5,10,15-triferrocenylcorrole complexes

Article abstract of DOI:10.1021/acs.inorgchem.5b01575

Complexes of 5,10,15-triferrocenylcorrole were synthesized from the crude free-base corrole product obtained by the reaction of ferrocenyl aldehyde and pyrrole. Direct formation of the complex in this manner leads to an increase of the reaction yield by protecting the corrole ring toward oxidative decomposition. The procedure was successful and gave the expected product in the case of the copper and triphenylphosphinecobalt complexes, but an unexpected result was obtained in the case of the nickel derivative, where metal insertion led to a ring opening of the macrocycle at the 5 position, giving as a final product a linear tetrapyrrole nickel complex bearing two ferrocenyl groups. The purified 5,10,15-triferrocenylcorrole complexes have been fully characterized by a combination of spectroscopic methods, electrochemistry, spectroelectrochemistry, and density functional theory calculations. Copper derivatives of 10-monoferrocenyl- and 5,15-diferrocenylcorrole were prepared to investigate how the number and position of the ferrocenyl groups influenced the spectroscopic and electrochemical properties of the resulting complexes. A complete assignment of resonances in the ¹H and ¹³C NMR spectra was performed for the cobalt and nickel complexes, and detailed electrochemical characterization was carried out to provide additional insight into the degree of communication between the meso-ferrocenyl groups on the conjugated macrocycle and the central metal ion of the ferrocenylcorrole derivatives.

Full text of DOI:10.1021/acs.inorgchem.5b01575

Article
pubs.acs.org/IC
5,10,15-Triferrocenylcorrole Complexes
Giuseppe Pomarico,† Pierluca Galloni,† Federica Mandoj,† Sara Nardis,† Manuela Stefanelli,†
Andrea Vecchi,† Sara Lentini,† Daniel O. Cicero,† Yan Cui,‡ Lihan Zeng,‡ Karl M. Kadish,*,‡
and Roberto Paolesse*,†
†Dipartimento di Scienze e Tecnologie Chimiche, Universita
̀
di Roma Tor Vergata, via della Ricerca Scientifica 1, 00133 Rome, Italy
‡Department of Chemistry, University of Houston, Houston, Texas 77204-5003, United States
S
* Supporting Information
ABSTRACT: Complexes of 5,10,15-triferrocenylcorrole were
synthesized from the crude free-base corrole product obtained by
the reaction of ferrocenyl aldehyde and pyrrole. Direct formation of
the complex in this manner leads to an increase of the reaction yield
by protecting the corrole ring toward oxidative decomposition. The
procedure was successful and gave the expected product in the case
of the copper and triphenylphosphinecobalt complexes, but an
unexpected result was obtained in the case of the nickel derivative,
where metal insertion led to a ring opening of the macrocycle at the
5 position, giving as a final product a linear tetrapyrrole nickel
complex bearing two ferrocenyl groups. The purified 5,10,15-triferrocenylcorrole complexes have been fully characterized by a
combination of spectroscopic methods, electrochemistry, spectroelectrochemistry, and density functional theory calculations.
Copper derivatives of 10-monoferrocenyl- and 5,15-diferrocenylcorrole were prepared to investigate how the number and
position of the ferrocenyl groups influenced the spectroscopic and electrochemical properties of the resulting complexes. A
1
complete assignment of resonances in the H and 13C NMR spectra was performed for the cobalt and nickel complexes, and
detailed electrochemical characterization was carried out to provide additional insight into the degree of communication between
the meso-ferrocenyl groups on the conjugated macrocycle and the central metal ion of the ferrocenylcorrole derivatives.
compounds in mixed-valence states that can be exploited for
optoelectronic applications.
INTRODUCTION
■
Porphyrins and related macrocycles are a family of compounds
widely distributed in biological systems, where they perform
functions essential for life.1 It is admirable how Nature is able to
finely tune the reactions of these species by simple
modifications in the molecular skeleton. For example, iron
porphyrins are involved in oxygen activation and transport in
animals, while magnesium chlorins are utilized in the
photosynthetic processes of plants. These features have long
attracted the interest of researchers, with one goal being the
ability to use synthetic models that can imitate the chemistry of
the natural systems while also allowing for the possibility of
using these compounds in the development of new catalysts,
electronic devices, drugs, and sensors.2−7 Following the
biological paradigm, a tuning of the metallomacrocycle
properties can be carried out in different ways, starting from
variation of the coordinated central metal ion to variation of the
peripheral substituents and, finally, to changes in the molecular
skeleton of the conjugated π-ring system.
From the point of view of porphyrin analogues, there has
been a growing interest in corroles,10,11 which are contracted
porphyrins that show particular coordination and photophysical
behavior and have promising applications in the areas of
catalysis,12−16 chemical sensors,17 solar cells18−20 and medi-
cine.21,22 Corroles act as trianionic ligands, which are
characterized by a low oxidation potential and the ability to
induce a facile ligand-to-metal electron transfer. These features
lead corrole to generally behave as a noninnocent ligand.23−25
Adding Fc as an electron-donor group at one or more meso
positions of the corrole macrocycle should enhance the
tendency of these compounds to undergo an oxidation process.
Also, because one of the main features of Fc-based compounds
is the formation of “mixed-valence state” derivatives,26,27 the
authors of this current manuscript wished to take advantage of
these properties and have recently demonstrated the marriage
of these two units with the preparation of a copper complex of
5,10,15-triferrocenylcorrole (TFcCorrH3).28
We have now turned our attention to investigating how the
number and positions of the meso-Fc groups on the corrole
macrocycle will influence the physicochemical properties of the
Looking at these different approaches in the wide arena of
different porphyrinoid structures, particular attention has been
focused on ferrocene (Fc) as a peripheral substituent on the
porphyrin ring.8,9 One impetus for these synthetic studies has
been the fact that Fc-containing porphyrins have shown, in the
past, a specific redox activity with the possibility of obtaining
Received: July 14, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.5b01575
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 1. Molecular structures of copper corrolates bearing one, two, or three Fc groups.
1
NMR Spectroscopy. H and 13C spectra were recorded at 300 K
either with a Bruker AV300 spectrometer operating at 300 MHz or
with a Bruker Avance 600 MHz spectrometer operating at 600 MHz
compounds and report in the present paper a full character-
ization of the previously synthesized 5,10,15-
triferrocenylcorrolatocopper(III) derivative, TFcCorrCu, to-
gether with the preparation and characterization of the
corresponding mono- and bis-substituted Fc analogues, 5,15-
diphenyl-10-ferrocenylcorrolatocopper(III) (FcDPhCorrCu)
and 5,15-diferrocenyl-10-phenylcorrolatocopper(III)
(DFcPhCorrCu), whose structures are shown in Figure 1.
Finally, we wished to explore the scope of the synthetic
pathway leading to the ferrocenylcorrole complexes and
examined the same procedures in an attempted synthesis of
the cobalt and nickel triferrocenylcorrole derivatives. We were
successful in the preparation and characterization of the
corresponding triphenylphosphinecobalt(III) complex
(TFcCorrCoPPh3), but a nonoxidative ring opening of the
TFcCorrH3 macrocycle was observed during metalation with
nickel acetate, giving as an isolated final product the linear
diferrocenyltetrapyrrole derivative, DFcLTPNi, whose structure
is shown in Figure 2. Both compounds were then characterized
as to their spectroscopic and electrochemical properties.
for H and 150 MHz for 13C, equipped with a 5 mm inverse broad-
1
band probe and z-axis gradients. All experiments were performed in
CDCl3 at 300 K. Chemical shifts are given in ppm relative to a residual
solvent (7.26 ppm for 1H and 77.3 ppm for 13C). NOESY spectra were
acquired with a spectral window of 3.4 kHz (carrier frequency at 6.55
ppm) using 8192 t2 data points (512 t1 increments) and four transients
with a relaxation delay of 2 s and a mixing time of 800 ms. COSY
spectra were collected with a spectral window of 5.4 kHz (carrier
frequency at 5 ppm) using 4096 t2 data points (256 t1 increments) and
four transients with a relaxation delay of 2 s. 1H−13C HSQC
experiments were acquired with a spectral window of 13.6 kHz and a
carrier frequency at 45 ppm (aliphatic region) or a spectral window of
9 kHz and a carrier frequency at 120 ppm (aromatic region) using
4096 t2 data points (256 t1 increments) and four transients with a
1
relaxation delay of 2 s. H−13C HMBC experiments were acquired
with a spectral window of 30 kHz (carrier frequency at 100 ppm)
using 4096 t2 data points (1024 t1 increments) and 16 transients with a
relaxation delay of 2 s. All data were processed with TopSpin.
Electrochemistry. Cyclic voltammetry was carried out with an
EG&G model 173 potentiostat/galvanostat. A homemade three-
electrode electrochemistry cell was used, which consisted of a platinum
or glassy carbon working electrode, a platinum wire counter electrode,
and a saturated calomel reference electrode (SCE). The SCE was
separated from the bulk of the solution by a fritted-glass bridge of low
porosity, which contained a solvent/supporting electrolyte mixture. All
potentials are referenced to the SCE.
Thin-layer UV−visible spectroelectrochemical experiments were
performed with a home-built thin-layer cell, which has a light-
transparent platinum net working electrode. Potentials were applied
and monitored with an EG&G PAR model 173 potentiostat. Time-
resolved UV−visible spectra were recorded with a Hewlett-Packard
model 8453 diode-array spectrophotometer. High-purity N2 from
Trigas was used to deoxygenate the solution and kept over the
solution during each electrochemical and spectroelectrochemical
experiment.
Synthesis. TFcCorrH3 and TFcCorrCu were prepared as
previously reported,28 starting from ferrocenecarboxaldehyde (400
mg, 1.87 mmol) and pyrrole (0.65 mL, 9.35 mmol). TFcCorrH3 was
partially purified by an alumina oxide (grade IV) plug eluted with
CHCl3 to remove the polypyrrolic impurities before the reactions to
insert the metal ions. Yields for the cobalt and nickel derivatives were
calculated with respect the amount of ferrocenecarboxaldehyde used
for the free-base corrole preparation.
(10-Ferrocenyl-5,15-diphenylcorrolato)copper (FcDPhCorr-
Cu). 5-Phenyldipyrromethane29 (400 mg, 1.80 mmol) and ferrocene-
carboxaldehyde (192 mg, 0.90 mmol) were dissolved in 72 mL of
methanol, after which 72 mL of aqueous HCl (0.6 M) was added to
the solution and the mixture stirred at room temperature for 3 h. The
suspension was extracted with CHCl3, washed twice with water, dried
Figure 2. Molecular structures of the cobalt and nickel derivatives.
EXPERIMENTAL SECTION
■
General Procedures. Reagents and solvents (Sigma-Aldrich,
Fluka, Merck, and Carlo Erba) were of synthetic grade and were
used without further purification. Silica gel 60 (70−230 mesh) or
neutral alumina (Brockmann grade III or IV) was used for
chromatography. UV−visible spectra were measured in CH2Cl2 with
a Varian Cary 50 spectrophotometer or a Hewlette-Packerd model
8453 diode-array spectrophotometer.
B
DOI: 10.1021/acs.inorgchem.5b01575
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
over Na2SO4, and diluted with CHCl3 to a volume of 270 mL, and 333
mg of chloranil (1.35 mmol) was added. This solution was stirred at
room temperature for 15 min, after which a methanolic solution of
Cu(AcO)2 was added and the mixture stirred to reflux for 20 min. The
solvent was removed and the mixture purified by a silica gel plug
eluted with CHCl3 to remove the decomposed material. Fractions
containing the corrole were collected and purified again by column
chromatography (silica gel and 1:1 CH2Cl2/hexane). The first eluted
band corresponds to 5,10,15-triphenylcorrrolatocopper (TPCorrCu),
followed by traces of copper porphyrin and then by the desired
FcDPhCorrCu, which was crystallized with CH2Cl2/CH3OH. Yield:
19 mg (3%). Mp: >300 °C. UV−vis [CH2Cl2; λmax, nm (log ε, M−1
cm−1)]: 415 (5.01), 527 sh, 634 (3.90). 1H NMR (300 MHz, CDCl3):
δ 8.49 (d, 2 H, J = 4.25 Hz, β-pyrrole), 7.90 (d, 2 H, J = 3.00, Hz, β-
pyrrole), 7.82 (d, 4 H, J = 7.35 Hz, phenyls), 7.68 (d, 2 H, J = 4.05 Hz,
β-pyrrole), 7.61 (d, 2 H, J = 6.93 Hz, phenyls), 7.56−7.51 (m, 4 H,
phenyls), 7.47 (d, 2 H, J = 4.23 Hz, β-pyrrole), 5.04 (br s, 2 H, α-Cp),
4.68 (br s, 2 H, β-Cp), 4.23 (s, 5 H, CpH). MS (FAB): m/z 696
(100%). Elem anal. Calcd for C41H27CuFeN4: C, 70.85; H, 3.91; N,
8.06. Found: C, 70.79; H, 3.85; N, 8.01.
column eluted with 1:1 CH2Cl2/hexane. A green fraction correspond-
ing to the title compound was crystallized by CH2Cl2/CH3OH. Yield:
9 mg (2%). Mp: >300 °C. UV−vis [CH2Cl2; λmax, nm (log ε, M−1
cm−1)]: 401 sh, 456 (4.48), 659 (4.03), 740 (sh), 818 (3.98), 900
(4.02). MS (FAB): m/z 711.05. Elem anal. Calcd for C38H28Fe2N4Ni:
C, 64.19: H, 3.97; N, 7.88. Found: C, 63.93; H, 3.91; N, 7.91.
Calculation Details. Computational calculations were performed
using Gaussian 09, revision A.02.31 Molecular orbitals (MOs) were
visualized using GaussView 5. Geometry optimizations were performed
with pure (BP86 and OLYP), pure meta (TPSSTPSS), and hybrid
(B3LYP and CAM-B3LYP) generalized gradient approximation
(GGA) functionals using a 6-31g(d,p) basis set.
RESULTS AND DISCUSSION
■
TFcCorrH3 was prepared by the reaction of pyrrole with
ferrocenyl aldehyde; this synthesis was carried out via two
routes: one a modified Lindsey approach optimized in our
laboratories for corrole synthesis32 and the other a water/
methanol route reported by Koszarna and Gryko.33 The yields
via these two approaches were generally comparable, and in
both cases, formation of the desired free-base triferrocenylcor-
role was accompanied by the corresponding free-base
tetraferrocenylporphyrin, usually in lower amounts when
following Koszarna and Gryko’s method. A key factor strongly
influencing the overall yield of TFcCorrH3 was, however, facile
oxidation of the macrocycle because, in addition to
decomposition products, both isocorrole species and open-
chain derivatives were observed in the reaction mixture. With
the aim of limiting this drawback, we modified the synthetic
protocol by adding copper acetate immediately after the
oxidation step. Insertion of the metal ion, with formation of the
copper complex, led to increased stability and reduced
overoxidation that affected the free-base corrole. Although
this modification does not allow one to directly obtain free-base
corrole, the possibility to demetallate the copper complexes
enables the preparation of free-base TFcCorrH3 and metal
derivatives upon treatment with the appropriate metal salt.
Spectroscopic characterization of TFcCorrCu indicated some
interesting optical properties of this complex. Copper 5,10,15-
triarylcorrolates exhibit UV−visible spectra that are strongly
influenced by substituents at the meso-phenyl rings of the
corrole and less affected by changes in substituents at the β-
pyrrolic positions on the macrocycle.34−38 Electron-donating
groups induce significant red shifts in the Soret band, and this
effect has been attributed to a phenyl (macrocycle)-to-metal
(5,15-Diferrocenyl-10-phenylcorrolato)copper (DFcPhCorr-
Cu). Ferrocenyldipyrromethane30 (330 mg, 1.00 mmol) and
benzaldehyde (51 μL, 0.50 mmol) were dissolved in 100 mL of
methanol, and then 50 mL of aqueous HCl (0.6 M) was added and the
mixture stirred at room temperature for 3 h. The suspension was
extracted with CHCl3, washed twice with water, dried over Na2SO4,
and diluted with CHCl3 to a volume of 250 mL, and 369 mg of
chloranil (1.50 mmol) was added. The solution was stirred at room
temperature for 15 min, after which a methanolic solution of
Cu(AcO)2 was added and the mixture further stirred to reflux for
20 min. The solvent was removed and the mixture purified by a silica
gel plug eluted with CHCl3 to remove any decomposed material.
Fractions containing corrole were collected and purified again by
column chromatography (silica gel and 3:2 CH2Cl2/hexane). The first
eluted band corresponds to TFcCorrCu, followed by traces of copper
porphyrin and then the desired DFcPhCorrCu, which was crystallized
with CH2Cl2/CH3OH. Yield: 8 mg (2%). Mp: >300 °C. UV−vis
1
[CH2Cl2; λmax, nm (log ε, M−1 cm−1)]: 420 (4.83), 666 (4.11). H
NMR (300 MHz, CDCl3): δ 8.65 (d, 2 H, J = 3.90 Hz, β-pyrrole), 8.36
(br s, 2 H, β-pyrrole), 8.02 (br s, 2 H, β-pyrrole), 7.73 (d, 2 H, J = 7.62
Hz, phenyls), 7.62−7.50 (m, 3 H, phenyls), 7.36 (d, 2 H, J = 4.38 Hz,
β-pyrrole), 5.07 (br s, 4 H, α-Cp), 4.66 (br s, 4 H, β-Cp), 4.22 (br s,
10 H, CpH). MS (FAB): m/z 805 (100%) Elem anal. Calcd for
C45H31CuFe2N4: C, 67.31; H, 3.89; N, 6.98. Found: C, 67.22; H, 3.91;
N, 6.93.
(Triphenylphosphine)(5,10,15-triferrocenylcorrolato)cobalt
(TFcCorrCoPPh3). The mixture containing TFcCorrH3 was dissolved
in CHCl3 and CH3OH (20 + 5 mL), and Co(AcO)2·4H2O (25 mg,
0.1 mmol) and PPh3 (26 mg, 0.1 mmol) were added to solution. The
mixture was refluxed for 20 min, while monitoring the course of the
reaction by thin-layer chromatography (TLC; silica gel/CH2Cl2) and
UV−visible spectroscopy. When the reaction was complete, following
the disappearance of the starting material as determined by TLC, the
solvent was removed under vacuum and the residue purified by a silica
gel plug eluted with CH2Cl2. Fractions containing cobalt corrole were
collected and purified again by a neutral Al2O3 plug eluted with 2:1
CH2Cl2/hexane. The green fraction, corresponding to TFcCorr-
CoPPh3, was crystallized by CH2Cl2/CH3OH. Yield: 22 mg (3%). Mp:
>300 °C. UV−vis [CH2Cl2; λmax, nm (log ε, M−1 cm−1)]: 405 (4.61),
598 (4.19). MS (FAB): m/z 904.6 (100%; M − PPh3). Elem anal.
Calcd for C67H50CoFe3N4P: C, 68.86; H, 4.31; N, 4.79. Found: C,
68.73; H, 4.11; N, 4.63.
2
2
(3dx −y ) charge-transfer character of the main absorbance band,
inducing a 10−11% ligand-to-metal charge-transfer character
for the B band.38 This feature provides further evidence for the
difficulty in characterizing electronic structures of such
noninnocent ligands. This phenyl (macrocycle)-to-metal charge
transfer also induces a saddlelike distortion of the tetrapyrrole
ring,39 even in the absence of peripheral substituent crowding, a
unique case among all of the metallocorrolates.
On the other hand, the observed spectrum of TFcCorrCu
(solid line in Figure 3) is quite different from spectra of the
copper triarylcorrole complexes (see, for example, the
TPCorrCu spectrum in Figure 3) and closely resembles that
of the corresponding copper tetraferrocenylporphyrin com-
plex.30 For this latter corrole, it has been reported that the UV−
visible spectroscopic pattern of the metalated macrocycle is
predominantly influenced by Fc-based orbitals lying in energy
between the occupied and unoccupied porphyrin-based
orbitals.30 For the corrole complexes, the situation is even
more complicated because it has been reported in the literature
Synthesis of a Nickel Complex. The mixture containing
TFcCorrH3 was dissolved in N,N-dimethylformamide (DMF; 8
mL), and Ni(AcO)2·4H2O (25 mg, 0.1 mmol) was added. The
mixture was stirred at 80 °C for 10 min, while monitoring the course
of the reaction by TLC (silica gel/CH2Cl2) and UV−visible
spectroscopy. After the disappearance of the starting material on a
TLC plate, water was added and the precipitate filtered. The residue
was dissolved in CHCl3, dried over Na2SO4, and purified on a silica gel
C
DOI: 10.1021/acs.inorgchem.5b01575
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
than does the substituent at the meso position 10.40 This
situation is quite different from what observed for similar
porphyrins bearing a variable number of Fc and phenyl units. In
this case, replacement of the phenyl rings with Fc moieties
induces a red shift of the Soret band because of increased
deviation from planarity of the porphyrin core.41
1
The H NMR spectra of FcDPhCorrCu and DFcPhCorrCu
show four doublets (or sometimes a low resolved broad singlet)
for the β-pyrrolic protons, which have resonance spread
between 7.3 and 8.8 ppm; in all cases, they can be easily
identified by the small coupling constant that is around 4 Hz,
while the signal for the phenyl rings appears as a multiplet at
around 7.5 ppm. There are three signals (1:1:2.5 ratio) for the
α- and β-substituted and unsubstituted cyclopentadienyl ring in
the 4−5 ppm region of the spectrum. The shape of the
absorption for the hydrogen atoms of the Fc rings suggests that
these substituents are free to rotate around, up and down with
respect to the main plane of the corrole macrocycle, as was
earlier reported for TFcCorrCu.28
Figure 3. UV−visible spectra (10−5 M in CH2Cl2) of TFcCorrCu
(solid line), DFcPhCorrCu (dotted line), FcDPhCorrCu (dashed
line), and TPCorrCu (dash-dotted line).
To gain further insight into both the electronic structure and
geometry of the complexes, density functional theory (DFT)
analysis on TFcCorrCu was performed. The results of these
calculations are described below.
that substituents at the meso positions 5 and 15 of the
macrocycle contribute more to variations of the corrole UV−
visible spectra than do substituents at the meso position 10.36
To verify whether this feature is observed in the case of Fc-
substituted corroles, copper derivatives were prepared with
corroles bearing one, two, or three meso-substituted Fc groups
(see the structures in Figure 1). FcDPhCorrCu and
DFcPhCorrCu were prepared following a method reported
for the synthesis of trans A2B corroles, starting from the
appropriate dipyrromethane and aldehyde.33 As was done for
TFcCorrCu, the free-base TFcCorrH3 was not isolated and
copper insertion was performed immediately after the oxidative
step in order to limit decomposition of the macrocycle. The
UV−visible spectrum of DFcPhCorrCu (dotted line in Figure
3) is very similar to that of TFcCorrCu (solid line in Figure 3),
both in the Soret and Q-band regions. On the other hand, the
spectrum of FcDPhCorrCu (dashed line in Figure 3) is quite
different, strongly resembling the published UV−visible
spectrum of TPCorrCu,38 although the Soret band of
FcDPhCorrCu is slightly blue-shifted with respect to the
bands of TFcCorrCu, while the Q band is significantly less
intense in the triferrocenylcorrole and positioned around 650
nm. It is also interesting to note that, with an increase in the
number of Fc substituents on the macrocycle, the intensity of
the Soret band decreases while the Q band becomes more
intense. This trend is seen in Figure 3 and also in Table 1,
which summarizes data for the three Fc-functionalized copper
corroles with those of the simple copper triphenylcorrole. The
similarity in the spectroscopic data for TPCorrCu and
FcDPhCorrCu compared to DFcPhCorrCu and TFcCorrCu
confirms that substituents at the 5 and 15 positions of the
macrocycle have a stronger effect on the UV−visible spectrum
Geometry optimizations and MO calculations were per-
formed first on TFcCorrCu in order to elucidate the
interactions, which occur between the meso-Fc groups and
the corrole core. Copper corroles have been studied in detail
from a theoretical point of view39,42−44 because of their
particular electronic configuration and geometric distortion of
the core induced by the metal and in order to discuss the
question of the noninnocence of the ligand. Different
approaches have been used, including DFT calculations,45−47
giving good results in terms of the geometric shape and extent
of the electronic interaction that occurs between the macro-
cycle and coordinated central metal ion. The pure and hydrid
GGA functionals give similar results to describe the geometry
of copper corrole, but a different reason has been found as
being responsible for saddling in the copper corrolates. In fact,
using pure OLYP and BP86 functionals, a covalent Cu 3d−
corrole π-type interaction was found, whereas using hybrid
DFT functionals (B3LYP or PBE0), an antiferromagnetic
coupling between an electron in the corrolate π and CuII 3d
orbitals was found. This latter interaction seems to be more
accurate in comparison with the ab initio calculations although
with a larger contribution.41 On the other hand, Fc-containing
porphyrins have been studied using DFT calculations, and the
best results, particularly those concerning the MO distributions,
have been obtained using the pure GGA functional with
Becke’s exchange functional and the Perdew 86 or Perdew
nonlocal correlation functional (BP86 or BPW91 GGA DFT
functionals).45−47
For all of these reasons, to investigate the geometry and
interaction between the Fc groups, the macrocycle core, and
the copper ion of the currently investigated compounds, we
decided to test first the different functionals on copper
derivatives: pure, hybrid, and meta-pure GGA combinations,
namely, BP86, OLYP, B3LYP, CAM-B3LYP, and TPSSTPSS.
All of the tested functionals give comparable results in terms of
the shape of the macrocycle, e.g., a saddled conformation as the
most stable geometry (Figure 4), with a up−down−up
arrangement of the meso-Fc substituents. Previous room
temperature NMR experiments indicate, however, that in
Table 1. Molar Absorption Coefficients (log ε) and Band
Maxima (λmax) of Copper Corrole Complexes in CH2Cl2
compound
Soret band (λmax
)
Q band (λmax)
TPCorrCu
5.09 (410)
5.01 (415)
4.83 (420)
4.75 (422)
3.89 (538)
3.90 (634)
4.11 (667)
4.14 (674)
FcDPhCorrCu
DFcPhCorrCu
TFcCorrCu
D
DOI: 10.1021/acs.inorgchem.5b01575
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
visible spectra (Figure 5). In fact, the HOMO−1 contributes
significantly to the Soret band absorption,45,46 and a strong
contribution from the Fc unit at the meso positions 5 and 15 is
seen. This indicates that the Fc groups at these two positions
mostly influence the electronic spectra compared to the Fc
group at position 10, which has a smaller effect. This is in
agreement with the similar absorption spectra of 5,15-
DiFcPhCorrCu and TFcCorrCu, whereas the 10-position
monosubstituted corrole FcDPhCorrCu has a UV−visible
spectrum similar to that of other corrole derivatives with
aromatic substituents. This behavior is in agreement with
published results that underline the strong contribution of
groups at the corrole meso positions 5 and 15 to the electronic
spectrum compared to the smaller contribution from groups at
the 10 position of the macrocycle.40,48
To further investigate the scope of the synthetic route for
obtaining triferrocenylcorroles, we attempted to prepare the
cobalt and nickel derivatives of TFcCorrH3. In the case of
TFcCorrCo(PPh3), the mixture containing TFcCorrH3 was
dissolved in CHCl3/CH3OH along with an excess of Co(AcO)2
and PPh3 and stirred to reflux, while monitoring the course of
the reaction by UV−visible spectroscopy and TLC analysis.
After the desired complex was formed, chromatographic
separation yielded a green fraction whose UV−visible spectrum
is shown in Figure 6. This spectrum is very similar to that of
Figure 4. Front and top views of TFcCorrCu [geometry optimization
at the BP86/6-31G(d,p) level of calculation].
solution the Fc units freely rotate around the meso-carbon
atoms.28
The main difference between the various functionals
concerns the contribution of different moieties to the orbital
density. In fact, the results of OLYP, B3LYP, and CAM-B3LYP
are inadequate to accurately describe the correct contribution
to the overall MOs from the corrole core, the Fc groups, and
the copper central metal. Using these functionals, the main
contribution to the highest occupied molecular orbital
(HOMO) is due to the Fc groups in positions 5 and 10, but
this result is not corroborated by the absorption spectrum,
where a strong contribution of the π-corrole core on the
HOMO is clearly visible; this result is very different from what
is seen for meso-ferrocenyl porphyrinoids, where the hybrid
functional overestimates the π-macrocycle contributions to the
HOMO and is quite unexpected for the OLYP ap-
proach.40,42,45−47 The correct distribution of frontier orbitals
is well described by the BP86 functional, as illustrated in Figure
5, while in Figure S6, the MO energy diagram is reported. Also,
Figure 6. UV−visible spectrum of TFcCorrCo(PPh3) in CH2Cl2.
related cobalt triarylcorrolates,49,50 and the formation of
TFcCorrCo(PPh3) was confirmed by the mass spectrum,
which exhibited a molecular peak at m/z 905, corresponding to
the target triferrocenylcorrole after loss of the axial PPh3 ligand.
Figure 5. Frontier MO calculated for TFcCorrCu at the BP86/6-
31G(d,p) level of calculation.
1
The H NMR spectrum of TFcCorrCoPPh3 is shown in
the meta GGA functional TPSSTPSS correctly allows for
describing the orbital shapes in comparison with the
experimental results. In fact, the HOMO is characterized by a
strong interaction between the core, the Fc units, and the metal,
as is seen from the experimentally measured absorption spectra.
On the other hand, the unoccupied orbitals [lowest unoccupied
molecular orbital (LUMO, LUMO+1, and LUMO+2)] are
mainly centered on the macrocycle core.
Figure 7, while Table S1 contains the observed chemical shifts
and coupling constants. The assignments were obtained by a
combination of 1H−1H COSY, 1H−1H ROESY, 1H−13C
1
HSQC, and H−13C HMBC. An interesting feature in the
spectra is the fact that two distinct signals are observed for the
hydrogen and carbon atoms in position 2 of the 5/15-position
Fc groups. The nonequivalence between these two positions is
likely due to a restricted rotation of the Fc substituents in
combination with a differentiation between the two faces of the
corrole ring caused by the PPh3 substituent. As a consequence,
the two H-2 protons couple with each other, which is readily
In particular, the electronic properties are strongly influenced
by substituents at the positions 5 and 15, and this can be clearly
observed by analyzing the orbital contribution to the UV−
E
DOI: 10.1021/acs.inorgchem.5b01575
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
1
Figure 7. H NMR spectrum of TFcCorrCo(PPh3) in CDCl3.
observed in the COSY spectrum (Figure S7) and yields two
quartets, instead of the expected triplets that are normally
observed for Fc protons of the substituted ring. On the other
hand, only one signal is observed for protons of the 10-Fc
group, for which a fast rotation is hypothesized. This suggests a
particular conformation of the PPh3 axial substituent that
causes a larger steric hindrance for rotation of the Fc groups in
positions 5 and 15 than in position 10. In fact, we observe in
the ROESY spectrum a very short distance between the o-PPh3
and β-pyrrolic protons in position 2/18. In this particular
conformation, with one of the three Fc rings pointing toward
the position 2/18, the 10-Fc is located in a relatively free region
and rotates faster than the other two meso-Fc groups.
These results led us to investigate the geometry optimiza-
tions and MO calculations on TFcCorrCo(PPh3), using the
functional that gave a correct description of the copper
complex, e.g., pure BP86. Quite surprisingly, the most stable
atropoisomer has all three Fc groups on the same side of the
molecule as the phosphine axial ligand, despite the steric
hindrance potentially connected to this geometrical shape (see
Figures 8, 9, and S6). This feature suggests some interaction
Figure 9. Frontier orbitals calculated for TFcCorrCo(PPh3) at the
BP86/6-31G(d,p) level of calculation.
calculations of the orbital distribution for TFcCorrCo(PPh3)
fit well the experimental results, and an interaction between the
different Fc groups and the macrocycle is clearly visible.
The reaction of TFcCorrH3 with nickel ions gave different
and quite unexpected results. In an initial synthetic attempt to
obtain the target molecule, the reaction mixture containing the
free-base corrole and excess Ni(AcO)2 was reacted at 80 °C in
DMF because Fc derivatives are not stable at temperature
higher than 90−100 °C. This procedure yielded mainly a green
fraction, whose UV−visible spectrum was not characteristic of
metalloporphyrin or metallocorrole (Figure S9). Formation of
the desired TFcCorrNi was definitely ruled out by the 1H NMR
spectrum, which revealed the presence of a diamagnetic
nickel(II) species, instead of a paramagnetic species, expected
for TFcCorrNi. An initially intriguing result was the presence of
only two Fc rings in the compound, as determined by signals in
the 4−5 ppm region of the 1H NMR spectrum. The presence of
10 signals (1 H each) between 5.9 and 8.2 ppm revealed the
low symmetry of the molecule. The presence of high-field
resonances of the β-pyrrolic protons indicated the absence of
an aromatic macrocyclic system, which could be either due to
Figure 8. Molecular geometry calculated for TFcCorrCo(PPh3) at the
BP86/6-31G(d,p) level of calculation.
among the axial ligand and Fc groups, but future studies are
needed to elucidate the nature of the interaction. A strong
interaction between the meso-Fc groups, the central cobalt ion,
and the corrole macrocycle can, however, be observed, and also
in this case, the MOs are well described with the BP86
functional. Like in the case of TFcCorrCu, theoretical
F
DOI: 10.1021/acs.inorgchem.5b01575
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 10. Structures of possible nickel isocorrole derivatives (a and b) and DFcLTPNi.
the formation of an isocorrole species,51,52 as shown by the
structures in Figure 10a,b, or due to a linear tetrapyrrolic
complex.53 The mass spectrum showed a molecular peak at
m/z+ 711, a value that did not match an isocorrole structure.
A definitive structure of the nickel derivative was then
confirmed by NMR analysis as the open-chain complex
DFcLTPNi illustrated in Figure 2 (NMR data are reported in
Table S2). The 1H NMR spectrum of DFcLTPNi (Figure S10)
shows the presence of only two Fc substituents. The COSY
spectrum of DFcLTPNi in the region of 5.9−8.2 ppm shows
the presence of four spin systems, two comprising two
hydrogen atoms, and two others with three hydrogen atoms
(Figure S11). This indicates that two of pyrrolic rings on the
voltammograms in this solvent are illustrated in Figure 11, and
a summary of the E1/2 values for each redox reaction under
1
compound bear three hydrogen atoms. Analysis of the H−1H
ROESY spectrum allowed for the assignment of all β-pyrrolic
signals and corroborated the proposed structure shown in
Figure 10. In this experiment, a very intense cross-peak was
observed between the hydrogen atoms in positions 14 and 15,
in line with their short distance in the proposed structure.
The measured coupling constants are shown in Table S3 and,
in the case of the two rings with three hydrogen atoms, follow
the expected trend for the J values observed for pyrrole (3J3,4
>
54
4
1
3J2,3 > J2,4). Chemical shifts for the H and 13C nuclei, based
on COSY, ROESY, 1H−13C HSQC, and 1H−13C HMBC
experiments, are shown in Table S2.
The formation of such a product is quite unexpected because
the opening of the corrole macrocycle is usually accompanied
by oxidation to an open-chain derivative, leading to the
formation of a biliverdine-like compound55,56 when it occurs at
the pyrrole−pyrrole direct link. When the ring opening has
been observed at the 5 position, a benzoyl group was the
substituent observed at the terminal α-pyrrolic position.57 In
our case, the ring opening is not oxidative and no oxygen
insertion has been observed at the peripheral position of the
linear tetrapyrrole. Attempts were made to coordinate nickel
ion by using mild reaction conditions and Ni(AcO)2 in
refluxing CH2Cl2/CH3OH, but the only isolable reaction
product was again DFcLTPNi and the expected triferrocenyl-
corrole complex was not isolated.
Figure 11. Cyclic voltammograms of (Fc)x(Ph)3−xCorrCu, where x =
0−3 in PhCN, containing 0.1 M TBAP at a scan rate of 0.1 V/s.
these solution conditions is given in Table 2. For comparison
purposes, potentials for oxidation and reduction of the related
non-Fc derivative, 5,10,15-triphenylcorrole (TPCorr), are also
given in the figure and table.
The reversible redox reactions in Figure 11 can be assigned
to three sites of electron addition or electron abstraction, as
indicated in the figure. One site of electron transfer involves the
macrocycle, which is reversibly oxidized at E1/2 values between
0.77 and 1.11 V. The second is oxidation of the Fc groups,
Electrochemistry of Copper Complexes. The electro-
chemistry of each copper ferrocenylcorrole was examined in
PhCN containing 0.1 M tetrabutylammonium perchlorate
(TBAP) as the supporting electrolyte. Examples of cyclic
G
DOI: 10.1021/acs.inorgchem.5b01575
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
FcDPhCorrCu (x = 1) is more difficult by 100 mV than the
measured E1/2 value for the oxidation of free Fc under the same
solution conditions (E1/2 = 0.48 V). The reversible oxidation of
Table 2. Half-Wave Potentials (V vs SCE) for Copper
Ferrocenylcorroles in PhCN, Containing 0.1 M TBAP
red
one Fc group on DFcPhCorrCu (x = 2) is located at E1/2
=
ox
(CuIII/
CuII)
0.50 V, and this potential is also more positive than the E1/2
value of unlinked Fc in this solvent. Only in the case of
TFcCorrCu is the Fc/Fc+ process easier than that of the free
Fc. This first oxidation of TFcCorrCu occurs at E1/2 = 0.45 V in
PhCN, and a plot of the half-wave potential for the first Fc-
centered oxidation versus the number of Fc groups on the
compound is again linear, as shown in Figure 12.
DFcPhCorrCu (x = 2) undergoes three one-electron
oxidations, the first two of which are ascribed to the oxidation
of Fc substituents at the 5 and 15 positions of the macrocycles.
The second Fc group oxidation on DFcPhCorrCu (x = 2) is
harder (occurs at a more positive potential) than the first by
160 mV, and the separation in potential between the two Fc/
Fc+ processes is consistent with an interaction between the two
equivalent redox-active groups of the complex.61−63
(macrocycle)
ox (Fc)
a
compound
#Fc
first
third second first
first
TPCorrCu
0
1
2
3
0.77
0.92
1.02
1.11
−0.19
−0.22
−0.27
−0.33
FcDPhCorrCu
DFcPhCorrCu
TFcCorrCu
0.58
0.66
0.67
0.50
0.45
0.67
a
Number of Fc groups on the macrocycle.
which occurs at E1/2 values between 0.45 and 0.67 V, and the
third site of electron transfer involves reduction of the
copper(III) central metal ion to its copper(II) form at E1/2
−0.19 to −0.33 V.
The neutral corrole formally contains copper(III), and the
first reduction process can then be described as electron
addition to the central metal ion, forming TFcCorrCuII as a
final reduction product. The alternate assignment of the neutral
compound as a copper(II) corrole π-cation radical39,40,42−44
cannot be ruled out, although the exact nature of the
assignment has no bearing on the product of the one-electron
reduction, which unambiguously contains copper(II) in each
compound.
=
Three reversible oxidations of TFcCorrCuIII can be observed
within the positive potential window of PhCN. These occur at
E1/2 = 0.45, 0.67, and 1.11 V, respectively. The current height
for the second oxidation of TFcCorrCu at E1/2 = 0.67 V is twice
that of the other processes (see Figure 11), indicating two
overlapping one-electron oxidations at the same potential. As
indicated above, the unlinked Fc molecule exhibits a reversible
one-electron oxidation at E1/2 = 0.48 V in PhCN, and this
potential can be compared to the E1/2 = 0.45 V for the first one-
electron oxidation of TFcCorrCu, where the process is
proposed to involve electron abstraction from the Fc group
located at the meso position 10 of the macrocycle.
The process at E1/2 = 0.67 for TFcCorrCu (x = 3) involves
two overlapping one-electron oxidations and is assigned to the
abstraction of one electron from the equivalent Fc groups at the
meso positions 5 and 15 of the macrocycle. The fact that the
half-wave potential for the first oxidation of TFcCorrCuIII (0.45
V) is very close to the E1/2 value of Fc itself in PhCN (0.48 V)
is consistent with the spectroscopic and theoretical data that
indicate little communication between this Fc group at the
meso position 10 of the macrocycle and the conjugated π-ring
system of the corrole. On the other hand, the potentials for
oxidation of the other two Fc groups at the meso positions 5
and 15 are positively shifted by 190 mV from the E1/2 value for
the oxidation of free Fc, and this is consistent with a strong
interaction occurring between the two unoxidized meso-Fc
substituents and the positively charged Fc+ group, generated in
the redox process at 0.45 V. From this point of view, Fc-based
copper and cobalt corroles (see the later section of the text)
behave similarly to the corresponding copper (and cobalt)
complex of TFcPorph, where a certain degree of electronic
communication has been observed by electrochemical meas-
ures.64 Anyway it should be noted that, in the case of porphyrin
derivatives, the potentials recorded for each Fc oxidation have
been obtained using a different solvent/supporting electrolyte
system.
A second reduction of TFcCorrCu is also observed at E1/2
=
−2.03 V in PhCN and −1.97 V in DMF (see Figure S12). This
process involves electron addition to the conjugated macro-
cycle, forming a copper(II) corrole π-anion radical. This
reduction behavior is similar to that of other previously
investigated copper corrole derivatives.35,58−60 Fc is a strong
electron-donating group, and thus the stepwise replacement of
each meso-phenyl group in TPCorrCu by a Fc substituent
results in a systematic negative shift in the potential for the first
reduction, whose E1/2 value is directly proportional to the
number of Fc groups on the compound. This is graphically
shown in Figure 12, where the reversible first reduction
potential shifts from −0.19 V for the copper(III)/copper(II)
reaction of TPCorrCu to −0.33 V for the same electrode
reaction of TFcCorrCu.
A systematic negative shift in E1/2 is also seen for the first
one-electron oxidation of each ferrocenylcorrole as the number
of Fc groups on the macrocycle is increased, from one in the
case of FcDPhCorrCu to three in the case of TFcCorrCu. The
potential of 0.58 V for the reversible conversion of Fc to Fc+ on
The first macrocycle-centered oxidation of the triphenylcor-
role TPCorrCu is observed at E1/2 = 0.77 V in PhCN, and this
potential can be compared to a similar macrocycle-centered
oxidation of the triferrocenylcorrole complex, TFcCorrCu, at
E1/2 = 1.11 V. The presence of the three electron donating
meso-Fc groups might be expected to lead to easier oxidations
of TFcCorrCu compared to those of TPCorrCu, but the effect
of the electron-donating substituents is counterbalanced by the
Figure 12. Relationship between the measured redox potentials of
copper corroles and the number of Fc groups on the molecule.
H
DOI: 10.1021/acs.inorgchem.5b01575
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
positive charges on the ferrocenium groups of the singly,
doubly, and triply oxidized complexes, thus resulting in much
harder ring-centered oxidations of the three ferrocenylcorrole
derivatives compared to the same electrode reaction of
TPCorrCu, which is uncharged when converted to a corrole
π-radical cation at 0.77 V. Each added positive charge shifts E1/2
for oxidation of the macrocycle toward a more positive
potential, and a plot of E1/2 for this process versus the number
of Fc groups is linear, as shown in Figure 12. The difference in
the potential between the macrocycle-centered oxidation of
TPCorrCu and the same redox reaction of TFcCorrCu
amounts to 340 mV, and as described later in the manuscript,
the same positive shift of the potential is observed for the two
structurally related cobalt complexes.
Spectra of Electroreduced and Electrooxidized Cop-
per Corroles. UV−visible spectra were monitored during the
first reduction and first oxidation of each copper corrole, in
order to further elucidate the site of electron transfer and also
evaluate the effect of the meso-Fc substituents on the UV−
visible spectra of the compounds in their neutral, singly reduced
and singly oxidized forms. Examples of the relevant cyclic
voltammograms for the four copper complexes are given in
Figure 11, and the relevant spectral changes, obtained in the
thin-layer cell during the first one-electron addition and first
one-electron abstraction, are shown in Figures 13 and 14,
respectively.
The spectral changes in Figure 13 are in each case consistent
with a copper(III)/copper(II) process. The formal conversion
of copper(III) corrole to its copper(II) form is accompanied by
a 22 nm red shift in the Soret band position for TPCorrCu
(Figure 13a), and a 18−23 nm red shift is also seen for the
three Fc derivatives (Figure 13b−d), with the main difference
between the four sets of spectral changes being the generation
of copper(II) corrole characterized by a decreasing intensity
and broadened Soret band with increase in the number of Fc
groups on the compound. All four electrogenerated copper(II)
corroles possess two moderately intense Q bands, as seen in the
figure, with the most intense band shifting more and more
toward the red with each added Fc group. The exact
wavelengths of the Soret and Q bands for the neutral and
electroreduced copper corroles are summarized in Table 3.
Of most significance are the spectral changes seen upon the
first one-electron oxidation of the same series of corroles in
PhCN. The first oxidation of each compound involves a
reversible one-electron abstraction (as seen by cyclic voltammo-
grams in Figure 11), but the site of electron transfer is different
in the two series of compounds, being ring-centered in the case
of TPCorrCu and Fc-centered in the case of FcDPhCorrCu,
DFcPhCorrCu, and TFcCorrCu. As such, one might anticipate
a radical-type UV−visible spectrum in the case of singly
oxidized TPCorrCu and very few spectral changes to occur for
the other singly oxidized compounds, where the reaction
involves a meso-Fc substituent on the molecule. However, this is
not what was experimentally observed, as seen in Figure 14,
where a significant loss of the Soret band intensity is observed
upon each one-electron oxidation, along with a loss of the Q
band in the case of DFcPhCorrCu and TFcCorrCu (Figure
14c,d). The large decrease of the Soret band intensity after
oxidation of a single Fc substituent indicates a significant
transfer of positive charge from the singly oxidized ferrocenium
group to the π-ring system of the corrole. The spectral changes
accompanying the oxidation of unlinked Fc are given in Figure
S13 and are characterized by the presence of a new absorption
Figure 13. UV−visible spectral changes during controlled potential
reduction of (a) TPCorrCu, (b) FcDPhCorrCu, (c) DFcPhCorrCu,
and (d) TFcCorrCu to their singly reduced forms in PhCN containing
0.1 M TBAP. The applied reducing potential was −0.60 V in each case.
at about 600 nm. This band is not observed in the spectrum of
singly oxidized FcDPhCorrCu, DFcPhCorrCu, or TFcCorrCu,
again indicating a significant coupling between the positive
charge on Fc+ and the π-ring system of the macrocycle.
Electrochemistry of Cobalt Complexes. The mechanism
for reduction of TFcCorrCo(PPh3) is similar to that of
previously investigated cobalt corroles containing a bound PPh3
axial ligand.60−64 The first reduction is quasi-reversible and
involves the conversion of TFcCorrCoIII(PPh3) to [TFcCorr-
CoII(PPh3)]−, followed by rapid dissociation of the axially
bound PPh3 ligand to generate the anionic four-coordinate
cobalt(II) corrole, [TFcCorrCoII]− (an electrochemical EC
mechanism). The four-coordinate cobalt(II) corrole can be
further reduced to its cobalt(I) form at −1.71 V, as shown by
I
DOI: 10.1021/acs.inorgchem.5b01575
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
chemical EC mechanism to give first a transient four-coordinate
TFcCorrCoIII species, followed by re-formation of the neutral
five-coordinate cobalt(III) corrole with a coordinated triphe-
nylphosphine axial ligand, as shown in Scheme 1.
Although the mechanism shown in Scheme 1 for
TFcCorrCoIII(PPh3) is the same as that previously reported
for TPCorrCoIII(PPh3),15,40,65−67 there is not a large effect of
the three Fc groups on the peak potential for conversion of
cobalt(III) to cobalt(II). As seen in Figure 16, the first
reduction peak potential has been shifted from Ep = −0.75 V
for the TPCorr complex in PhCN to −0.78 V for the TFcCorr
derivative in the same solvent. The second metal-centered
reduction of cobalt corrole is also insensitive to the electron-
donating Fc substituents, as indicated by the fact that
TFcCorrCoII and TPCorrCoII are both reversibly reduced to
their CoI form at an identical potential of E1/2 = −1.71 V. In
contrast, the expected substituent effect is seen in the peak
potential for reoxidation of the singly reduced corrole. This
potential shifts from Epc = −0.10 V for TPCorr to −0.26 V for
TFcCorr.
It is important to note that potentials for the three Fc-
centered oxidations of TFcCorrCo(PPh3) are almost identical
with the E1/2 values for the three Fc oxidations of TFcCorrCu.
The first electron abstraction from TFcCorrCo(PPh3) is
proposed to involve the Fc group at position 10 and occurs
at E1/2 = 0.42 V; the following two-electron abstractions are
then proposed to occur at the Fc groups located at positions 5
and 15, and this redox process occurs at E1/2 = 0.62 V. The
separation between the E1/2 values for the two sets of Fc-
centered oxidations is 200 mV, which is very similar to ΔE1/2
=
220 mV for the same two Fc-centered redox processes of
TFcCorrCu. However, as seen in Figure 15, the first and
second oxidations of TFcCorrCo(PPh3) have similar current
heights, thus suggesting that the first oxidation process actually
involves two overlapping one-electron transfers. We propose
that one electron in the first step is abstracted from the Fc
group (at the 10 position) and the other from the metal center
of the cobalt corrole [a cobalt(III)/cobalt(IV) process] at the
same potential of 0.42 V. Two additional overlapping one-
electron transfers, both involving the unoxidized Fc groups, are
then observed at E1/2 = 0.62 V, and this is followed by the
stepwise one-electron transfers of the conjugated macrocycle at
0.92 and 1.28 V, respectively. It should be noted that the 330−
360 mV shift in E1/2 between the two macrocycle-centered
oxidations of TPCorrCo(PPh3) and TFcCorrCo(PPh3) is
similar to the earlier described 340 mV separation between
the related redox reactions of TPCorrCu and TFcCorrCu.
Spectra of Electroreduced and Electrooxidized Cobalt
Corroles. A comparison of the spectral changes that occur
Figure 14. UV−visible spectral changes during the first one-electron-
controlled potential oxidation of (a) TPCorrCu, (b) FcDPhCorrCu,
(c) DFcPhCorrCu, and (d) TFcCorrCu in PhCN containing 0.1 M
TBAP. The site of electron transfer is the macrocycle for TPCorrCu
(#Fc = 0) and a meso-Fc substituent for the other compounds.
the cyclic voltammogram in Figure 15. The unligated cobalt(II)
corrole can also be irreversibly reoxidized via an electro-
Table 3. UV−Visible Spectral Data (λmax and log ε) of Neutral, Singly Oxidized and Singly Reduced CorrCuIII in PhCN
Containing 0.1 M TBAP
λmax, nm (log ε, M−1 cm−1
)
a
[CorrCuIII]+
Soret band
CorrCuIII
[CorrCuII]−
Q band
compound
#Fc
Soret band
Q band
Soret band
TPCorrCu
0
1
2
3
426 (4.89)
418 (4.51)
425 (4.59)
428 (4.67)
414 (5.15)
421 (4.91)
425 (4.87)
428 (4.80)
624 (3.86)
634 (4.05)
671 (4.09)
680 (4.19)
436 (5.23)
444 (4.98)
443 (4.80)
451 (4.73)
579 (4.29)
586 (4.00)
600 (4.02)
610 (3.97)
610 (4.47)
626 (4.22)
654 (4.35)
671 (4.32)
FcDPhCorrCu
DFcPhCorrCu
TFcCorrCu
a
In the absence of Fc substituents (#Fc = 0), the first oxidation occurs on the macrocycle, while in all other cases (#Fc = 1, 2, or 3), a Fc substituent
is the site of electron abstraction.
J
DOI: 10.1021/acs.inorgchem.5b01575
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 15. Cyclic voltammograms of TPCorrCoIII(PPh3) and TFcCorrCoIII(PPh3) in PhCN containing 0.1 M TBAP at a scan rate of 0.1 V/s.
Scheme 1. Proposed Mechanisms for Reduction of
TFcCorrCoIII(PPh3) in PhCN Containing 0.1 M TBAP
products possess a Q band that is also red-shifted by 23−24 nm
compared to the initial compounds in their cobalt(III) form.
These spectral changes are as expected for a metal-centered
electron transfer, in this case cobalt(III) to cobalt(II).
Similar types of spectral changes are seen during the global
two-electron oxidation of TFcCorrCo(PPh3) at E1/2 = 0.42 V
and the one-electron oxidation of TPCorrCo(Ph3) at 0.56 V. A
metal-centered CoIII/CoIV process has been suggested in the
case of the triphenylcorrole,49 and the spectral changes shown
in Figure 17 are consistent with this assignment. The fact that
similar spectral changes are observed upon the two-electron
Figure 16. UV−visible spectral changes during the controlled
potential reduction of (a) TPCorrCoIII(PPh3) and (b) TFcCorr-
CoIII(PPh3) in PhCN containing 0.1 M TBAP.
upon conversion of cobalt(III) to cobalt(II) in the
triferrocenyl- and triphenylcorrole derivatives is shown in
Figure 16. The Soret band shifts to the red by 35−36 nm for
both compounds upon reduction, and both cobalt(II) corrole
Figure 17. UV−visible spectral changes during the controlled
potential oxidation of (a) TPCorrCoIII(PPh3) and (b) TFcCorr-
CoIII(PPh3) in PhCN containing 0.1 M TBAP.
K
DOI: 10.1021/acs.inorgchem.5b01575
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Scheme 2. Proposed Mechanism for the oxidation of TFcCorrCu and TFcCorrCo(PPh3) in PhCN Containing 0.1 M TBAP
oxidation of TFcCorrCo(PPh3) and the one-electron oxidation
of TPCorrCo(PPh3) can be interpreted in terms of one
electron being abstracted from the central cobalt ion and the
other from one of the three Fc groups on the triferrocenyl
compound. A band close to 600 nm, which can be assigned to
Fc+, is not detected in the spectrum of the electrooxidized
cobalt corrole, but the loss of the intense Q band in the
electrooxidized product provides further evidence for con-
version of one Fc group to its Fc+ form.
AUTHOR INFORMATION
Corresponding Authors
*Tel: +1 (713) 743-2740. Fax: +1 (713) 743-2745. E-mail:
kkadish@uh.edu (K.M.K.).
*Tel: +39 (06) 72594752. Fax: +39 (06) 72594328. E-mail:
roberto.paolesse@uniroma2.it (R.P.).
■
Notes
The authors declare no competing financial interest.
In summary, the proposed overall mechanism for the
oxidation of TFcCorrCo(PPh3) and TFcCorrCu is given in
Scheme 2.
ACKNOWLEDGMENTS
■
This work is dedicated to Professor Barbara Floris on the
occasion of her retirement. Thanks are given to PRIN 2010−
2011 (to P.G.) and to the Robert A. Welch Foundation (Grant
E-680 to K.M.K.).
CONCLUSIONS
■
We report herein a synthetic pathway that allows the
preparation of metal derivatives of TFcCorr, together with a
detailed study of the corresponding electrochemical and
spectral properties of the compound in their neutral, singly
oxidized and singly reduced forms. Theoretical, spectroscopic,
and electrochemical characterization of the copper and cobalt
ferrocenyl derivatives demonstrates that Fc groups at the 5 and
15 positions of the macrocycle interact more strongly with the
corrole π-ring system than does the Fc substituent at the 10
position. DFT calculations indicated the up−down−up con-
former as the most stable atropoisomer of TFcCorrCu,
although free rotation of the Fc groups around the meso-
carbon atoms is observed by NMR. On the other hand, DFT
calculations of TFcCorrCo(PPh3) indicate that the most stable
conformer has all three Fc groups on the same side of the
macrocyclic plane as the PPh3 ligand, while NMR character-
ization of the compound in solution shows that steric hindrance
due to the PPh3 group hampers rotation of the Fc units at the 5
and 15 positions of the macrocycle. Attempts to prepare the
nickel complex of TFcCorr following the same synthetic
pathway as that for the copper or cobalt derivatives gave a
different result, showing an opening of the corrole ring, with a
loss of a methine bridge to give a diferrocenyl-substituted linear
nickel tetrapyrrole complex.
REFERENCES
■
(1) Battersby, A. R.; McDonald, E. In Porphyrins and Metal-
loporphyrins; Smith, K. M., Ed.; Elsevier: Amsterdam, The Nether-
lands, 1975; pp 61−122.
(2) Che, C. M.; Lo, V. K.-Y.; Zhou, C.-Y.; Huang, J.-S. Chem. Soc. Rev.
2011, 40, 1950−1975.
(3) Paolesse, R.; Monti, D.; Nardis, S.; Di Natale, C. In The
Handbook of Porphyrin Sciences; Kadish, K. M., Smith, K. M., Guilard,
R., Eds.; World Scientific Publishing Co.: Singapore, 2011; Vol 12, pp
121−225.
(4) Guldi, D. M. Chem. Soc. Rev. 2002, 31, 22−36.
́
(5) Lang, K.; Mosinger, J.; Wagnerova, D. M. Coord. Chem. Rev.
2004, 248, 321−350.
(6) Ormond, A. B.; Freeman, H. S. Materials 2013, 6, 817−840.
(7) Walter, M. G.; Rudine, A. B.; Wamser, C. C. J. Porphyrins
Phthalocyanines 2010, 14, 759−792.
(8) Vecchi, A.; Galloni, P.; Floris, B.; Nemykin, V. N. J. Porphyrins
Phthalocyanines 2013, 17, 165−196.
(9) Vecchi, A.; Galloni, P.; Floris, B.; Dudkin, S.; Nemykin, V. N.
Coord. Chem. Rev. 2015, 291, 95−171.
(10) Nardis, S.; Monti, D.; Paolesse, R. Mini-Rev. Org. Chem. 2005, 2,
355−374.
(11) Barata, J. F. B.; Santos, C. I. M.; Neves, M. G. P. M. S.; Amparo,
M. A. F.; Cavaleiro, J. A. S. Synthesis and Modification of
PorphyrinoidsTopics in Heterocyclic Chemistry; Paolesse, R., Ed.;
Springer: New York, 2014; Vol. 33, pp 79−114.
(12) Liu, H. Y.; Mahmood, M. H. R.; Qiu, S. X.; Chang, C. K. Coord.
Chem. Rev. 2013, 257, 1306−1333.
ASSOCIATED CONTENT
■
(13) Lemon, C. M.; Dogutan, D. K.; Nocera, D. G. In The Handbook
of Porphyrin Sciences; Kadish, K. M., Smith, K. M., Guilard, R., Eds.;
World Scientific Publishing Co.: Singapore, 2011; Vol 21, pp 1−144.
(14) Schechter, A.; Stanevsky, M.; Mahammed, A.; Gross, Z. Inorg.
Chem. 2012, 51, 22−24.
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.5b01575.
(15) Ou, Z.; Lu, A.; Meng, D.; Huang, S.; Fang, Y.; Lu, G.; Kadish, K.
̈
Electrochemistry of DFcLTPNi, 1H NMR, FAB MS data,
M. Inorg. Chem. 2012, 51, 8890−8896.
(16) Lei, H.; Fang, H.; Han, Y.; Lai, W.; Fu, X.; Cao, R. ACS Catal.
2015, 5, 5145−5153.
1
2D H−1H COSY and UV−visible spectra, and cyclic
voltammograms of compounds (PDF)
L
DOI: 10.1021/acs.inorgchem.5b01575
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(17) Tortora, L.; Pomarico, G.; Nardis, S.; Martinelli, E.; Catini, A.;
D’Amico, A.; Di Natale, C.; Paolesse, R. Sens. Actuators, B 2013, 187,
72−77.
(18) Walker, D.; Chappel, S.; Mahammed, A.; Brunschwig, B. S.;
Winkler, J. R.; Gray, H. B.; Zaban, A.; Gross, Z. J. Porphyrins
Phthalocyanines 2006, 10, 1259−1262.
(19) Flamigni, L.; Gryko, D. T. Chem. Soc. Rev. 2009, 38, 1635−1646.
(20) Salvatori, P.; Amat, A.; Pastore, M.; Vitillaro, G.; Sudhakar, K.;
Giribabu, L.; Soujanya, Y.; De Angelis, F. Comput. Theor. Chem. 2014,
1030, 59−66.
(43) Alemayehu, A. B.; Conradie, J.; Ghosh, A. Eur. J. Inorg. Chem.
2011, 1857−1864.
(44) Thomas, K. E.; Alemayehu, A. B.; Conradie, J.; Beavers, C. M.;
Ghosh, A. Acc. Chem. Res. 2012, 45, 1203−1214.
(45) Vecchi, A.; Erickson, N. R.; Sabin, J. R.; Floris, B.; Conte, V.;
Venanzi, M.; Galloni, P.; Nemykin, V. N. Chem. - Eur. J. 2015, 21,
269−279.
(46) Nemykin, V. N.; Hadt, R. G. J. Phys. Chem. A 2010, 114,
12062−12066.
(47) Zhang, L.; Qi, D.; Zhang, Y.; Bian, Y.; Jiang, J. J. Mol. Graphics
Modell. 2011, 29, 717−725.
(21) Barata, J. F. B.; Zamarron
M. A. F.; Tome,
Rodriguez, F. Eur. J. Med. Chem. 2015, 92, 135−144.
(22) Haber, A.; Gross, Z. Chem. Commun. 2015, 51, 5812−5827.
(23) Ghosh, A.; Wondimagegn, T.; Parusel, A. B. J. J. Am. Chem. Soc.
2000, 122, 5100−5104.
́
, A.; Neves, M. G. P. M. S.; Faustino,
́
́
(48) Canard, G.; Gao, D.; D'Aleo, A.; Giorgi, M.; Dang, F.-X.;
́
A. C.; Cavaleiro, J. A. S.; Roder, B.; Juarranz, A.; Sanz-
̈
Balaban, T. S. Chem. - Eur. J. 2015, 21, 7760−7771.
(49) Paolesse, R.; Licoccia, S.; Bandoli, G.; Dolmella, A.; Boschi, T.
Inorg. Chem. 1994, 33, 1171−1176.
(50) Huang, S.; Fang, Y.; Lu, A.; Lu, G.; Ou, Z.; Kadish, K. M. J.
̈
Porphyrins Phthalocyanines 2012, 16, 958−967.
(24) Stefanelli, M.; Nardis, S.; Tortora, L.; Fronczek, F. R.; Smith, K.
M.; Licoccia, S.; Paolesse, R. Chem. Commun. 2011, 47, 4255−4257.
(25) Palmer, J. H. Struct. Bonding (Berlin, Ger.) 2011, 142, 49−90.
(26) Day, P.; Hush, N. S.; Clark, R. J. H. Philos. Trans. R. Soc., A
2008, 366, 5−14.
(51) Nardis, S.; Pomarico, G.; Fronczek, F. R.; Vicente, M. G. H;
Paolesse, R. Tetrahedron Lett. 2007, 48, 8643−8646.
(52) Pomarico, G.; Xiao, X.; Nardis, S.; Paolesse, R.; Fronczek, F. R.;
Smith, K. M.; Fang, Y.; Ou, Z.; Kadish, K. M. Inorg. Chem. 2010, 49,
5766−5774.
(27) Hadt, R. G.; Nemykin, V. N. Inorg. Chem. 2009, 48, 3982−3992.
(28) Pomarico, G.; Vecchi, A.; Mandoj, F.; Bortolini, O.; Cicero, D.
O.; Galloni, P.; Paolesse, R. Chem. Commun. 2014, 50, 4076−4078.
(53) Wojaczyn
2013, 69, 10445−10449.
(54) Balazs, Y. S.; Saltsman, I.; Mahammed, A.; Tkachenko, E.;
Golubkov, G.; Levine, J.; Gross, Z. Magn. Reson. Chem. 2004, 42, 624−
635.
́ ́
ski, J.; Duszak, M.; Latos-Grazynski, L. Tetrahedron
̇
́ ̌
́
(29) Kral, V.; Vasek, P.; Dolensky, B. Collect. Czech. Chem. Commun.
2004, 69, 1126−1136.
(30) Nemykin, V. N.; Galloni, P.; Floris, B.; Barrett, C. D.; Hadt, R.
G.; Subbotin, R. I.; Marrani, A. G.; Zanoni, R.; Loim, N. M. Dalton
Trans. 2008, 4233−4246.
(55) Tardieux, C.; Gros, C. P.; Guilard, R. J. Heterocycl. Chem. 1998,
35, 965−970.
(56) Paolesse, R.; Sagone, F.; Macagnano, A.; Boschi, T.; Prodi, L.;
Montalti, M.; Zaccheroni, N.; Bolletta, F.; Smith, K. M. J. Porphyrins
Phthalocyanines 1999, 3, 364−370.
(31) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci,
B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H.
P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima,
T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin,
K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega,
N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.;
Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.;
Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.;
Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09;
Gaussian, Inc.: Wallingford, CT, 2009.
(57) Nardis, S.; Mandoj, F.; Paolesse, R.; Fronczek, F. R.; Smith, K.
M.; Prodi, L.; Montalti, M.; Battistini, G. Eur. J. Inorg. Chem. 2007,
2345−2352.
(58) Kadish, K. M.; Adamian, V. A.; Van-Caemelbecke, E.; Gueletii,
E.; Will, S.; Erben, C.; Vogel, E. J. Am. Chem. Soc. 1998, 120, 11986−
11993.
(59) Stefanelli, M.; Mandoj, F.; Mastroianni, M.; Nardis, S.; Mohite,
P.; Fronczek, F. R.; Smith, K. M.; Kadish, K. M.; Xiao, X.; Ou, Z.;
Chen, P.; Paolesse, R. Inorg. Chem. 2011, 50, 8281−8292.
(60) Lu, G.; Lin, W.; Fang, Y.; Zhu, W.; Ji, X.; Ou, Z. J. Porphyrins
Phthalocyanines 2011, 15, 1265−1274.
(61) Guilard, R.; Gros, C. P.; Barbe, J.-M.; Espinosa, E.; Jerome, F.;
Tabard, A.; Latour, J.-M.; Shao, J.; Ou, Z.; Kadish, K. M. Inorg. Chem.
2004, 43, 7441−7455.
(62) Barbe, J.-M.; Habermeyer, B.; Khoury, T.; Gros, C. P.; Richard,
P.; Chen, P.; Kadish, K. M. Inorg. Chem. 2010, 49, 8929−8940.
(63) Fang, Y.; Mandoj, F.; Zeng, L.; Pudi, R.; Stefanelli, M.; Paolesse,
R.; Kadish, K. M. J. Porphyrins Phthalocyanines 2015, 19, 388−397.
(64) Rohde, G. T.; Sabin, J. R.; Barrett, C. D.; Nemykin, V. N. New J.
Chem. 2011, 35, 1440−1448.
(65) Adamian, V. A.; D'Souza, F.; Licoccia, S.; Di Vona, M. L.;
Tassoni, E.; Paolesse, R.; Boschi, T.; Kadish, K. M. Inorg. Chem. 1995,
34, 532−540.
(66) Li, B.; Ou, Z.; Meng, D.; Tang, J.; Fang, Y.; Liu, R.; Kadish, K.
M. J. Inorg. Biochem. 2014, 136, 130−139.
(67) Kadish, K. M.; Koh, W.; Tagliatesta, P.; Sazou, D.; Paolesse, R.;
Licoccia, S.; Boschi, T. Inorg. Chem. 1992, 31, 2305−2313.
(32) Paolesse, R.; Marini, A.; Nardis, S.; Froiio, A.; Mandoj, F.;
Nurco, D. J.; Prodi, L.; Montalti, M.; Smith, K. M. J. Porphyrins
Phthalocyanines 2003, 7, 25−36.
(33) Koszarna, B.; Gryko, D. T. J. Org. Chem. 2006, 71, 3707−3717.
(34) Steene, E.; Dey, A.; Ghosh, A. J. Am. Chem. Soc. 2003, 125,
16300−16309.
(35) Ou, Z.; Shao, J.; Zhao, H.; Ohkubo, K.; Wasbotten, I. H.;
Fukuzumi, S.; Ghosh, A.; Kadish, K. M. J. Porphyrins Phthalocyanines
2004, 8, 1236−1247.
(36) Bruckner, C.; Brinas, R. P.; Bauer, J. A. K. Inorg. Chem. 2003, 42,
̈
4495−4497.
(37) Luobeznova, I.; Simkhovich, L.; Goldberg, I.; Gross, Z. Eur. J.
Inorg. Chem. 2004, 1724−1732.
(38) Wasbotten, I. H.; Wondimagegn, T.; Ghosh, A. J. Am. Chem.
Soc. 2002, 124, 8104−8116.
(39) Alemayehu, A. B.; Gonzalez, E.; Hansen, L. K.; Ghosh, A. Inorg.
Chem. 2009, 48, 7794−7799.
(40) Alemayehu, A.; Conradie, M. M.; Ghosh, A. J. Porphyrins
Phthalocyanines 2012, 16, 695−704.
(41) Nemykin, V. N.; Rohde, G. T.; Barrett, C. D.; Hadt, R. G.;
Sabin, J. R.; Reina, G.; Galloni, P.; Floris, B. Inorg. Chem. 2010, 49,
7497−7509.
(42) Pierloot, K.; Zhao, H.; Vancoillie, S. Inorg. Chem. 2010, 49,
10316−10329.
M
DOI: 10.1021/acs.inorgchem.5b01575
Inorg. Chem. XXXX, XXX, XXX−XXX