Phenoxazinone synthase-like catalytic activity of novel mono- And tetranuclear copper(ii) complexes with 2-benzylaminoethanol

Article abstract of DOI:10.1039/d0dt00222d

Three novel coordination compounds, [Cu(ca)₂(Hbae)₂] (1), [Cu(va)₂(Hbae)₂] (2) and [Cu₄(va)₄(bae)₄]·H₂O (3), have been prepared by self-assembly reactions of copper(ii) chloride (1 and 2) or tetrafluoroborate (3) and CH₃OH (1 and 3) or CH₃CN (2) solution of 2-benzylaminoethanol (Hbae) and cinnamic (Hca, 1) or valeric (Hva, 2 and 3) acid. Crystallographic analysis revealed that both 1 and 2 have mononuclear crystal structures, wherein the complex molecules are H-bonded forming extended supramolecular chains. The tetranuclear structure of 3 is based on the {Cu₄(μ₃-O)₄} core, wherein the metal atoms are bound together by μ₃ oxygen bridges from 2-benzylaminoethanol forming an overall cubane-like configuration. The strong hydrogen bonding in 1-3 leads to the joining of the neighbouring molecules into 1D chains. Concentration-dependent ESI-MS studies disclosed the equilibria between di-, tri- and tetranuclear species in solutions of 1-3. All three compounds act as catalysts for the aerobic oxidation of o-aminophenol to the phenoxazinone chromophore (phenoxazinone synthase-like activity), with the maximum reaction rates of 4.0 × 10^-7, 2.5 × 10^-7 and 2.1 × 10^-7 M s^-1 for 1, 2 and 3, respectively, supported by the quantitative yield of the product after 24 h. The dependence of the reaction rates on catalyst concentrations is evidence of reaction orders higher than one relative to the catalyst. Kinetic and ESI-MS data allowed us to assume that the tetranuclear species, originating from 1, 2 and 3 in solution, possess considerably higher activity than the species of lower nuclearity. Mechanistic and isotopic ¹⁸O-labelling experiments suggested that o-aminophenol coordinates to Cu^II species with the formation of reactive intermediates, while the oxygen from ¹⁸O₂ is not incorporated into the phenoxazinone chromophore.

Full text of DOI:10.1039/d0dt00222d

Dalton
Transactions
View Article Online
View Journal
PAPER
Phenoxazinone synthase-like catalytic activity of
novel mono- and tetranuclear copper(II)
complexes with 2-benzylaminoethanol†
Cite this: DOI: 10.1039/d0dt00222d
Oksana V. Nesterova, a Olena E. Bondarenko,a Armando J. L. Pombeiro a and
a,b
Dmytro S. Nesterov
*
Three novel coordination compounds, [Cu(ca)2(Hbae)2] (1), [Cu(va)2(Hbae)2] (2) and [Cu4(va)4(bae)4]·H2O
(3), have been prepared by self-assembly reactions of copper(II) chloride (1 and 2) or tetrafluoroborate (3)
and CH3OH (1 and 3) or CH3CN (2) solution of 2-benzylaminoethanol (Hbae) and cinnamic (Hca, 1) or
valeric (Hva, 2 and 3) acid. Crystallographic analysis revealed that both 1 and 2 have mononuclear crystal
structures, wherein the complex molecules are H-bonded forming extended supramolecular chains. The
tetranuclear structure of 3 is based on the {Cu4(µ3-O)4} core, wherein the metal atoms are bound
together by µ3 oxygen bridges from 2-benzylaminoethanol forming an overall cubane-like configuration.
The strong hydrogen bonding in 1–3 leads to the joining of the neighbouring molecules into 1D chains.
Concentration-dependent ESI-MS studies disclosed the equilibria between di-, tri- and tetranuclear
species in solutions of 1–3. All three compounds act as catalysts for the aerobic oxidation of o-amino-
phenol to the phenoxazinone chromophore (phenoxazinone synthase-like activity), with the maximum
reaction rates of 4.0 × 10−7, 2.5 × 10−7 and 2.1 × 10−7 M s−1 for 1, 2 and 3, respectively, supported by the
quantitative yield of the product after 24 h. The dependence of the reaction rates on catalyst concen-
trations is evidence of reaction orders higher than one relative to the catalyst. Kinetic and ESI-MS data
allowed us to assume that the tetranuclear species, originating from 1, 2 and 3 in solution, possess con-
siderably higher activity than the species of lower nuclearity. Mechanistic and isotopic 18O-labelling
experiments suggested that o-aminophenol coordinates to CuII species with the formation of reactive
intermediates, while the oxygen from 18O2 is not incorporated into the phenoxazinone chromophore.
Received 19th January 2020,
Accepted 13th March 2020
DOI: 10.1039/d0dt00222d
rsc.li/dalton
of the O2 radical (superoxide dismutase).7–9 Phenoxazinone
synthase (PHS) is a copper-containing enzyme that catalyses
−
Introduction
Coordination compounds of copper are of continuous interest the oxidative coupling of substituted o-aminophenols to
due to their broad applications, including biomimetic catalytic produce the phenoxazinone chromophore (Fig. 1).10,11 PHS is
systems.1–6 Being one of the essential trace elements found in involved in the last stage of the biosynthesis of the antitumor
living organisms, copper, as part of active centres of many drug Actinomycin D, a naturally occurring antineoplastic agent
enzymes, catalyses a range of important oxidation–reduction clinically used in the treatment of a range of cancers.10,12 The
bioprocesses, such as hydroxylation of methane and other phenoxazine scaffold itself is considered as a promising
hydrocarbons (particulate methane monooxygenase, pMMO), radical-trapping antioxidant.13 This demonstrates the impor-
oxidation of o-diphenols (catechol oxidase) or decomposition tance of phenoxazinone synthetic reaction, which has great
potential for biocatalytic applications and drug design.
Furthermore, coupling of two o-aminophenol moieties could
aCentro de Química Estrutural, Instituto Superior Técnico, Universidade de Lisboa,
be considered as a novel click reaction for chemical and bio-
Av. Rovisco Pais, 1049-001 Lisboa, Portugal.
chemical studies.14
E-mail: dmytro.nesterov@tecnico.ulisboa.pt
bPeoples’ Friendship University of Russia (RUDN University), 6 Miklukho-Maklaya
st., Moscow 117198, Russia
The necessity of a multinuclear copper environment for
efficient PHS activity was clearly demonstrated in an early
research study17 and later was confirmed by an X-ray analysis
of PHS single crystals.15 The X-ray data showed that the whole
PHS exists in an oligomeric form wherein every oligomer con-
tains various copper active centres, from mono- to pentanuc-
†Electronic supplementary information (ESI) available: Selected bond lengths
and angles for crystal structures 1–3, figure of the structure 4, packing diagrams
for 1 and 2, IR spectra of 1–3, reaction chromatograms, and UV-Vis and EI mass-
spectra. CCDC 1912946-1912948, 1966765. For ESI and crystallographic data in
CIF or other electronic format see DOI: 10.1039/D0DT00222D
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
tion of homo- and heterometallic molecular complexes.6,26–30
These ligands, due to their polydentate nature and variety of
coordination modes, facilitate the formation of close-packed
molecular structures with a broad variety of nuclearities and
topologies, using the spontaneous self-assembly synthetic
approach.31,32 Although this approach somewhat lacks predict-
ability, it allows the synthesis of large cores which could be
difficult (or impossible) to assemble via other methods.6,26
The nature of substituents within the aminoalcohol ligand
structure serves as a “handle” that allows the tuning of the pro-
perties of the ligand.32 The use of bulky aliphatic substituents
is an efficient way to achieve steric isolation of the coordi-
nation cores and facilitate their solubility in non-polar sol-
vents, which is of great importance for catalytic applications.
Following this strategy, recently we have obtained a series of
unique copper compounds,33–36 including a Cu8 compound
having a paramagnetic ground spin state (which is not
common for high-nuclear CuII complexes)34 and heterometal-
lic Cu/Mn and Cu/Fe compounds.37
In this paper we report the synthesis of three novel mono-
and tetranuclear Cu(II) complexes [Cu(ca)2(Hbae)2] (1), [Cu
(va)2(Hbae)2] (2) and [Cu4(va)4(bae)4]·H2O (3) using mixed-
ligand reaction systems (2-benzylaminoethanol (1–3) and cin-
namic (1) or valeric (2 and 3) acid). Single crystal X-ray diffrac-
tion analysis reveals that 1 and 2 possess similar 1D supramo-
lecular structures formed by mononuclear moieties, while
supramolecular chains of 3 are created with the participation
of cubane-like tetranuclear molecules. To our knowledge, 1–3
Fig. 1 Top: overall view of the hexameric form in the crystal structure
of PHS (left) and the structure of the active centre containing four
copper atoms (RCSB PDB code 3GYR15,16). Bottom: schematic represen-
tation of the Actinomycin D biosynthesis catalysed by PHS (R = O(Thr-
Val-Pro-Sar-MeVal)) and the respective model reaction (oxidation of
o-aminophenol to the phenoxazinone chromophore).
lear ones (Fig. 1).15 Although the crystal structure of PHS was are the first mixed-ligand complexes containing simul-
solved in 2006, reports on successful artificial systems mimick- taneously 2-benzylaminoethanol and cinnamic or valeric acid
ing the PHS activity are not numerous.11 Just one year after the (basic forms) with any metal ions. The ability to catalytically
PHS structure discovery, Chaudhuri reported the catalytic generate the phenoxazinone chromophore from o-amino-
activity of a tetranuclear cubane-like copper core, bearing an phenol through its aerobic oxidation and thus mimic the func-
N,O-donor radical ligand, in the oxidation of o-aminophenol tion of the enzyme PHS has been studied in detail for 1–3. It
(OAP) to the phenoxazinone chromophore (see Fig. 1 for this has been found that 1 possesses an exceptionally high catalytic
model process).18 It was suggested that the presence of four performance, while the other complexes are less active.
copper atoms is essential and proposed that all of them par-
ticipate in the catalytic cycle by coordinating and oxidizing the
substrate and intermediates. Similar results were obtained for
Experimental section
General
other tetranuclear copper complexes with aminoalcohol
ligands.19,20 Coordination of OAP to metal centres to produce
reactive intermediates was suggested by EPR and ESI-MS spec- All chemicals were of reagent grade and used as received. All
trometry using Cu/Mn and Cu/Ln heterometallic complexes as experiments were carried out in air. Elemental analyses for C,
catalysts.21–23 Despite some diversity of coordination cores H and N were carried out by the Microanalytical Service of the
being tested as catalysts for OAP oxidation,11 the respective Instituto Superior Técnico. Infrared spectra (4000–400 cm−1
)
catalytic tests are typically limited to measuring the yields of were recorded using a BX-FT IR “PerkinElmer” instrument in
products only for one substrate (non-substituted OAP) and the KBr pellets. Mass spectra were obtained using an LCQ Fleet
study of W0 vs. [OAP]0 kinetic dependence (where W0 is the mass spectrometer with an ESI source (Thermo Scientific).
initial reaction rate) that tends to show a common behaviour. UV-Vis spectra were recorded using a PerkinElmer Lambda 35
Studies on other dependencies, particularly the one with spectrometer in a 330–600 nm spectral range, unless stated
respect to the determination of the reaction order relative to otherwise.
the catalyst, are rare.18,24,25 Thus, knowledge regarding the
catalytic pathways and intermediates in the OAP oxidation is 2 mmol), 2-benzylaminoethanol (0.85 mL, 6 mmol) and tri-
limited and requires further improvement.
ethylamine (0.14 mL, 1 mmol) were dissolved in CH3OH
Synthesis of [Cu(ca)2(Hbae)2] (1). CuCl2·2H2O (0.34 g,
This work stems from our ongoing interest in the utilization (30 mL) forming a blue-green solution which was magnetically
of simple and flexible aminoalcohol ligands for the construc- stirred at 50–60 °C (40 min). Then, cinnamic acid (0.3 g,
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
2 mmol) was added to this solution and the mixture was sparingly soluble in CH3CN and DMF, insoluble in water and
stirred for 1 h. Precipitation occurred immediately, and a soluble in CH3OH.
mixture of dark-green and dark-blue precipitates was obtained
Crystallography
at the end of the reaction. The resulting solution was filtered
and dark-blue microcrystals suitable for the X-ray crystallo-
graphic study were formed after two weeks. Yield of dark-blue
crystals: 0.11 g, 8% (based on copper chloride). Anal. calc. for
C36H40CuN2O6 (M = 660.24): C, 65.49; N, 4.24; H, 6.12%.
Found: C, 65.05; N, 4.21; H, 6.32%. The compound is sparingly
soluble in CH3CN and DMF, insoluble in water and soluble in
CH3OH.
Synthesis of [Cu(va)2(Hbae)2] (2). The complex was prepared
in a way similar to that of 1, but using valeric acid (0.22 mL,
2 mmol) and CH3CN (30 mL) instead of cinnamic acid and
CH3OH. Similar to 1, precipitation occurred immediately after
addition of carboxylic acid to the solution and a mixture of
dark-green and dark-blue precipitates was obtained at the end
of the reaction. The resulting solution was filtered and dark-
blue crystals suitable for X-ray crystallography were isolated
from the filtrate within two days. Yield of dark-blue crystals:
0.31 g, 27% (based on copper chloride). Anal. calc. for
C28H44CuN2O6 (M = 568.19): C, 59.18; N, 4.93; H, 7.82%.
Found: C, 58.96; N, 4.86; H, 7.78%. The compound is sparingly
soluble in CH3CN and DMF, insoluble in water and soluble in
CH3OH.
X-ray diffraction data were collected using a Bruker D8 Quest
(for 1–3) or AXS KAPPA APEX II diffractometer (for 4) with
graphite-monochromated Mo Kα radiation. Cell parameters
were retrieved and refined using Bruker SAINT and
CELL_NOW.38 Absorption corrections were applied using
SADABS.39 The structures were solved by direct methods and
refined against F2 using the programs package SHELX-2018/
3.40 The crystal structure of 4 has been reported earlier.41,42
The carbon atoms of the aliphatic groups C11–C14 (in 2), C9–
C12, C25–C28 and C53–C56 (in 3) were refined as being dis-
ordered over two positions with site occupancies equal to 0.51
(2), 0.602(19), 0.682(17) and 0.552(19), respectively (only main
site occupancies are given). The bond distances and angles in
these groups were restrained to ideal values. The hydrogen
atoms H1 and H2 (in 1 and 2) were localized in a difference
Fourier map and refined freely. Positions of the H atoms of the
water molecules O25 and O27 in 3 were not localised. All other
hydrogen atoms in 1–3 were placed at calculated positions and
refined using the model with Uiso = nUeq (n = 1.5 for H atoms
of the methyl group and n = 1.2 for other H atoms). The hydro-
gen atoms in 4 were located and refined freely by applying geo-
metrical restraints for the H atoms of the methyl group C7.
Details of data collection and processing, structure solution
and refinement are summarized in Table 1.
Synthesis of [Cu4(va)4(bae)4]·H2O (3). The complex was pre-
pared in a way similar to that of 2, but using Cu(BF4)2·2H2O
(0.55 g, 2 mmol) and CH3OH (30 mL) instead of CuCl2·2H2O
and CH3CN. Dark-blue crystals suitable for X-ray analysis were
obtained after one month from the resulting blue solution.
Yield: 0.22 g, 9% (based on copper tetrafluoroborate). Anal.
calc. for C56H84Cu4N4O13 (M = 1277.49): C, 52.65; N, 4.39; H,
6.79%. Found: C, 52.54; N, 4.38; H, 6.65%. The compound is
CCDC 1912947 (1), 1912946 (2), 1912948 (3) and 1966765 (4).†
Catalytic reactions
The phenoxazinone synthase activity of 1–3 was monitored
spectroscopically by recording the UV-Vis spectra in
a
Table 1 Crystal data and structure refinement for 1–4
Complex
1
2
3
4
Empirical formula
Formula weight
C36H40CuN2O6
660.24
C28H44CuN2O6
568.19
C56H84Cu4N4O13
1275.43
C7H9NO
123.15
Crystal system
Space group
Monoclinic
P21/n
Monoclinic
P21/c
Triclinic
P1
Orthorhombic
Pbca
ˉ
a/Å
b/Å
c/Å
α/°
7.0928(16)
15.026(4)
16.003(4)
90
92.895(9)
90
1703.4(7)
2
1.287
14.310(4)
14.800(4)
7.2165(18)
90
102.584(8)
90
1491.7(6)
2
1.265
17.161(5)
18.613(5)
19.954(6)
87.895(10)
84.097(10)
83.089(10)
6292(3)
4
1.346
296(2)
1.393
2672
7.5254(16)
7.7391(15)
22.763(5)
90
90
90
1325.7(5)
8
1.234
150(2)
0.083
528
β/°
γ/°
V/Å3
Z
Calculated density/g cm−3
T, K
296(2)
0.687
694
296(2)
1.265
606
μ(Mo-Kα)/mm−1
F(000)
Reflections collected/unique
Rint
89 053/4954
0.0410
4073
2.549, 30.000
0.0408
0.1134
23 639/3075
0.1620
1476
2.752, 26.498
0.0679
0.2069
305 943/28 877
0.0729
19 070
2.273, 27.500
0.0665
0.2002
26 838/1922
0.0694
1534
3.580, 29.939
0.0412
0.1177
1.029
Reflections with I2 > 2σ(I2)
Θmin, Θmax/°
R1, I2 > 2σ(I2)
wR2 (all data)
GoF (all data)
1.070
1.048
1.139
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
330–600 nm region at certain periods of time. Typically, 5 mL
of the methanol solution of the complex of desired concen-
tration (calculated as a full concentration of copper cations)
was added to a 10 mL flask containing 5.5 mg of o-amino-
phenol (OAP). After 10 s of vigorous stirring a portion of the
resulting mixture was transferred into
a quartz cuvette
(10 mm; Labbox MISQ-010-002) of the spectrophotometer,
without further stirring. Catalytic reactions under an 18O2
atmosphere were performed in a Schlenk tube. The reagents
were introduced in the same order as for a normal catalytic
reaction. After that the Schlenk tube was closed with the
septum, and the mixture was immediately frozen with liquid
nitrogen; the gas atmosphere was pumped and filled with N2 a
few times to remove air. The frozen mixture was left to warm
under vacuum (to degasify) until becoming a liquid, and the
above procedure was repeated. Finally, the Schlenk tube was
filled with 18O2 through the septum, and the reaction mixture
was stirred for 24 h at room temperature.
Scheme 1 Schematic representation of the structures of complexes
1–3 and their syntheses.
blue solution (for 3) were obtained. Dark-blue crystals of 1 and
2 suitable for the X-ray crystallographic study were formed in
two weeks (for 1) or two days (for 2) from the filtrates without
any additional procedures, while crystals of 3 grew within one
month after slow evaporation of CH3OH.
Gas chromatography
A PerkinElmer Clarus 600 gas chromatograph, equipped with
two nonpolar capillary columns (Phenomenex ZB-5; 30 m ×
0.25 mm × 25 μm), one having an electron impact (EI) MS
detector and the other one with an FID detector, was used for
detailed analyses of the reaction mixtures. The following
general GC conditions were used: 80 °C (3 min), 80–330 °C
(20° min−1), 330 °C (0.5 min), 16 min total run time; 250 °C
injector temperature. For determination of 16O/18O compo-
sition of water the program starting from 50 °C was employed.
Helium was used as the carrier gas (constant 1 mL min−1
flow). All EI mass spectra were recorded with 70 eV energy. The
16O/18O compositions of the product (phenoxazinone chromo-
phore) were determined by the relative abundances of m/z
mass peaks 212/214, respectively.
The general reactions can be written as follows:
CuCl2 þ 2Hca þ 2Hbae þ 2Et3N
Â
Ã
ð1Þ
ð2Þ
ð3Þ
! CuðcaÞ2ðHbaeÞ2 þ 2Et3N ꢀ HCl
CuCl2 þ 2Hva þ 2Hbae þ 2Et3N
Â
Ã
! CuðvaÞ2ðHbaeÞ2 þ 2Et3N ꢀ HCl
4CuðBF4Þ2 þ 4Hva þ 4Hbae þ H2O þ 8Et3N
Â
Ã
! Cu4ðvaÞ4ðbaeÞ4 ꢀ H2O þ 8Et3N ꢀ HBF4
The IR spectra of 1–3 confirm that both carboxylate and
aminoalcohol ligands are present in the compounds (Fig. S1–
S3†).
The broad band in the 3400–3500 cm−1 region is attributed
to the ν(OH) frequencies of the coordinated Hbae ligand (for 1
and 2) and ν(OH) frequencies of the solvated water molecule
(for 3). In addition, the ν(CN) stretching vibration is detected
Results and discussion
Synthesis and spectroscopic analysis
Complexes 1–3 were formed through the self-assembly reaction in the 1070−1080 cm−1 region in the spectra of all com-
of a copper(II) salt (CuCl2 for 1 and 2 or Cu(BF4)2 for 3) with a pounds. In the spectrum of 1, the strong absorption peak at
methanol (1, 3) or acetonitrile (2) solution of 2-benzylami- 1641 cm−1 corresponding to the ν(CvC) stretching frequencies
noethanol (Hbae) and carboxylic acid (cinnamic (Hca) for 1 of aromatic rings is also easily observed. Although the differ-
and valeric (Hva) for 2 and 3) in the presence of triethylamine, ence in wavenumbers Δ (204 cm−1 for 1, 159 cm−1 for 2 and
using the molar synthetic ratio of CuX2 : Hbae : carboxylic acid 193 cm−1 for 3) between the antisymmetric νas(CO) (1579 cm−1
= 1 : 3 : 1, where the carboxylic acid was the last reagent to be for 1, 1561 cm−1 for 2 and 1580 cm−1 for 3) and symmetric
added (Scheme 1). Such a ratio, where an excess of amino- νs(CO) (1375 cm−1 for 1, 1402 cm−1 for 2 and 1387 cm−1 for 3)
alcohol is added prior to the addition of carboxylic acid, pre- stretching frequencies suggests the bridging coordination
vents the formation of stable “paddlewheel” dinuclear copper mode of the carboxylate ligands,44 this is not in agreement
complexes. The opposite procedure (addition of carboxylic with the X-ray analysis results which indicate the presence of a
acid first) with less excess of aminoalcohol is known to lead to monodentate coordination mode of the carboxylate groups.
the abovementioned dinuclear species, as some of us reported Such an inconsistency could be explained by the participation
earlier.43 The reactions were initiated and brought to com- of monodentate carboxylates in the formation of strong intra-
pletion by stirring in open air and heating (50–60 °C) for 1 h (1, 2) and inter-molecular (1–3) hydrogen bonds resulting in a
40 min in all cases. At the end of the reactions, mixtures of decrease of the νas(CO) and an increase of the νs(CO) stretching
dark-green and dark-blue precipitates (for 1 and 2) and a dark- frequencies.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
The difference in the syntheses of 1, 2 and 3 lies in the
solvent and copper salt anion. Since the aminoalcohol ligands
in the crystal structures of 1 and 2 were found to be in the pro-
tonated (original) form, we assume that under the conditions
of experiments the BF4−/CH3OH combination (for 3) promotes
the deprotonation of aminoalcohol, while the Cl−/CH3CN com-
bination favours the protonated form that prevents the coup-
ling of copper fragments, finally resulting in the mononuclear
compounds 1 or 2. The syntheses using equimolar ratios of
copper chloride and tetrafluoroborate salts, for both Hca and
Hva ligands, afforded crystalline materials with unit cell para-
meters corresponding to the structures 1 and 2, respectively.
Hence, the presence of the chloride anion favours the for-
mation of mononuclear complexes.
Crystal structure description
Single crystal X-ray analysis reveals that [Cu(ca)2(Hbae)2] (1)
and [Cu(va)2(Hbae)2] (2) are mononuclear compounds with
similar crystal structures (Fig. 2), wherein the complex mole-
cules are linked together by strong hydrogen bonding forming
extended supramolecular chains (Fig. 3). The important differ-
ences show up as relatively small changes in bond lengths and
angles (Table S1†). In both compounds the 2-benzylami-
noethanol ligands remain in the protonated form and show
bidentate-chelating (N,O) coordination, while deprotonated
carboxylate groups are monodentate and responsible for the
metal ion charge compensation.
Fig. 3 The representations of the supramolecular chain in 1 (a) and 2
(b) with the enlarged fragments showing H-bonding interactions
between copper units. H atoms are omitted for clarity. Colour scheme:
Cu, cyan; O, red; N, blue; C, grey.
The copper(II) ions in 1 and 2 have slightly distorted
elongated octahedral coordination environments with an O4N2
donor set formed by oxygen and nitrogen atoms from cin-
namic (1) or valeric (2) acid and aminoalcohol at the equatorial
positions and oxygen atoms from Hbae at the apical ones. The
Cu–O(N) distances in the plane assume values from 1.987(4) to
2.0276(13) Å, while the longer values of 2.3956(15) and 2.411
(4) Å pertain to the apical Cu–O bonds in 1 and 2, respectively
(Table S1†). In 1 and 2 the cis O–Cu–O(N) bond angles vary
from 79.18(15) to 100.82(15)° and the trans O–Cu–O and N–
Cu–N ones are equal to 180°.
Two sets of strong hydrogen bonds, intermolecular N–H⋯O
and intramolecular O–H⋯O, involving monodentate carboxy-
lates and bidentate Hbae ligands (Table S3†) are responsible
for the formation of the overall 1D supramolecular structures
of 1 and 2 (Fig. 3), wherein the chain building fragments are
the 12-membered {Cu2(OH)2O2(NH)2} supramolecular syn-
thons (Fig. 3, in the enlargement).
The subsequent increase of the dimensionality of the
hydrogen-bonded polymer does not occur due to steric restric-
tions: the bulky substituents in both aminoalcohol and carbox-
ylate ligands prevent the formation of the hydrogen bonds
between adjacent chains (Fig. S4–S6†). The nearest non-
Fig. 2 The ball-and-stick representations of the crystal structures of 1
(a) and 2 (b) with atom numbering. H atoms are omitted for clarity.
Colour scheme: Cu, cyan; O, red; N, blue; C, grey.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
bonded Cu⋯Cu separations within the supramolecular chains
are 7.039(2) Å for 1 and 7.217(2) Å for 2.
Interestingly, in spite of the quite similar crystal structures
of 1 and 2, but due to the presence of carboxylates bearing
linear aliphatic/aromatic (1) or linear aliphatic (2) bulky substi-
tuents, the packings of the supramolecular chains viewed
along the Cu⋯Cu⋯Cu lines are different (Fig. S5 and S6†). In
the case of 1, the chains are located in the staggered order
thus forming small-scale channels (Fig. 4). In contrast, the
hydrogen-bonded chains in 2 are placed over each other creat-
ing columns which further are arranged in a parallel manner
with shortest Cu⋯Cu distances of 14.310(4) Å between them
(Fig. 5).
Fig. 6 The ball-and-stick representation of the molecular structure of 3
with atom numbering. H atoms are omitted for clarity. Colour scheme:
Cu, cyan; O, red; N, blue; C, grey.
Crystallographic analysis shows that compound 3 consists
of two crystallographically independent tetranuclear
[Cu4(va)4(bae)4] molecules having slightly different geometries
and uncoordinated molecules of water (Fig. 6). Each complex
molecule is based on the {Cu4(µ3-O)4} core revealing the overall
cubane-like configuration, wherein the metal atoms are bound
together by µ3 oxygen bridges from 2-benzylaminoethanol
(Fig. 7). The aminoalcohol ligands in 3 are deprotonated and
show bidentate-chelating (N,O) coordination. The auxiliary
deprotonated valeric acids have a monodentate coordination
mode and, together with 2-benzylaminoethanol, account for
the metal ion charge compensation.
Fig. 4 The perspective view of the packing of supramolecular chains in
1 down the a axis showing the channels. H atoms are omitted for clarity.
Colour scheme: Cu, cyan; O, red; N, blue; C, grey.
All copper(II) atoms in 3 are crystallographically indepen-
dent and have five-coordinated distorted square-pyramidal
coordination environments with O4N donor sets which are
formed by the donor atoms of carboxylate and aminoalcohol
ligands. The equatorial Cu–O(N) distances range from 1.919(3)
to 2.032(4) Å, while the apical Cu–O bond lengths are in the
2.402(3)–2.461(3) Å range. The cis angles around the Cu atoms
Fig. 5 The perspective view of the packing of supramolecular chains in
2 down the c axis. H atoms are omitted for clarity. Colour scheme: Cu,
cyan; O, red; N, blue; C, grey.
Fig. 7 The ball-and-stick representation of the central parts of 3 with
atom numbering and the cubane-like {Cu4(µ3-O)4} molecular cores are
highlighted in dark-green.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
vary from 76.74(11) to 111.01(13)° and the trans bond angles
span from 162.74(14) to 178.68(15)° (Table S2†). The non-
bonded Cu⋯Cu distances within the tetranuclear cores of 3
are in the range of 2.8804(11)–3.4616(10) Å.
The strong intramolecular hydrogen bonding of types N–
H⋯O and O–H⋯O, formed by N and O atoms, from the depro-
tonated form of 2-benzylaminoethanol and valeric acids as
well as from the solvated water molecules join the tetranuclear
molecules of 3 into supramolecular polymeric chains (Fig. 8).
But, in spite of all protons involved in the N–H⋯O bond for-
mation between the NH groups of aminoalcohols and oxygens
of carboxylate groups being easily crystallographically identi-
fied (Table S3†), the protons that belong to the water mole-
cules and participate in the O–H⋯O bonding cannot be loca-
lized for sure. Indeed, two protons of each of the uncoordi-
nated water molecules (H2O25 and H2O27) are delocalized and
participate at the same time in four hydrogen bonds with
donor⋯acceptor, OH O⋯Ocarboxylate, distances in the range of
2
3.082(9)–3.203(9) Å. The observed hydrogen bonding affords
the 1D packing of the tetranuclear molecules, wherein neigh-
bouring cubanes are located side to side with solvated mole-
cules of water inserted between them (Fig. 8). Such an arrange-
ment led to a quite short intermolecular Cu⋯Cu distance of
5.8444(18) Å between the tetranuclear molecules (8.58 Å
between the Cu4 unit centres) and a compressed mode of the
supramolecular packing. According to the Cambridge
Structural Database (CSD),45 the shortest Cu⋯Cu separations
between neighbouring cubane-like units less than 6 Å are
found only in 12 cases, while in the majority of the reported
data (in total, 194 hits of tetranuclear cubane-like complexes
with the {Cu4(µ3-O)4} core are found in the CSD) these dis-
tances are longer, with an average separation of ca. 7.5 Å
(Fig. S7†). Similar to 1, the resulting 1D supramolecular chains
of 3 are packed in a staggered order, but due to the disposition
Fig. 9 Representation of the packing of supramolecular chains in 3
down the a axis. H atoms are omitted for clarity. Colour scheme: Cu,
cyan; O, red; N, blue; C, grey.
of the alternate bulky substituents of the ligands no channel
formation is observed (Fig. 9).
Despite the molecular structural type of 3 belonging to a wide
family of {Cu4(µ3-O)4} tetranuclear cubane-like complexes,
wherein the metal centres are linked by four bridging oxygen
atoms, complex 3 represents a rare example of a crystal structure
consisting of two crystallographically independent tetranuclear
molecules. According to the CSD, there are only four compounds
showing similar structural peculiarities, namely [Cu
(H2L3)]4(NO3)4·2H2O (H2L3 = (2,3-dihydroxybenzylidene)benzohy-
drazide),46 [Cu4(L3OMe)4]·10H2O·CH3OH (L3OMe = 3-methoxy-2-oxy-
benzaldehyde 4-hydroxybenzhydrazonato),47 [Cu(R-L)(CH3OH)]8
(R/S-H2L = R/S-3-phenyl-2-(2-hydroxy-5-chlorobenzylideneamino)
propane-1-ol)48 and [Cu{OCH2CH2N(C4H9)2}NCO]4.49
ESI-MS spectrometry and solution behaviour
The electrospray mass-spectra (ESI-MS) of methanol solutions
of 1–3 were investigated. Surprisingly, all spectra at concen-
trations higher than 1 mM were found to show similar pat-
terns, independent of the nuclearity of the complexes in the
solid state, with the top intensity peaks associated with the
tetrameric units [Cu4(L)3(bae)4]+ (L = ca or va for 1 or 2 and 3,
respectively) (Fig. 10). These units lose one proton and one
electron and appear as [Cu4(L)3(bae)4 – H]+ particles at 1294.7
(for 1) and 1156.8 m/z (for 2 and 3) (Fig. 10 and S8†). Such
fragmentation with deprotonation and copper oxidation state
change is possible in ESI-MS spectrometry.50–54 The peaks
associated with further elimination of one or two {Cu(L)(bae)}
units were detected for all compounds (Fig. 10). The group of
peaks appearing at m/z higher than those of tetranuclear units
presumably originate from further aggregation of polynuclear
Fig. 8 The fragment of the supramolecular chain in 3 viewed down the
b axis. H atoms are omitted for clarity. Colour scheme: Cu, cyan; O, red;
N, blue; C, grey.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Fig. 11 The intensities of the signals of the respective particles in
ESI-MS spectra plotted as a function of copper concentrations in metha-
nol solutions of 2 and 3.
tion of a smaller cell (a = 17.11 Å, b = 26.74 Å, c = 27.67 Å, α =
93.85°, β = 98.86°, γ = 90.05°) which forms a twin by 180°
rotation about the real axis [1 0 0]. No relationship with the
initial unit cell parameters of 2 was found and, moreover, a
tentative solution disclosed the tetranuclear cubane-like struc-
ture (only preliminary solution was accessible due to the low
data quality). Furthermore, the volume of the unit cell
(12 479 Å3) is twice larger than the respective volume of the
crystal structure of 3 (6292 Å3, Table 1), thus suggesting that
these crystals correspond to complex 3. The unit cell para-
meters of structure 3 were acceptable, revealing the tetranuc-
lear structure, after application of the same twin law. Hence,
the mononuclear complex 2 can be converted into the tetra-
nuclear complex 3 by recrystallization from methanol, in this
way confirming the equilibria disclosed by ESI-MS.
Fig. 10 The ESI-MS spectra of methanol solutions of 1–3 with a con-
centration of ca. 2 mM.
species in solution and/or in the ionization chamber of the
mass spectrometer.
Although the interpretation of the major peaks observed in
the ESI-MS spectra of 1–3 looks chemically reasonable, the
question appears if these peaks correspond to di-, tri- and tet-
ranuclear species truly existing in solutions or they just result
from the electrospray fragmentation or association of the
samples. We attempted to answer this question by studying
Catalysis
how ESI-MS spectra depend on the concentrations of the Complexes 1–3 were tested as catalysts for the aerobic oxi-
respective complexes in solutions. It was found that all com- dation of o-aminophenol (OAP) to 2-aminophenoxazine-3-one
pounds behave in the same way, starting from the comparable (Fig. 1). Methanol was chosen as the reaction medium due to
intensities of all signals at the lowest concentration of 0.1 mM the solubility properties of the complexes and due to their
and then showing the dominating signals of the tetranuclear ability to maintain polynuclear species in methanol solutions.
species [Cu4(L)3(bae)4]+ at 0.9 mM concentration (Fig. 11 and The reactions were monitored by means of UV-Vis spec-
S9–S11†). Although the peak intensities in ESI-MS spectra may trometry at 435 nm which is known to be a characteristic
depend on various factors apart from concentration,52,55 the absorption band of the phenoxazinone chromophore
data obtained account for the presence of equilibria between (Fig. 12).17 The absorbance was converted into the molar con-
di-, tri- and tetranuclear species in methanol solutions of 1–3 centration of the product using the molar extinction coefficient
and, furthermore, for the possibility of interconversions (ε = 2.4 × 104 M−1 cm−1).18 The initial tests with 2 mM 1–3 and
between the samples 2 and 3 that have the same ligands.
10 mM OAP in methanol afforded the quantitative conversion
Slow evaporation of a methanol solution of 3 afforded a of OAP after 24 h, as evidenced by GCMS (Fig. S12†). Here and
microcrystalline material with the crystal habit identical to further the concentration of a complex means the full concen-
that of 3 (acicular needles). Single crystal X-ray analysis con- tration of copper ions in the respective solution. The typical
firmed the unit cell of the recrystallized sample corresponding UV spectra of the catalytic reaction solution are depicted in
to that of the crystal structure of 3 (Table 1). The crystals Fig. 12. In general, the intensity of the 435 nm absorption
obtained during slow evaporation of a solution of 2 were found increases linearly with time in all studied cases. The initial
to be twinned, tending to show large unit cells with lengths reaction rates W0 determined under these conditions are 2.3 ×
more than 50 Å. Careful analysis of the data allowed the extrac- 10−7, 1.0 × 10−7 and 6.8 × 10−8 M s−1 for 1, 2 and 3, respect-
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
Fig. 12 Increase in the phenoxazine band (435 nm) over time (the
spectra were recoded with 5 min intervals) in the oxidation of OAP
(10 mM) catalysed by 3 ([Cu]0 = 0.64 mM) in CH3OH at room tempera-
ture. Blue and green lines show the absorption spectra of pure 3 and
OAP, respectively, in CH3OH taken at the above concentrations. Inset
shows the evolution of 435 nm absorption with time.
ively. In the absence of any catalysts a rate of 9 × 10−10 M s−1
was observed, which is of two orders lower compared to those
exhibited by 1–3.
The dependence of the reaction rates on the concentrations
of 1–3 is shown in Fig. 13. All complexes exhibit a non-linear
growth of W0 with the concentration increase. A comparative
test was performed using copper acetate as a catalyst (Fig. 13).
In contrast to 1–3, the W0 vs. [Cu(OAc)2]0 dependence was found
Fig. 13 Top: dependence of the initial reaction rates W0 on the initial
to be linear, allowing the calculation of the respective rate con- concentrations of catalysts 1–3 or copper acetate (total copper concen-
stant keff = 3.6 × 10−5 s−1 (calculated as the slope of W0 vs. [Cu
trations are used) in the oxidation of OAP (10 mM) with air dioxygen in
methanol at room temperature. Symbols are experimental data and solid
lines are theoretical fits (for 1–3) or linear interpolation (for copper
(OAc)2]0 dependence). At the lowest concentrations, [Cu]0
<
0.2 mM, the reaction rates exhibited by all three complexes as
acetate). Inset shows the low concentration range for 3 in detail.
well as copper acetate are very close (ca. 1 × 10−8 M s−1). These
1/2
Bottom: the same data plotted in [Cu]0/W0
vs. W1/2 coordinates,
observations suggest that at low concentrations compounds 1–3 showing the presence of two quasilinear phases in all curves. Symbols
are experimental data and dotted lines are linear fits. The fits belonging
exist in the form of mono- or dinuclear species with catalytic
activities comparable to that of Cu(OAc)2. The latter is known to
exist as the dimer Cu2(OAc)4 in the solid state as well as in
methanol solution.56 The catalytic tests for some other copper
compounds were performed for comparative purposes
1/2
to the higher W0 region were used to generate the curves on the top
plot. The top inset plot uses the respective linear fit deducted from the
lower W01/2 region.
(Table S4†) at 1 mM total copper concentration. It was found chloride in methanol) rather than mononuclear solvated
that [CuI(CH3CN)2](PF6), [CuI8IO(tbdea)6(H2O)2](BF4)2·3CH3OH copper cations.57,58 This may lead to incorrect conclusions
and
[CuI8IO(tbdea)5(Htbdea)Cl2][CuICl2]·2H2O
complexes when performing the comparative catalytic investigations
(H2tbdea = N-tert-butyldiethanolamine)34,35 show the reaction using the copper salts.57,58 The ESI-MS spectra of 1 mM
rates ranging from 7.7 × 10−9 to 8.6 × 10−8 M s−1 (Table S4†). methanol solutions of CuCl2 and Cu(NO3)2 are shown in
The very low rates exhibited by octanuclear complexes34,35 could Fig. S14.† For copper chloride the formation of [CunCl2n+1
]
−
be explained by the low concentration of polynuclear species oligomeric species (where n ranges from 1 to 6) was detected,
(ca. 0.13 mM), in case the cores retain their integrities in cata- in accord with reported results.57 A similar behaviour was
lytic solutions. Divalent copper chloride and nitrate as catalysts exhibited by copper nitrate which shows prominent peaks at
revealed the reaction rates of up to 4 × 10−7 M s−1 (Table S4†).
454.76 and 485.66 m/z assigned to trinuclear assemblies [Cu3O
It is known that copper salts in the absence of chelating (NO3)4]− and [Cu3O(NO3)4(CH3O)]−, respectively (Fig. S14†).
ligands exist in solution in the form of stable polynuclear As the catalytic activity of copper salts could be associated
aggregates (such as [Cu4IIOCl6(CH3OH)4] for the case of copper with the presence of a range of polynuclear species in solution,
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
1/2
a comparative test was performed by adding 1 equivalent of a
Dividing this equation by W0 one can get the linearized
rigid chelating ligand, 2,2′-bipyridine (bipy), to prevent their version:
uncontrolled formation. In accord with expectations, the reac-
tion rate exhibited by the Cu(NO3)2/bipy catalytic system was
found to be roughly one order lower than that observed for
pure Cu(NO3)2 (Table S4†). In contrast to copper nitrate, the
catalytic activity of complex 3 was enhanced in the presence of
½Cuꢁ0
¼ aW01=2 þ b
ð6Þ
W01=2
4
keff
2
where a ¼
and b ¼
.
1=2
1 equivalent of bipy to W0 = 1.3 × 10−7 M s−1 (Table S4†).
ðkeff KÞ
The acute growth of W0 at [Cu]0 > 1 mM for 1–3 could be
The linearized plots show two distinct linear phases
(Fig. 13, bottom). These results demonstrate that the whole
range of concentrations cannot be adequately described by the
single model (2), but at the highest catalyst loading the model
gives a reasonable fit. Presumably, the point of interception of
the two lines corresponds to the point at which the concen-
tration of [Cu4] becomes comparable to [Cu2] and thus the
initial assumptions are no more satisfactory for lower [Cu]0
explained by the aggregation of catalytically less active mono-
and dinuclear species into highly active tri- and tetranuclear
ones.59,60 This hypothesis is in accord with ESI-MS obser-
vations, where the peaks corresponding to tetranuclear par-
ticles start to dominate only at [Cu]0 = 0.9 mM concentration.
However, from this point of view complexes 2 and 3 could be
catalytically similar, while the mononuclear complex 2 demon-
strates a bit higher activity (Fig. 13). This difference suggests
that the initial solid state form of the compound (mono- and
tetranuclear for 2 and 3, respectively) somehow influences the
equilibria in solution.
The dependence of the initial reaction rates W0 on the con-
centrations of complexes is evidence of a reaction order higher
than one relative to the catalyst concentration. Let us write the
equilibria between di-, tri- and tetranuclear species in the fol-
lowing way, where [Cu1] is a mononuclear copper species:
W
−1/2. For all complexes this point was found to be of nearly
the same value of [Cu]0 = 0.66 mM. The keff and K values
obtained from the fit of the [Cu]0 > 0.66 mM region are 8 ×
10−3 and 74 (for 1), 3 × 10−3 and 103 (for 2) and 8 × 10−3 s−1
and 36 M−1 (for 3), respectively.
The dependence of W0 on OAP concentration was found to
be linear with no saturation up to [OAP]0 = 0.06 M (Fig. 14).
The first-order behaviour suggests the active participation of a
copper catalyst at only one step of the catalytic process.
The overall mechanism of the OAP oxidation is depicted in
Scheme 2.11,18,61 Air dioxygen is assumed to be a terminal
oxidant because the amount of Cu(II) cations in catalytic solu-
tions is not enough for direct oxidation of the substrate with
quantitative conversion. The oxidative activity of Cu(II) towards
OAP was studied in the following way. Stirring of OAP (10 mM)
in the presence of 3 (1.1 mM of total copper amount) and in
the absence of O2 resulted in a yellow solution after 24 h.
Measurement of its absorption at 435 nm (Fig. S13†) gave the
concentration of phenoxazinone chromophore of 1.9 × 10−4
M. This value agreed perfectly with that expected (1.8 × 10−4
M) for full conversion of OAP by six-electron oxidation with the
Cu(II) → Cu(I) process. Considering the reaction rate of the
!
!
½Cu2ꢁ þ 2½Cu1ꢁ
½Cu3ꢁ þ ½Cu1ꢁ
½Cu4ꢁ
ð1Þ
Assuming [Cu4] as tetranuclear species having the highest
catalytic activity and [Cu2] as dinuclear ones having the lowest
activity, and equating 2[Cu1] to [Cu2] (mononuclear species
make negligible contribution, according to ESI-MS) one can
rewrite the above equilibria in a simplified manner:
!
2½Cu2ꢁ
½Cu4ꢁ
ð2Þ
The equilibrium constant associated with the above
equation is K = [Cu4]/[Cu2]2, and the material balance of the
whole system is as follows, where [Cu]0 is the full concen-
tration of copper ions:
½Cuꢁ0 ¼ 4½Cu4ꢁ þ 2½Cu2ꢁ
ð3Þ
Let us assume that the catalytic activity of [Cu2] is negligible
compared to that of [Cu4] and the non-first-order behaviour of
W0 is stipulated only by the equilibrium (2). Hence, the initial
reaction rate can be written as W0 = keff[Cu4] where keff includes
all quasistationary parameters, such as [OAP]. From this, the
concentration of the dimeric species can be written as a func-
tion of W0:
ꢀ
ꢁ
1=2
W0
keff
½Cu2ꢁ ¼
ð4Þ
K
The material balance equation can be rewritten as follows:
Fig. 14 Dependence of the initial reaction rate W0 in the oxidation of
OAP, catalysed by 3 (0.5 mM), with air dioxygen in methanol at room
temperature, on the initial concentration of OAP. Squares are experi-
mental points and the solid line is the linear fit.
ꢀ
ꢁ
1=2
W0
keff
W0
keff
½Cuꢁ0 ¼ 4
þ 2
ð5Þ
K
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
The ESI-MS spectra of the catalytic mixture (1 mM 3 and
10 mM OAP) revealed a one-order increase of the intensity of
the peak at 152 m/z (H3bae+), compared to the spectrum of
pure 3 (Fig. 11 and S11†), while the intensities of other peaks
were suppressed (Fig. S15†). A partial substitution of the
aminoalcohol ligand by OAP in the coordination sphere of
copper was confirmed by the presence of numerous peaks,
attributed to various Cu/OAP adducts, from [Cu(bea)(OAP)]+ at
320.97 m/z to [Cu3(bae)2(va)(OAP)2
–
H]+ at 807.42 m/z
(Fig. S15†). The low overall intensity of these peaks could be
explained by the presence of uncharged species, which are not
detectable by the ESI-MS method. Reduction of divalent
copper in ESI-MS spectrometry is a common behavior of its
coordination compounds.64,65
The proposed mechanism (Scheme 2) does not foresee an
oxygen atom transfer from O2 to form the final product. To
confirm such assumption, we performed a test using 18O2 as
an oxidant and monitored the 18O/16O isotopic composition of
the product using its characteristic molecular ion peak of 212
m/z. A separate test demonstrated that neither OAP nor phe-
noxazine exchange oxygen with H218O directly. The reaction
catalysed by 3 under 18O2 gave quantitative conversion of OAP
after 24 h. No 18O-labelled phenoxazinone was observed, while
the formation of H218O was accounted even in the beginning
of the reaction (Fig. S17†), as expected from the proposed reac-
tion mechanism (Scheme 2). Reaction conducted in the pres-
ence of 1.5 M H218O showed very low conversion of less than
10% after 24 h, presumably due to the low solubility of 3 in a
water-containing solvent.
The presence of free radical species was checked by using
the stable radical TEMPO (2,2,6,6-tetramethylpiperidin-1-yl)
oxyl, which is known to affect catalytic systems by generating
or scavenging free radicals. A search through the literature
shows conflicting data on the influence of TEMPO on OAP oxi-
dation. Some authors reported that TEMPO can be used as a
reaction initiator and even as a catalyst,66 while others demon-
strated its catalysis suppression character.67 We first studied
the accumulation of products in the presence of 7 mol%
TEMPO (Fig. 15). This loading was reported to be enough to
produce a reaction rate of ca. 1 × 10−7 M s−1 under similar con-
ditions.66 However, our observations do not support these
data. The product accumulation exhibited by TEMPO shows
the initial stage with W0 = 1.4 × 10−8 M s−1, and then the reac-
tion slows down to Wlin = 2.0 × 10−9 M s−1, a rate comparable
to that for a blank test reaction (Fig. 15). Complex 3, taken at
the same concentration (calculated as the total copper amount
in solution), shows a constant reaction rate of 1.8 × 10−8 M s−1
for 1 h. An equimolar mixture of 3 and TEMPO catalyses OAP
oxidation with a rate of 2.4 × 10−8 M s−1, which is very close to
the simple sum of reaction rates exhibited by 3 and TEMPO
independently (Fig. 15). Thus, under the used experimental
conditions, the TEMPO radical did not interfere with the
copper-catalysed process. No by-products that could be associ-
ated with compounds of TEMPO with OAP were detected by
GCMS. These results support the general mechanism as
shown in Scheme 2, with no participation of C-centred free
Scheme 2 Top: equilibria between the o-aminopheno anion and
o-iminobenzoquinone. Bottom: general mechanism of OAP oxidation.
catalyst-free reaction (9 × 10−10 M), one can expect the respect-
ive maximum product concentration after 24 h as 7.8 × 10−5
M. Thus, the rate of OAP oxidation with Cu(II) is higher than
that with O2 in the absence of a catalyst.
It is commonly accepted that the rate-determining step in
the OAP oxidation is the initial activation of the OAP molecule
with the formation of the aminophenoxyl radical or its next
derivative, iminobenzoquinone. The ESI-MS spectra of catalytic
mixtures (Fig. S15†) suggest that OAP may coordinate to a
copper centre, presumably replacing the aminoalcohol ligand
in a reversible manner. The presence of a weak absorption at
620 nm in the UV-Vis spectra of the catalytic mixtures
(Fig. S16†) is in accord with this assumption, since this band
is associated with complexes of copper with OAP.62 After
coordination, OAP is oxidized directly by Cu(II) or by inter-
action with molecular oxygen via the copper–dioxygen–OAP
intermediate. Hence, one may suppose that in the catalytic
cycle copper serves as an OAP coordination species. The Cu(I)
species formed is restored to Cu(II) by dioxygen.
The principal feature of catalysts 1–3 is the acute increase
of W0 with increase of catalyst concentration (Fig. 13). The
reported copper catalysts typically exhibit the first order
(linear) W0 vs. [catalyst]0 dependence,18,24,25,63 and our tests
using copper acetate as the catalyst follow this tendency.
Complexes of copper are known to catalyse OAP oxidation by
coordinating OAP with the subsequent formation of radical
species,21–23 as observed from the OAP equilibrium62 depicted
in Scheme 2. As it was proposed that at higher concentrations
complexes 1–3 exist mainly in a tetranuclear form, we can ten-
tatively assume that the higher activity of the tetranuclear
species is associated with their coordination to more than one
OAP molecule simultaneously. The presence of two (or more)
activated OAP molecules at a close distance could facilitate
their joining into the final phenoxazinone chromophore.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
methylcyclohexane, cis-1,2-dimethylcyclohexane and triphenyl-
methane were tested), catalysed by 3 (1.5 mM), revealed no
suppression of OAP conversion after 24 h, which was quantitat-
ive. No products that could be associated with the oxidation of
the secondary substrates were detected, except for cis-1,2-di-
methylcyclohexane (DMCH), for which weak peaks of cis- and
trans-tertiary alcohols in ca. 1 : 1 ratio were visible (Fig. S20†).
The loss of stereoconfiguration of DMCH (formation of both
stereoisomers from
a pure cis-substrate) may tentatively
account for the presence of small amounts of H2O2 as a reac-
tion by-product, because interaction of H2O2 with a copper
catalyst typically affords the hydroxyl radical which abstracts
the H atom from a C–H bond with the formation of a long-
lived alkyl radical.6 The latter leads to the epimerization of the
substrate stereoconfiguration.69–71 The presence of the H2O2
by-product is expected for the aerobic oxidation of OAP.18
An interesting observation was the ability of 3 to catalyse
the aerobic oxidation of DMCH in the absence of OAP.
Although the conversion level was low, the products were
easily detectable (Fig. S20†) and, moreover, the products
pattern was quite unusual. Mild catalytic oxidation of DMCH
with various peroxides typically affords tertiary and secondary
alcohols as main products along with 2,7-octanedione as a
main ring cleavage product.37,69,71–74 Such distribution is seen
when 10 mM DMCH is oxidized in the presence of 10 mM
OAP, catalysed by 3 (Fig. S20c†). However, in the absence of
OAP, the formation of 2,7-octanedione was suppressed, while
6- and 7-octen-2-ones were clearly seen. The latter compounds
have never been observed earlier (or observed in negligible
quantities) under similar conditions.37,69,71–74
Fig. 15 Accumulation of the product (phenoxazinone chromophore) in
the oxidation of OAP (10 mM), catalysed by TEMPO (6.4 × 10−4 M), 3 (6.4
× 10−4 M, total copper concentration), their mixture, or in the absence
of any catalyst. The filled symbols show the sum of the independent
accumulation curves for 3 and TEMPO.
radicals, while formation of short-lived O-centred radicals
cannot be excluded as their reaction with TEMPO would be
less significant.68
Oxidation of an equimolar mixture of OAP and 2-amino-p-
cresol (APC), 5 mM each, was performed to determine how the
presence of a methyl substituent at the C–H activation position
can influence the reactivity. The respective reaction, catalysed
by 1, afforded both possible products (Scheme 3 and Fig. S18†)
in a 1 : 1 ratio, after 24 h. This result agrees with the proposed
reaction mechanism. In contrast to pure OAP, oxidation of an
OAP/APC mixture did not afford complete conversion even
after 24 h. An attempt to crystallise the reaction product by
slow evaporation of the solvent afforded dark brown crystals,
found to be pure APC after X-ray analysis (structure 4,
Fig. S19†).
As this reaction may account for the catalytic activity
towards the aerobic oxidation of alcohols (and keeping in
mind the possibility of H2O2 formation as a by-product11), we
attempted to check what happens in case the reaction is con-
ducted in the presence of an easy oxidizable substrate. Aerobic
oxidation of a mixture of OAP (10 mM) and an equimolar
amount of another substrate (cyclohexanol, 1-phenylethanol,
The reaction rates of OAP oxidation to phenoxazinone
exhibited by complexes 1–3 (a W0 of up to 4 × 10−7 M s−1 and a
TOF of up to 2.4 × 10−4 −1) are close to the highest ones
s
reported for this reaction, whose typical rates are in the range
from 1 × 10−8 to 1 × 10−6 M s−1, under related conditions.11
The highest reported rates are exhibited by the manganese
complexes bearing polydentate N-donor ligands. Among the
cubane-like Cu4 compounds the highest reported catalytic rate
of 9.4 × 10−7 M s−1 is shown by the complex [Cu4(L)4] (where
H2L
=
(E)-4-bromo-2-((thiazol-2-ylimino)methyl)phenol).19
Finally, we found only one study reporting the W0 vs. [catalyst]0
dependence for a polynuclear copper catalyst.18 In that case
the catalyst was a copper cubane-like complex [Cu4(L•)4] with a
radical-type ligand (H3L = N-(2-hydroxy-3,5-di-tert-butylphenyl)-
2-aminobenzylalcohol) and the reaction rate exhibited satur-
ation at [catalyst]0 > 0.2 mM, while the higher catalyst concen-
trations were not investigated. This is in contrast with our
results (Fig. 13), where no saturation is observed for 1–3 in a
much wider concentration range.
Conclusions
In the present work, mono- and tetranuclear Cu(II) complexes
containing both carboxylate and aminoalcohol ligands were
synthesized in a quick and efficient manner using self-assem-
Scheme 3 Comparison of the reaction products formed by the com-
petitive oxidation of 2-amino-p-cresol (APC) and o-aminophenol (OAP),
5 mM each, catalysed by 1 (1 mM), in methanol, after 24 h.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
bly reactions. Single crystal X-ray analysis shows that 1 and 2 Program "5-100". The authors acknowledge the IST Node of
have similar molecular mononuclear structures, while the the Portuguese Network of Mass-spectrometry for ESI-MS
structure of 3 represents a cubane-like configuration bearing measurements.
the {Cu4(µ3-O)4} molecular core. The strong hydrogen bonding
in 1–3 is responsible for the formation of chain supramolecu-
lar structures in all cases. The stability of the complexes in
methanol solutions was studied by ESI-MS spectrometry.
References
Surprisingly, the ESI-MS patterns of all complexes were found
to be similar, accounting for the possibility of interconversion
between complexes 2 and 3. Such a possibility was confirmed
by the recrystallization of 2 from methanol: although the
quality of new crystals was not sufficient for complete structure
refinement, the unit cell parameters and tentative solution
revealed that they correspond to the tetranuclear complex 3.
The ESI-MS spectra obtained for methanol solutions of 1–3
clearly demonstrated that the aggregation of tetranuclear
species is favoured by the higher concentration of the
complexes.
All compounds show phenoxazinone synthase like activity
in methanol solvent with the maximum reaction rate for 1.
These rates are among the highest ones reported for this reac-
tion. Evaluation of W0 vs. [catalyst]0 disclosed an unexpected
effect of non-first-order reaction acceleration for all complexes.
Considering ESI-MS observations, a kinetic model was devel-
oped where the catalytic activity of tetranuclear species is
1 P. Gandeepan, T. Muller, D. Zell, G. Cera, S. Warratz and
L. Ackermann, Chem. Rev., 2019, 119, 2192–2452.
2 R. Trammell, K. Rajabimoghadam and I. Garcia-Bosch,
Chem. Rev., 2019, 119, 2954–3031.
3 S. D. McCann and S. S. Stahl, Acc. Chem. Res., 2015, 48,
1756–1766.
4 C. E. Elwell, N. L. Gagnon, B. D. Neisen, D. Dhar,
A. D. Spaeth, G. M. Yee and W. B. Tolman, Chem. Rev.,
2017, 117, 2059–2107.
5 S. E. Allen, R. R. Walvoord, R. Padilla-Salinas and
M. C. Kozlowski, Chem. Rev., 2013, 113, 6234–6458.
6 D. S. Nesterov, O. V. Nesterova and A. J. L. Pombeiro, Coord.
Chem. Rev., 2018, 355, 199–222.
7 V. C. C. Wang, S. Maji, P. R. Y. Chen, H. K. Lee, S. S. F. Yu
and S. I. Chan, Chem. Rev., 2017, 117, 8574–8621.
8 I. Bento, M. A. Carrondo and P. F. Lindley, J. Biol. Inorg.
Chem., 2006, 11, 539–547.
9 I. A. Koval, P. Gamez, C. Belle, K. Selmeczi and J. Reedijk,
Chem. Soc. Rev., 2006, 35, 814–840.
much higher than those of tri- and dinuclear ones. With this
1/2
model, linearization of the kinetic data in the [Cu]0/W0 vs. 10 M. Le Roes-Hill, C. Goodwin and S. Burton, Trends
1/2
W0 coordinates allowed the fitting of the non-linear W0 vs.
[Cu]0 dependence. Oxidation of o-aminophenol under an 18O2 11 S. K. Dey and A. Mukherjee, Coord. Chem. Rev., 2016, 310,
atmosphere resulted in the formation of H218O, while the
80–115.
Biotechnol., 2009, 27, 248–258.
main product, the phenoxazinone chromophore, was found to 12 U. Hollstein, Chem. Rev., 1974, 74, 625–652.
contain 16O only. Furthermore, complex 3 can oxidize o-amino- 13 L. A. Farmer, E. A. Haidasz, M. Griesser and D. A. Pratt,
phenol under an inert atmosphere, revealing the yield of the
product corresponding to the six-electron oxidation of the sub- 14 C. D. Hein, X. M. Liu and D. Wang, Pharm. Res., 2008, 25,
strate with CuII. These data suggest that (a) tetranuclear
2216–2230.
species are essential for a highly efficient PHS-like activity and 15 A. W. Smith, A. Camara-Artigas, M. T. Wang, J. P. Allen and
(b) o-aminophenol is likely oxidized by CuII, while the role of
W. A. Francisco, Biochem., 2006, 45, 4378–4387.
J. Org. Chem., 2017, 82, 10523–10536.
O2 is to oxidize the intermediates and the CuI formed to regen- 16 Image of the PHS enzyme (PDB code 3GYR) was created
erate the active CuII species. It is expected that these results
would help in understanding the biomimetic oxidation mecha-
nisms of PHS and inspire further developments in the field.
using Mol* software; D. Sehnal, A. S. Rose, J. Kovca,
S. K. Burley and S. Velankar, Mol*: Towards a common
library and tools for web molecular graphics, MolVA/
EuroVis Proceedings, 2018.
17 J. C. Freeman, P. G. Nayar, T. P. Begley and J. J. Villafranca,
Biochemistry, 1993, 32, 4826–4830.
Conflicts of interest
18 C. Mukherjee, T. Weyhermuller, E. Bothe, E. Rentschier
and P. Chaudhuri, Inorg. Chem., 2007, 46, 9895–9905.
19 S. Sagar, S. Sengupta, A. J. Mota, S. K. Chattopadhyay,
A. E. Ferao, E. Riviere, W. Lewis and S. Naskar, Dalton
Trans., 2017, 46, 1249–1259.
There are no conflicts to declare.
Acknowledgements
20 M. Mitra, T. Kundu, G. Kaur, G. Sharma, A. R. Choudhury,
Y. Singh and R. Ghosh, RSC Adv., 2016, 6, 58831–58838.
This work was supported by the Foundation for Science and
Technology (FCT), Portugal (projects UIDB/00100/2020 of 21 P. Mahapatra, S. Ghosh, S. Giri, V. Rane, R. Kadam,
Centro de Química Estrutural, PTDC/QEQ-QIN/3967/2014,
PTDC/QUI-QIN/29778/2017, contracts IST-ID/086/2018, IST-ID/
M. G. B. Drew and A. Ghosh, Inorg. Chem., 2017, 56, 5105–
5121.
117/2018 and fellowship BL/CQE-2018-021). The publication 22 A. Das, S. Goswami, R. Sen and A. Ghosh, Inorg. Chem.,
has been prepared with the support of the RUDN University
2019, 58, 5787–5798.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.

View Article Online
Paper
Dalton Transactions
23 P. Mahapatra, M. G. B. Drew and A. Ghosh, Cryst. Growth 46 M. Sutradhar, M. V. Kirillova, M. F. C. Guedes da Silva,
Des., 2017, 17, 6809–6820.
24 T. M. Simandi, L. I. Simandi, M. Gyor, A. Rockenbauer and
A. Gomory, Dalton Trans., 2004, 1056–1060.
25 L. I. Simandi, T. Barna and S. Nemeth, J. Chem. Soc., Dalton
Trans., 1996, 473–478.
26 D. S. Nesterov, O. V. Nesterova, V. N. Kokozay and
A. J. L. Pombeiro, Eur. J. Inorg. Chem., 2014, 4496–4517.
27 D. S. Nesterov, V. G. Makhankova, V. N. Kokozay and
B. W. Skelton, Inorg. Chim. Acta, 2005, 358, 4519–4526.
28 D. S. Nesterov, C. Graiff, A. Tiripicchio and
A. J. L. Pombeiro, CrystEngComm, 2011, 13, 5348–5353.
C. M. Liu and A. J. L. Pombeiro, Dalton Trans., 2013, 42,
16578–16587.
47 V. Vrdoljak, G. Pavlovic, N. Maltar-Strmeckic and
M. Cindric, New J. Chem., 2016, 40, 9263–9274.
48 M. J. Niu, Z. Li, X. Li and X. Q. Huang, RSC Adv., 2016, 6,
98171–98179.
49 J. Sassmannshausen, H. G. von Schnering and L. Walz, Z.
Kristallogr., 1997, 212, 656–661.
50 D. P. Demarque, A. E. M. Crotti, R. Vessecchi,
J. L. C. Lopes and N. P. Lopes, Nat. Prod. Rep., 2016, 33,
432–455.
29 V. V. Semenaka, O. V. Nesterova, V. N. Kokozay, 51 K. L. Vikse, Z. Ahmadi and J. S. McIndoe, Coord. Chem.
R. I. Zybatyuk, O. V. Shishkin, R. Boca, C. J. Gomez-Garcia, Rev., 2014, 279, 96–114.
J. M. Clemente-Juan and J. Jezierska, Polyhedron, 2010, 29, 52 J. S. McIndoe and K. L. Vikse, J. Mass Spectrom., 2019, 54,
1326–1336. 466–479.
30 D. S. Nesterov, V. N. Kokozay, V. V. Dyakonenko, 53 A. Steckel and G. Schlosser, Molecules, 2019, 24, 11.
O. V. Shishkin, J. Jezierska, A. Ozarowski, A. M. Kirillov, 54 H. Lavanant, H. Virelizier and Y. Hoppilliard, J. Am. Soc.
M. N. Kopylovich and A. J. L. Pombeiro, Chem. Commun.,
2006, 4605–4607.
31 P. Seppala, R. Sillanpaa and A. Lehtonen, Coord. Chem.
Rev., 2017, 347, 98–114.
32 J. W. Sharples and D. Collison, Coord. Chem. Rev., 2014,
260, 1–20.
33 O. V. Nesterova, D. S. Nesterov, J. Jezierska,
A. J. L. Pombeiro and A. Ozarowski, Inorg. Chem., 2018, 57,
12384–12397.
34 D. S. Nesterov, J. Jezierska, O. V. Nesterova,
A. J. L. Pombeiro and A. Ozarowski, Chem. Commun., 2014,
50, 3431–3434.
35 D. S. Nesterov, E. C. B. A. Alegria and J. Jezierska, Inorg.
Chim. Acta, 2017, 460, 83–88.
36 O. V. Nesterova, M. N. Kopylovich and D. S. Nesterov,
Inorganics, 2019, 7, 19.
37 O. V. Nesterova, D. S. Nesterov, B. Vranovicova, R. Boca and
A. J. L. Pombeiro, Dalton Trans., 2018, 47, 10941–10952.
Mass Spectrom., 1998, 9, 1217–1221.
55 A. Kiontke, A. Oliveira-Birkmeier, A. Opitz and
C. Birkemeyer, PLoS One, 2016, 11, 16.
56 H. Grasdalen, J. Chem. Soc., Faraday Trans. 2, 1973, 462–
470.
57 S. Becker, M. Durr, A. Miska, J. Becker, C. Gawlig,
U. Behrens, I. Ivanovic-Burmazovic and S. Schindler, Inorg.
Chem., 2016, 55, 3759–3766.
58 S. Low, J. Becker, C. Wurtele, A. Miska, C. Kleeberg,
U. Behrens, O. Walter and S. Schindler, Chem. – Eur. J.,
2013, 19, 5342–5351.
59 D. S. Nesterov, E. N. Chygorin, V. N. Kokozay, V. V. Bon,
R. Boca, Y. N. Kozlov, L. S. Shul’pina, J. Jezierska,
A. Ozarowski, A. J. L. Pombeiro and G. B. Shul’pin, Inorg.
Chem., 2012, 51, 9110–9122.
60 G. B. Shul’pin, M. V. Kirillova, L. S. Shul’pina,
A. J. L. Pombeiro, E. E. Karslyan and Y. N. Kozlov, Catal.
Commun., 2013, 31, 32–36.
38 G. M. Sheldrick, CELL_NOW, Georg-August-Universität, 61 J. Kaizer, G. Barath, R. Csonka, G. Speier, L. Korecz,
Göttingen, Germany, 2008.
39 Bruker, APEX2 & SAINT, AXS Inc., Madison, WI, 2004.
A. Rockenbauer and L. Parkanyi, J. Inorg. Biochem., 2008,
102, 773–780.
40 G. M. Sheldrick, Acta Crystallogr., Sect. C: Struct. Chem., 62 M. Wakizaka, T. Matsumoto, R. Tanaka and H. C. Chang,
2015, 71, 3–8. Nat. Commun., 2016, 7, 11.
41 A. Dey, M. T. Kirchner, V. R. Vangala, G. R. Desiraju, 63 T. Horath, J. Kaizer and G. Speier, J. Mol. Catal. A: Chem.,
R. Mondal and J. A. K. Howard, J. Am. Chem. Soc., 2005,
127, 10545–10559.
2004, 215, 9–15.
64 C. Y. Hao and R. E. March, J. Mass Spectrom., 2001, 36,
509–521.
65 L. Gianelli, V. Amendola, L. Fabbrizzi, P. Pallavicini and
G. G. Mellerio, Rapid Commun. Mass Spectrom., 2001, 15,
2347–2353.
66 J. Kaizer, R. Csonka and G. Speier, J. Mol. Catal. A: Chem.,
2002, 180, 91–96.
42 S. Kashino, M. Tomita and M. Haisa, Acta Crystallogr., Sect.
C: Cryst. Struct. Commun., 1988, 44, 730–732.
43 O. V. Nesterova, M. V. Kirillova, M. F. C. Guedes da Silva,
R. Boca and A. J. L. Pombeiro, CrystEngComm, 2014, 16,
775–783.
44 K. Nakamoto, Infrared and Raman Spectra of Inorganic and
Coordination Compounds, Part B, Applications in 67 A. K. Dhara, U. P. Singh and K. Ghosh, Inorg. Chem. Front.,
Coordination, Organometallic and Bioinorganic Chemistry,
Wiley, Chichester, 5th edn, 1997.
45 C. R. Groom, I. J. Bruno, M. P. Lightfoot and S. C. Ward,
2016, 3, 1543–1558.
68 R. Amorati, G. F. Pedulli, D. A. Pratt and L. Valgimigli,
Chem. Commun., 2010, 46, 5139–5141.
Acta Crystallogr., Sect. B: Struct. Sci., Cryst. Eng. Mater., 69 O. V. Nesterova, M. N. Kopylovich and D. S. Nesterov, RSC
2016, 72, 171–179. Adv., 2016, 6, 93756–93767.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Dalton Transactions
Paper
70 E. A. Buvaylo, V. N. Kokozay, O. Y. Vassilyeva, B. W. Skelton, 72 O. V. Nesterova, K. V. Kasyanova, E. A. Buvaylo,
O. V. Nesterova and A. J. L. Pombeiro, Inorg. Chem.
Commun., 2017, 78, 85–90.
O. Y. Vassilyeva, B. W. Skelton, D. S. Nesterov and
A. J. L. Pombeiro, Catalysts, 2019, 9, 15.
71 G. S. Astakhov, A. N. Bilyachenko, A. A. Korlyukov, 73 O. V. Nesterova, K. V. Kasyanova, V. G. Makhankova,
M. M. Levitsky, L. S. Shul’pina, X. Bantreil, F. Lamaty,
V. N. Kokozay, O. Y. Vassilyeva, B. W. Skelton, D. S. Nesterov
A. V. Vologzhanina, E. S. Shubina, P. V. Dorovatovskii,
and A. J. L. Pombeiro, Appl. Catal., A, 2018, 560, 171–184.
D. S. Nesterov, A. J. L. Pombeiro and G. B. Shul’pin, Inorg. 74 D. S. Nesterov, O. V. Nesterova, M. N. Kopylovich and
Chem., 2018, 57, 11524–11529.
A. J. L. Pombeiro, Mol. Catal., 2018, 459, 8–15.
This journal is © The Royal Society of Chemistry 2020
Dalton Trans.