution (green dashed
line) in acetonitrile at room temperature. Grey lines represent spectral changes
over the time. Right Panel: ESI-MS spectrum of the final product solution
(upper panel) and simulated mass spectrum showing expected isotope pattern
(lower panel).
14

ACCEPTED MANUSCRIPT
Figure 5. ESI-MS spectra of demetallated solution from complex 2 (final green
chromophore). (In red: Software simulation of expected mass peaks).
In summary, these observations suggest that the reaction of complexes
1 and 2 with H2O2 at room temperature resulted in the oxidation of the ligand
framework (Figure S1). This is an important finding, as there are very few
reports of ligand oxidation with mononuclear copper(II) complexes and
H2O2/NEt3.3, 8, 27 In order to corroborate the source of the oxygen atom inserted
18
in the ligand, the reactions were performed in the presence of H2 O2. Use of
this isotopically labeled oxidant elicited a shift of the peak at m/z = 347 to m/z
= 349, consistent with the matching simulation (Figure S2). This result
establishes that H2O2 is the source of the oxygen inserted in the ligand.
In order to establish the nature of the metallic species present in the
product solution, we undertook additional spectroscopic studies of the final
product solutions generated from 1 and 2. Continuous wave (CW) X-band (9-
10 GHz) EPR of frozen product solutions of 1 and 2 displayed axial EPR
15

ACCEPTED MANUSCRIPT
spectra, with g║= 2.238, g┴ = 2.055 and A║=181 x 10-4 cm-1 (63Cu) for the 1-
Product solution and g║= 2.242, g┴ = 2.054 and A║=170 x 10-4 cm-1 (63Cu) for
the 2-Product solution (Figure 6). These values are typical of mononuclear
Cu(II) complexes in square pyramidal geometry (g║ > 2.1 > g┴ > 2.00 and
A║=158-201 x 10-4 cm-1) suggesting that the major species in the final solution
is also mononuclear in nature.28
Figure 6. CW X-band EPR spectra (at T = 50 K) of final product solutions from
reactions of complexes 1 and 2 with H2O2/NEt3 at room temperature in a 1:3
mixture of MeCN/DCM (see methods for details). Black – experimental
spectra; Red – simulated spectra. Simulation parameters are g║= 2.238, g┴ =
2.055 and A║=181 x 10-4 cm-1 for 1-Product solution and g║= 2.242, g┴ = 2.054
and A║=170 x 10-4 cm-1 for 2-Product solution. Cu hyperfine coupling
63
constants are given for the Cu isotope. The perpendicular value A┴ is not
given, since the hyperfine structure is not clearly resolved. The simulations were
16

ACCEPTED MANUSCRIPT
performed assuming axial g-tensor and Cu A-tensors. In addition, it was
assumed for the simulation that the principal g-tensor axes are collinear with the
Cu A-tensor principal axes.
This was further corroborated by X-ray absorption spectroscopy. The X-
ray absorption near edge structure (XANES) spectra (Figure 7) of all four
complexes show an extremely weak 1s-to-3d transition at 8977.8 eV and
minimal structure along the rising edge, consistent with a Cu(II) center.
Notably, spectra for both of the 1-Product and 2-Product solutions show
somewhat weakened pre-edges and broadened edges relative to their respective
precursors.
17

ACCEPTED MANUSCRIPT
Figure 7. Normalized XANES spectra of (a) 1 (——) and 1-Product (- - -), (b)
2 (——) and 2-Product (- - -). The insets show expansions of the pre-edge
region.
EXAFS analysis (Figure 8, Tables S1-S4) establishes that all four
complexes are best fit with a set of 5 nitrogen ligands at 1.98 – 2.02 Å. Complex
2 exhibits a slightly altered coordination environment, with 4N at 2.03 Å and 1
N/O (presumably bound counter ion or solvent molecule) at 2.22 Å, in line with
an apically distorted square pyramidal geometry. There is evidence that the
18

ACCEPTED MANUSCRIPT
other three complexes (1, 1-Product, and 2-Product) have a similarly distorted
structure, however they could not be defensibly fit with two subshells of
scatterers in the first coordination sphere.
19

ACCEPTED MANUSCRIPT
Figure 8. Representative best fits (bolded entries in Tables S1-S4) to k3-
weighted EXAFS data of 1 (a), 1-Product solution (b), 2 (c), and 2-Product
solution (d). Experimental data is shown as a dotted line while the best fit is
shown as a solid line.
The observation of ligand oxidation by 1 and 2 in the presence of H2O2
led us to investigate the reactivity of these complexes toward external substrates
under similar conditions. We first examined oxo-transfer reactivity using
triphenylphosphine (PPh3). Addition of H2O2 to the solution of 1 in MeCN in
the presence of excess PPh3 (40 equiv.) resulted in the complete loss of the 624
nm peak assigned to 1. No growth of the features at 483 nm and 641 nm
associated with ligand oxidation was observed (Figure 9). ESI-MS analysis of
the faint-yellow product solution showed only a single peak at m/z = 333.25
corresponding to [(BPMEN)Cu]+. No peak at m/z = 347.17 assignable to
[(BPMEN)Cu+O-2H]+ was observed (Figure 9), which suggests that ligand
oxidation was inhibited in the presence of phosphine. Analysis of the product
solution using gas chromatography – mass spectrometry (GC-MS) showed the
presence of triphenylphosphine oxide (Ph3PO), indicating that 1 was able to
oxidize PPh3 in the presence of H2O2 (Table 1). Complex 2, when allowed to
react with H2O2 in the presence of PPh3, exhibited similar reactivity (Figure 9
and Table 1).
20

ACCEPTED MANUSCRIPT
Figure 9. UV-Vis absorption spectral changes of complex 1 (2mM, upper
panel) and complex 2 (2mM, lower panel) with H2O2/NEt3 in the presence of
excess PPh3 (40 equivalents) in acetonitrile at room temperature in a cuvette
with a 1-cm path length. On right side, ESI-MS spectra of the final product
solution (Upper panel: complex 1 +PPh3 +H2O2/NEt3, Lower panel: complex 2
+PPh3 +H2O2/NEt3).
With this intriguing result in hand, we then evaluated whether 1 and 2
were capable of reacting with hydrocarbons as exogenous substrates. It is
noteworthy that no mononuclear Cu-complexes reported thus far showed
external hydrocarbon oxidation with H2O2 except under catalytic conditions.29
We explored the reactivity of 9,10-dihydroanthracene (DHA, 77 kcal/mol) and
1,4-cyclohexadiene (CHD, 78 kcal/mol) under our conditions. Both substrates
have weak C-H bonds, but access toward the weakest C-H bond is more
sterically demanding for DHA compared to CHD.30 Oxidation of 1 and 2 in the
21

ACCEPTED MANUSCRIPT
presence of excess hydrocarbons (40 equiv.) resulted in absorption spectra
(Figures 10 and 11) that were significantly different than those observed without
substrate (Figures 2 and 4).
Analyses of the product solutions using GC-MS established that both
substrates were oxidized. In the case of DHA, both anthraquinone and
anthracene were identified as products, while benzene was obtained as the major
product of CHD oxidation (Table 1). The product yields were calculated with
respect to complexes 1 and 2. When the hydrocarbon oxidation reactions were
attempted with copper(II) perchlorate or with tetrakis(acetonitrile)copper(I)
tetrafluoroborate, addition of H2O2 resulted in precipitation, suggesting that the
reactivity displayed by 1 and 2 arises because of their unique ligand
environment. Furthermore, no measurable amount of hydrocarbon oxidation
was observed using only oxidant in the absence of 1 and 2 (with PPh3 we found
2% of product Ph3PO under this condition). Collectively, our reactivity studies
show that 1 and 2 are able to oxidize the ligand, perform oxo-transfer to a
phosphine substrate, and oxidize external hydrocarbons using H2O2 as the
oxidant.
22

ACCEPTED MANUSCRIPT
Figure 10. Optical changes of complex 1 (Panel A) and 2 (Panel B) with
H2O2/NEt3 in the presence of excess DHA in acetonitrile at room temperature.
Spectrum of starting material (in black solid line), final spectrum (in green
dashed line). Gray lines represent spectral changes over time in a 1-cm path
length cuvette.
23

ACCEPTED MANUSCRIPT
Figure 11. UV-Vis absorption spectrum of complex 1 (A) and 2 (B) with
H2O2/NEt3 in presence of excess CHD in acetonitrile at room temperature;
starting material (in black solid line), final spectra (in brown dashed line). Gray
lines represent spectral changes over the time in a 1-cm path length cuvette.
Table 1. Product analysis with different substrates
Substrate
1
2
Product (%Yield)
O=PPh3
(23%)
Product (%Yield)
O=PPh3
(35%)
Triphenylphosphine
(PPh3)
9,10-Dihydro-anthracene
(DHA)
Anthraquinone
(8%)
Anthraquinone
(3%)
Anthracene
(8%)
Anthracene
(4%)
1,4-Cyclo-hexadiene
(CHD)
Benzene
(68%)
Benzene
(67%)
Density functional theory (DFT) calculations at the M06-L level of
theory18 with the SMD continuum solvation model23 for acetonitrile were
performed to investigate the possible pathways for ligand and external substrate
24

ACCEPTED MANUSCRIPT
oxidation by complex 1. A proposed mechanism for ligand oxidation with H2O2
starting from [1-MeCN]2+ is presented in Scheme S1. Here we will focus on the
critical step of hydrogen atom abstraction by proposed CuII-OOH ([1-OOH]+)
and CuIIO•− ([1-O]+) species.
The optimized geometries of [1-MeCN]2+ and [1-OOH]+ at the M06-L
level of theory and computed UV-Vis spectra at the TD-B3LYP level of
theory24 confirm the square pyramidal geometry around the copper center and
the nature of the associated electronic bands (Figures S3-S4). The intermediate
[1-OOH]+ features three distinct sites for C-H activation, specifically, benzylic
CH2, bridging CH2 and amine CH3 groups. The oxidation kinetics of the ligand
at these distinct sites was probed by examining the free energy of activation
(ΔG‡) associated with hydrogen atom abstraction by either (i) the distal oxygen
atom (with respect to the Cu center) or (ii) the proximal oxygen atom of [1-
OOH]+ with concomitant O-O bond cleavage. The latter pathway (ii) was found
to be consistently more favorable, and optimized transition state (TS) structures
along with the associated ΔG‡ values for hydrogen atom abstraction from
distinct sites of the ligand by [1-OOH]+ are listed in Figure S5 along other TS
structures. We have provided further details in the supporting information on
the description and energetics of the optimized transition states, and we will
focus only on the relevant ones for the discussion here. Among the several sites
considered, H-abstraction from CH3 groups on the amine nitrogens features the
lowest activation free energy, with ΔG‡ of 22.7 kcal/mol (TS-1a, Figure 12),
whereas ΔG‡ is computed to be 37.1 kcal/mol for the benzylic CH2 groups (TS-
1b, Figure 12). Furthermore, we found H-abstraction from CHD as an external
substrate (TS-1c, Figure 12) to be higher in energy (ΔG‡ = 29.1 kcal/mol)
25

ACCEPTED MANUSCRIPT
compared to H-abstraction from the CH3 groups on the amine nitrogens, but still
lower than other possible sites of H-abstraction from the ligand (Figure S5).
Figure 12. C-H activation pathways via [1-OOH]+ intermediate. Free energy
changes are in units of kcal/mol.
Next, we considered the reactivity of a possible reduced copper species, [1-
OOH] (Figure S7), which could form via hydroperoxide coordination to a CuI
center upon homolytic cleavage of the Cu-O bond in [1-OOH]+ (ΔE‡ = 21.0
kcal/mol, Figure S6), as was recently proposed.9 The computed ΔG‡ values
indicate that if [1-OOH] forms in solution, CuIIO•− species could be generated
with heterolytic cleavage of the O-O bond in the presence of Et3NH+ as the
Bronsted acid (ΔG‡ = 14.8 kcal/mol) (Figure S7). Subsequent H-abstraction
26

ACCEPTED MANUSCRIPT
from the CH3 groups (TS-2a, Figure 13) or the benzylic CH2 groups (TS-2b,
Figure 13) of the ligand are equally feasible, with ΔG‡ values of 13.2 and 13.7
kcal/mol, respectively (see Figure S8 for alternative C-H activation pathways
with [1-O]+). Furthermore, including quantum tunneling effects via the Skodje-
Truhlar approximation yields activation barriers in a narrow range of (~ 5
kcal/mol) indicating that H-abstraction from different sites is quite feasible
(Figure S8).31 Interestingly, the reaction of the CuIIO•− intermediate with CHD
proceeds with an even lower ΔG‡ of 6.6 kcal/mol, in contrast to the CuII-OOH
case. This is in accord with the experimental observation that the external
substrate CHD can successfully intercept ligand oxidation. This latter finding
implies that CuIIO•− , but not CuII-OOH,could be the reactive species in solution.
27

ACCEPTED MANUSCRIPT
Figure 13. C-H activation pathways via the [1-O]+ intermediate. Free energy
changes are in units of kcal/mol.
4. Conclusion
In summary, a new mononuclear Cu-system that reacts with H2O2/NEt3
and performs both internal and external substrate oxidation is reported (Figure
14). Our findings show that both 1 and 2 give higher yields of product with
CHD substrate compared to DHA, likely as a consequence of the steric
requirements of the latter. Spectroscopic characterization of 1 and 2 and their
respective product solutions unequivocally establishes the mononuclear nature
of both the starting material and the major product species. Computational
calculations performed to understand the reaction mechanism suggest that a
cupric-oxyl intermediate is the more likely reactive intermediate responsible for
both ligand and external substrate oxidation.
Figure 14. Reactivity of 1 with H2O2 in MeCN in the presence and absence of
substrate. (2 exhibits similar reactivity).
28

ACCEPTED MANUSCRIPT
These noteworthy results represent, to our knowledge, the first report of
a copper system with H2O2 as the oxidant where all three modes of reactivity
are observed: ligand oxidation, oxo-transfer to phosphine, and external
hydrocarbon substrate oxidation. Clearly, the mechanistic underpinning of this
system requires a thorough investigation at low temperature. Work in our
laboratory is ongoing to trap and spectroscopically investigate the reactive
intermediate in these transformations and establish its electronic structure.
List of Abbreviations
BPMEN
N,N‘-dimethyl-N,N‘-bis-(pyridine-2-ylmethyl)-1,2-diaminoethane
BQDMEN
N,N‘-dimethyl-N,N‘-bis-(quinoline-2-ylmethyl)-1,2-diaminoethane
DHA
9,10-dihydroanthracene
UV-Vis
Ultraviolet-Visible
ESI-MS
Electrospray ionization mass spectrometry
29