All Four Atropisomers of Iron Tetra(o- N, N, N-trimethylanilinium)porphyrin in Both the Ferric and Ferrous States

Article abstract of DOI:10.1021/acs.inorgchem.1c00236

Electrostatic effects are key to many biological and (electro)chemical transformations, especially those that involve charged species. The position and orientation of the electric field with respect to the molecules undergoing charge rearrangement are often crucial to the progress of the reaction. Recently, several molecular (electro)catalysts have been designed to contain spatially positioned charged groups that can engage in specific intramolecular electrostatic interactions. For instance, iron complexes of the tetra(o-N,N,N-trimethylanilinium)porphyrin ligand, which has four cationic groups, have been used to great effect for both CO2 and O2 reduction. Because of the ortho-substitution pattern on the porphyrin ligand, there are four possible atropisomers - such as the αβαβ isomer with trimethylanilinium groups on alternating faces of the porphyrin - and thus four unique electrostatic environments. This study details the synthesis and characterization (1H NMR spectroscopy, single crystal X-ray diffraction, and cyclic voltammetry) of these four metalloporphyrin isomers in both the ferric (FeIII) and ferrous (FeII) forms by using a synthetic route that preserves atropisomeric purity. The atropisomers are different in some respects but show remarkable similarities in others, such as their reduction potentials. This study also shows that the widely-cited literature method used previously to prepare the molecular electrocatalyst for CO2 and O2 reduction yields a mixture of atropisomers rather than a single one, as was previously assumed. These results identify the ways in which intra- and intermolecular electrostatic effects affect both solution and solid-state properties as well underscoring the challenges associated with preparing metalloporphyrins with high atropisomeric purity.

Full text of DOI:10.1021/acs.inorgchem.1c00236

pubs.acs.org/IC
Article
All Four Atropisomers of Iron
Tetra(o‑N,N,N‑trimethylanilinium)porphyrin in Both the Ferric and
Ferrous States
Daniel J. Martin, Brandon Q. Mercado, and James M. Mayer*
Cite This: Inorg. Chem. 2021, 60, 5240−5251
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: Electrostatic effects are key to many biological and (electro)-
chemical transformations, especially those that involve charged species. The
position and orientation of the electric field with respect to the molecules
undergoing charge rearrangement are often crucial to the progress of the
reaction. Recently, several molecular (electro)catalysts have been designed to
contain spatially positioned charged groups that can engage in specific
intramolecular electrostatic interactions. For instance, iron complexes of the
tetra(o-N,N,N-trimethylanilinium)porphyrin ligand, which has four cationic
groups, have been used to great effect for both CO2 and O2 reduction. Because
of the ortho-substitution pattern on the porphyrin ligand, there are four
possible atropisomerssuch as the αβαβ isomer with trimethylanilinium
groups on alternating faces of the porphyrinand thus four unique
electrostatic environments. This study details the synthesis and character-
ization (1H NMR spectroscopy, single crystal X-ray diffraction, and cyclic voltammetry) of these four metalloporphyrin isomers in
both the ferric (FeIII) and ferrous (FeII) forms by using a synthetic route that preserves atropisomeric purity. The atropisomers are
different in some respects but show remarkable similarities in others, such as their reduction potentials. This study also shows that
the widely-cited literature method used previously to prepare the molecular electrocatalyst for CO2 and O2 reduction yields a
mixture of atropisomers rather than a single one, as was previously assumed. These results identify the ways in which intra- and
intermolecular electrostatic effects affect both solution and solid-state properties as well underscoring the challenges associated with
preparing metalloporphyrins with high atropisomeric purity.
binding15effects not often emphasized in studies that used
INTRODUCTION
■
highly charged molecular designs.
Electrostatic and electric field effects are increasingly recognized
as key to the success of many challenging, multistep reactions,
Traditionally, charged groups have been added to macrocyclic
ligand designs to improve solubility, especially solubility in
aqueous solutions.16,17 Pyridinium-, carboxylate-, and sulfonate-
especially those that involve charged intermediates or significant
charge redistribution.1−3 Such complex reactions are common
derivatized macrocycles are the most common examples of these
in molecular electrocatalysis, which often involves single
designs, where solubility is controlled, at least in part, by the pH
electron or proton transfer steps and the formation of charged
intermediates.4−7 Stabilizing these intermediates and decreasing
kinetic barriers are required to improve reaction rates and
efficiencies. Recently, these goals have prompted the design of
molecular (electro)catalysts that contain spatially positioned
charged groups which can stabilize charged intermediates via
electrostatic interactions.6,8−14
One such design is the tetracationic 5,10,15,20-tetra(o-
N,N,N-trimethylanilinium)porphyrin ligand (o-TMA), which
has four positive charges positioned around the porphyrin ring.
Iron complexes of this ligand are among the leading molecular
electrocatalysts for both CO2 and O2 reduction in terms of
reaction rates and efficiencies.10,12 The success of this catalyst is
due in part to the stabilization of pre-equilibria that involve
anionic ligands or ligands that become sufficiently anionic upon
of the solution. Alkylated pyridinium and ammonium functional
groups offer a more permanent form of charge installation and
have been used in porphyrin designs to facilitate aqueous O2/
CO binding and superoxide dismutase studies.18−22 There are a
few examples of highly charged ligand designs (8+/8−) affecting
basic physicochemical properties of metalloporphyrins, but
these studies have only probed charge-symmetric systems in
aqueous solvents.23−27 There are no prior systems insofar as we
Received: January 24, 2021
Published: March 22, 2021
https://doi.org/10.1021/acs.inorgchem.1c00236
© 2021 American Chemical Society
Inorg. Chem. 2021, 60, 5240−5251
5240

Inorganic Chemistry
pubs.acs.org/IC
Article
a
Scheme 1. Four Different Atropisomers Available to [FeIII(o-TMA)](OTf)5
a
The stick figures in the second row represent the different orientations of the o-[N(CH3)3]+ groups.
Scheme 2. Synthesis Route Used to Prepare the [FeIII(o-TMA)](OTf)5 Atropisomers
can find that analyze metal−macrocycles with different
asymmetric charge distributions in nonaqueous solvents.19
For this reason, the (o-TMA) ligand and corresponding metal
complexes are highly unusual. By nature of the mono-ortho
substitution pattern on the aryl rings and the restricted
rotational freedom at the porphyrin meso-carbons, there are
four atropisomers available to the (o-TMA) ligandαβαβ,
ααββ, αααβ, and ααααand thus four unique electrostatic
the reported method;12 however, bulk atropisomeric purity was
not established.
Here, we report improved synthetic and separation
procedures for isolating all four atropisomers of Fe(o-TMA)
in both the ferric (FeIII) and ferrous (FeII) forms. Each of the
atropisomers was fully characterized by using 1H NMR
spectroscopy, high-resolution mass spectrometry, and cyclic
voltammetry. Seven of the eight molecules were characterized by
single-crystal X-ray diffraction. The ways in which intra- and
intermolecular electrostatics affect and do not affect the
molecular and solid-state properties are identified and discussed.
Moreover, we show that the previous synthetic route reported
́
environments (Scheme 1). Saveant and co-workers reported the
first preparation of Fe(o-TMA) for CO2 reduction and claimed
the successful synthesis of the ferric αβαβ atropisomer, in which
the cationic functional groups alternate on either side of the
porphyrin ring.10 The characterization data, however, were
limited only to infrared spectroscopy, UV−vis absorbance,
elemental analysis, and mass spectrometry techniques that
cannot necessarily distinguish between the atropisomers.
Crystallographic data from our group showed that the αβαβ
isomer was indeed a component of the product synthesized via
́
by Saveant et al. led to unwanted rotamerization of the aryl
groups and scrambling of the isomers, resulting in a mixture of
atropisomeric catalysts.
RESULTS
■
Synthesis. The [FeIII(o-TMA)](OTf)5 atropisomers can be
obtained in six steps by using commercially available reagents
5241
https://doi.org/10.1021/acs.inorgchem.1c00236
Inorg. Chem. 2021, 60, 5240−5251

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 1. (A) Partial 1H NMR spectra in CD3CN of the αβαβ, ααββ, αααβ, and αααα atropisomers of [FeIII(o-TMA)](OTf)5 (see the Supporting
Information for full spectra). The downfield regions are enhanced by 8× for clarity. The resonances for the β-pyrrolic and N(CH3)3+ protons are
identified with black circles and triangles, respectively; the sharp peak at 3.5 ppm is from Et2O. (B) Partial 1H NMR spectra of the ferrous compounds,
showing the region containing the N(CH3)3+ protons. The αααβ spectra show ca. 10% of the ααββ atropisomer, as shown by the peaks at ∼4.0 ppm in
(A) and 3.18 ppm in (B). (C) High-resolution mass spectra and simulated spectrum of the [Fe(o-TMA)(OTf)]4+ cation (C57H60N8FeO3SF3) in the
samples from (A). Full spectra are available in the Supporting Information.
and without heating later-stage intermediates (Scheme 2). The
synthesis reported below is different than the more cumbersome
and lower-yielding method previously used to prepare the ferric
and ferrous forms of the αβαβ atropisomer in ref 15. Following
the literature,10 the target products first require the atropisomers
of tetra(o-aminophenyl)porphyrin, H2(o-AMP) (Figures S5−
S8). These were obtained from the acid-catalyzed condensation
of o-nitrobenzaldehyde and pyrrole, followed by reduction of the
corresponding tetra(o-nitrophenyl)porphyrin with stannous
chloride/hydrochloric acid and repeated chromatography.28
Several chromatography conditions and eluents have been
reported for isolating the various atropisomers of H2(o-
AMP).10,28−31 After trying several of these conditions, we
found that the most consistent method of obtaining
atropisomerically pure samples (>95%) required a minimum
of three separate columns with various eluent mixtures
(conditions reported in the Experimental Methods section).
The atropisomeric purity of the target [FeIII(o-TMA)](OTf)5
salts was dictated by this early stage chromatography, so it was
imperative that the atropisomers of tetra(o-aminophenyl)-
porphyrin were carefully separated and that they were not
heated to avoid isomerization to the other atropisomers (see
below).28
The FeIIICl(o-DMA) compounds were then converted to the
respective hydroxo complexes, FeIIIOH(o-DMA), by dissolving
them in DCM and stirring with 1 M NaOH(aq) for 30 min.33
This ligand substitution more consistently yielded the desired
pentatriflate salts in the final methylation step. The hydroxo
form of the αααα atropisomer rapidly hydrolyzed to form the
1
corresponding μ-oxo dimer (by H NMR spectroscopy and
MS), which was the isolated product after chromatography. The
other isomers did not form μ-oxo dimers, presumably due to
steric bulk on both sides of the porphyrin ring.
Finally, the FeIIIOH(o-DMA) complexes (and μ-oxo dimer of
the αααα isomer) were quaternized to the target [FeIII(o-
TMA)](OTf)5 molecules by using excess methyl triflate in
trimethyl phosphate containing a few drops of 2,6-di-tert-
butylpyridine.15 The sterically bulky base was key to the success
of the reaction, as it sequestered disadvantageous triflic acid
present in the commercial methyl triflate.18 After stirring (12 h
at 20 °C), excess methyl triflate was quenched with methanol,
and the products were precipitated by adding the reaction
mixture dropwise into stirring Et2O. Quenching the methyl
triflate with methanol generated triflic acid, which protonated
the hydroxo ligands and hydrolyzed the μ-oxo dimer of the
αααα isomer (see the Supporting Information). The product of
this reaction was quantitative and yielded the corresponding
pentatriflate salts (Figures S20−S23). The crude solids were
slowly recrystallized from MeCN/Et2O mixtures in a glovebox,
and the crystalline samples of the [FeIII(o-TMA)](OTf)5
The individual H2(o-AMP) atropisomers were then methy-
lated by reductive amination using formaldehyde and sodium
cyanoborohydride to yield the respective tetra(o-N,N-dimethyl-
1
aminophenyl)porphyrins, H2(o-DMA). The H NMR spectra
1
for these molecules were diagnostic but typically contained
minor components (<5%) that could not be separated by
chromatography (see Figures S10−S13).
isomers were characterized by H NMR spectroscopy, single-
crystal X-ray crystallography, and cyclic voltammetry (see below
and the Supporting Information).
The corresponding iron(III) chloride tetra(o-N,N-dimethyl-
aminophenyl)porphyrins, FeIIICl(o-DMA), were prepared in
70−80% yields at 20 °C via transmetalation of the
corresponding dilithium porphyrin complexes.32 The forest
green dithilium materials were (i) generated in situ by reacting
the H2(o-DMA) isomers with 2 equiv of lithium hexamethyl-
disilazide in THF and then (ii) reacted with ferrous bromide
(FeBr2·2THF). Iron insertion proceeds more readily with the
lithiated porphyrins than with the free base analogues, likely a
result of forming LiCl or LiBr, which are poorly soluble in THF.
The corresponding iron(II) atropisomers were prepared by
stirring the iron(III) salts over Zn(Hg) amalgam in the
glovebox.15,34−36 Within an hour, the reactions were complete,
and the solutions had lightened in color from maroon to cherry
red. After filtering and rinsing the amalgams, we recrystallized
the iron(II) porphyrin-containing solutions by vapor diffusion,
and the products were isolated as blocky purple crystals. These
crystalline solids were also characterized by 1H NMR spectros-
copy (Figures S29−S32) and single-crystal X-ray crystallog-
raphy (see below and the Supporting Information).
1
The H NMR spectra of the product metalloporphyrins were
1H NMR Spectra of the Fe(III) and Fe(II) Porphyrin
Salts. The ferric (FeIII) pentatriflate atropisomers each had a
unique, paramagnetic 1H NMR spectrum, with signals
corresponding to the β-pyrrolic, aromatic, and trimethyl-
anilinium protons (Figure 1A and Figures S20−S23). The β-
broad due to the paramagnetism of the iron center. Chloride
binding introduced additional asymmetry, which was identified
in the diagnostic pyrrole region of the spectra (75−85 ppm; see
the Supporting Information).
5242
https://doi.org/10.1021/acs.inorgchem.1c00236
Inorg. Chem. 2021, 60, 5240−5251

Inorganic Chemistry
pubs.acs.org/IC
Article
pyrrolic protons were assigned by integration (8H) and were
typically the most downfield signals (ranging from 13 to 50
ppm). These signals appeared as broad singlets for the αβαβ and
αααα atropisomers and as a set of two overlapping singlets for
the ααββ isomer, in which there are two different sets of
pyrroles. The β-pyrrolic signal for the αααβ atropisomer was
broad and unsymmetric. The aromatic protons were less
downfield (ranging from 7 to 16 ppm) and were generally
sharper. The spectra of the αβαβ, ααββ, and αααα atropisomers
each contained four unique aromatic peaks (4H per peak; one
signal was broad and downfield), while the spectrum of the
αααβ atropisomer contained a more complicated set of peaks
(16H total). The 36 protons that corresponded to the
trimethylanilinium groups were the most upfield signals in the
spectra (1−5 ppm). These protons appeared as broad singlets
for each of the αβαβ, ααββ, and αααα atropisomers and as three
broad singlets (1:1:2 ratio) in the αααβ spectrum.
Figure 2. Isomerization profiles for the rotamerization of [FeIII(o-
TMA)](OTf)5, H2(o-DMA), and H2(o-AMP) at 80 °C. The solvent
was CD3CN for [FeIII(o-TMA)](OTf)5 and CDCl3 for H2(o-AMP)
and H2(o-DMA) porphyrins. H NMR spectra are available in the
Supporting Information.
1
1
The ferrous (FeII) atropisomers had H NMR spectra in
atropisomer was also stable at these higher temperatures (Figure
S37).
The rapid isomerization rate for the o-N(CH3)2-substituted
porphyrin was surprising and led us to question the
atropisomeric fidelity of the original synthesis reported by
CD3CN that were diamagnetic with no evidence of remaining
paramagnetic impurities, showing complete reduction (Figures
1
S29−S32). Like the ferric complexes, the H NMR spectra
contained signals that corresponded to the β-pyrrolic (8H),
aromatic (16H), and trimethylanilinium (36H) protons, which
were assigned by relative integrations (Figure 1B).
́
Saveant et al. The original synthesis required extended heating
of αβαβ FeIIICl(o-DMA) during the final methylation step (24 h
at 100 °C in DMF).10 These conditions far exceed the
temperatures and times that were shown to cause rotamerization
in this work. To probe whether atropisomeric purity could be
preserved under these harsh conditions, the reported synthetic
procedures were repeated by using an isolated sample of the
1
The H NMR spectra of the product porphyrins show their
atropisomeric purity. For the αβαβ, ααββ, and αααα
porphyrins, the synthesis method described in this report
allowed for preparation with >95% isomeric purity. While the
actual purities are likely higher for these atropisomers, we report
1
these values as lower limits due to limitations of H NMR
1
αβαβ FeIIICl(o-DMA) precursor. After work-up, the H NMR
integration. The αααβ complex was isolated with >90% purity,
with the ααββ atropisomer accounting for nearly all the
spectrum of the product was compared to the genuine spectra of
the atropisomers isolated in this work. Rather than the singular
αβαβ atropisomer, as was reported, the product was a mixture of
isomers (Figure 3). The αβαβ and αααβ isomers made up
1
remaining signal in the H NMR spectrum. As expected, the
mass spectra are indistinguishable and cannot be used to identify
any individual isomer (Figure 1C).
Thermal Atropisomer Rotamerization. The rates of
tetraarylporphyrin rotamerization have been documented for
several ortho-substituted porphyrins, including the H2(o-AMP)
isomers used in this work.30,31,37−39 Generally, rotamerization
rates increase with temperature and decrease when sterically
bulky groups are added at the ortho-position of the aryl rings.37,39
Here, the relative rates of isomerization for the αβαβ
atropisomers of H2(o-AMP), H2(o-DMA), and [FeIII(o-TMA)]-
1
(OTf)5 were measured using a H NMR time course (Figures
S34−S36). A solution of each molecule was prepared in
deuterated solvent, loaded into a J. Young tube, and heated to 80
°C using a preheated oil bath for 48 h with regular spectra being
collected. A portion of the aromatic region of the spectra was fit
by using MestReNova to yield the percent αβαβ isomer
remaining at each time point (see the Supporting Information),
which are plotted in Figure 2.
1
Figure 3. Partial H NMR spectrum of the [FeIII(o-TMA)](OTf)5
́
product obtained by using the synthesis conditions reported by Saveant
et al.10 The isomers are identified by the respective pyrrolic peaks (8H)
which were integrated by using MestReNova. Spectrum was recorded
in CD3CN at 400 MHz. Full spectra are available in the Supporting
Information.
As shown in Figure 2, both the αβαβ H2(o-AMP) and H2(o-
DMA) porphyrins rotamerize with similar time profiles and
approach the theoretical limit (12.5%) expected for the
statistical mixture of isomers.28 At 80 °C, the half-life of both
reactions is <0.5 h, and complete isomerization was reached
within 6 h. While rapid isomerization of H2(o-AMP) was
expected at 80 °C, it was surprising that the more sterically
encumbered H2(o-DMA) isomerized just as quickly. In contrast,
there was no evidence of isomerization for the αβαβ
atropisomer of [FeIII(o-TMA)](OTf)5 under these conditions,
even with additional heating to 100 °C for 48 h (Figure S36).
The more sterically encumbered αααα [FeIII(o-TMA)](OTf)5
approximately equal fractions (40% and 38%, respectively),
followed by the ααββ (17%), and αααα (5%) atropisomers
(Figures S27 and S28). These data show that original synthesis
does not yield a single isomer, as was assumed, but rather a
mixture of all four atropisomers. The ramifications of this are
discussed below.
Single-Crystal X-ray Characterization. Crystals suitable
for single-crystal X-ray diffraction were obtained for both the
ferric and ferrous forms of the αβαβ, ααββ, and αααα
atropisomers and for the ferrous-only form of the αααβ isomer
5243
https://doi.org/10.1021/acs.inorgchem.1c00236
Inorg. Chem. 2021, 60, 5240−5251

Inorganic Chemistry
pubs.acs.org/IC
Article
Figure 4. Single-crystal X-ray structures of the αβαβ, ααββ, and αααα atropisomers in both the ferric and ferrous forms and the αααβ isomer in the
ferrous form only. In the ferric structures on the top, the multiple disordered orientations of the single bound triflate ligand are shown. Fe, orange; N,
blue; C, white; H atoms and triflates that were not bound were omitted for clarity; thermal ellipsoids shown at 50% probability. The αβαβ [FeII(o-
TMA)(CH3CN)2] structure is reported in ref 15.
by vapor diffusion of Et2O into MeCN (Figure 4). The ferrous
αβαβ structure has already been reported15 and is repeated here
for reference. A bis-aquo complex of the ferric αβαβ porphyrin
was reported in ref 12, but the triflate-bound structure reported
here is new. Five triflate anions were identified in each of the
ferric porphyrin crystal structures, and four were identified in
each of the ferrous structures. Many of the triflate anions were
disordered and had to be modeled, as described in the
Supporting Information. The crystals obtained for the ferric
αααβ atropisomer were too disordered for single-crystal studies.
All seven structures have ligands bound to the iron center. For
the three ferric structures, a single triflate was bound to the
metal. The α and β faces of the αβαβ and ααββ structures are
symmetry equivalent; thus, there is no site selectivity for the
bound triflate. The αααα atropisomer, however, has inequiva-
lent sides with different steric and electrostatic environments. In
the solid-state structure, a triflate ligand was bound to the more
crowded, more cationic α face, and a water molecule was bound
to the β face (Figure 4). The triflate being bound to the α face is
the opposite of what one might expect based on sterics. The α
face of the αααα atropisomer is by far the most congested site
Figure 5. (left) Packing structure of the ferric αααα [FeIII(o-
TMA)(H2O)(OTf)](OTf)4 complex. (right) Packing structure of
the ferrous αααα [FeII(o-TMA)(OTf)](OTf)3 complex. The layered
structure has the α-faces (with the anilinium groups) oriented toward
one another in a repeating βα|αβ pattern, with the triflates concentrated
between the α faces. All the atoms of the different orientations of the
disordered triflates are shown. Color coding: C, gray; N, blue; O, red; S,
orange; F, light green; Fe, dark orange; H atoms omitted for clarity.
across the series of structures in this work and is intuitively the
least likely site for a large anion to bind. The structures of the
ferrous αβαβ, ααββ, and αααβ complexes each had two
acetonitrile ligands bound to the iron. The αααα isomer did not
have any acetonitrile ligands; rather, a triflate ligand was bound
to the metal, again on the α-face.
The packing of anions and cations in the αβαβ, ααββ, and
αααβ structures showed repeating units of metalloporphyrin
with triflate molecules distributed near the trimethylanilinium
groups. In contrast, both the ferric and ferrous forms of the
αααα atropisomer packed as a bilayer structure with a densely
packed layer of triflates appearing between the α-faces (Figure
5). In the ferric structure, the β faces are parallel and separated
by 5.17 Å, longer than typical porphyrin π−π interaction
distances.40 The ferrous structure has nonparallel β faces and no
evidence of π−π interactions.
Electrochemistry. Cyclic voltammograms (CVs) of the four
atropisomers were measured in acetonitrile (MeCN) containing
0.1 M tetrabutylammonium hexafluorophosphate [n-Bu4N]-
[PF6]. The voltammograms were internally referenced to
decamethylferrocene (Me10Fc), which was independently
5244
https://doi.org/10.1021/acs.inorgchem.1c00236
Inorg. Chem. 2021, 60, 5240−5251

Inorganic Chemistry
pubs.acs.org/IC
Article
a
Table 1. Reduction Potentials for the Four Atropisomers of Fe(o-TMA) for Fe(p-TMA) and for Fe(TPP)
structural isomer
atropisomer
solvent
E
1/2(FeIII/FeII)
E
1/2(FeII/FeI)
E
1/2(FeI/Fe0)
Fe(o-TMA)
αβαβ
ααββ
αααβ
αααα
average
αβαβ
ααββ
αααβ
αααα
average
MeCN
MeCN
MeCN
MeCN
MeCN
DMF
DMF
DMF
DMF
DMF
0.142
0.143
0.130
0.135
0.14 0.01
−0.351
−0.337
−0.341
−1.194
−1.201
−1.200
−1.187
−1.20 0.01
−1.166
−1.198
−1.194
−1.195
−1.594
−1.640
−1.632
−1.635
−1.63 0.02
−1.683
−1.705
−1.698
−1.695
−1.70 0.01
−0.329
−0.34 0.01
−0.089
−1.19 0.01
−1.316
a
[Fe(p-TMA)]
MeCN
DMF
ca. −1.8
b
b
b
−0.55
−1.40
−1.92
d
Fe(TPP)
PrCN
DMF
−0.259
−0.61
e
e
e
−1.45
b
−2.07
a
In MeCN or DMF containing 0.1 M [n-Bu4N][PF6]. Potentials ( 0.005 V) referenced vs Fc+/Fc. This redox feature was broad and poorly
c
reversible. Reported value equal to the midpoint potential from the maximum and minimum current responses. Reported values in DMF for
[Fe(p-TMA)](Cl)5.46 Reported in n-butyronitrile containing 0.1 M [n-Bu4N][PF6] for [Fe(TPP)]OTf.15 Reported for Fe(TPP)Cl in ref 46.
d
e
referenced to ferrocene (Fc) by using a separate solution (to
avoid overlaps between E1/2(Fc+/0) and metalloporphyrin redox
features). Each of the atropisomers showed three reversible
reductions, which were assigned to the corresponding FeIII/FeII,
“FeII/FeI”, and “FeI/Fe0” redox couples, as is typical for iron
porphyrins in nonaqueous solvent (Table 1 and Figure 6).35,41
positive than the corresponding values for Fe(p-TMA), a control
molecule bearing para-[N(CH3)3]+ groups (Table 1). Yet, while
the magnitudes of the positive shifts are unusual, the Fe(o-
TMA) E1/2 values were quite similar between the four
atropisomers. The range of E1/2(FeIII/FeII) and E1/2(FeII/FeI)
values was only ∼10 mV across the series. The range was ∼40
mV for the E1/2(FeI/Fe0) values; however, this larger deviation
was due only to the positively shifted E1/2(FeI/Fe0) of the αβαβ
atropisomer. In N,N-dimethylformamide (DMF) containing 0.1
́
M electrolyte, the same conditions reported by Saveant et al. in
ref 10, the E1/2(FeI/Fe0) values span only 22 mV. Under these
conditions, three of the four isomers have E1/2(FeI/Fe0) values
that are within the uncertainty of the measurement (Table 1).
DISCUSSION
■
Effects of Oriented Charges on the Properties of Fe(o-
TMA). The spectroscopic, electrochemical, and structural data
for the four Fe(o-TMA) atropisomers provide unusual insights
into the role of the orientation of electrostatic groups on
molecular properties. In contrast to most studies of electrostatics
in small molecule and inorganic chemistries, this work yields
information about the effects of the relative position of charges
without changing their type or number or distance from the
metal center.
By nature of the o-TMA ligand design, there are four potential
atropisomers available to the single iron porphyrin complex,
each of which has a unique symmetry and electrostatic
environment. The αβαβ (D2d) isomer has the highest point
group symmetry of the series, followed by the ααββ (C2h), αααα
(C4v), and αααβ (Cs) atropisomers. The following discussion is
divided into two sections that highlight how these unique
electrostatic environments affect or do not affect the electro-
chemical and solid-state structural properties.
Effects of Oriented Charges on Electrochemical Data. The
electrochemistry of the four atropisomers was almost completely
unaffected by the orientation of the o-[N(CH3)3]+ groups. The
isomers have very similar potentials, both in MeCN and in DMF.
This is a surprising result. For instance, the αααα isomer should
have a substantial dipole moment along the 4-fold axis while the
αβαβ isomer has no net dipole moment (ignoring any
counterion or ligand binding). These results do not imply the
Figure 6. Cyclic voltammograms for the four atropisomers of [FeIII(o-
TMA)](OTf)5 in MeCN containing 0.1 M [n-Bu4N][PF6]. All
voltammograms were collected at 0.1 V s−1 and referenced to Fc+/Fc.
The iron(I) and iron(0) labels are in quotation marks to indicate
that these complexes may involve a significant amount of ligand-
centered reduction.42,43 Characterization of the low-valent
species was not pursued in this work.
As a result of the four cationic, electron-withdrawing o-
[N(CH3)3]+ groups, each of the respective redox couples has
E
1/2 values that are several hundred millivolts more positive than
those typical for iron porphyrin complexes in polar organic
solvents (acetonitrile, n-butyronitrile, and N,N-dimethylforma-
mide, among others).41,44,45 Despite similar inductive effects,
the various E1/2 values are also more than 0.1−0.2 V more
5245
https://doi.org/10.1021/acs.inorgchem.1c00236
Inorg. Chem. 2021, 60, 5240−5251

Inorganic Chemistry
pubs.acs.org/IC
Article
charged groups are unimportant, only that their relative
orientations with respect to the metal center do not significantly
affect the electrochemistry.
The inductive effects of the [N(CH3)]+ groups should be similar
in Fe(o-TMA) and Fe(p-TMA), yet the ortho isomers have
substantially more positive reduction potentials, by +0.23 V
(ΔE1/2[FeIII/FeII]), +0.12 V (ΔE1/2[FeII/FeI]), and +0.17 V
(ΔE1/2[FeI/Fe0]) in MeCN. Similar trends and values were also
observed in DMF (Table 1): +0.21 V (ΔE1/2[FeIII/FeII]), +0.21
V (ΔE1/2[FeII/FeI]), and +0.22 V (ΔE1/2[FeI/Fe0]).15,41
An electrostatic, charged sphere model offers a partial
explanation of the higher potential for Fe(o-TMA) vs Fe(p-
TMA). For such a model, we consider each of these complexes
as positively charged spheres and the electron being added to the
surface of the sphere. While this is a very crude model, it is a
good first approximation because a spherically symmetric charge
density behaves as a point charge concentrated at the center.
The change in potential energy (ΔU) to move a charge (q1)
from infinite distance to the surface of a sphere of charge q2 is
given by eq 1, where ε0 is the permittivity of the vacuum and ε is
the static dielectric constant of the medium. The distance r is the
radius of the sphere, the distance from the center to the surface.
For the chemistry being analyzed here, q1 = −1e for the electron
being added, and q2 for the ferric complexes is +5e, for both
Fe(o-TMA) and Fe(p-TMA). The only difference between
these compounds in this model is the size of the sphere that
encloses most of the charge, which is clearly smaller for Fe(o-
TMA) and Fe(p-TMA) (Scheme 3). By approximating the
diameter of each sphere as the distance between the N atoms of
the [N(CH3)3]+ groups on the phenyl rings in the 5- and 15-
meso positions, we estimate that the radius for the Fe(o-TMA)
and Fe(p-TMA) complexes is rortho ≅ 5.5 Å and rpara ≅ 9.3 Å (see
Supporting Information Section VI for all distances). Thus, the
more compact charges of the ortho isomer should yield a larger
electrostatic effect than the four charges in the larger para
isomer, as observed experimentally.
The lack of variation between the atropisomer E1/2 values
indicates that the energy required to bring a negative point
charge (e.g., the e−) from infinity to a polycationic, quasi-
spherical species is largely unaffected by the precise orientation
of the charges within the cation. Rather, the overall energetics
chiefly concern the addition of a monoanion to a compact
polycation. These data also indicate that the electrochemical
double layers surrounding the respective atropisomers (e.g.,
[OTf]− or [PF6]− anions from the electrolyte) are not
sufficiently different that they affect the reduction thermochem-
istry.
The most important function of the four o-[N(CH3)3]+
groups on the electrochemistry is to make the molecules easier
to reduce (more positive E1/2 values). For applications in
electrocatalysis, more positive E1/2 values imply smaller
overpotentials.12 This increase in E1/2 over related compounds
like the neutral Fe(TPP) complex is observed for all three redox
couples: FeIII/FeII, “FeII/FeI”, and “FeI/Fe0” (Table 1).
The more positive E1/2 values could result from either
electrostatic, through-space effects or inductive, through-bond
effects, or some combination of the two. In a previous study, we
estimated the inductive effects from the four [N(CH3)3]+
groups account for roughly half of the difference between the
E
1/2(FeIII/FeII) values of the αβαβ isomer and Fe(TPP) in n-
butyronitrile.15 To understand the electrostatic component, we
compare the E1/2 values of the Fe(o-TMA) isomers with the
structural para-substituted isomer, Fe(p-TMA) (Scheme 3).
Scheme 3. Electrostatic, Charged Sphere Model for the
Reduction of Fe(o-TMA) and Fe(p-TMA) Structural Isomers
q1q2
1
ΔU =
4πϵϵ0 r
(1)
This electrostatic picture provides a qualitative estimate of the
expected effect. Using q2 = +5, rortho ≅ 5.5 Å, rpara ≅ 9.3 Å, and
εMeCN = 3847 (though it might be larger with 0.1 M
[nBu4N][PF6]48), we estimate that the FeIII/FeII reduction
potential is larger by ∼0.13 V for the ortho-isomer because of its
higher charge density. For such a simple model, this is
remarkably close to the 0.2−0.1 V observed experimentally.
This simple model is not as successful at predicting other
trends in the electrochemical data. One might have expected
that the αααα isomer would have a more compact sphere of
charge than the αβαβ isomer and therefore a larger electrostatic
contribution. However, the potentials for these isomers are quite
close, and the αααα isomer has less positive than the larger αβαβ
isomer. Of course, a key piece missing from this model is the
direct binding of anions such as triflate to the iron center, as
observed crystallographically. It seems likely that such anion
binding occurs to a different extent for the different
atropisomers. As another example, eq 1 and intuition predict
that the electrostatic effects should be greater for reduction of
the FeIII complex, a 5+ cation, than for the 3+ FeI species (q2 =
5+ vs 3+ in eq 1). Yet such a trend is hard to discern in the data in
Table 1, either in MeCN or in DMF. Perhaps this effect is too
small to be observed comparing different redox couples, where
some of the assumptions in the simple spherical charge model
may not hold very well.
We advocate using this simple model not to make quantitative
predictions, but as a starting point to build intuition and to
5246
https://doi.org/10.1021/acs.inorgchem.1c00236
Inorg. Chem. 2021, 60, 5240−5251

Inorganic Chemistry
pubs.acs.org/IC
Article
identify when other effects such as anion binding play important
roles. It is remarkable, with all the different effects involved, that
the reduction potentials are almost invariant to the atropisomer
identity in both MeCN and DMF (Table 1). This is the case for
the FeIII/FeII, “FeII/FeI”, and “FeI/Fe0” transformations even
though the lower-valent complexes may involve significant
charge transfer to the redox noninnocent porphyrin ring. The
similar potentials suggest that the simple electrostatic hard-
sphere model may capture a significant part of the physics
involved. Future work will explore the role(s) of ion pairing and
the “double layer” around these polycationic complexes.
Effects of Oriented Charges on Crystallographic Data. In
contrast to the electrochemical data, the crystallographic
structures were distinct for each isomer and indicate that
electrostatic interactions control both the primary coordination
environment and packing structure of the solids. The ferric and
ferrous structures of the αααα atropisomer are the most
indicative of these effects. In both the ferric and ferrous
complexes, the αααα isomer shows a bound triflate ligand to the
more crowded α face and a layered packing structure. These data
contrast the structures obtained for the three remaining isomers,
which have triflate ligands bound only to the ferric structures
and do not pack in layers.
The difference in primary coordination environments for the
different atropisomers, and between the respective ferric and
ferrous structures, is evidence to the strength of local,
intramolecular electrostatic interactions between the bound
triflate ligand and the o-[N(CH3)3]+ groups. For the αβαβ,
ααββ, and αααβ atropisomers, only the ferric structures have a
bound triflate ligand. The ferrous structures, which have a
smaller overall charge and a softer metal ion,49 have two bound
acetonitrile ligands. In contrast, both the ferric and ferrous forms
of the αααα atropisomer have a triflate ligand that is bound to
the sterically crowded α-face. These data indicate not only that
local, short-range interactions exist between the triflate ligand
and the nearby o-(N(CH3)3)+ cations on the α-face of the αααα
atropisomer but also that this electrostatic attraction overcomes
the intrinsic preference of the ferrous ion for soft acetonitrile
ligands. Yet the αααβ isomer, with three o-[N(CH3)3]+ groups
on the same side and lower steric hindrance, does not sufficiently
stabilize triflate as to overcome the intrinsic ligand preferences.
Local electrostatic interactions also impact the packing
structure of the solids. In both the ferric and ferrous forms of
the αβαβ, ααββ, and αααβ atropisomers, the polycationic iron
porphyrin complex[FeIII(o-TMA)(OTf)]4+ or [FeII(o-
TMA)(CH3CN)2]4+is surrounded by four triflate anions
that are evenly distributed around each 4+ cation. The 3D lattice
of alternating ions is reminiscent of crystal packing for common
ionic solids. In contrast, the ferric and ferrous αααα solids pack
in dense, 2D layers of cations and anions. This unique packing
emphasizes the strength of the local electrostatic interactions
that exist between the alternating layers of Fe(o-TMA) α-faces
and triflate anions, which persists despite the change in overall
charge of the cation: [FeIII(o-TMA)(OTf)]4+ vs [FeII(o-
TMA)(OTf)]3+.
used to prepare this catalyst results in a mixture of Fe(o-TMA)
atropisomers and not just the claimed αβαβ product.
The unwanted rotamerization is caused by the extended
heating during the final methylation step: the published
procedure involves heating the αβαβ FeCl(o-DMA) precursor
at 100 °C for 24 h. The studies reported here show that these
conditions result in a mixture that contains only 40% of the
desired αβαβ isomer. Still, this 40% is substantially more than
the 12.5% expected from a statistical mixture, and we have
shown that αβαβ [FeIII(o-TMA)](OTf)5 does not isomerize at
100 °C. Therefore, rotamerization of the αβαβ FeCl(o-DMA)
precursor must occur with rates that are commensurate with the
methylation reaction under the reported conditions. Rotameri-
zation can only be avoided by using milder conditions, such as
those described in this work.
The data reported here rationalize why this misidentified
catalyst product was not detected in ref 10. In that study, the
final metalloporphyrin catalyst was characterized by IR, UV−vis,
mass spectrometry, and elemental analysis, none of which
distinguishes between the atropisomers [FeIII(o-TMA)](OTf)5.
Electrochemically, the E1/2 values for the atropisomers are
sufficiently similar in both MeCN and DMF that broadening of
cyclic voltammograms of the mixture might not have been
evident. The largest difference in potentials is 46 mV between
the αβαβ and ααββ isomers in MeCN for the “FeI/Fe0” couple.
Our group was also guilty of this oversight in ref 12, following the
original synthesis and characterization of the material. We have
moved to using milder synthetic conditions to avoid isomer-
ization, as for the compounds used in ref 15.
The correct catalyst identification is directly relevant to the
electrocatalysis reported by both the original studies and our
subsequent paper. Both refs 10 and 12 report (electro)catalysis
using a mixture of atropisomers rather than with a single catalyst
species. Our study,12 as noted above, used a catalyst mixture
with a very narrow range of E1/2(FeIII/FeII) values and
considered only how E1/2 changed with buffer pKa. These
values were obtained by using in situ experiments that (in
retrospect) showed that atropisomeric purity of the catalyst was
not required. Reference 10 hypothesized that the enhanced
catalysis is “most likely [due to] the stabilization of the initial
Fe(0)−CO2 adduct by the interaction of the negative charge
borne by the oxygens of CO2 in this adduct with the nearby
positive charges [specifically of the αβαβ isomer] borne by the
trimethylanilinium substituents.” Computational results from
our group support this argument for the αβαβ isomer.15
However, the results from this work suggest that the specific
orientation of the [N(CH3)3]+ cations may be less significant
than the overall charge and electron-withdrawing nature of the o-
[N(CH3)3]+ groups.50
CONCLUSIONS
■
The addition of well-positioned charged groups to molecular,
inorganic complexes is an increasingly popular topic and has
garnered significant attention in the molecular electrocatalysis
literature. Here we report the synthesis and characterization of
all four atropisomers of iron(III) tetra(o-N,N,N-trimethyl-
anilinium)porphyrin pentatriflate and the corresponding
reduced iron(II) tetratriflate salts. Each of these complexes
contains four spatially resolved, cationic functional groups that
are uniquely arranged around a redox-active iron. The single-
Catalyst Identity in Prior Electrocatalytic CO2 and O2
Reduction Studies. The αβαβ atropisomer of Fe(o-TMA)
was first designed and reported as a CO2 reduction electro-
catalyst.10 To date, it remains one of the leading molecular
CO2RR catalysts in both rates and overpotentials and has gained
significant attention in the literature (>200 references as of 12/
2020). Yet, as shown above, duplicating the reported conditions
1
crystal X-ray structures and H NMR spectra show the nature
and high purity of the separate atropisomers that are available
from this revised synthesis. The previously reported synthesis is
5247
https://doi.org/10.1021/acs.inorgchem.1c00236
Inorg. Chem. 2021, 60, 5240−5251

Inorganic Chemistry
pubs.acs.org/IC
Article
(ααββ). 1H NMR (CD3Cl, ppm): 8.76 (s, 4H), 8.76 (s, 4H), 7.90 (d,
4H), 7.68 (t, 4H), 7.40 (d, 4H), 7.29 (t, 4H), 2.25 (s, 24H), and −2.32
(s, 2H). HRMS (ESI/Q-TOF): m/z [M + H]+ calcd for [C52H51N8]+,
787.424; found 787.43.
shown to form a mixture of atropisomers because rotamerization
occurred upon heating in one of the steps after the atropisomer
separation. Material from the prior synthesis, incorrectly
assumed to be the single αβαβ isomer, was used in an earlier
study of CO2 reduction electrocatalysis10 and in our earlier
paper on O2 reduction.12 Because the CO2 reduction study is
currently the leading molecular CO2-to-CO catalyst in
combined turnover frequency and overpotential and has been
highly cited, the actual multiple-isomer nature of the catalyst
present in those solutions is of significance.
(αααβ). 1H NMR (CD3Cl, ppm): 8.76 (s, 4H), 8.75 (s, 4H), 7.90
(m, 4H), 7.68 (m, 4H), 7.41 (m, 4H), 7.28 (m, 4H), 2.30 (s, 12H), 2.27
(s, 6H), 2.23 (s, 6H), and −2.30 (s, 2H). HRMS (ESI/Q-TOF): m/z
[M + H]+ calcd for [C52H51N8]+, 787.424; found 787.42.
(αααα). 1H NMR (CD3Cl, ppm): 8.75 (s, 8H), 7.86 (d, 4H), 7.70 (t,
4H), 7.46 (d, 4H), 7.28 (t, 4H), 2.37 (s, 24H), and −2.25 (s, 2H).
HRMS (ESI/Q-TOF): m/z [M + H]+ calcd for [C52H51N8]+, 787.424;
found 787.43.
The impact of unique charge positioning around the iron
center was probed by examining the properties of these
atropisomers. The single-crystal X-ray structures suggest that
triflate binding to the iron center is enhanced by electrostatics
much more strongly in the αααα isomer, where the charges all
lie on the same side of the molecule. The ferric and ferrous αααα
structures also have layered packing arrangements in the solid
state, different from the more typical 3D packing seen in the
other structures. In contrast, the electrochemistry of the
atropisomers was almost unaffected by the orientation of the
charged groups, with their FeIII/FeII reduction potentials all
being within ∼20 mV, in both MeCN and DMF. Yet the ortho-
positioning of the cationic groups is clearly important, as these
potentials are hundreds of millivolts more positive than the E1/2
values for the structural isomer containing the same cationic
groups in the para position. These trends are rationalized, at
least in part, by a simple electrostatic conductive-sphere model.
Overall, the studies described here show the varied effects of
positioned charges in a metal complex, and they provide
guidelines for future catalyst designs.
The iron(III) chloride tetra(o-N,N-dimethylaminophenyl)-
porphyrins were prepared by transmetalation of the corresponding
dilithium porphyrin saltsgenerated in situwith FeBr2(THF)2.32
1
The H NMR spectra are far more complicated for these iron(III)
chloride metalloporphyrins due to slow chloride exchange but are
qualitatively unique and are reported in Figures S15−S18.
The iron(III) tetra(o-N,N,N-trimethylanilinium)porphyrin pentatri-
flate salts were prepared by using methyl triflate in trimethyl phosphate
following a modified literature procedure (Figures S20−S23; see the
Supporting Information and discussion above).15,18 After recrystalliza-
tion by vapor diffusion of Et2O into MeCN solutions containing the
porphyrins, lustrous purple crystals were collected for all four products.
The crystals of the αβαβ, ααββ, and αααα atropisomers were suitable
for single-crystal X-ray diffraction.
Iron(III) Tetra(o-N,N,N-trimethylanilinium)porphyrin Pentatri-
1
flate. (αβαβ). H NMR (CD3CN, ppm): 15.0 (4H, Ar−H), 14.6
(8H, pyrr−H), 10.53 (4H, Ar−H), 10.22 (4H, Ar−H), 9.90 (4H, Ar−
H), and 2.22 (36H, −(CH3)12). HRMS (ESI/Q-TOF): m/z [M −
4(OTf)]4+ calcd for [C57H60N8FeO3SF3]4+, 262.345; found 262.34.
(ααββ). 1H NMR (CD3CN, ppm): 46.0 (4H, pyrr−H), 45.6 (4H,
pyrr−H), 13.49 (4H, Ar−H), 10.70 (4H, Ar−H), 10.47 (4H, Ar−H),
9.66 (4H, Ar−H), and 4.03 (36H, −(CH3)12). HRMS (ESI/Q-TOF):
m/z [M − 4(OTf)]4+ calcd for [C57H60N8FeO3SF3]4+, 262.345; found
262.34.
EXPERIMENTAL METHODS
See the Supporting Information for more complete descriptions and for
experiments not discussed here.
■
(αααβ). 1H NMR (CD3CN, ppm): 34.0 (8H, pyrr−H), 14.57−9.32
(16H, Ar−H), 5.33 (9H, −(CH3)3), 3.26 (18H, −(CH3)6), and 2.60
(9H, −(CH3)3). HRMS (ESI/Q-TOF): m/z [M − 4(OTf)]4+ calcd for
[C57H60N8FeO3SF3]4+, 262.345; found 262.34.
Synthesis. An atropisomeric mixture of 5,10,15,20-tetra(o-amino-
phenyl)porphyrin was prepared from the sequential (i) condensation of
pyrrole and 2-nitrobenzaldehyde and (ii) reduction of the resulting
5,10,15,20-tetra(o-nitrophenyl)porphyrin, following literature meth-
ods.28 Each of the four atropisomers was isolated by repeated
chromatography on silica with 1H NMR spectra that matched reported
spectra (Figures S5−S8).28
(αααα). 1H NMR (CD3CN, ppm): 47.6 (8H, pyrr−H), 13.36 (4H,
Ar−H), 10.80 (4H, Ar−H), 10.48 (4H, Ar−H), 9.62 (4H, Ar-H), and
4.29 (36H, −(CH3)12). HRMS (ESI/Q-TOF): m/z [M − 4(OTf)]4+
calcd for [C57H60N8FeO3SF3]4+, 262.345; found 262.34.
The reduced iron(II) tetra(o-N,N,N-trimethylanilinium)porphyrin
tetratriflate complexes were prepared by stirring the ferric porphyrin
salts with solid Zn(Hg) amalgam in the glovebox, following a reported
procedure. The porphyrin products were then precipitated by vapor
diffusion of Et2O into the collected MeCN solutions (Figures S29−
S32). As before, purple crystals were collected, all of which were
suitable for single-crystal X-ray diffraction. The 1H NMR spectrum for
the αβαβ atropisomer matched the reported spectrum.15
5,10,15,20-Tetra(o-aminophenyl)porphyrin. (αβαβ). 1H NMR
(CD3Cl, ppm): 8.91 (s, 8H), 7.87 (d, 4H), 7.60 (t, 4H), 7.16 (t,
4H), 7.11 (d, 4H), 3.50 (s, 8H), and −2.67 (s, 2H). HRMS (ESI/Q-
TOF): m/z [M + H]+ calcd for [C44H35N8]+, 675.298; found 675.30.
(ααββ). 1H NMR (CD3Cl, ppm): 8.90 (s, 8H), 7.84 (d, 4H), 7.60 (t,
4H), 7.16 (t, 4H), 7.11 (d, 4H), 3.55 (s, 8H), and −2.68 (s, 2H).
HRMS (ESI/Q-TOF): m/z [M + H]+ calcd for [C44H35N8]+, 675.298;
found 675.30.
Iron(II) Tetra(o-N,N,N-trimethylanilinium)porphyrin Tetratriflate.
(αβαβ). 1H NMR (CD3CN, ppm): 8.56 (s, 8H, pyrr−H), 8.54 (d, 4H,
Ar−H), 8.24 (d, 4H, Ar−H), 8.09 (t, 4H, Ar−H), 7.99 (t, 4H, Ar−H),
and 3.05 (s, 36H, −(CH3)12).
(αααβ). 1H NMR (CD3Cl, ppm): 8.90 (s, 8H), 7.85 (m, 4H), 7.60
(t, 4H), 7.17 (m, 4H), 7.11 (m, 4H), 3.54 (br s, 8H), and −2.68 (s, 2H).
HRMS (ESI/Q-TOF): m/z [M + H]+ calcd for [C44H35N8]+, 675.298;
found 675.30.
(ααββ). 1H NMR (CD3CN, ppm): 8.86 (s, 4H, pyrr−H), 8.83 (s,
4H, Ar−H), 8.30 (d, 4H, Ar−H), 8.28 (d, 4H, Ar−H), 8.08 (t, 4H, Ar−
H), 7.87 (t, 4H, Ar−H), 3.15 (s, 36H, −(CH3)12).
(αααα). 1H NMR (CD3Cl, ppm): 8.92 (s, 8H), 7.89 (d, 4H), 7.62 (t,
4H), 7.21 (m, 4H), 7.13 (d, 4H), 3.54 (s, 8H), and −2.66 (s, 2H).
HRMS (ESI/Q-TOF): m/z [M + H]+ calcd for [C44H35N8]+, 675.298;
found 675.30.
(αααβ). 1H NMR (CD3CN, ppm): 8.87 (m, 8H, pyrr−H), 8.51 (d,
1H, Ar−H), 8.44 (d, 2H, Ar−H), 8.36 (d, 1H, Ar−H), 8.26 (d, 4H, Ar−
H), 8.07 (m, 4H, Ar−H), 7.96 (t, 1H, Ar−H), 7.93 (t, 2H, Ar−H), 7.77
(t, 1H, Ar−H), 3.32 (s, 9H, −(CH3)3), 3.07 (s, 18H, −(CH3)6), 3.04 (s,
9H, −(CH3)3).
The atropisomers were methylated by reductive amination using
formaldehyde and sodium cyanoborohydride (4 h at 15 °C) and
purified by chromatography on silica (Figures S10−S13). The 1H NMR
spectrum for the αβαβ isomer matched the literature.10
5,10,15,20-Tetra(o-N,N-dimethylaminophenyl)porphyrin.
(αβαβ). 1H NMR (CD3Cl, ppm): 8.75 (s, 8H), 8.00 (d, 4H), 7.69 (t,
4H), 7.41 (d, 4H), 7.30 (t, 4H), 2.23 (s, 24H), and −2.30 (s, 2H).
HRMS (ESI/Q-TOF): m/z [M + H]+ calcd for [C52H51N8]+, 787.424;
found 787.43.
(αααα). 1H NMR (CD3CN, ppm): 10.80 (br s, 8H, pyrr−H), 8.30
(d, 4H, Ar−H), 8.22 (d, 4H, Ar−H), 8.06 (t, 4H, Ar−H), 7.82 (4H,
Ar−H), 3.22 (s, 36H, −(CH3)12).
Column Conditions for Isolating the Atropisomers of
5,10,15,20-Tetra(o-aminophenyl)porphyrin. αβαβ: Column 1
was an 8 in. × 3 in. column of silica, slurry loaded with DCM. The
5248
https://doi.org/10.1021/acs.inorgchem.1c00236
Inorg. Chem. 2021, 60, 5240−5251

Inorganic Chemistry
pubs.acs.org/IC
Article
αβαβ and ααββ atropisomers were separated from the αααβ and αααα
isomers by flash chromatography using 80:20 DCM/Et2O eluent (Rf =
0.8 and 0.7). The ratio of αβαβ to ααββ in the collected fractions was
approximately 1:2, consistent with the statistical mixture of isomers.
Column 2 was an 8 in. × 2 in. column of silica, slurry loaded with DCM.
Using a 90:10 DCM/Et2O eluent, the bulk of the αβαβ atropisomer
was separated from the ααββ (Rf = 0.7 and 0.5). The ratio of αβαβ to
ααββ in the collected fractions was atropisomerically enriched, though
often still impure (9:1). The dimensions and eluent mixture for column
3 was the same as column 2. Only the first few fractions were collected
and carefully monitored by TLC for contamination by the ααββ
atropisomer. After combining fractions and removing the solvent, we
obtained the αβαβ atropisomer with high purity.
Brandon Q. Mercado − Department of Chemistry, Yale
University, New Haven, Connecticut 06520-8107, United
States
Complete contact information is available at:
https://pubs.acs.org/10.1021/acs.inorgchem.1c00236
Author Contributions
D.J.M. and J.M.M. conceived the project, constructed the
scientific arguments, and wrote the paper. B.Q.M. performed X-
ray crystallography and solved the structures. D.J.M. performed
all other experiments and analyzed/interpreted the data.
ααββ: Columns 1, 2, and 3 were the same as described above for the
isolation of the αβαβ atropisomer. Fractions of ααββ were collected
and carefully monitored by TLC for trace αααβ and αβαβ
contamination in the second and third columns (Rf = 0.7, 0.5, and
0.2, respectively). It is worth noting that the solubility of this porphyrin
in DCM is lowest of the atropisomers, and so care should be taken to
avoid overloading the columns. In general, the ααββ isomer was the
easiest atropisomer to purify.
αααβ: The dimension of column 1 is the same as described above.
After eluting the αβαβ and ααββ atropisomers by using 80:20 DCM/
Et2O, the αααβ could be obtained by eluting with 50:50 DCM/Et2O
(Rf = 0.8). Columns 2 and 3 were the same dimensions as described
above and used the same two-stage eluant mixtures (80:20 followed by
50:50 DCM/Et2O). Care should be taken to avoid contamination by
the ααββ atropisomer.
αααα: The dimension of column 1 is the same as described above.
After eluting the column with 50:50 DCM/Et2O, we eluted the αααα
aminophenylporphyrin using 50:50 acetone/Et2O (Rf = 0.9). Columns
2 and 3 were the same dimensions as those used above but were loaded
with 50:50 DCM/Et2O and eluted with the same mixture until the
eluent was clear. The αααα porphyrin was obtained in high
atropisomeric purity by final elution with 50:50 acetone/Et2O. Of
note, the αααα isomer could be conveniently enriched prior to column
1 by refluxing the statistical mixture of atropisomers in the presence of
silica.51
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This research was supported as part of the Center for Molecular
Electrocatalysis, an Energy Frontier Research Center funded by
the U.S. Department of Energy (DOE), Office of Science, Office
of Basic Energy Sciences. D.J.M. gratefully acknowledges
support from a National Science Foundation Graduate Research
Fellowship.
REFERENCES
■
(1) Fried, S. D.; Boxer, S. G. Electric Fields and Enzyme Catalysis.
Annu. Rev. Biochem. 2017, 86 (1), 387−415.
(2) Warshel, A.; Sharma, P. K.; Kato, M.; Xiang, Y.; Liu, H.; Olsson, M.
H. Electrostatic basis for enzyme catalysis. Chem. Rev. 2006, 106 (8),
3210−35.
(3) Shaik, S.; Mandal, D.; Ramanan, R. Oriented electric fields as
future smart reagents in chemistry. Nat. Chem. 2016, 8 (12), 1091−
1098.
(4) Elgrishi, N.; Rountree, K. J.; McCarthy, B. D.; Rountree, E. S.;
Eisenhart, T. T.; Dempsey, J. L. A Practical Beginner’s Guide to Cyclic
Voltammetry. J. Chem. Educ. 2018, 95 (2), 197−206.
(5) Rountree, E. S.; McCarthy, B. D.; Eisenhart, T. T.; Dempsey, J. L.
Evaluation of Homogeneous Electrocatalysts by Cyclic Voltammetry.
Inorg. Chem. 2014, 53 (19), 9983−10002.
(6) Costentin, C.; Robert, M.; Savéant, J. M. Current Issues in
Molecular Catalysis Illustrated by Iron Porphyrins as Catalysts of the
CO2-to-CO Electrochemical Conversion. Acc. Chem. Res. 2015, 48
(12), 2996−3006.
(7) Costentin, C.; Savéant, J.-M. Multielectron, Multistep Molecular
Catalysis of Electrochemical Reactions: Benchmarking of Homoge-
neous Catalysts. ChemElectroChem. 2014, 1 (7), 1226−1236.
(8) Chantarojsiri, T.; Reath, A. H.; Yang, J. Y. Cationic Charges
Leading to an Inverse Free-Energy Relationship for N-N Bond
Formation by Mn(VI) Nitrides. Angew. Chem., Int. Ed. 2018, 57 (43),
14037−14042.
(9) Nichols, A. W.; Machan, C. W. Secondary-Sphere Effects in
Molecular Electrocatalytic CO2 Reduction. Front. Chem. 2019, 7, 397.
(10) Azcarate, I.; Costentin, C.; Robert, M.; Savéant, J. M. Through-
Space Charge Interaction Substituent Effects in Molecular Catalysis
Leading to the Design of the Most Efficient Catalyst of CO2-to-CO
Electrochemical Conversion. J. Am. Chem. Soc. 2016, 138 (51), 16639−
16644.
ASSOCIATED CONTENT
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/acs.inorgchem.1c00236.
Extended methods, 1H NMR spectra, mass spectra, UV−
vis spectra, and tabulated crystallographic data (PDF)
Accession Codes
CCDC 2056956−2056961 contain the supplementary crystallo-
graphic data for this paper. These data can be obtained free of
charge via www.ccdc.cam.ac.uk/data_request/cif, or by email-
ing data_request@ccdc.cam.ac.uk, or by contacting The Cam-
bridge Crystallographic Data Centre, 12 Union Road, Cam-
bridge CB2 1EZ, UK; fax: +44 1223 336033.
AUTHOR INFORMATION
■
Corresponding Author
(11) Zhang, R.; Warren, J. J. Controlling the Oxygen Reduction
Selectivity of Asymmetric Cobalt Porphyrins by Using Local Electro-
static Interactions. J. Am. Chem. Soc. 2020, 142 (31), 13426−13434.
(12) Martin, D. J.; Mercado, B. Q.; Mayer, J. M. Combining scaling
relationships overcomes rate versus overpotential trade-offs in O2
molecular electrocatalysis. Sci. Adv. 2020, 6 (11), No. eaaz3318.
(13) Margarit, C. G.; Asimow, N. G.; Gonzalez, M. I.; Nocera, D. G.
Double Hangman Iron Porphyrin and the Effect of Electrostatic
Nonbonding Interactions on Carbon Dioxide Reduction. J. Phys. Chem.
Lett. 2020, 11 (5), 1890−1895.
James M. Mayer − Department of Chemistry, Yale University,
New Haven, Connecticut 06520-8107, United States;
orcid.org/0000-0002-3943-5250; Email: james.mayer@
yale.edu
Authors
Daniel J. Martin − Department of Chemistry, Yale University,
New Haven, Connecticut 06520-8107, United States;
orcid.org/0000-0003-2039-592X
5249
https://doi.org/10.1021/acs.inorgchem.1c00236
Inorg. Chem. 2021, 60, 5240−5251

Inorganic Chemistry
pubs.acs.org/IC
Article
(14) Sampson, M. D.; Kubiak, C. P. Manganese Electrocatalysts with
Bulky Bipyridine Ligands: Utilizing Lewis Acids To Promote Carbon
Dioxide Reduction at Low Overpotentials. J. Am. Chem. Soc. 2016, 138
(4), 1386−93.
(15) Martin, D. J.; Johnson, S. I.; Mercado, B. Q.; Raugei, S.; Mayer, J.
M. Intramolecular Electrostatic Effects on O2, CO2, and Acetate
Binding to a Cationic Iron Porphyrin. Inorg. Chem. 2020, 59 (23),
17402−17414.
(16) Kalyanasundaram, K.; Neumannspallart, M. Photophysical and
Redox Properties of Water-Soluble Porphyrins in Aqueous-Media. J.
Phys. Chem. 1982, 86 (26), 5163−5169.
(29) Nishino, N.; Kobata, K.; Mihara, H.; Fujimoto, T. Efficient
Preparation of αβαβ-Atropisomer of meso-Tetra(o-aminophenyl)-
porphyrin. Chem. Lett. 1992, 21 (10), 1991−1994.
(30) Rose, E.; Cardonpilotaz, A.; Quelquejeu, M.; Bernard, N.;
Kossanyi, A.; Desmazieres, B. Efficient Preparation of the α,α,β,β-
Atropoisomer of meso-Tetrakis(o-aminophenyl)porphyrin. J. Org.
Chem. 1995, 60 (12), 3919−3920.
(31) Rose, E.; Quelquejeu, M.; Pochet, C.; Julien, N.; Kossanyi, A.;
Hamon, L. Large-scale preparation of α,β,α,β atropisomer of meso-
Tetrakis(o-aminophenyl)porphyrin. J. Org. Chem. 1993, 58 (19),
5030−5031.
(17) Kano, K.; Minamizono, H.; Kitae, T.; Negi, S. Self-Aggregation of
Cationic Porphyrins in Water. Can π-π Stacking Interaction Overcome
Electrostatic Repulsive Force? J. Phys. Chem. A 1997, 101 (34), 6118−
6124.
(18) La, T.; Richards, R.; Miskelly, G. M. Synthesis and Character-
ization of the Cationic Porphyrin Meso-Tetrakis(2,3,5,6-Tetrafluoro-
N,N,N-Trimethyl-4-Aniliniumyl)Porphyrin. Inorg. Chem. 1994, 33
(14), 3159−3163.
(32) Arnold, J.; Dawson, D. Y.; Hoffman, C. G. Synthesis and
Characterization of Lithium, Sodium, and Potassium Porphyrin
Complexes - X-Ray Crystal-Structures of Li2(C6H12O2)2TMPP,
Na2(THF)4OEP, and K2(pyridine)4OEP. J. Am. Chem. Soc. 1993,
115 (7), 2707−2713.
(33) Boersma, A. D.; Goff, H. M. Multinuclear magnetic resonance
spectroscopy of spin-admixed S = 5/2, 3/2 iron(III) porphyrins. Inorg.
Chem. 1982, 21 (2), 581−586.
(19) Miskelly, G. M.; Webley, W. S.; Clark, C. R.; Buckingham, D. A.
Acidity and dimerization of three water-soluble iron(III) porphyrin
cations: (meso-α,β,α,β-tetrakis(o-(N-methylnicotinamido)phenyl)-
porphyrinato)iron(III), (meso-α,α,α,α-tetrakis(o-(N-
methylnicotinamido)phenyl)porphyrinato)iron(III), and (meso-
tetrakis(1-methylpyridinium-4-yl)porphyrinato)iron(III). Inorg.
Chem. 1988, 27 (21), 3773−3781.
(20) Tovmasyan, A.; Weitner, T.; Sheng, H.; Lu, M.; Rajic, Z.; Warner,
D. S.; Spasojevic, I.; Reboucas, J. S.; Benov, L.; Batinic-Haberle, I.
Differential coordination demands in Fe versus Mn water-soluble
cationic metalloporphyrins translate into remarkably different aqueous
redox chemistry and biology. Inorg. Chem. 2013, 52 (10), 5677−91.
(34) Safo, M. K.; Nesset, M. J. M.; Walker, F. A.; Debrunner, P. G.;
Scheidt, W. R. Models of the Cytochromes. Axial Ligand Orientation
and Complex Stability in Iron(II) Porphyrinates: The Case of the
Noninteracting dπ Orbitals. J. Am. Chem. Soc. 1997, 119 (40), 9438−
9448.
(35) Walker, F. A.; Simonis, U. Iron Porphyrin Chemistry; John Wiley
& Sons, Ltd.: Chichester, UK, 2006.
(36) Pegis, M. L.; Martin, D. J.; Wise, C. F.; Brezny, A. C.; Johnson, S.
I.; Johnson, L. E.; Kumar, N.; Raugei, S.; Mayer, J. M. Mechanism of
Catalytic O2 Reduction by Iron Tetraphenylporphyrin. J. Am. Chem.
Soc. 2019, 141, 8315−8326.
(37) Tomé, A. C.; Silva, A. M. S.; Alkorta, I.; Elguero, J.
Atropisomerism and conformational aspects of meso-tetraarylporphyr-
ins and related compounds. J. Porphyrins Phthalocyanines 2011, 15 (01),
1−28.
(38) Zimmer, B.; Bulach, V.; Drexler, C.; Erhardt, S.; Hosseini, M. W.;
De Cian, A. Design, synthesis, structural analysis and atropoisomerisa-
tion studies of polynucleating ligands based on porphyrins bearing
catechol units. New J. Chem. 2002, 26 (1), 43−57.
(39) Crossley, M. J.; Field, L. D.; Forster, A. J.; Harding, M. M.;
Sternhell, S. Steric Effects on Atropisomerism in Tetraarylporphyrins. J.
Am. Chem. Soc. 1987, 109 (2), 341−348.
(40) Nozawa, R.; Tanaka, H.; Cha, W. Y.; Hong, Y.; Hisaki, I.;
Shimizu, S.; Shin, J. Y.; Kowalczyk, T.; Irle, S.; Kim, D.; Shinokubo, H.
Stacked antiaromatic porphyrins. Nat. Commun. 2016, 7, 13620.
(41) Kadish, K. M.; Morrison, M. M.; Constant, L. A.; Dickens, L.;
Davis, D. G. A study of solvent and substituent effects on the redox
potentials and electron-transfer rate constants of substituted iron meso-
tetraphenylporphyrins. J. Am. Chem. Soc. 1976, 98 (26), 8387−90.
(42) Romelt, C.; Song, J.; Tarrago, M.; Rees, J. A.; van Gastel, M.;
Weyhermuller, T.; DeBeer, S.; Bill, E.; Neese, F.; Ye, S. Electronic
Structure of a Formal Iron(0) Porphyrin Complex Relevant to CO2
Reduction. Inorg. Chem. 2017, 56 (8), 4746−4751.
(43) Romelt, C.; Ye, S.; Bill, E.; Weyhermuller, T.; van Gastel, M.;
Neese, F. Electronic Structure and Spin Multiplicity of Iron
Tetraphenylporphyrins in Their Reduced States as Determined by a
Combination of Resonance Raman Spectroscopy and Quantum
Chemistry. Inorg. Chem. 2018, 57 (4), 2141−2148.
́
́
(21) Batinic-Haberle, I.; Spasojevic, I.; Hambright, P.; Benov, L.;
Crumbliss, A. L.; Fridovich, I. Relationship among Redox Potentials,
Proton Dissociation Constants of Pyrrolic Nitrogens, and in Vivo and in
Vitro Superoxide Dismutating Activities of Manganese(III) and
Iron(III) Water-Soluble Porphyrins. Inorg. Chem. 1999, 38 (18),
4011−4022.
(22) Weitner, T.; Budimir, A.; Kos, I.; Batinic-Haberle, I.; Birus, M.
Acid-base and electrochemical properties of manganese meso(ortho-
and meta-N-ethylpyridyl)porphyrins: potentiometric, spectrophoto-
metric and spectroelectrochemical study of protolytic and redox
equilibria. Dalton Trans. 2010, 39 (48), 11568−76.
(23) Franke, A.; Scheitler, A.; Kenkel, I.; Lippert, R.; Zahl, A.;
Balbinot, D.; Jux, N.; Ivanovic-Burmazovic, I. Positive Charge on
Porphyrin Ligand and Nature of Metal Center Define Basic
Physicochemical Properties of Cationic Manganese and Iron
Porphyrins in Aqueous Solution. Inorg. Chem. 2019, 58 (15), 9618−
9630.
(24) Brausam, A.; Eigler, S.; Jux, N.; van Eldik, R. Mechanistic
investigations of the reaction of an iron(III) octa-anionic porphyrin
complex with hydrogen peroxide and the catalyzed oxidation of
diammonium-2,2′-azinobis(3-ethylbenzothiazoline-6-sulfonate). Inorg.
Chem. 2009, 48 (16), 7667−78.
(25) Jee, J. E.; Eigler, S.; Hampel, F.; Jux, N.; Wolak, M.; Zahl, A.;
Stochel, G.; van Eldik, R. Kinetic and mechanistic studies on the
reaction of nitric oxide with a water-soluble octa-anionic iron(III)
porphyrin complex. Inorg. Chem. 2005, 44 (22), 7717−31.
(26) Jee, J. E.; Wolak, M.; Balbinot, D.; Jux, N.; Zahl, A.; van Eldik, R.
A comparative mechanistic study of the reversible binding of NO to a
water-soluble octa-cationic Fe(III) porphyrin complex. Inorg. Chem.
2006, 45 (3), 1326−37.
(44) Pegis, M. L.; McKeown, B. A.; Kumar, N.; Lang, K.; Wasylenko,
D. J.; Zhang, X. P.; Raugei, S.; Mayer, J. M. Homogenous
Electrocatalytic Oxygen Reduction Rates Correlate with Reaction
Overpotential in Acidic Organic Solutions. ACS Cent. Sci. 2016, 2 (11),
850−856.
(45) Nichols, E. M.; Derrick, J. S.; Nistanaki, S. K.; Smith, P. T.;
Chang, C. J. Positional effects of second-sphere amide pendants on
electrochemical CO2 reduction catalyzed by iron porphyrins. Chem. Sci.
2018, 9 (11), 2952−2960.
(46) Bonin, J.; Maurin, A.; Robert, M. Molecular catalysis of the
electrochemical and photochemical reduction of CO2 with Fe and Co
(27) Jee, J. E.; Eigler, S.; Jux, N.; Zahl, A.; van Eldik, R. Influence of an
extremely negatively charged porphyrin on the reversible binding
kinetics of NO to Fe(III) and the subsequent reductive nitrosylation.
Inorg. Chem. 2007, 46 (8), 3336−52.
(28) Collman, J. P.; Gagne, R. R.; Reed, C. A.; Halbert, T. R.; Lang, G.;
Robinson, W. T. Picket fence porphyrins. Synthetic models for oxygen
binding hemoproteins. J. Am. Chem. Soc. 1975, 97 (6), 1427−39.
5250
https://doi.org/10.1021/acs.inorgchem.1c00236
Inorg. Chem. 2021, 60, 5240−5251

Inorganic Chemistry
pubs.acs.org/IC
Article
metal based complexes. Recent advances. Coord. Chem. Rev. 2017, 334,
184−198.
(47) Conway, B. E. The Electrolyte Factor in Supercapacitor Design and
Performance: Conductivity, Ion Pairing and Solvation BT - Electrochemical
Supercapacitors: Scientific Fundamentals and Technological Applications;
Conway, B. E., Ed. Springer US: Boston, MA, 1999; pp 335−375.
(48) Wang, P.; Anderko, A. Computation of dielectric constants of
solvent mixtures and electrolyte solutions. Fluid Phase Equilib. 2001,
186 (1−2), 103−122.
(49) Parr, R. G.; Pearson, R. G. Absolute Hardness - Companion
Parameter to Absolute Electronegativity. J. Am. Chem. Soc. 1983, 105
(26), 7512−7516.
(50) Mao, Y.; Loipersberger, M.; Kron, K. J.; Derrick, J. S.; Chang, C.
J.; Sharada, S. M.; Head-Gordon, M. Consistent inclusion of continuum
solvation in energy decomposition analysis: theory and application to
molecular CO2 reduction catalysts. Chem. Sci. 2021, 12, 1398.
(51) Lindsey, J. Increased Yield of a Desired Isomer by Equilibria
Displacement on Binding to Silica-Gel, Applied to Meso-Tetrakis(o-
aminophenyl)Porphyrin. J. Org. Chem. 1980, 45 (25), 5215−5215.
5251
https://doi.org/10.1021/acs.inorgchem.1c00236
Inorg. Chem. 2021, 60, 5240−5251