Fe-complex of a tetraamido macrocyclic ligand: Spectroscopic characterization and catalytic oxidation studies

Article abstract of DOI:10.1016/j.cplett.2010.09.002

This work presents the spectroscopic characterization and reaction studies of a Fe^III-complex (2) of a tetraamido macrocyclic ligand (1, 15,15-dimethyl-5,8,13,17-tetrahydro-5,8,13,17-tetraaza-dibenzo[a,g] cyclotridecene-6,7,14,16-tetraone). 2 was characterized primarily by means of EPR. In agreement with the magnetic moment (μ_eff = 3.87 BM), EPR spectroscopy of 2 shows signals consistent with S = 3/2 intermediate-spin ferric-iron. Besides EPR, mass spectrometry, UV/vis spectroscopy and cyclic voltammetry were used to further characterize 2. 2 is soluble in water and activates hydrogen peroxide under ambient conditions. 2 catalytically bleaches dyes, pulp and paper effluents and oxidizes several amines to their corresponding N-oxides with high turnover number and good yields.

Full text of DOI:10.1016/j.cplett.2010.09.002

Chemical Physics Letters 498 (2010) 359–365
Contents lists available at ScienceDirect
Chemical Physics Letters
journal homepage: www.elsevier.com/locate/cplett
Fe-complex of a tetraamido macrocyclic ligand: Spectroscopic characterization
and catalytic oxidation studies
Shane Z. Sullivan a, Anindya Ghosh a, ,1, Alexandru S. Biris b, Sharon Pulla a, Anna M. Brezden a,
⇑
Samulel L. Collom a, Ross M. Woods b, Pradip Munshi c, Laura Schnackenberg d, Brad S. Pierce e,1
,
Ganesh K. Kannarpady b
a Department of Chemistry, University of Arkansas at Little Rock, Little Rock, AR 72204, USA
b University of Arkansas at Little Rock, Nanotechnology Center, Little Rock, AR 72204, USA
c Reliance Industries Limited, Research Center, Vadodara, India
d National Center for Toxicological Research, Jefferson, AR 72079, USA
e Department of Chemistry and Biochemistry, The University of Texas at Arlington, Arlington, TX 76019-0065, USA
a r t i c l e i n f o
a b s t r a c t
This work presents the spectroscopic characterization and reaction studies of a FeIII-complex (2) of a tet-
raamido macrocyclic ligand (1, 15,15-dimethyl-5,8,13,17-tetrahydro-5,8,13,17-tetraaza-dibenzo[a,g]cyclot-
ridecene-6,7,14,16-tetraone). 2 was characterized primarily by means of EPR. In agreement with the
Article history:
Received 29 June 2010
In final form 1 September 2010
Available online 6 September 2010
magnetic moment (leff = 3.87 BM), EPR spectroscopy of 2 shows signals consistent with S = 3/2 interme-
diate-spin ferric-iron. Besides EPR, mass spectrometry, UV/vis spectroscopy and cyclic voltammetry were
used to further characterize 2. 2 is soluble in water and activates hydrogen peroxide under ambient con-
ditions. 2 catalytically bleaches dyes, pulp and paper effluents and oxidizes several amines to their cor-
responding N-oxides with high turnover number and good yields.
Ó 2010 Elsevier B.V. All rights reserved.
1. Introduction
oxidations [3,17] in water. These complexes utilize a tetraamido
macrocyclic ligand to coordinate a mononuclear ferric-iron.
Annually, several tons of hydrogen peroxide (H2O2) [1,2] are
used for industrial oxidation purposes. By itself, H2O2 is a reason-
ably moderate oxidant but its oxidizing activity can be significantly
enhanced by variety of metal complexes. The major challenge in
the design of H2O2-activating catalysts is selecting a suitable metal
ligand which can stabilize high-valent metal oxidation states while
simultaneously remaining resistant to self-oxidation [3]. With this
caveat in mind, a major research effort has evolved which focuses
on biologically inspired metal complexes mimicking the structures
and function of H2O2 or oxygen (O2) activating metalloenzymes [4–
8]. Generally speaking, this is accomplished by carefully control-
ling the geometry of the metal coordination site and the electron
donating ability of the ligand(s) to modulate the H2O2- or O2-acti-
vating capacity of the catalyst. In recent years various metal com-
plexes have been investigated, each employing a myriad of donor
ligands to effectively catalyze either O2- or H2O2-dependent oxida-
tion reactions [1,9–16]. Among these catalysts, a series of H2O2-
activating iron-complexes developed by Collins and co-workers is
particularly noteworthy as ‘green catalysts’ for performing various
Ellis et al. [18] recently reported a new generation of tetraamido
macrocyclic ligands and their FeIII-complexes including ligand 1
and complex 2 (Figure 1). The iron-complexes were found to be
highly active in activating H2O2 rapidly in aqueous solutions in
the pH range of 7–9 and subsequently tested for dye bleaching (Or-
ange II). However, reactions of 2 were limited only to bleaching of
one water soluble dye as reported by the authors. This has made us
curious to study the Fe-complex additionally for various oxida-
tions. Beside this, spectroscopic characteristics such as ESI-MS,
EPR, UV/vis spectroscopy, cyclic voltammetry, and magnetic sus-
ceptibility of 2 provided us additional scope for investigating 2 in
terms of further characterizing the Fe-complex. Therefore, the
work presented here describes the bleaching of pulp and paper
effluents, various water soluble organic dyes and synthesis of
N-oxides from their corresponding amines along with detailed
spectroscopic investigation of the FeIII-complex containing the
new variant of the tetraamido macrocyclic ligands (1, Figure 1)
which were not studied previously.
As compared with other tetraamido macrocyclic ligands, the
synthesis of 1 is much simpler and economic and does not require
the use of an amino acid starting material [18]. This FeIII-oxalamide
complex (2, Figure 1) was found to be amenable to physicochemi-
cal characterization by ESI-MS, UV–visible, EPR spectroscopy, cyc-
lic voltammetry, and magnetic susceptibility. Furthermore,
⇑
Corresponding author. Fax: +1 501 569 8838.
E-mail address: axghosh@ualr.edu (A. Ghosh).
1
Equal work was contributed by these authors.
0009-2614/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.cplett.2010.09.002

360
S.Z. Sullivan et al. / Chemical Physics Letters 498 (2010) 359–365
Figure 1. Tetradentate amidomacrocyclic ligand (1) and its Fe-Complex (2). The counter cation is not shown.
catalytic studies of 2 using H2O2 as a primary oxidant are provided.
In these experiments 2 rapidly catalyzes the H2O2-dependent oxi-
dation of various water soluble organic dyes and paper pulp efflu-
ent under mild reaction conditions and ambient temperature. 2
also behave as a catalyst for the synthesis of small organic N-oxides
from their corresponding organic amines. These results demon-
strate the overall chemical versatility of tailored tetraamido mac-
rocyclic ligands for a variety of industrial applications.
spectral line-width is dominated by D and E-strain. Therefore sim-
ulations employ GAUSSIAN distributions in the D-values and the ratio
of E/D to produce the correct line-width, specified as rD and rE/D
,
respectively. Least squares and decovolution analysis of the spectra
were combined to allow relevant parameters to vary while main-
taining a sum of multiple species that best-fit the experimental
data. The simulations were generated with consideration of all
intensity factors, both theoretical and experimental, to allow con-
centration determination of species. The only unknown factor
relating the spin concentration to signal intensity was an instru-
mental factor that depended on the microwave detection system.
However, this was determined by the spin standard, Cu(EDTA),
prepared from a copper atomic absorption standard solution pur-
chased from Sigma–Aldrich.
2. Experimental
2.1. Materials and methods
All chemicals, reagents, and solvents were obtained either Sig-
ma–Aldrich (USA) or Fisher Scientific (USA) and used without fur-
ther purifications unless otherwise noted. Tetrahydrofuran (THF)
was purified using sodium metal and benzoephenone. 1H NMR
spectra were obtained using a 600 MHz Bruker Avance instrument
equipped with a 5 mm triple resonance inverse probe. Infrared (IR)
spectra were obtained using a Nicolet Magna IR 500 spectrometer.
Electrospray ionization mass spectra (ESI-MS) were obtained using
an Applied Biosystem API-4000 spectrometer or using an Agilent
100 series MSD VL spectrometer. Gas Chromatography Mass spec-
tra (GC/MS) were obtained using an Agilent G1701DA GC/MSD.
UV–visible spectroscopy was performed using a Varian Cary 5000
UV–visible-NIR spectrophotometer. Electrochemical studies were
performed using a Pine Instrument Company Biopotentiostat
(model no. AFCBP1). Elemental analysis was performed at Midwest
Microlab LLC., Indianapolis. Magnetic susceptibility was measured
using Alfa Aesar magnetic susceptibility balance-mark 1.
The half-power microwave saturation (P1/2) was determined by
least-squares fitting the intensity of the derivative signal (I) as a
function of microwave power (P) using Eq. (1). The variable b is
an inhomogeneity factor which varies between 1.0 for inhomoge-
neous and 3.0 for homogeneous line broadening.
3. Results and discussion
3.1. Preparation of catalyst 2
Synthesis of 2 was performed according to the method reported
by Ellis et al. with minor modifications [18]. The synthetic details
for both 1 and 2 are described in the supplementary information.
3.2. Electron spray ionization mass spectrum (ESI-MS)
Section 2 for bleaching studies, N-oxide synthesis, cyclic vol-
tammetry, and demetallation study are given in the supporting
information.
If the iron within 2 has a + 3 oxidation state, the resulting com-
plex would be monoanionic and thus observable in the negative
ion mode when analyzed by ESI-MS. Indeed, as observed in Figure
2 the ESI-MS spectrum of 2 has a pronounced peak at 418 m/z.
Moreover, the theoretical isotopic distribution shown within the
inset of Figure 2 is consistent with the experimentally observed
isotope distribution. Taken together, these results support the for-
mation of the ferric-iron complex shown in Figure 2. Elemental
analysis was performed to confirm the expected elemental compo-
sition of the desired product.
While the axial water ligand indicated in the Figure 1 could not
be observed by ESI-MS it is possible that this labile ligand is disso-
ciated within the harsh conditions of the mass spectrometer. In
fact, an axial water ligand was observed in the X-ray crystallo-
graphic structure of analogous Fe-tetraamido macrocyclic ligand
2.2. Electron paramagnetic resonance (EPR) spectroscopy
X-band (9 GHz) EPR spectra were recorded on a Bruker EMX
Plus spectrometer equipped with a bimodal resonator (Bruker
model 4116DM). Low-temperature measurements were made
using an Oxford ESR900 cryostat and an Oxford ITC 503 tempera-
ture controller. A modulation frequency of 100 kHz was used for
all EPR spectra. All experimental data used for spin-quantitation
were collected under non-saturating conditions. EPR spectra were
simulated and quantified using Spin Count (ver.3.0.0), created by
Professor M.P. Hendrich at Carnegie Mellon University. The

S.Z. Sullivan et al. / Chemical Physics Letters 498 (2010) 359–365
361
in acetonitrile (ACN) at high concentrations (1–2 mM). While this
phenomenon was not observed in UV–visible, electrochemical, or
kinetic studies, the working concentration for typical EPR experi-
ments is several orders of magnitude higher. As a result, EPR spec-
tra of 2 in ACN show only weak (S = 3/2) signals typically observed
for FeIII-tetraamido macrocyclic ligand complexes. Furthermore, as
observed in Figure 3, at elevated temperature (above 50 K) a broad
signal can be seen at g ꢁ 2.3. As indicated by the inset (Figure 3),
the signal at g ꢁ 2.3 does not follow Curie-law behavior as in that
its intensity increases quadratically with temperature. This qua-
dratic response to increased temperature is not consistent with a
change in the Boltzman population of a well defined spin-manifold,
and therefore cannot be attributed to a simple dimeric or multi-
meric molecular aggregate of complex 2. Moreover, as indicated
by the dashed lines (Figure 3), this resonance is initially observed
at g ꢁ 2.6 at temperatures below 50 K. However, at temperatures
above 50 K, the resonant position for this signal shifts up-field to
g ꢁ 2.3. Taken together, these observations are consistent with a
signal originating from superparamagnetic relaxation of ferric
nanoparticles, likely from aggregation of complex 2 in ACN [21–
24]. The line shape and temperature dependence for this class of
paramagnetic signals are highly dependent on the size, geometry,
and composition of the nanoparticles, therefore no meaningful
interpretation can be made regarding these spectra. For this reason
several glassing solvents and glassing mixtures were evaluated to
prevent aggregation of complex 2 in frozen samples [25]. The
aggregation of the iron complex (2) at high concentration in ACN
is due to the unique nature of the ligand (1). The ligand (1) is much
less sterically hindered compared to previously reported tetraam-
ido macrocyclic ligands [3,17] and as a result metal complexes (2)
can come close to each other to form aggregates presumably via
either a water or hydroxo bridge. Unlike ligand 1, conventional tet-
raamido macrocyclic ligands [3,17] contain several methyl groups
which render the ligands and their corresponding iron-complexes
Figure 2. ESI-MS of Fe-Complex (2) and its theoretical isotope distribution (inset).
complexes [18,19]. Furthermore, as will be discussed in greater de-
tail below, the EPR spectra of 2 is similar to analogous FeIII-tetra-
amido macrocyclic ligand complexes in the presence of an axial
aqua- or hydroxo-ligand. Therefore the presence of an axial water
ligand on 2 cannot be completely ruled out at this point.
3.3. UV–visible spectroscopy and electrochemical studies
Figure S3 (supporting information) shows the UV–visible spec-
tra of 1 and 2 dissolved in acetonitrile. Ligand 1 is a white solid and
thus is not expected to exhibit any peaks within the visible range.
Within the UV region a single peak is observed at 202 nm with a
shoulder at 223 nm. Alternatively, complex 2 is a dark red solid
and thus exhibits several absorption features within the visible
spectrum (268, 346 nm, and a relatively weak but broad shoulder
at 500 nm). The molar extinction coefficient at 450 nm (4050
Mꢀ1 cmꢀ1) was used for quantitative purposes. As discussed below,
both EPR spectroscopy and magnetometry confirms the oxidation
state of 2 is ferric. The bands within the visible region are largely
attributable to ligand to metal charge transfer bands. However,
the lower extinction coefficient (
e
ꢁ 500 Mꢀ1 cmꢀ1) observed for
the features above 400 nm could be from d–d transitions within
an intermediate-spin d5-ion assuming tetragonal distortion.
The cyclic voltammetry of 2 was performed in acetonitrile con-
taining 0.1 M n-tetrabutylammonium hexafluorophosphate as
electrolyte with respect to Ag/AgCl electrode at 298 K. As illus-
trated in Figure S3 (supporting information), the voltamogram
for 2 exhibits one peak at E1/2 = 0.50 V (DEp = 100 mV) within the
potential range of 0.2–1 V. Previous FeIII-tetraamido macrocyclic li-
gand complexes show peaks within the E1/2-range of 0.44–1.0 V.
The broad range or redox potentials available for FeIII-tetraamido
macrocyclic ligand complexes is largely a result of substitution of
electron donating or withdrawing functional groups on the tetra-
amido macrocyclic ligand backbone [20]. For example, Fe-tetraam-
ido macrocyclic ligand complex without any modification on the
benzene ring exhibit a redox potential of E1/2 = 0.76 V where as
with Fe-tetraamido macrocyclic ligand with dimethyl substitution
on the benzene ring shows E1/2 = 0.44 V [20]. By analogy, the peak
observed at E1/2 = 0.50 V for complex 2 can be similarly assigned a
one electron oxidation for a FeIII/FeIV couple. However, while unli-
kely, the possibility that a FeIII radical cation or a dimeric species is
being produced cannot be completely ruled out.
Figure 3. Temperature dependent X-band EPR spectra of 1.32 mM complex 2 in
pure acetonitrile. Selected EPR spectra were collected at 4, 7, 10, 15, 20, 30, 50, and
80 K and are normalized to correct for Curie-law dependence. (Inset) Normalized
signal intensity of the g ꢁ 2.3 features with increasing temperature. A simple
quadratic function (dashed line), y = y0 + ax + bx2 is overlaid on the data for clarity.
3.4. Electron paramagnetic resonance spectroscopy
The initial EPR characterization of complex 2 was complicated
by the tendency of this molecule to form intermolecular aggregates

362
S.Z. Sullivan et al. / Chemical Physics Letters 498 (2010) 359–365
more sterically hindered and results in less or almost no
aggregation.
result, the intensity of the transition within the ms = 3/2 doublet
observed at g = 5.89 shows significant intensity. As discussed
previously, it was not possible to observe the axial ligand as sug-
gested for the complex 2. However based on the EPR signals with
g-anisotropy and E/D-strain of the complex 2 and observing the
similarity of the EPR signal of previously reported similar com-
plexes [19], the existence of the axial ligand(s) in the complex 2
can be surmised.
Upon correction of the observed signals for Curie-law depen-
dence, the observed signal intensity of these transitions is inversely
correlated as a function of temperature. Thus the intensity of the
g = 5.89 signal is greatest at lower temperature whereas the signal
observed at g = 4.67 increases with temperature. The assignment of
these features are further corroborated by fitting the temperature-
normalized signal intensity to a Boltzman population distribution
for a 2-level system; Figure 5A(s), (solid line). From the best-fit
to each transition, the value for the axial-zero field splitting term
(D) for 2a was determined to be D = ꢀ1.9 0.2 cmꢀ1. Thus the
ms = 3/2 doublet represents the ground state of the 2a S = 3/2
spin-system.
Species 2b represents a minority fraction of EPR spectra I repre-
senting only 14% of the total S = 3/2 species. This component shows
considerably less g-anisotropy and a more axial magnetic symme-
try, E/D-value of 0.07. Given the low concentration of this species it
could easily be overlooked. However, the sharp spectral features
observed in spectra I at g = 4.35 and 1.99 cannot be simulated
within SI while simultaneously maintaining the features associ-
ated with 2a at g = 5.89 and 4.67. As observed in Figure 3, the signal
intensity observed at g = 1.99 is largely dominated by species 2b,
and very little intensity originates from 2a due to its broad features
within this region of the spectrum. Conversely, the signal intensity
observed at g = 4.67 (and 5.89) is nearly completely attributed to
species 2a. As illustrated by Figure 5A, the temperature depen-
dence of the feature at g = 1.99 associated with 2b is significantly
different than that observed at g = 4.67 for 2a. Since there are only
two doublet within a S = 3/2 spin-system, these two spectroscopic
features cannot be attributed to the same S = 3/2 species. As illus-
trated in Figure 5A(d), (dashed line), the same procedure outlined
for 2a can be used to determine the D-value for 2b;
D = ꢀ3.6 0.3 cmꢀ1. While the magnitude of the 2b D-value is sig-
nificantly larger than that of 2a, both species have a ground state
ms = 3/2 doublet.
By altering the solvent composition from pure ACN to a 1:1
mixture of toluene/ethanol nearly all indications of molecular
aggregation (as observed by EPR spectroscopy) were eliminated. As
illustrated in Figure 4 the 10 K EPR spectra (solid line, I) of complex
2 (in 1:1 toluene/EtOH) shows signals consistent with a ferric-iron
in an intermediate axial-ligand field, S = 3/2 [26,27]. Based on the
effect of the solvent on the EPR spectra it is possible that EtOH is
acting as an axial ligand to the Fe-tetraamido macrocyclic ligand
complex. This could potentially decrease molecular aggregation
via a bridging water or hydroxide ligand.
The temperature dependence, microwave power saturation
behavior, and spectral features associated with spectra I can only
be understood if the observed EPR spectra is fit to two spectroscop-
ically distinct S = 3/2 species, termed 2a and 2b henceforth. A
quantitative component sum simulation (dashed line, S1) for the
spectra associated with species 2a and 2b was utilized to generate
the simulation SI for complex 2. Species 2a is the dominate compo-
nent of EPR spectra I representing 86% of the total S = 3/2 species
observed by EPR spectroscopy. The signals observed at g-values
of 4.67, 4.18, and 1.88 originate from a transition within the
ms = 1/2 doublet of the S = 3/2 spin-system with an E/D-value of
0.15. Similar signals have been previously observed in Fe-tetraam-
ido macrocyclic ligand complexes with one axial water ligand and
one axial hydroxide ligand [19].
The EPR line-width of 2b is dominated by g-anisotropy and E/D-
strain which is consistent with other ferric-tetraamido macrocyclic
ligand complexes [19]. The increased rhombicity of this species al-
lows for significant mixing of the ms = 1/2 and 3/2-eigenstates. As a
The microwave power saturation behavior of 2a and 2b was
determined at 10 K as a final confirmation for the presence of
two species. As with the temperature dependence, the signal inten-
sity as a function of microwave power shown in Figure 5B was
measured at g ꢁ 4.67 for 1a and g = 1.99 for 2b. The half-saturation
microwave power (P1/2) for species 2a and 2b was determined by
least-squares fit to Eq. (1) [28,29].
ꢀ
ꢁ
ꢀ
ꢁ
pffiffiffi
log I=
P
¼ log I0 ꢀ b=2 ꢂ log 1 þ p=p1=2
:
ð1Þ
Within this temperature regime the line-width of 2a and 2b are
not significantly broadened as a function of temperature and thus
both species exhibit inhomogeneous saturation behavior (b = 1)
and are best-fit to a P1/2-value of 85 10 and 40 5 mW, respec-
tively. Given the additional spectral feature observed at g = 5.89
for 2b and the differential temperature dependence and micro-
wave saturation behavior observed at g = 4.67 and g = 1.99, the ob-
served EPR spectra I can be quantitatively simulated using two
spectroscopically distinct S = 3/2 species, 2a and 2b.
The S = 3/2 spin-state observed for 2a and 2b is consistent with
the magnetic moment (leff = 3.87 Bohr Magneton) determined by
the Gouy method. This value for the spin-only magnetic moment
is consistent with three unpaired electrons, S = 3/2. Furthermore,
as stated in the Section 2, the concentration of each species was
Figure 4. X-band EPR spectra (solid line) of 1.0 mM complex 2 at 10 K in 1:1
(toluene:EtOH). A component simulation for the spectra associated with species 2a
and 2b was utilized to generate the simulation (dashed line, SI). Using this
technique, the relative concentration for species 2a and 2b was determined by
least-squares analysis to be 86% mM and 14%, respectively. Instrumental
parameters: microwave frequency, 9.64 GHz; microwave power, 0.63 mW;
modulation amplitude, 0.92 mT; temperature, 10 K. Simulation parameters: (2a)
S = 3/2; gxyz = 2.28 1.95, 2.01;
r
gxyz = 0.03; D = ꢀ1.90 cmꢀ1 D = 0.04 cmꢀ1; E/D =
; r
0.15;
E/D = 0.07;
r
E/D = 0.05;
r
B = 1.5 mT; (2b) S = 3/2; gxyz = 2.11, 2.00, 2.02; D = ꢀ3.60 cmꢀ1
;
r
gxyz = 0.02, 0.03, 0.004; r ; rE/D = 0.02; rB = 1.5 mT.
D = 0.6 cmꢀ1

S.Z. Sullivan et al. / Chemical Physics Letters 498 (2010) 359–365
363
Figure 5. A. Normalized temperature dependence of S = 3/2 species 2a (s) and 2b (d). The values for the axial-zero field splitting term (D) for each species (2a,
D = ꢀ1.9 0.2 cmꢀ1; 2b, D = ꢀ3.6 0.3 cmꢀ1) were determined by fitting the data to a Boltzman-dependent population distribution for a 2-level system. B. Microwave power
saturation behavior of the two S = 3/2 species 2a (s) and 2b (d) at 10 K. The half-saturation microwave power (P1/2) for species 2a and 2b were determined by least-squares
regression according to published methods.22, 23Within this temperature regime the line-width of 2a and 2b are not significantly broadened as a function of temperature and
thus both species exhibit inhomogeneous saturation behavior (b = 1) and are best-fit to a P1/2-value of 85 10 and 40 5 mW, respectively.
determined by component simulation. Using this approach the
concentration of species 2a and 2b within sample I (Figure 3)
was determined to be 0.61 and 0.10 mM, respectively. This value
represents ꢁ71% of the total iron in the sample as determined by
UV–visible spectroscopy.
in the presence of catalytic amounts of 2 (0.5 lM). The reaction
was initiated by addition of H2O2 to a final concentration of
3 mM. A small amount (10 ppm) of ethylenediamine tetraacetate
(EDTA) was added to each reaction mixture to minimize Fenton-
based hydroxyl radical chemistry associated with free transition
metals. As indicated by Figure 6A the bleaching for all dyes showed
nearly complete bleaching within 9-min at ambient temperature.
The rate of dye bleaching was measured at both pH 10 and 11.5
to demonstrate the tolerance of 2 under basic conditions. As illus-
trated in Table 1 the bleaching time for each dye is essentially the
same at pH 10 and 11.5 indicating the efficacy of 2 within this pH
range. The deviations observed in the rate of dye bleaching cata-
lyzed by 2 can likely be attributed to differences in the nature
and accessibility of oxidizable groups in each dye. For example,
H2O2-dependent bleaching of Methyl Orange by 2 was consider-
ably slower than observed for other organic dyes. However, dye
bleaching by H2O2 alone under similar conditions shows no
bleaching within the time observed (15-min). As with previously
reported Fe–TAML catalysts complex 2 is slowly inactivated under
multiple turnover conditions resulting in decreased bleaching effi-
ciency [30].
3.5. Stability of the complex
Experimental sections are given in the supporting information.
The complex is stable in neutral to alkaline aqueous solutions for
several days at moderately high temperature (40–50 °C). However,
heating the aqueous solutions of 2 to 90 °C causes the catalyst to
demetallate rapidly as indicated by changes in the UV–visible spec-
tra (Figure S5, supporting information). Demetallation gives rise to
the free ligand, which was verified by 1H NMR. This is a limitation
of using 2 at very high temperature. Macrocyclic ring size and
amide planarity are critical for hydrolytic stability of iron-com-
plexes of deprotonated amide ligands. The Fe-complex of a tetra-
amido macrocyclic ligand with a ring size of fourteen atoms was
found to be extremely unstable in water [19]. With this in mind,
2 was synthesized with a ring size of thirteen atoms with the
intention that this size should provide adequate stability to the
Fe-complex in aqueous solution. However, the instability of
the complex at high temperature is not currently understood.
One potential explanation for the thermal instability of 2 may be
attributed to the increased rigidity of this complex. Unlike other
Fe–TAML complexes, 2 has two aromatic rings which makes the
complex very rigid and thus may decrease the strength of the li-
gand chelation to the iron ion.
The ability complex 2 to catalyze H2O2-dependent bleaching of
paper pulp and paper effluent was directly measured spectropho-
tometrically using a similar procedure as for dye-bleaching studies.
In these experiments 6 mg of catalyst 2 was added per liter of efflu-
ent and the absorbance was monitored at 466 nm for 4 h upon
addition of 28.2 mM H2O2 at pH 9.5 [31,3]. As shown in Figure
6B, under these conditions complex 2 (13 lM) decreased the
absorbance by 52% within 4 h at ambient temperature. While
H2O2 alone is also capable of bleaching paper pulp effluent under
similar conditions, the rate and extent of bleaching is significantly
lower than observed for 2 catalyzed bleaching. The rates observed
here for dye bleaching are consistent with the kinetic findings re-
cently reported by Ellis et al. [1–10 ꢂ 105 Mꢀ1sꢀ1] [18].
3.6. Oxidation studies
Experimental sections are given in the supporting information.
Textile industries produce a large quantity of water effluent which
requires treatment to remove colored compounds prior to release.
Activated H2O2 provides an effective and inexpensive means to
treat water effluent. For this purpose the catalytic efficiency of
complex 2 as an activator of H2O2 was evaluated by UV–visible
spectroscopy. In these experiments orange IV (a), clayton yellow
(b), methyl orange (c), methyl violet (d), and napthol B green (e)
were used as a substrate-surrogate to illustrate the efficacy of 2
2, H2O2
r.t., pH 10
ð2Þ
N
O
N
In addition to the bleaching of organic dyes and paper effluents,
complex 2 was evaluated as a synthetic catalyst to oxidize a variety
of tertiary amines to their corresponding N-oxides (Eq. (2)). This
simple oxidation chemistry is remarkably useful for both synthetic
and biological applications [32,33]. In all instances these reactions
were performed at room temperature and pH 10. After 2 h each
reaction was quenched and products analyzed by GC/MS as
ꢀ
catalyzed H2O2-bleaching within an aqueous CO32ꢀ/HCO3 buffer
(pH 10) [30]. As a control, one dye (f, Orange IV) was monitored
in the absence of 2. Figure 6A shows the relative change in absor-
bance for each dye (12 lM) at the wavelength specified in Table 1

364
S.Z. Sullivan et al. / Chemical Physics Letters 498 (2010) 359–365
Figure 6. A. The rate of bleaching for the water soluble dyes listed in Table 1. Each dye was monitored by UV–visible spectroscopy at its corresponding absorption maximum.
Reaction conditions: temperature, 25 °C; 10 ppm EDTA; pH 10; 12 M dye; 31.3 mM H2O2; and 0.5 M catalyst 2. B. Room temperature bleaching of paper pulp effluent
M 2 and 28.2 mM H2O2 for 4 h at 25 °C.
l
l
following treatment with 13
l
Table 1
were found to be the less reactive substrates to furnish N-oxides in
lower yields (39–40%) and lower turnover numbers (404–415) due
to the difficulty in oxidizing these amines. Both steric and elec-
tronic factors must be considered in order to justify the difference
in reactivity of various amines. Alternatively, reactions carried out
in the absence of 2 and using only H2O2 as a primary oxidant re-
sulted in nearly no (or very little) N-oxide product formation.
These results demonstrate the versatility of 2 and its ability to acti-
vate H2O2 under relatively mild reaction conditions. Free metal
ions are also known to catalyze oxidation reactions [34], but that
is not the case here as reactivity will drop dramatically if free metal
ion in such low concentration is the sole reactive species.
Dye bleaching reactions using 2 in H2O.
Bleaching timeb
a
Entry
Dye
kmax
(nm)
pH 10 (s)
pH 11.5 (s)
a/f
b
c
d
e
Orange IV
444
403
464
584
718
310
307
>600
235
325
305
295
>600
225
335
Clayton yellow
Methyl orange
Methyl violet
Napthol B Green
a
kmax was the wavelength used to determine bleaching time.
Bleaching time is defined to be the time required for the absorbance to reach
b
half its initial value.30 All reactions performed in pH 10 or 11.5 carbonate buffer
with 10 ppm EDTA, dye concentration of 12 M, H2O2 concentration 31.3 mM, and
catalyst (2) concentration of 0.5 M at 25 °C. As a control the bleaching of Orange IV
l
l
by H2O2 in the absence of catalyst 2 (f) is shown in Figure 6.
4. Conclusion
described within the Section 2. As shown in Table 2 the reactions
catalyzed by 2 show yields approaching stoichiometric conversion
in some instances. The catalytic efficiency of 2 was determined by
the number of moles or product produced per mol of catalyst. For
clarity, this ratio is termed the turnover number (TON) henceforth.
For the amines selected (Table 2), the TON observed for 2 range
from 404 to 608. Among several amines triethylamine showed
the highest reactivity to form the N-oxide followed by N,N-diiso-
propylamine and N,N-diethylaniline. Understandably, these
amines are relatively easy to oxidize thus resulting in the high
yields and TON observed. Pyridine and N,N-diisopropylethylamine
In this work a water soluble Fe-tetraamido macrocyclic ligand
complex was synthesized and characterized by UV–visible, EPR
spectroscopy, magnetometry, and ESI-MS. The synthesis of this
Fe-complex is relatively straight forward and can be produced in
large quantities. EPR spectroscopy reveals two spectroscopically
distinct forms of this complex, termed 2a and 2b. The heterogene-
ity of complex 2 observed by EPR spectroscopy can likely be attrib-
uted to differential protonation states of one or both axial water
ligands. In agreement with magnetic susceptibility measurements
complex 2 exhibits an S = 3/2 spin-state consistent with ferric-iron
within a strong-axial-ligand field. As a cautionary note, complex 2
showed a propensity for molecular aggregation at concentrations
amenable to spectroscopic characterization. Therefore, some care
should be employed in the selection of an appropriate solvent sys-
tem prior to future spectroscopic investigation of reactive
intermediates.
Table 2
N-oxide formation of amines catalyzed by 2a.
Entry
Amine
TON
Yieldb
The experiments presented here demonstrate that 2 rapidly
activate H2O2 under ambient reaction conditions to catalyze the
oxidation of water soluble organic dyes and paper pulp effluent.
Moreover, this catalyst can be utilized in the synthesis of small
N-oxides from their corresponding organic tertiary amines. Indeed,
in many instances the synthesis of N-oxides proceeded with a high
turnover number and high yield. These observations attest to the
superior catalytic properties of this complex and its chemical ver-
satility. While several attempts to crystallographically characterize
2 were made, results obtained from EPR spectroscopy suggest this
complex has a tendency to form molecular aggregates. This
1
2
3
4
5
6
7
N,N-diisopropylethylamine
N,N-diisopropylamine
N,N-dimethylaniline
N,N-diethylaniline
Triethylamine
415 50
514 51
481 36
468 54
608 83
404 20
–
40.2 (0)
49.7 (0)
46.5 (15.8)
45.3 (18.9)
58.8 (0)
39.1 (22.9)
–
Pyridine
4-dimethylaminopyridinec
a
All the reactions were performed at room temperature at pH 10 in carbonate
buffer. H2O2 was added as a primary oxidant for all reactions. Product yield was
determined after 2 h.
b
Values in parenthesis indicate % yields of N-oxides by hydrogen peroxide only.
Product formation was not quantified.
c

S.Z. Sullivan et al. / Chemical Physics Letters 498 (2010) 359–365
365
[2] C.W. Jones, Applications of Hydrogen Peroxide and Derivatives, The Royal
Society of Chemistry, Cambridge, 1999.
[3] T.J. Collins, Acc. Chem. Res. 27 (1994) 279.
property appears to be highly solvent dependent and may inhibit
homogenous crystallization.
One of the major advantages of using 2 is simple and economic
synthesis of the ligand (1) of the complex compared to the conven-
tional tetraamido macrocyclic ligands. Although 2 is soluble in
water and rapidly activates H2O2 for bleaching, the activity of the
complex is low compared to conventional Fe-complexes of the tet-
raamido macrocyclic ligands. However, 2 was used to synthesize
small molecules such as N-oxides, which has never been attempted
using other Fe-complexes of tetraamido macrocyclic ligands. This
certainly opens new direction for further investigations. Hydrolytic
stability of the complex 2 compared to the similar Fe-complexes at
high temperature is low but the complex is safe to use at room
temperature or slightly higher ranges. Future work will also focus
on the development of additional catalysts with increased thermo-
dynamic stability and improved catalytic activity for benign
oxidations.
[4] B. Meunier, J. Bernadou, Struct. Bonding 97 (2000) 1.
[5] M. Costas, M.P. Mehn, M.P. Jensen, L.J. Que, Chem. Rev. 104 (2004) 939.
[6] M. Sono, M.P. Roach, E.D. Coulter, J.H. Dawson, Chem. Rev. 96 (1996) 2841.
[7] E.Y. Tshuva, S.J. Lippard, Chem. Rev. 104 (2004) 987.
[8] T. Funabiki, in: L.I. Simandi (Ed.), Catalysis by Metal Complexes (Advances in
Catalytic Activation of Dioxygen by Metal Complexes), Kluwer Academic
Publishers., Dordrecth/Boston/London, 2003, pp. 157–226. vol. 26.
[9] K. Wieghardt et al., J. Am. Chem. Soc. 110 (1988) 7398.
[10] M.C. White, A.G. Doyle, E.N. Jacobsen, J. Am. Chem. Soc. 123 (2001) 7194.
[11] G. Yin et al., Inorg. Chem. 45 (2006) 3467.
[12] G. Roelfes, M. Lubben, R.A. Hage Jr., L. Que, B.L. Feringa, Chem. Eur. J. 6 (2000)
2152.
[13] G. Yin et al., J. Am. Chem. Soc. 127 (2005) 171170.
[14] F.A. Chavez, P.K. Mascharak, Acc. Chem. Res. 33 (2000) 539.
[15] A. Ghosh et al., J. Am. Chem. Soc. 127 (2005) 2505.
[16] C.E. MacBeth, A.P. Golombek, V.G. Young Jr., C. Yang, K. Kuczera, M.P. Hendrich,
A.S. Borovik, Science 289 (2000) 938.
[17] T.J. Collins, Acc. Chem. Res. 35 (2002) 782.
[18] W.C. Ellis, C.T. Tran, M.A. Denardo, A. Fischer, A.D. Ryabov, T.J. Collins, J. Am.
Chem. Soc. 131 (2009) 18052.
[19] A. Ghosh et al., J. Am. Chem. Soc. 302 (2003) 359.
[20] A. Ghosh et al., J. Am. Chem. Soc. 130 (2008) 15116.
[21] R. Miao et al., Biochemistry 47 (2008) 9888.
Acknowledgements
[22] E. Wajnberg, L.J. El-Jaick, M.P. Linhares, D.M. Esquivel, J. Magn. Reson. 153
(2001) 69.
[23] T.G. St Pierre, P. Chan, K.R. Bauchspiess, J. Webb, S. Betteridge, S. Walton, D.P.E.
Dickson, Coord. Chem. Rev. 151 (1996) 125.
[24] E.A. Zhillinskaya, G. Delahay, M. Mauvezin, B. Coq, A. Aboukais, Langmuir 19
(2003) 3596.
[25] R.S. Drago, Physical Methods in Chemistry, Saunders, 1992.
[26] H. Rupp, K.K. Rao, D.O. Hall, R. Cammack, Biochim. Biophys. Acta. 537 (1978)
255.
[27] M.W. Makinen, M.B. Yim, Proc. Natl. Acad. Sci. USA 78 (1981) 6221.
[28] M. Sahlin, A. Gräslund, A. Ehrenberg, J. Magn. Reson. 67 (1986) 135.
[29] A. Davydov, A. Liu, A. Graslund, J. Inorg. Biochem. 80 (2000) 213.
[30] C.P. Horwitz, D.R. Fooksman, L.D. Vuocolo, S.W. Gordon-Wylie, N.J. Cox, T.J.
Collins, J. Am. Chem. Soc. 120 (1998) 4867.
We thank the department of Chemistry (and Biochemistry) at
UALR and UTA, respectively for financial assistance (AG and BSP).
AG also likes to thank Department of Energy (Grant number DE-
FG36-06GO86072) for financial assistance to complete the work
too.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.cplett.2010.09.002.
[31] C.P. Horwitz et al., ACS Symp. Ser. 921 (2006) 156.
[32] A. Albini, Synthesis (1993) 263.
[33] A. Albini, S. Pietra, Heterocyclic N-oxides, CRC Press, Boca Raton, 1991.
[34] E. Ember, S. Rothbart, R. Puchta, R. van Eldik, New J. Chem. 33 (2009) 34.
References
[1] R. Hage, A. Lienke, Angew. Chem. Int. Ed. 45 (2006) 206.