Modulation of electronic and redox properties in phenolate-rich cobalt(iii) complexes and their implications for catalytic proton reduction

Article abstract of DOI:10.1039/c4dt03337j

We investigate the redox, spectroscopy and catalytic reactivity of new cobalt(iii) complexes based on phenolate-rich [N₂O₃] ligands. These complexes are described as [Co^III(L^X)MeOH], where X indicates the presence of chloro (1), bromo (2), iodo (3), or tert-butyl (4) substituents in the 3^rd and 5^th positions of each phenolate ring. These substituents modulate the Co(iii) ← PheO^- LMCT bands of the parent complexes with 1 (451) > 2 (453) > 3 (456) > 4 (468 nm) and the redox potentials involved with the Co(iii)/Co(ii) and ligand reduction and with the phenolate/phenoxyl oxidation processes. The influence of the substituents on the phenolate pendant arms was also observed on the kinetic parameters; 1 presented a rate constant of 1.0 × 10^-3 s^-1 whereas 4 showed a considerably slower rate (5.3 × 10^-5 s^-1). Species 1 and 4 are electrocatalysts towards proton reduction in the presence of weak acid in acetonitrile. A TON of 10.8 was observed for 1 after 3 h of bulk electrolysis at -2.20 V_Fc/Fc⁺ using a mercury pool as the working electrode. This journal is

Full text of DOI:10.1039/c4dt03337j

Dalton
Transactions
PAPER
Modulation of electronic and redox properties in
phenolate-rich cobalt(III) complexes and their
implications for catalytic proton reduction†
Cite this: Dalton Trans., 2015, 44,
3454
Debashis Basu, Marco M. Allard,‡ Fernando R. Xavier,§ Mary Jane Heeg,
H. Bernhard Schlegel and Claudio N. Verani*
We investigate the redox, spectroscopy and catalytic reactivity of new cobalt(III) complexes based on
phenolate-rich [N2O3] ligands. These complexes are described as [CoIII(LX)MeOH], where X indicates the
presence of chloro (1), bromo (2), iodo (3), or tert-butyl (4) substituents in the 3rd and 5th positions of
each phenolate ring. These substituents modulate the Co(III) ← PheO− LMCT bands of the parent com-
plexes with 1 (451) > 2 (453) > 3 (456) > 4 (468 nm) and the redox potentials involved with the Co(III)/Co(II)
and ligand reduction and with the phenolate/phenoxyl oxidation processes. The influence of the substitu-
ents on the phenolate pendant arms was also observed on the kinetic parameters; 1 presented a rate con-
stant of 1.0 × 10−3 s−1 whereas 4 showed a considerably slower rate (5.3 × 10−5 s−1). Species 1 and 4 are
electrocatalysts towards proton reduction in the presence of weak acid in acetonitrile. A TON of 10.8 was
observed for 1 after 3 h of bulk electrolysis at −2.20 VFc/Fc+ using a mercury pool as the working
electrode.
Received 31st October 2014,
Accepted 2nd January 2015
DOI: 10.1039/c4dt03337j
www.rsc.org/dalton
reduction because of the expected high overpotentials associ-
ated with the Co(II)/Co(I) couple, but the search for new archi-
Introduction
Considering the impending limitations of fossil fuels and the tectures that either yield new catalysts or allow for
increasing demand for energy, the search for earth-abundant rationalizations about catalyst design6 is impending.
catalysts capable of hydrogen production is at the core of an
Our group is engaged in the chemistry of cobalt complexes
economy based on renewable fuels.1 The cobalt ion has been with redox-active phenolate-rich environments focusing on tri-
used in several proton-reduction electrocatalysts because of its dentate [NN′O]7 and pentadentate [N2O3]8 ligands to predict
energetically viable conversion from 3d6 LSCo(III) or 3d7 Co(II) structural, electronic and redox properties of metallosurfac-
into the procatalytic nucleophile 3d8 LSCo(I) species.2 This tants, and the reactivity of metallodrugs. In [N2O3] environ-
state captures a proton to yield a reactive Co(III)H species ments with three t-butyl-decorated phenolates8 four reversible
which, pending on specific conditions can stay as Co(III)H or redox-accessible states are present; three involving distinct
undergo reduction to generate a related Co(II)H; These hydride phenolato/phenoxyl radical couples between 0.3 and 1.0 VFc+/Fc
species are able to react with another proton, thus yielding and one process associated with the Co(III)/Co(II) couple at
H2.3 Several ligands have been studied to accommodate the −0.6 VFc+/Fc. The redox processes were cycled 30 times without
cobalt ion; chief designs include oximes, pending-arm polypyri- major decomposition at the surface of the electrode. These
dines, macrocycles, and iminopyridines.4 Although phenolate- results are encouraging because reversibility of redox processes
based cobalt complexes have been used in olefin and epoxide along with modulation of the Co(III)/Co(II) potential may be
polymerization,5 these complexes are less studied for proton relevant for the development of proton reduction catalysts.
With the previous arguments in mind, we hypothesize that
the nature of phenolate-linked substituent will modulate the
potential by which cobalt(III)/(II) reduction takes place. In this
Department of Chemistry, Wayne State University, Detroit, MI 48202, USA.
E-mail: cnverani@chem.wayne.edu; Fax: +313577 8022; Tel: +313 577 1076
article we investigate the structural, electronic and redox
behavior of four complexes, shown in Scheme 1, by means of
experimental and computational methods. We also evaluate
their behavior as electron acceptors in ascorbic acid-mediated
reductions and in proton reduction in acidic media. The
results follow.
†Electronic supplementary information (ESI) available. CCDC 1012180 and
1012179. For ESI and crystallographic data in CIF or other electronic format see
DOI: 10.1039/c4dt03337j
‡Current address: Department of Chemistry, La Sierra University, Riverside, CA
92515, USA.
§Current address: Department of Chemistry, Santa Catarina State University,
Joinville 89219, Brazil.
3454 | Dalton Trans., 2015, 44, 3454–3466
This journal is © The Royal Society of Chemistry 2015

Dalton Transactions
Paper
for 1–4. Species 4 shows intense peaks between 0 to 2 ppm
due to the presence of tert-butyl substituents. Selected
1H-NMR spectra for 1 and 4 are shown in Fig. 1 while the
spectra for 2 and 3 are shown in the ESI as Fig. S4 and S5.†
Species 1, and 2 had their molecular structures solved by X-ray
analysis, while complex 4 was described previously.8
Molecular structures
Complexes 1, and 2 yielded well-formed brown crystals used
for single crystal X-ray diffractometric analysis. The atomic
labeling schemes of 1, and 2 are shown in Fig. 2 while selected
bond lengths and angles are shown in Table 1.
Complexes 1 and 2 belong to a C2/c monoclinic space
group. Both crystal structures reveal a mononuclear [CoIIILX-
(MeOH)] neutral complex unit (X = Cl or Br) where each of the
cobalt(III) centers is in an [N2O4] coordination environment. A
distorted octahedral geometry is imposed over the metallic ion
by the three phenolate arms, the rigid azomethine, the amine
moieties, and the coordinated methanol molecule, as summar-
ized by the selected bond lengths and angles shown in
Table 1. In all three structures the O1 oxygen atoms associated
with the iminophenolates are coordinated to the Co1 center in
a trans position to the N2 amine nitrogen atom, while the O2
and O3 aminophenolate oxygen atoms are trans to each other.
The O4 oxygen atom of the weakly bound protonated molecule
of methanol is trans to the N1 imine nitrogen atoms. This Co–
O(4) is the longest, reaching 1.95 to 2.01 Å. The Co–Ophenolate
bonds vary between 1.87 and 1.92 Å, with the Co–O2 bonds
being slightly longer than Co–O1 and Co–O3. The Co–N bonds
show considerable length variability, with a short Co-N1 ≈
1.87 Å associated with the CvN group (N1–C7 ≈ 1.29 Å) and
longer Co–N2 ≈ 1.98 Å associated with the amine group. The
average N1–C8 bond length at 1.41 Å for 1 and 2 is slightly
shorter than the equivalent N2–C13 (av. 1.47 Å), thus
suggesting electron delocalization along with the 1,2-diamino-
benzene ring and the iminophenolate arm (Fig. S6†). Shorter
O1–C1 (av. 1.30 Å) and C6–C7 (av. 1.42 Å) bonds reinforce the
argument while the C6–C7–N1–C8 dihedral angles of 177.7(3)°
for 1 and 174.3(8)° for 2 reveal planarity indicative of a conju-
gated π framework. The geometrical arrangement, as well as
the bond lengths and angles are isostructural with the pre-
Scheme 1 Cobalt(III) complexes.
Results and discussion
Syntheses and characterizations
A series of asymmetrical pentadentate ligands was synthesized
by treatment of 1,2-diaminobenzene with 2,4-disubstituted-
chloromethylphenol precursors bearing chloro (1), bromo (2),
iodo (3) and t-butyl (4) functionalities in presence of triethyl-
amine in dichloromethane (Fig. S1†). Pale yellow solids were
obtained for each ligand after brine extraction and character-
1
ized with spectroscopic (FTIR and H-NMR) and ESI-MS spec-
trometric methods. The ligands were dissolved in methanol
and treated with cobalt(II) perchlorate under aerobic con-
ditions and in the presence of sodium methoxide to afford the
desired cobalt(III) complexes 1–4 (Fig. S2†). During the course
of the reaction, the cobalt(II) ion is oxidized into cobalt(III) and
the ligand is oxidized from its amine to the imine form. This
amine/imine conversion has been observed in iron complexes
with related ligands.9
Compounds 1–4 were characterized by spectroscopic and
spectrometric methods. The FTIR spectra for each complex
confirms a CvN stretching around 1585 cm−1 resulting from
ligand oxidation. The absence of peaks associated with the
perchlorate anions confirms 1–4 as neutral Co(III) species, in
excellent agreement with the resulting elemental analyses.
High-resolution ESI mass spectra show the species [M + H+]+,
[M + Li+]+, and [M + Li+]+, for 1–3, respectively (Fig. S3†).
Experimental and simulated isotopic distributions are in
agreement with the proposed molecular composition. Due to
the diamagnetic nature of the low spin cobalt(III) complexes
the 1H-NMR spectra for all the complexes were taken in d6-
DMSO. The spectra have shown appropriate line-splitting pat-
terns for 1–4 with 10 typical aromatic protons, the azomethine
proton resonance (over the aromatic region) and four methy-
lene hydrogens between 4.0 and 5.0 ppm. A methanol molecule
bound to the sixth coordination position of the Co(III) center
was assigned with a resonance line between 3.0 and 3.3 ppm
viously reported 4 and other species in similar [N2O3OMeOH
]
environments.8
Electronic and electrochemical properties
Calculated electronic structures. Insight into the nature of
the molecular orbitals of complexes 1, 2, and 4 were obtained
by means of DFT calculations. Complexes 4 were modeled
replacing the t-butyl groups of the phenolate rings by methyl
1
groups (4Me). In agreement with the H-NMR data, the orbital
occupancies are shown in Fig. 3 plotted as comparative mole-
cular orbital ladders of diamagnetic closed shell singlets of
3d6 LSCo(III) complexes. The calculated bond lengths are con-
sistent with the experimental values determined by X-ray dif-
fraction with differences between 0.002–0.043 Å, well within
acceptable range. An energy difference of 0.6 eV (∼14 kcal
This journal is © The Royal Society of Chemistry 2015
Dalton Trans., 2015, 44, 3454–3466 | 3455

Paper
Dalton Transactions
Fig. 1 1H-NMR spectra for compounds 1 and 4 in d6-DMSO.
Fig. 2 ORTEP10 representations of 1, and 2 (left to right).
mol−1) becomes evident, when comparing the highest occu- 2 is the energy lowering of the first three unoccupied orbitals
pied molecular orbitals (HOMOs) of the electron-withdrawing LUMO, LUMO+1, and LUMO+2 which correspond to the
chloro-substituted
1 with those of the electron-donating empty π* orbital of the imine, and two empty Co-based orbitals
methyl-substituted 4. The first four HOMOs, namely HOMO, respectively.
HOMO−1, HOMO−2, and HOMO−3 are predominantly based
Effect of ligand substitution and solvents in the electronic
on the phenolato groups, in good agreement with previous spectra. The electronic spectra of 1–4 were measured in aceto-
assignments.7d,e,8,11,12 Species 1, 2, and 4 are comparable to nitrile and dichloromethane (Fig. 4 and S8†) and summarized
each other in MO occupancy and orbitals energies (Fig. 3 and in Table 2. UV-visible spectra were also measured in other
Fig. S7†). A noticeable effect of halogen substitution on 1 and coordinating solvents (Fig. S9†). The complexes show charge-
3456 | Dalton Trans., 2015, 44, 3454–3466
This journal is © The Royal Society of Chemistry 2015

Dalton Transactions
Paper
Table 1 Selected bond lengths and bond angles
1
2
Co1–N1⋯1.878(2)
Co1–N1⋯1.879(8)
Co1–N2⋯1.981(8)
Co1–O1⋯1.886(7)
Co1–O2⋯1.925(7)
Co1–O3⋯1.870(7)
Co1–O4⋯1.955(7)
N1–C7⋯1.305(12)
N2–C14⋯1.520(13)
N2–C21⋯1.501(11)
N1–C8⋯1.406(12)
N2–C13 1.473(14)
O1–C1⋯1.302(12)
O2–C20⋯1.357(11)
O3–C27⋯1.336(10)
N1–Co1–O2⋯92.6(3)
O1–Co1–O2⋯86.3(3)
O2–Co1–O4⋯89.4(3)
O2–Co1–N2⋯93.4(3)
O1–Co1–N1⋯95.4(3)
O1–Co1–O3⋯88.4(3)
O1–Co1–O4⋯86.5(3)
N1–Co1–N2⋯85.7(3)
N1–Co1–O3⋯87.8(3)
O3–Co1–N2⋯91.9(3)
O4–Co1–N2⋯92.5(3)
O3–Co1–O4⋯90.5(3)
Co1–N2⋯1.986(2)
Co1–O1⋯1.871(2)
Co1–O2⋯1.9077(19)
Co1–O3⋯1.8834(19)
Co1–O4⋯1.9627(19)
N1–C7⋯1.296(4)
N2–C14⋯1.517(4)
N2–C21⋯1.498(4)
N1–C8⋯ 1.426(4)
N2–C13⋯1.470(3)
O1–C1⋯1.297(3)
O2–C20⋯1.331(3)
O3–C27⋯1.316(3)
N1–Co1–O2⋯92.88(9)
O1–Co1–O2⋯85.41(9)
O2–Co1–O4⋯87.93(8)
O2–Co1–N2⋯94.81(9)
O1–Co1–N1⋯96.04(10)
O1–Co1–O3⋯87.58(9)
O1–Co1–O4⋯87.27(8)
N1–Co1–N2⋯85.65(10)
N1–Co1–O3⋯87.65(9)
O3–Co1–N2⋯92.21(9)
O4–Co1–N2⋯⋯91.03(9)
O3–Co1–O4⋯91.95(8)
Fig. 4 UV-Visible spectra of 1–4 (1.0 × 10−4 M; acetonitrile).
Table 2 UV-Visible spectroscopic data for 1–4 in dichloromethane and
acetonitrile
λmax, nm/(ε, M−1 cm−1
Compound in acetonitrile
)
λmax, nm/(ε, M−1 cm−1) in
dichloromethane
1
2
3
4
451 (8034), 475 (7602),
697 (594)
453 (8160), 476 (7751),
696 (755)
456 (8592), 479 (8053),
697 (796)
468 (9011), 492 (8305),
757 (656)
455 (7431), 486 (7323),
712 (565)
456 (8077), 486 (7782),
712 (582)
457 (8684), 487 (8012),
712 (634)
471 (8798), 496 (8174),
759 (958)
the phenolate-installed substituents in the order 1 < 2 < 3 < 4.
Similarly, the LMCT bands for 1–3 in acetonitrile are hypso-
chromically shifted by 8–16 nm, when compared to dichloro-
methane, thus indicating that acetonitrile replaces the weakly
bound methanol molecule in solution. Although changes are
more subtle for 4, the ability to replace methanol is relevant
for proton-reduction catalysis.
Time-dependent DFT methods were used to model and
assign the bands observed in acetonitrile spectra of 1, and 4
(Fig. 5 and S10†); The simulated spectrum of 1 indicates a
lower energy band at 683 nm associated with an aminopheno-
late-to-Co(III) CT and in good agreement with the experi-
Fig. 3 MO ladders and plots for 1 (left) and 4 (right).
transfer absorptions between 400 and 800 nm along with pro- mentally observed band at 697 nm. The main CT contribution
cesses below 400 nm, typically associated with ligand-centered at higher energy originates from amino/iminophenolate-
charge transfer.
to-phenyleneimine
and
amino/iminophenolate-to-Co(III)
The two strong but ill-solved bands between 400–500 nm (Table S1†). Two distinct bands result from multiple processes
are attributed to a phenolate-to-azomethine intraligand charge calculated between 390–520 nm; one at ca. 430 nm and
transfer (ILCT, 450–475 nm) and a ligand-to-metal charge another at 500 nm. Both bands are in relatively good agree-
transfer process (LMCT) involving an in-plane pπphenolate ment with the experimentally observed processes at 451 and
→dσ*Co(III) transition13 (475–500 nm). The shoulders seen at 475 nm and are relevant to indicate the origin of the experi-
lower energy (550–800 nm) are attributed to out-of-plane pπphenolate mental CT processes. Calculated TD-DFT spectra for 4 also
14
→dσ*Co(III)
LMCT processes. A modest variation in the show similar behavior (Fig. S10 and Tables S2†).
Effect of ligand substituents on the redox behavior. Cyclic
associated with the electron withdrawing or donating nature of voltammograms (CV) for 1, 2, and 4 were recorded in aceto-
position of the ILCT and in-plane LMCT processes can be
This journal is © The Royal Society of Chemistry 2015
Dalton Trans., 2015, 44, 3454–3466 | 3457

Paper
Dalton Transactions
Fig. 5 Absorption spectra for 1 in CH3CN.TD-DFT calculated spectra (red trace) in CH3CN as solvent model and experimental spectra (black trace)
in the same solvent. The inset indicates the lowest energy transition.
Table 3 Cyclic voltammetric data for 1–4 in CH3CN
3rd PhO−/PhO•
2nd PhO−/PhO•
1st PhO−/PhO•
Co(III)/Co(II)
CvN/C•–N−
E1/2, V (ΔE, V)
E1/2, V (ΔE, V)
E1/2, V (ΔE, V)
|ipc/ipa
|
|ipc/ipa
|
|ipc/ipa
|
E1/2, [Epc; Epa], V (ΔE, V) |ipc/ipa
|
E1/2, V (ΔE, V) |ipc/ipa
|
1
2
+1.33 (0.13) |0.82|
+1.33 (0.05)
+1.09 (0.08) |0.95|
+1.06 (0.09) |0.53|
+0.73 (0.08) |1.00|
+0.70 (0.08) |0.30|
−0.63 [−0.72; −0.52] (∼0.20)
|1.7|
Spike at Epa = −0.15
−0.68 [−0.80; −0.56] (0.22) |1.8| −2.06 (0.11) |7.29| −2.20 (0.09)
−2.03 (0.06) |1.43|
|2.53|
3a
4
—
—
—
Spike at Epa = −0.16
−0.64 [−0.72; −0.56] (0.12)
|1.06|
Epc = −2.07; Epc = −2.25
−2.55 (0.08) |1.09|
+0.99 (0.16) |0.82|
+0.58 (0.07) |0.98|
+0.21 (0.07) |0.97|
a 3 in DMF.
nitrile using TBAPF6 (0.1 M) as supporting electrolyte. Com- we infer that two electrons are transferred at slightly different
pound 3 was measured in N,N′-dimethylformamide due to potentials. We did not pursue this issue. The metal-centered
solubility restrictions. Potentials were measured against Ag/ process was quasi-reversible with Epc = −0.80 and Epa = −0.56
AgCl and are reported versus the Fc/Fc+ redox couple. Table 3 VFc/Fc+ and E1/2 ≈ −0.68 VFc/Fc+. The CV for 3 was taken in DMF
summarizes the electrochemical data and Fig. 6 (1 and 4) and is shown in Fig. S12† over a narrower voltage window in the
Fig. S11 and S12† (2 and 3) display the results. Voltammo- anodic region that prevents further discussion of the pheno-
grams taken in dichloromethane for 1 and 4 are displayed in late/phenoxyl processes. An irreversible cobalt(III)/(II) reduction
Fig. S13† and data tabulated in Table S3,† show comparable was observed at Epc = −1.04 VFc/Fc+, along with two irreversible
behaviors to those in acetonitrile (Fig. S13†).
processes attributed to the phenylene-imine observed with Epc
Compound 1 (Fig. 6, top) shows three reversible phenolate/ values of −2.07 and −2.25 VFc/Fc+. Complex 4 (Fig. 6, middle)
phenoxyl processes between 0.50 and 1.50 VFc/Fc+ while a showed three reversible phenolate/phenoxyl processes between
reversible process at −2.03 V Fc/Fc+ is attributed to the reduction 0.10 and 1.25 VFc/Fc+ and one reversible phenylene-imine
of the phenylene-imine moiety.15 The process attributed to the process at −2.55 VFc/Fc+. Additionally—contrary to the barely
cobalt(III)/(II) couple is quasi-reversible with Epc and Epa reversible behavior observed for 1–3—the Co(III)/Co(II) redox
respectively at −0.72 and ca. −0.52 VFc/Fc+, corresponding to an couple observed at −0.64 VFc/Fc+ exhibited excellent reversibility
E1/2 ≈ −0.63 VFc/Fc+ with |ipc/ipa| = 1.7.7d,16 This process is with ΔE = 0.12 and |ipc/ipa| = 1.06.
obfuscated by a sharp spike at −0.15 VFc/Fc+ associated to an
The quasi-reversible behavior observed for 1–3 is associated
unidentified product. Compound 2, (Fig. S11†) shows a less with the reorganizational energy barrier for the Co(II)/Co(III)
reversible profile with three phenolate/phenoxyl processes couple.17 The Epc values for the Co(III)/(II) couple in 1 and 4
between 0.70 and 1.50 VFc/Fc+, whereas the phenylene-imine vary by ∼0.72 V, whereas the imine based process (2nd
process is split in two irreversible waves between −1.80 and reduction) shifted to more negative potentials by 0.5 V. This
−2.50 V. As the phenylene-imine moiety can accept 2 electrons, variation is attributed to electronic effects of the different
3458 | Dalton Trans., 2015, 44, 3454–3466
This journal is © The Royal Society of Chemistry 2015

Dalton Transactions
Paper
selected systems. These MOs are directly related to the reactiv-
ity of complexes and have proven useful in comparing the rele-
vant redox sequences between species 1 and 4. In order to
make calculations more affordable, we modeled the tert-butyl
substituents in 4 using methyl groups. All five electrochemical
processes were probed and are summarized in Fig. 7. Spin
density plots resulting from reduction, namely [CoL]−, [CoL]2−
,
as well as oxidation, namely [CoL], [CoL]+ [CoL]2+ and [CoL]3+
were obtained. Interestingly, the second phenolate-based oxi-
dation, [CoL]2+, occurs at the amino-phenolate moiety for the
4CH3, whereas for the chloro-substituted 1 this oxidation origi-
nates from the imino-phenolate (Fig. S15,† 7). An observed
extended conjugation for 1 and other halogeno-substituted
complexes is most likely the reason behind this phenomenon.
The first reduction yields spin density on the cobalt center,
consistent with a high spin Co(II) while the second reduction
is solely based on the imine moiety with most of the added
electron being shared between the carbon and nitrogen atoms.
Upon the addition of this second electron, we observed a sig-
nificant bond length increase of 0.06 Å in the CvN bond. This
is consistent with our previous assignment of this reduction as
exclusively ligand-based. As reduction processes are relevant
for proton reduction catalysis, usual attributions include the
transformations Co(III) → Co(II) → Co(I). For species 1–4 we did
not observe the Co(II)/Co(I) couple, instead we observed a
ligand-based reduction that follows the Co(III)/Co(II) couple.
Better understanding of the redox processes observed for
1–4 was achieved via spectropotentiostatic experiments in
acetonitrile (or dimethylformamide for 3) using TBAPF6 as
supporting electrolyte. When the potential was fixed in −1.40
VFc/Fc+ spectral changes in the visible region were observed as
shown in Fig. 8. The LMCT absorption bands between 450 and
800 nm decrease, in agreement with the proposed Co(III)/Co(II)
redox pair. Isosbestic points were observed at ca. 430–440 nm
reflecting a decrease in the LMCT bands at 475–500 nm, while
new absorption bands appeared at 420–440 nm for all com-
plexes. The disappearance of the LMCT bands upon reduction
suggests these new bands do not require the involvement of
the metal center, thus supporting the TD-DFT assignment of
intraligand charge transfer (πphenolate→π*imine).
Fig. 6 Cyclic Voltammograms of 1 (top), and 4 (bottom) in acetonitrile.
Conditions: 0.1 M TBAPF6 as supporting electrolyte; Glassy carbon
(working), Pt wire (counter) and Ag/AgCl (reference); Scan rate:
100 mV s−1
.
substituents attached to the imino-phenolate rings, and the
more electronegative the substituent, the less electron density
will be centered in the imine unit, resulting in less negative
reduction potentials. For 1—containing electron-withdrawing
–Cl substituents—the lowest reduction potential for the imine
process is observed. For the oxidative couples, the electro-
chemical potentials for the phenolates shifted to less positive
values when substitution changes from the more electron-
withdrawing chloro to the more donating tert-butyl group.
Higher electron density upon the phenolate rings facilitates
electrochemical oxidation and so, potentials tend to be less
positive with a ∼0.5 VFc/Fc+ shift from 1 to 4. The potentials of
1–4 were recorded 5 times using different stock solutions to
confirm reproducibility of the data.
When a fixed potential was applied at −2.40 VFc/Fc+ (for 1–3)
targeting the azomethine process, the initial response within
ca. 10 min was similar to that shown in Fig. 8a. However after
the full reduction of the cobalt(III) center, the ILCT absorption
Redox potentials were calculated via DFT methods and
showed comparable values to the experimental data (Table S4
and Fig. S14†). In order to confirm the redox loci observed, we
have evaluated the frontier molecular orbitals (MOs) for
Fig. 7 Spin density plots for redox processes in 1.
This journal is © The Royal Society of Chemistry 2015
Dalton Trans., 2015, 44, 3454–3466 | 3459

Paper
Dalton Transactions
Fig. 8 (a) Spectral changes upon electrochemical reduction of the
Co(III)/Co(II) process in 1, 2 and 4 in acetonitrile and 3 in DMF. The
applied potential was −1.40 VFc/Fc+ over a period of 10 min; (b) Spectral
changes upon second reduction of 1 in acetonitrile. The applied poten-
tial was −2.40 VFc/Fc+ and the graph represents the behavior after
10 min. TBAPF6 (0.1 M) was used as supporting electrolyte.
Fig. 9 (a) Time-dependent UV-Visible spectral changes in acetonitrile–
water (90 : 10% v/v) for complex 1; (b) Chemical reduction tests for 1–4
with the reducing agent ascorbic acid. Conditions: [C]final = 1.80 ×
10−4 M, [AA] = 2.00 × 10−2 M and pH ∼ 3.0.
gradually starts to decrease (Fig. 8b and Fig. S16†). This is
indicative that the azomethine moiety becomes reduced. As for
oxidations, when the potential was fixed to 0.85 VFc/Fc+ for 1–3
(and 0.30 VFc/Fc+ for 4) a band appeared around 1000 nm,
which has been ascribed to a phenolate to phenoxyl ILCT
process12,18 (Fig. S17†). This is in good agreement with our
previous observations in similar systems.8 TD-DFT calculations
were used to model the experimentally observed changes in
the visible spectrum of 1, both upon cobalt(III) reduction and
phenolate oxidation. The first reduction results in disappear-
ance of the LMCT process whereas the ILCT band remains
place due to the timescale of the experiment, chemical
reductions tend to be faster and more complete. Spontaneous
electron transfer reaction from ascorbic acid to 1–4 is assured
by the electrochemical potentials observed for the Co(III)/Co(II)
couple,19 and the experiments were performed in CH3CN–H2O
(90 : 10% v/v). Results were evaluated spectrophotometrically
following the disappearance of the phenolate-to-cobalt(III)
LMCT processes in the parent compounds at room tempera-
ture (Fig. 9a, Fig. S20†). This disappearance is due to the con-
version of the low-spin 3d6 Co(III) into a 3d7 Co(II) species that
precludes the Co(III) ← PhO− LMCT process. Plots of complex
concentrations vs. time were obtained and fitted using a first
order exponential decaying equation (Fig. S21†) and linearized
using a first order rate equation (Fig. 9b). From the linearized
data, the rate constants (kobs) and half-life values (t1/2) were
obtained for the complexes 1–4 and are listed in Table 4.
intact. Similarly, oxidation results in
a
peak around
600–800 nm, along with another band at very low energy
(∼1500 nm). Both observations are in excellent agreement with
the experimental findings (Fig. 8, S18 and S19†).
Reactivity studies
Behavior as electron acceptors. As phenolates stabilize the
trivalent cobalt in complexes 1–4, conversion of Co(III) into
Co(II) is a fundamental step relevant for catalysis. In order to
understand the electron acceptance ability of 1–4, chemical
reduction tests were carried out using excess ascorbic acid as
the reductant. Opposite to an electrochemical reduction, in
which slow and sometimes incomplete redox processes take
The electronic influence of the substituents present on the
phenolate pendant arms was clearly observed on the kinetic
parameters. Compound 1 presented a rate constant of 1.0 ×
10−3 s−1 whereas compound 4 showed a considerably slower
rate of 5.3 × 10−5 s−1. Based on these results we conclude that
the rate constants (kobs) tend to decrease when electron-with-
3460 | Dalton Trans., 2015, 44, 3454–3466
This journal is © The Royal Society of Chemistry 2015

Dalton Transactions
Paper
Table 4 Kinetic parameters obtained for the chemical reduction reac-
tion between complexes 1–4 and ascorbic acida
Complexes
Parameter
kobs (s−1
1
2
3
4
)
1.0 × 10−3
12
4.4 × 10−4
26
1.3 × 10−4
89
5.3 × 10−5
220
t1/2 (min)
a Acetonitrile–water (90 : 10% v/v); [C]final = 1.80 × 10−4 M, [AA] = 2.00 ×
10−2 M, pH ∼ 3.0 at room temperature.
drawing groups such as chloro are replaced by electron-donat-
ing groups such as tbutyl. Half-lives (t1/2) for 1–4 were also cal-
culated. For these electron transfer reactions, the t1/2 values
follow the same trend observed for the rate constants. The
results show clearly the time-wise efficiency of the complex 1,
where in 12 minutes 50% of that species had been reduced to
1
its Co(II) counterpart. For complex 4, on the other hand, the t
2
value is approximately 20-fold lower. This behavior can be
explained in terms of electron density over the metallic center:
electron deficiency is triggered on the metal when electron-
withdrawing groups are present, making the cobalt(III) centers
more positive and electron acceptance easier. Overall, the
trend can be summarized as follows: 1 (–Cl) > 2 (–Br) > 3 (–I) >
4 (tBu).
Behavior as proton-reduction catalysts. We tested the
activity of compounds 1–4 towards proton reduction in pres-
ence of a weak acid such as acetic acid (HOAc). We focused our
attention on complexes 1 and 4 due to their inherently
different redox and electronic natures. These species show
their catalytic wave in close vicinity to the ligand-based imine
process (Fig. 10a and Fig. S22†). Overpotential was calculated
to be 0.60 V and 0.90 V for 1 and 4, respectively. Species 1 gen-
erated H2 at lower overpotentials than those observed for the
t-butyl substituted 4. This is tentatively associated with the pres-
ence of the electron-withdrawing chloro substituents. Com-
parison of the cyclic voltammogram between 1 and a control
blank was also performed. The blank contained ten equivalent
of acetic acid in absence of the catalyst. The system containing
1 generated dihydrogen at a lower overpotential of 0.35 V when
compared to the blank (Fig. 10b and Fig. S23†). Hydrogen
Fig. 10 (a) Electrocatalytic activity of complex
1
towards proton
reduction in presence of acetic acid. (b) Comparison of cyclic voltam-
mogram between 1 and blank in the presence of 10 equivalents of
acetic acid. [Conditions for (a) and (b): Glassy carbon (WE); Pt-wire (AE);
Ag/AgCl (RE)]. (c) Charge vs. time plot over 3 h during bulk-electrolysis
at −2.2 VFc/Fc+ of 1 in comparison to the blank. [Conditions: Hg-pool
(WE); Pt-coil (AE); Ag/AgCl (RE)].
generation was confirmed from bulk-electrolysis measure- experiment, no deposition of any solids was observed on the
ments for 1 at an applied potential of −2.20 VFc/Fc+, and an electrode surface. Moreover, analysis of the post-catalysis solu-
associated turnover number (TON) of 10.8 was determined tion suggested the presence of a CoII-species within a similar
after three hours with a faradaic efficiency of 85% using a Hg- ligand environment, as shown by the UV-Visible spectra
pool as working electrode, Ag/AgCl as the reference electrode, (Fig. S24†). The attribution of the spectral profile is based on
and a platinum wire as the auxiliary electrode. Charge con- similarities with that of the Co(II) spectra obtained during
sumption over time is observed to be considerably higher spectroelectrochemistry. Similar bulk-electrolysis measure-
when the cobalt complex is present (Fig. 10c), thus validating ments were not possible for 4 due to its high overpotential.
species 1 as a viable catalyst. Furthermore, the catalyst gener- Appearance of the catalytic peak close to the imine-based
ates 660 μmol of hydrogen in three hours of bulk-electrolysis process suggested the doubly reduced state as the catalytic
at −2.2 VFc/Fc+. This is a 16 fold increase in comparison to a species. On the other hand, protonation of the phenolates in
negligible 40 μmol of hydrogen observed for the control under these conditions was ruled out due to unchanged UV-Visible
identical conditions. Each set of experiments was repeated in spectra for either 1 or 4 upon addition of much stronger nitric
triplicate using fresh solutions. During the controlled potential acid to acetonitrile.
This journal is © The Royal Society of Chemistry 2015
Dalton Trans., 2015, 44, 3454–3466 | 3461

Paper
Dalton Transactions
Fig. S23† shows a comparison of the redox-catalytic data to cm−1. All the spectra were recorded in acetonitrile, dichloro-
the blank sample and confirms lower overpotential and higher methane and dimethylformamide (HPLC-grade) solutions (for
increase of current in presence of the complexes. This behavior CV, UV-Visible) and in CH3CN–H2O (90 : 10% v/v) at pH 3.0
validates the complexes as the catalysts. Species 1 displays with final [complex] = 1.8 × 10−4 M (for chemical reduction).
chloro-substituted phenolates accepted to withdraw electronic
Electrochemistry and spectroelectrochemistry. The electro-
density from the metal center, thus turning the metal ion chemical behavior of 1–4 was investigated with a BAS 50W
more electrophilic and helping stabilize the bivalent and potentiostat/galvanostat. Cyclic voltammograms were obtained
monovalent states involved in catalysis. The Co(I) species at room temperature in acetonitrile, dichloromethane or N,N′-
shows some electronic density spreading over one of the dimethylformamide solutions containing 0.1 M of n-Bu4NPF6
phenolate rings, reinforcing this idea. As these overpotentials as supporting electrolyte under argon atmosphere. The electro-
and TONs are encouraging in terms of substituents effect for chemical cell employed was comprised of three electrodes:
proton reduction, they point out to the possibility of incorpor- glassy-carbon (working), platinum wire (auxiliary) and Ag/AgCl
ating stronger electron-withdrawing nitro and sulfonic groups (reference). The ferrocene/ferrocenium redox couple Fc/Fc+
to the ligand framework as a means to optimize catalysis. This (E° = 400 mV vs. NHE)20 was used as the internal standard.
work is currently under development in our labs.
Peak to peak potential separations (ΔEp = |Ep,c − Ep,a|) and
|ipa/ipc| values were measured to evaluate the reversibility of
the redox processes. Spectroelectrostatic measurements were
carried out in
a
optically transparent thin-layer cell
Conclusions
(ca. 0.1 mm) constructed according to a procedure described
as follows: a flat platinum wire in a “u” shape was sandwiched
between two glass slides where the inner parts were coated
with indium-tin oxide (ITO) (8–12 Ω sq−1). The flat wire acted
as the working electrode and extended outside of the slides for
electrical contact. The solutions were prepared and degassed
under an inert atmosphere (argon) and introduced into the
cell through a capillary action. The working electrode was
located within 4–6 mm of the cell bottom to minimize ohmic
potential (iR) drop. All potentials were measured vs. a Ag/AgCl
reference electrode and a second platinum wire (counter).
Potentials were applied to the cell by a BAS 50W potentiostat/
galvanostat, and the spectra were collected with a Varian Cary
50 apparatus at the room temperature within a typical time
interval of 30 s until the equilibrium between oxidized/
reduced species was achieved.
X-ray structural determination. Diffraction data for 1 and 2
were measured on a Bruker X8 APEX-II kappa geometry dif-
fractometer with Mo radiation and a graphite monochromator
and are summarized in Table 5. Frames were collected at
100 K with the detector at 40 mm and 0.3 degrees between
each frame and were recorded for 10 s. APEX-II21 and SHELX22
software were used in the collection and refinement of the
models. Crystals of 1 (CCDC # 1012180) appeared as dark
In this study we investigated the effect of phenolate substitu-
ents on the redox and electronic processes of a series of cobalt(III)
complexes. These complexes showed the expected Co(III)/
Co(II) couple along with multiple ligand-centered redox pro-
cesses, and the nature of the substituent modulates the poten-
tials by which ligand-based reduction takes place. The visible
region of the spectrum dominated by phenolate-to-cobalt(III)
and phenolate-to-phenyleneimine charge transfer bands.
Upon reduction of the metal center, the LMCT disappears, evi-
dencing the LLCT. The chloro-substituted complex showed
promising electron accepting ability. Kinetic plots for chemical
reduction favored a relative order as –Cl > –Br > –I > –t-Bu. The
chloro and t-butyl substituted species also showed potential as
catalysts for proton reduction in acidic media. Again the
chloro-substituted species yielded lower overpotentials than its
t-Bu counterpart. These overpotentials are considerably nega-
tive, therefore, limiting practical use of this framework for
efficient proton reduction. Nonetheless they support the notion
that incorporation of strong electron-withdrawing groups to the
ligand framework may lead to optimized catalytic properties.
These efforts are currently under development in our labs.
Materials, methods and synthetic procedures.
General. All reagents and solvents were used as received needles. A total of 106 022 reflections were measured, yielding
from commercial sources. Infrared spectra were recorded from 7148 unique data (Rint = 0.074). Hydrogen atoms were placed
4000 to 650 cm−1 as KBr pellets on a Bruker Tensor 27 FTIR in calculated positions. These neutral molecules crystallized
spectrophotometer. 1H-NMR spectroscopy was carried out without ions, solvent or appreciable disorder. Crystals of 2
within a Mercury FT-NMR 400 MHz setup using CDCl3 or d6- (CCDC # 1012179) were dark rods. A total of 58 214 reflections
DMSO as solvents, at 25 °C. ESI-(+) mass spectrometry were were counted, which averaged to 7895 independent data (Rint
=
measured in a triple quadrupole Micromass Quattro LC equip- 0.089). Hydrogen atoms were placed at calculated positions.
ment where experimental mass patterns were fitted with The solvate molecules occupied sites of symmetry and this did
theoretical isotopic distribution. Elemental analysis (C, H, and not refine without disorder in the space group C2/c. The
N) determinations were performed using Exeter analytical PLATON program SQUEEZE23 was utilized to include the
CHN analyzer by Midwest Microlab: Indianapolis, Indiana. solvate electrons resulting in an empirical formula of 1
UV–visible spectra were obtained at room temperature using a complex: 1 dichloromethane molecule.
Varian Cary 50 spectrophotometer operating in the range of
Computational methods. Electronic structure calculations
200 to 1000 nm with quartz cells. Values of ε are given in M−1 were carried out with the Gaussian 09 suite of programs24
3462 | Dalton Trans., 2015, 44, 3454–3466
This journal is © The Royal Society of Chemistry 2015

Dalton Transactions
Paper
tration at a given time (M); [C]0 is the initial concentration (M),
k is the rate constant (s−1) and t is the time (s). The half-lives
for the compounds 1–4 were calculated using the expression
t1/2 = ln 2/k where t1/2 is the half-life and k is the rate
constant (s−1).
Table 5 Crystal Data for the complexes 1 and 2
1
2
Empirical formula
Formula weight
Temperature (K)
Wavelength (Å)
C28H19Cl6CoN2O4 C29H21Br6Cl2CoN2O4
719.08
100(2)
1070.77
100(2)
0.71073
Catalytic activity. Proton reduction electrocatalysis was
tested for 1 and 4 via cyclic voltammetry in presence of acetic
acid (HOAc, pKa: 22.3 in CH3CN).28a Glassy carbon was used as
the working electrode, platinum wire as the auxiliary electrode
(15 cm length, from which 13 cm coiled and submerse in solu-
tion), and Ag/AgCl as the reference electrode, with tetra-butyl
ammonium hexafluorophosphate (TBAPF6) as supporting elec-
trolyte. Overpotentials (η) were calculated from the observed
changes in cyclic voltammogram followed by subtracting the
thermodynamic standard potential for H+/H2 in CH3CN in the
presence of HOAc (after considering the homoconjugation
effect28) from the experimental half-wave potential for the cata-
lytic peak in the presence of particular amount of HOAc. To
determine the amount of hydrogen release, bulk electrolysis
has been done in a custom-made air-tight H-type cell using a
mercury-pool as working electrode, Ag/AgCl as reference, both
in the main chamber. A coiled Pt wire was used as the auxiliary
electrode isolated on another compartment by a frit. TBAPF6
was used as the supporting electrolyte. The main chamber was
filled with an electrolyte solution and a proton source
(TBAPF6: 1.56 g; HOAc: 0.24 g [4 mmol], 20 mL MeCN),
whereas the smaller compartment was filled with electrolyte
solution (TBAPF6: 0.39 g; 5 mL MeCN). The applied potential
for bulk electrolysis is measured against Ag/AgCl and poten-
tials are converted against Fc/Fc+ upon subtracting the electro-
chemical potential of Fc/Fc+ from the applied potential versus
Ag/AgCl in similar condition. In a typical test, the cell was
purged with N2 gas for 20 min followed by sampling of the
head space gas (100μl) to ensure an O2 free environment in the
gas-chromatograph. The GC is a Gow-Mac 400 equipped with a
thermal conductivity detector and a 8′ × 1/8″ long 5 Å mole-
cular sieve column operating at 60 °C used with N2 as the
carrier gas. The calibration was carried out with hydrogen (H2)
gas (Hydrogen GC grade 99.99 + %, Scotty analyzed gases,
Sigma Aldrich). The solution (no catalyst) was electrolyzed for
three hours at −1.8 VAg/AgCl (−2.2 VFc/Fc+) and the head space
gas was injected into the GC to record the amount of dihydro-
gen generated. Then the cell was purged with N2 gas for
another 20 min followed by injection of the catalyst
(0.04 mmol) dissolved in MeCN. Bulk electrolysis was con-
ducted for three hours at −1.8VAg/AgCl (−2.2 VFc/Fc+) and the
head space gas (100 μL) was injected to the GC instrument.
The amount of dihydrogen produced was measured. The turn-
over number was calculated after background subtraction as
the ratio between moles of dihydrogen produced over moles of
catalyst used. Faradic efficiency was calculated from GC
measurements.
0.71073
Crystal system, space group Monoclinic, C2/c Monoclinic, C2/c
a (Å)
b (Å)
c (Å)
α (°)
22.0260(10)
12.9174(6)
22.6310(12)
90
28.251(3)
13.5819(10)
22.091(2)
90
β (°)
116.109(3)
90
5781.9(5)
8
125.407(7)
90
6908.9(10)
8
γ (°)
Volume (Å3)
Z
Calculated density
1.652
2.059
(Mg m−3
)
Absorption coefficient
1.187
7.629
(mm−1
)
F (000)
R(F) (%)
Rw(F) (%)
2896
4.39
6.70
4096
8.22
11.29
R(F) = ∑||Fo| − |Fc||/|Fo|; Rw(F) = [∑w(Fo − Fc2)2/∑w(Fo2)2]1/2 for
I > 2σ(I).
2
using Density Functional Theory (DFT). The B3PW91/6–31G(d)
level of theory25 was employed throughout. Geometries were
fully optimized without symmetry constraints, and stationary
points were verified via frequency analysis. Solvent effects in
dichloromethane and or acetonitrile were estimated using the
built-in default settings of the IEF-PCM polarizable continuum
model.26 Cartesian coordinates of all optimized structures are
provided as ESI.† Vertical electronic excitation energies and
intensities were calculated using time-dependent density func-
tional theory (TD-DFT).27 Molecular orbitals were plotted with
GaussView and UV-visible spectral plots were prepared using
SWizard with a full width at half-height of 2000 cm−1. All opti-
mized geometries agreed well with the crystallographic data
available for all relevant structures.
Chemical reductions. These experiments were performed in
CH3CN–H2O (90 : 10% v/v) upon ascorbic acid (reductant) for
complexes 1–4 were evaluated spectrophotometrically on a
Varian Cary 50 spectrophotometer, at their λmax, respectively
by the disappearance/decrease of their ligand-to-metal charge
transfer (LMCT) at room temperature. The absorption values
were converted into concentration using their absorptivity
molar coefficient values (ε, M−1 cm−1). All reactions were moni-
tored throughout the time until the absorbance at each λmax
value was constant. Reactions were performed using the fol-
lowing conditions: In a 4 mL cuvette (1 cm optical path) were
added 2700 μL of a freshly prepared and degassed acetonitrile
solution of the complex ([C]final 1.80 × 10−4 M), and the reac-
tion was initiated by the addition 300 μL of a aqueous ascorbic
acid (pH ∼ 3.0 adjusted with HNO3; [AA]final = 1.80 × 10−2 M).
The concentration of the complexes 1–4 towards the elapsed
time were plotted and fitted using a typical first order exponen-
tial decaying equation [C] = [C]0 × ekt and linearized using the
first order rate law ln[C] = −kt + ln[C]0 where [C] is concen-
Synthetic procedures.
The 2,4-substituted chloromethylphenols (pendant arms)
where –R can be Cl or Br were synthesized under similar
This journal is © The Royal Society of Chemistry 2015
Dalton Trans., 2015, 44, 3454–3466 | 3463

Paper
Dalton Transactions
conditions (vide infra). The precursor 2-(chloromethyl)-4,6-
diiodo-phenol, 2-(chloromethyl)-4,6-di-tert-butyl-phenol, and
the ligand 6,6′-(((2-((3,5-di-tert-butyl-2-hydroxybenzyl)amino)-
phenyl)azanediyl)bis(methylene)bis(2,4-di-tert-butyl-phenol)
6,6′-(((2-((3,5-Dichloro-2-hydroxybenzyl)amino)phenyl)azanediyl)-
bis(methylene)bis(2,4-dichlorophenol) – H3L′Cl. Yield: 75%. IR
(KBr, cm−1) 3505(w), 3414(w) (OH); 3263(w) (NH); 3078(w) (Ar–
CH); 2983(w) (alkyl–CH); 1599(m), 1467(s) (Ar–C–C); 1165(m)
1
(H3LtBu
)
were synthesized according methods already
(C–O). H-NMR [400 MHz, CDCl3, 300 K] δ/ppm = 4.093 [s, 4H
described in the literature.7c,8
(CH2)]; 4.263 [s, 2H (CH2)]; 6.61 [d, 1H (aryl)]; 6.776 [t, 1H
(aryl)]; 6.887 [s, 2H (aryl)]; 7.011 [t, 1H (aryl)]; 7.068 [s, 1H
(aryl)]; 7.167 [d, 1H (aryl)]; 7.212 [s, 2H (aryl)]; 7.3 [s, 1H (aryl)].
ESI pos. in MeOH: m/z = 630.9670 for [H3L1 + H+]+.
6,6′-(((2-((3,5-Dibromo-2-hydroxybenzyl)amino)phenyl)azanediyl)-
bis(methylene)bis(2,4-dibromophenol) – H3L′Br. Yield: 70%. IR
(KBr, cm−1) 3490(w) (OH); 3252(w) (NH); 3070(w) (Ar–CH);
2982(w) (alkyl–CH); 1598(m), 1455(s) (Ar–C–C); 1144(m) (C–O).
1H-NMR [400 MHz, CDCl3, 300 K] δ/ppm = 4.096 [s, 4H (CH2)];
4.249 [s, 2H (CH2)]; 6.644 [d, 1H (aryl)]; 6.798 [t, 1H (aryl)];
7.023 [t, 3H (aryl)]; 7.157 [d, 1H (aryl)]; 7.268 [s, 1H (aryl)];
7.475 [s, 2H (aryl)]; 7.542 [s, 1H (aryl)] ESI pos. in MeOH: m/z =
894.7 for [H3L2 + H+]+.
The precursor 2,4-substituted-hydroxymethylphenols. A
70 mL methanol solution of the 3,5-di-substituted-2-hydroxy-
benzaldehyde (25.0 mmol) was reduced by NaBH4 (60 mmol)
at 0 °C overnight. Then, the solvent was removed under
reduced pressure and the residual white solid was dissolved in
water and the pH was adjusted to ∼5.0 with 2.0 M HCl. The
product was extracted from water using dichloromethane and
dried over anhydrous Na2SO4. The solution was filtered and
the solvent was removed by rotary evaporation yielding a white
colored solid. The crude products were used with no further
purification. 2,4-Dichloro-6-(hydroxymethyl)phenol: Yield:
70%. IR (KBr, cm−1) 3281(s) (OH); 3050(w) (Ar–CH); 2948(w)
(alkyl–CH); 1596(w), 1580(w), 1474(s) (Ar–C–C); 1166(m) (C–O).
1H-NMR [400 MHz, CDCl3, 300 K] δ/ppm = 2.259 [s, 1H (ali-
phatic-OH)]; 4.779 [s, 2H (CH2)]; 6.654 [s, 1H (aryl–OH)]; 7.129
[s, 1H (aryl)]; 7.289 [s, 1H (aryl)]. 2,4-Dibromo-6-(hydroxy-
methyl)phenol: Yield: 70%. IR (KBr, cm−1) 3508(m), 3241(m)
(OH); 3075(w) (Ar–CH); 2934(w) (alkyl–CH); 1456(s) (Ar–C–C);
1141(s) (C–O). 1H-NMR [400 MHz, CDCl3, 300 K] δ/ppm = 2.632
[s, 1H (aliphatic-OH)]; 4.786 [s, 2H (CH2)]; 4.8 [s, 1H (aryl–
OH)]; 6.921 [s, 1H (aryl)]; 7.590 [s, 1H (aryl)].
6,6′-(((2-((2-Hydroxy-3,5-diiodobenzyl)amino)phenyl)azanediyl)-
bis(methylene))bis(2,4-diiodophenol)
–
H3L′I. Yield: 73%. IR
(KBr, cm−1) 3385(w) (OH); 3226(w) (NH); 3061(w) (Ar–CH);
2920(w) (alkyl–CH); 1598(m), 1451(s) (Ar–C–C); 1147(m) (C–O).
1H-NMR [400 MHz, CDCl3, 300 K] δ/ppm = 4.07 [s, 4H (CH2)];
4.238 [s, 2H (CH2)]; 6.733[t, 2H (aryl)]; 6.867 [t, 1H (aryl)];
7.077 [t, 2H (aryl)]; 7.183 [d, 2H (aryl)]; 7.526 [s, 1H (aryl)];
7.809 [s, 1H (aryl)]; 7.961 [s, 1H (aryl)] ESI pos. in MeOH: m/z =
1182.582 for [H3L3 + H+]+.
The complexes. Caution: Perchlorate salts are potentially
explosive and should be handled in small quantities and with
care.
The precursor 2,4-substituted-chloromethylphenols. Thio-
nyl chloride (50 mmol) was added to the solution of the
previously synthesized alcohols (20 mmol) in 30 mL of
dichloromethane. After stirring it overnight, the solvent was
removed by rotary evaporation and washed five times with
n-pentane. The 2,4-disubstituted-6-(chloromethyl)phenols were
isolated as a white solid and used with no further purification.
2,4-Dichloro-6-(chloromethyl)phenol: Yield: 94%. IR (KBr,
cm−1) 3474(s) (OH); 3081(w) (Ar–CH); 2978(w) (alkyl–CH); 1597
(w), 1580(w), 1469(s) (Ar–C–C); 1161(s) (C–O). 1H-NMR
[400 MHz, CDCl3, 300 K] δ/ppm = 4.618 [s, 2H (CH2)]; 5.775 [s,
1H (OH)]; 7.281 [s, 1H (aryl)]; 7.322 [s, 1H (aryl)]. 2,4-Dibromo-
6-(chloromethyl)phenol: Yield: 90%. IR (KBr, cm−1) 3462(s)
(OH); 3066(w) (Ar–CH); 2971(w) (alkyl–CH); 1461(s) (Ar–C–C);
1143(s) (C–O). 1H-NMR [400 MHz, CDCl3, 300 K] δ/ppm = 4.618
[s, 2H (CH2)]; 5.761 [s, 1H (aryl–OH)]; 7.445 [s, 1H (aryl)]; 7.577
[s, 1H (aryl)].
The proligands. Phenylene diamine (2 mmol) was treated
with the appropriate 2,4-disubstituted-6-(chloromethyl)-
phenol (6.2 mmol) in presence of triethylamine (8 mmol) in
80 mL of dichloromethane for 3 days under reflux to yield a
yellow colored solution. The mixture was washed three times
with brine solution (3 × 200 mL) to remove excess triethyl-
amine, dried over anhydrous sodium sulfate and the crude
product was isolated by solvent rotoevaporation. Unreacted
chloride was removed by washing the solid with cold hexane
to yield a yellow-colored solid. Upon coordination to the
metal these amine proligands are stabilized as imine
ligands.8
Complexes 1–4 have been synthesized under aerobic con-
ditions using the general procedure described as follows: to a
30 mL solution of ligand in dichloromethane (1 mmol) were
added sodium methoxide (0.162 g, 3 mmol) in 30 mL of
methanol and the mixture was stirred for 10 minutes. A 20 mL
methanol solution of [Co(H2O)6](ClO4)2 (0.365 g, 1 mmol)
were added dropwise to the mixture in a period of 5 minutes.
After the addition was complete, the solution was refluxed
for 4 hours to ensure the completion of reaction and
the mixture was concentrated to 10 mL. Slow evaporation
of the solvent gives rise to brown colored precipitate which
has been collected by vaccum filtration. Further recrystali-
zation form different solvent mixtures give crystalline pure
product.
[CoIII(LCl)MeOH] (1). Recrystallized from MeOH–di-ethyl
ether (1 : 1). Yield. 80%. IR (KBr, cm−1) 3444(w) (OH); 3063(w)
(Ar–CH); 2950(w) (alkyl–CH); 1612(s), 1450(s) (Ar–C–C); 1584
(m) (CvN); 1177(m) (C–O); No ClO4−. H-NMR [400 MHz, d6-
1
DMSO, 300 K] δ/ppm = 3.150 [d, 3H (CH3)]; 3.967 [d, 2H
(CH2)]; 4.083 [q, 1H (OH)]; 4.767 [d, 2H (CH2)]; 6.734 [s, 2H
(aryl)]; 6.820 [s, 2H (aryl)]; 7.138 [t, 1H (aryl)]; 7.300 [t, 1H
(aryl)]; 7.552 [s, 1H (aryl)]; 7.645 [s, 1H (aryl)]; 7.686 [d, 1H
(aryl)]; 7.955 [s, 1H (NvCH)]; 8.056 [d, 1H (aryl)]. ESI pos. in
MeOH: m/z = 684.8624 for [CoIII(L′Cl) + H+]+. Anal. Calcd for
C28H19Cl6CoN2O4: C, 46.77; H, 2.66; N, 3.90. Found: C, 46.46;
H, 2.64; N, 4.00.
3464 | Dalton Trans., 2015, 44, 3454–3466
This journal is © The Royal Society of Chemistry 2015

Dalton Transactions
Paper
[CoIII(LBr)MeOH] (2). Recrystallized from MeOH–DCM (1 : 1).
Yield. 85%. IR (KBr, cm−1) 3263(w) (OH); 3050(w) (Ar–CH);
(e) N. S. Lewis and D. G. Nocera, Proc. Natl. Acad. Sci.
U. S. A., 2006, 103, 15729; (f) J. A. Turner, Science, 2004,
305, 972.
2 (a) W. R. McNamara, Z. Han, C. J. Yin, W. W. Brennessel,
P. L. Holland and R. Eisenberg, Proc. Natl. Acad. Sci.
U. S. A., 2012, 109, 15594; (b) M. T. D. Nguyen,
M. F. Charlot and A. Aukauloo, J. Phys. Chem. A, 2011, 115,
911; (c) B. D. Stubbert, J. C. Peters and H. B. Gray, J. Am.
Chem. Soc., 2011, 133, 18070.
3 (a) S. C. Marinescu, J. R. Winkler and H. B. Gray, Proc. Natl.
Acad. Sci. U. S. A., 2012, 109, 15127; (b) J. T. Muckermann
and E. Fujita, Chem. Commun., 2011, 47, 12456;
(c) B. H. Solis and S. Hammes-Schiffer, Inorg. Chem., 2011,
50, 11252.
2945(w) (alkyl–CH); 1612(m), 1442(s) (Ar–C–C); 1584(m)
−
(CvN); 1157(m) (C–O); No ClO4
.
1H-NMR [400 MHz, d6-
DMSO, 300 K] δ/ppm = 3.149 [d, 3H (CH3)]; 3.977 [d, 2H
(CH2)]; 4.077 [q, 1H (OH)]; 4.8 [d, 2H (CH2)]; 6.877 [s, 2H
(aryl)]; 7.032 [s, 2H (aryl)]; 7.145 [t, 1H (aryl)]; 7.295 [t, 1H
(aryl)]; 7.693 [d, 1H (aryl)]; 7.724 [s, 1H (aryl)]; 7.843 [s, 1H
(aryl)]; 7.981 [s, 1H (NvCH)]; 8.054 [d, 1H (aryl)]. ESI pos. in
MeOH: m/z = 954.5676 for [CoIII(L′Br) + Li+]+. Anal. Calcd for
C28H19Br6CoN2O4: C, 34.11; H, 1.94; N, 2.84. Found: C, 33.60;
H, 1.81; N, 2.95.
i
[CoIII(LI)MeOH] (3). PrOH Recrystallized from DCM–isopro-
panol (1 : 1). Yield. 90%. IR (KBr, cm−1) 3443(w) (OH); 3047(w)
(Ar–CH); 2925(w) (alkyl–CH); 1610(m), 1427(s) (Ar–C–C); 1581
4 (a) C. C. McCrory, L. C. Uyeda and J. C. Peters, J. Am. Chem.
Soc., 2012, 134, 3164; (b) W. M. Singh, T. Baine, S. Kudo,
S. Tian, X. A. N. Ma, H. Zhou, N. J. DeYonker, T. C. Pham,
J. C. Bollinger, D. L. Baker, B. Yan, C. E. Webster and
X. Zhao, Angew. Chem., Int. Ed., 2012, 51, 5941;
(c) W. R. McNamara, Z. Han, C. J. Yin, W. W. Brennessel,
P. L. Holland and R. Eisenberg, Proc. Natl. Acad. Sci.
U. S. A., 2012, 109, 15594; (d) L. M. Utschig, S. C. Silver,
K. L. Mulfort and D. M. Tiede, J. Am. Chem. Soc., 2011, 133,
16334; (e) T. M. McCormick, Z. Han, D. J. Weinberg,
W. W. Brennessel, P. L. Holland and R. Eisenberg, Inorg.
Chem., 2011, 50, 10660; (f) C. F. Leung, Y. Z. Chen,
H. Q. Yu, S. M. Yiu, C. C. Ko and T. C. Lau, Int. J. Hydrogen
Energy, 2011, 36, 11640; (g) Y. Sun, J. P. Bigi, N. A. Piro,
M. L. Tang, J. R. Long and C. J. Chang, J. Am. Chem. Soc.,
2011, 133, 9212; (h) B. Probost, A. Rodenberg,
M. Guttentag, P. Hamm and R. Alberto, Inorg. Chem., 2010,
49, 6453; (i) P. A. Jacques, V. Artero, J. Pecaut and
M. Fontecave, Proc. Natl. Acad. Sci. U. S. A., 2009, 106,
20627; ( j) P. Du, J. Schneider, G. Luo, W. W. Brennessel
and R. Eisenberg, Inorg. Chem., 2009, 48, 4952; (k) C. Li,
M. Wang, J. Pan, P. Zhang, R. Zhang and L. Sun, J. Organo-
met. Chem., 2009, 694, 2814; (l) A. Fihri, V. Artero,
A. Pereira and M. Fontecave, Dalton Trans., 2008, 5567;
(m) C.-F. Leung, S.-M. Ng, C.-C. Ko, W.-L. Man, J. Wu,
L. Chen and T.-C. Lau, Energy Environ. Sci., 2012, 5, 7903;
(n) S. Varma, C. E. Castillo, T. Stoll, J. Fortage,
A. G. Blackman, F. Moltan, A. Deronzier and
M.-N. Collomb, Phys. Chem. Chem. Phys., 2013, 15, 17544;
(o) L. Tong, R. Zong and R. P. Thummel, J. Am. Chem. Soc.,
2014, 136, 4881; (p) C. Bachmann, B. Probst, M. Guttentag
and R. Alberto, Chem. Commun., 2014, 50, 6737;
(q) M. Natali, A. Luisa, E. Lengo and F. Scandola, Chem.
Commun., 2014, 50, 1842.
1
(m) (CvN); 1150(m) (C–O); No ClO4−. H-NMR [400 MHz, d6-
DMSO, 300 K] δ/ppm = 3.147 [d, 3H (CH3)]; 3.945 [d, 2H
(CH2)]; 4.01 [q, 1H (OH)]; 4.808 [d, 2H (CH2)]; 6.978 [s, 2H
(aryl)]; 7.065 [t, 1H (aryl)]; 7.116 [t, 1H (aryl)]; 7.271 [s, 2H
(aryl)]; 7.718 [d, 1H (aryl)]; 7.843 [d, 1H (aryl)]; 7.945 [s, 1H
(NvCH)]; 8.044 [s, 2H (aryl)]. ESI pos. in MeOH: m/z =
1242.4844 for [CoIII(L′I)
+
Li+]+. Anal. Calcd for
C31H27I6CoN2O5: C, 28.04; H, 2.05; N, 2.11 Found: C, 28.55; H,
1.55; N, 2.52.
[CoIII(Lt-Bu)MeOH].8 (4). Recrystallized from MeOH–DCM.
1H-NMR [400 MHz, d6-DMSO, 300 K] δ/ppm = 0.86 [s, 18H (t-
butyl)]; 1.00 [s, 18H (t-butyl)]; 1.24 [s, 9H (t-butyl)]; 1.62 [s, 9H
(t-butyl)]; 3.145 [d, 3H (CH3)]; 3.855 [d, 2H (CH2)]; 4.085 [q, 1H
(OH)]; 4.865 [d, 2H (CH2)]; 6.46 [s, 2H (aryl)]; 6.53 [s, 2H (aryl)];
6.88 [t, 1H (aryl)]; 7.08 [t, 1H (aryl)]; 7.10 [s, 1H (aryl)]; 7.28
[s, 1H (aryl)]; 7.53 [s, 1H (NvCH)]; 7.645 [d, 1H (aryl)]; 8.035
[d, 1H (aryl)].
Acknowledgements
This material is based upon work supported by the U.S.
Department of Energy, Office of Science, Office of Basic Energy
Sciences under award number DE-SC0001907 to C. N. V. and
H. B. S., including financial support for D. B. (synthesis,
characterizations, catalysis), M. M. A. (DFT calculations), and
F. R. X. (spectroelectrochemistry, kinetics). Equipment use at
the WSU-Lumigen Instrument Center and computer time allo-
cated at the WSU-Grid System for DFT calculations is also
acknowledged.
References
5 (a) R. M. Thomas, P. C. B. Widger, S. M. Ahmed,
R. C. Jeske, W. Hirahata, E. B. Lobkovsky and G. W. Coates,
J. Am. Chem. Soc., 2010, 132, 16520; (b) E. N. Jacobsen, Acc.
Chem. Res., 2000, 33, 421.
6 (a) L. Chiang, L. E. N. Allan, J. Alcantara, M. C. P. Wang,
T. Storr and M. P. Shaver, Dalton Trans., 2014, 43, 4295;
(b) B. Solis, Y. Yu and S. Hammes-Schiffer, Inorg. Chem.,
2013, 52, 6994; (c) A. Kochem, F. Thomas, O. Jarjayes,
1 (a) P. D. Tran, V. Artero and M. Fontecave, Energy Environ.
Sci., 2010, 3, 727; (b) M. G. Walter, E. L. Warren,
J. R. McKone, S. W. Boettcher, Q. Mi, E. A. Santori and
N. S. Lewis, Chem. Rev., 2010, 110, 6446; (c) T. R. Cook,
D. K. Dogutan, S. Y. Reece, Y. Surendranath, T. S. Teets and
D. G. Nocera, Chem. Rev., 2010, 110, 6474; (d) A. J. Esswein
and D. G. Nocera, Chem. Rev., 2007, 107, 4022;
This journal is © The Royal Society of Chemistry 2015
Dalton Trans., 2015, 44, 3454–3466 | 3465

Paper
Dalton Transactions
G. Gellon, C. Philouze, T. Weyhermüller, F. Neese and 18 L. Benisvy, E. Bill, A. J. Blake, D. Collison, E. S. Davies,
M. Van Gastel, Inorg. Chem., 2013, 52, 14428;
(d) A. Kochem, H. Kanso, B. Baptiste, H. Arora, C. Philouze,
O. Jarjayes, H. Vezin, D. Luneau, M. Orio and F. Thomas,
Inorg. Chem., 2012, 51, 10557.
C. D. Garner, C. I. Guindy, E. J. L. McInnes, G. McArdle,
J. McMaster, C. Wilson and J. Wolowska, Dalton Trans.,
2004, 3647.
19 A. Neves, W. Herrmann and K. Wieghardt, Inorg. Chem.,
1984, 23, 3435.
7 (a) F. D. Lesh, R. L. Lord, M. J. Heeg, H. B. Schlegel and
C. N. Verani, Eur. J. Inorg. Chem., 2012, 463; (b) D. Tomco, 20 R. Gagne, C. Koval and G. Licenski, Inorg. Chem., 1980, 19,
F. R. Xavier, M. M. Allard and C. N. Verani, Inorg. Chim. 2854.
Acta, 2012, 393, 269; (c) D. Tomco, S. Schmitt, B. Ksebati, 21 APEX II collection and processing programs are distributed
M. J. Heeg, Q. P. Dou and C. N. Verani, J. Inorg. Biochem.,
2011, 105, 1759; (d) R. Shakya, S. S. Hindo, L. Wu,
by the manufacturer, Bruker AXS Inc., Madison, WI, USA,
2009.
M. M. Allard, M. J. Heeg, H. P. Hratchian, B. R. McGarvey, 22 G. M. Sheldrick, Acta Crystallogr., Sect. A: Fundam. Crystal-
S. R. P. D. Rocha and C. N. Verani, Inorg. Chem., 2007, 46, logr., 2008, 64, 112.
9808; (e) R. Shakya, C. Imbert, H. P. Hratchian, 23 A. L. Spek, J. Appl. Crystallogr., 2003, 36, 7.
M. Lanznaster, M. J. Heeg, B. R. McGarvey, M. M. Allard, 24 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
H. B. Schlegel and C. N. Verani, Dalton Trans., 2006, 2517.
8 M. M. Allard, F. R. Xavier, M. J. Heeg, H. B. Schlegel and
C. N. Verani, Eur. J. Inorg. Chem., 2012, 4622.
9 (a) M. Lanznaster, M. J. Heeg, G. T. Yee, B. R. McGarvey
and C. N. Verani, Inorg. Chem., 2007, 46, 72;
(b) M. M. Allard, J. A. Sonk, M. J. Heeg, B. R. McGarvey,
H. B. Schlegel and C. N. Verani, Angew. Chem., Int. Ed.,
2012, 51, 3178; (c) L. D. Wickramasinghe, M. M. Perera,
L. Li, G. Mao, Z. Zhou and C. N. Verani, Angew. Chem., Int.
Ed., 2013, 52, 13346.
10 (a) L. J. Farrugia, ORTEP for windows, J. Appl. Crystallogr.,
1997, 30, 565; (b) L. Zsolnai, ZORTEP: an interactive ortep
program, University of Heidelberg, Germany, 1997.
11 G. M. Zats, H. Arora, R. Lavi, D. Yufit and L. Benisvy,
Dalton Trans., 2011, 40, 10889.
12 A. Sokolowski, B. Adam, T. Weyhermuller, A. Kikuchi,
K. Hildenbrand, R. Schnepf, P. Hildebrandt, E. Bill and
K. Wieghardt, Inorg. Chem., 1997, 36, 3702.
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato,
X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,
J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,
H. Nakai, T. Vreven, J. A. Montgomery, J. E. Peralta,
F. Ogliaro, M. Bearpark, J. Heyd, J. E. Brothers,
K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand,
K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,
J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,
J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin,
R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
J. J. Dannenberg, S. Dapprich, P. V. Parandekar,
N. J. Mayhall, A. D. Daniels, O. Farkas, J. B. Foresman,
J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian G09,
Wallingford, CT, 2009.
13 (a) A. A. Khandar, B. Shaabani, F. Belaj and A. Bakhtiari, 25 (a) A. D. Becke, J. Chem. Phys., 1993, 98, 5648;
Inorg. Chim. Acta, 2007, 360, 3255; (b) A. H. Sarvestani and
S. J. Mohebbi, Chem. Res., 2006, 257; (c) M. S. Shongwe,
S. K. M. Al-Hatmi, H. M. Marques, R. Smith, R. Nukada
(b) J. P. Perdew, Phys. Rev. B: Condens. Matter, 1986, 33,
8822; (c) J. P. Perdew, K. Burke and Y. Wang, Phys. Rev.,
1996, 54, 16533.
and M. Mikuriya, J. Chem. Soc., Dalton Trans., 2002, 4064; 26 (a) S. Miertus, E. Scrocco and J. Tomasi, Chem. Phys., 1981,
(d) D. G. McCollum, G. P. A. Yap, A. L. Rheingold and
B. Bosnich, J. Am. Chem. Soc., 1996, 118, 1365.
14 M. I. Davis, A. M. Orville, F. Neese, J. M. Zaleski,
J. D. Lipscomb and E. I. Solomon, J. Am. Chem. Soc., 2002,
124, 602.
55, 117; (b) J. Tomasi, B. Mennucci and R. Cammi, Chem.
Rev., 2005, 105, 2999; (c) G. Scalmani, M. J. Frisch,
B. Mennucci, J. Tomasi, R. Cammi and V. Barone, J. Chem.
Phys., 2006, 124, 9410; (d) G. Scalmani and M. J. Frisch,
J. Chem. Phys., 2010, 132, 114110; (e) J. Tomasi,
B. Mennucci and R. Cammi, Chem. Rev., 2005, 105,
2999.
15 (a) S. Koyuncu, I. Kaya, F. B. Koyuncu and E. Ozdemir,
Synth. Met., 2009, 159, 1034; (b) S. A. M. Refaey,
A. A. Hassan and H. S. Shehata, Int. J. Electrochem. Sci., 27 (a) E. Runge and E. K. U. Gross, Phys. Rev. Lett., 1984, 52;
2008, 3, 325.
(b) R. E. Stratmann, G. E. Scuseria and M. J. Frisch,
J. Chem. Phys., 1998, 109, 8218.
16 A. Bottcher, T. Takeuchi, K. I. Hardcastle, T. J. Meade,
H. B. Gray, D. Cwikel, M. Kapon and Z. Dori, Inorg. Chem., 28 (a) V. Fourmond, P. A. Jacques, M. Fontecave and
1997, 36, 2498.
V. Artero, Inorg. Chem., 2010, 49, 10338; (b) J. A. S. Roberts
and R. M. Bullock, Inorg. Chem., 2013, 52, 3823;
(c) A. M. Appel and M. L. Helm, ACS Catal., 2014,
4, 630.
17 (a) P. D. Metelski, Y. Fu, K. Khan and T. W. Swaddle, Inorg.
Chem., 1999, 38, 3103; (b) J. F. Endicott and T. Ramasami,
J. Phys. Chem., 1986, 90, 3740.
3466 | Dalton Trans., 2015, 44, 3454–3466
This journal is © The Royal Society of Chemistry 2015

Copyright of Dalton Transactions: An International Journal of Inorganic Chemistry is the
property of Royal Society of Chemistry and its content may not be copied or emailed to
multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.