A Synthetic Strategy for Cofacial Porphyrin-Based Homo- and Heterobimetallic Complexes

Article abstract of DOI:10.1002/chem.202002394

We present a straightforward and generally applicable synthesis route for cofacially linked homo- and heterobimetallic porphyrin complexes. The protocol allows the synthesis of unsymmetrical aryl-based meso-meso as well as β-meso-linked porphyrins. Our method significantly increases the overall yield for the published compound known as o-phenylene-bisporphyrin (OBBP) by a factor of 6.8. Besides the synthesis of 16 novel homobimetallic complexes containing Mn^III, Fe^III, Ni^II, Cu^II, Zn^II, and Pd^II, we achieved the first single-crystal X-ray structure of an unsymmetrical cofacial benzene-linked porphyrin dimer containing both planar-chiral enantiomers of a Ni^II₂ complex. Additionally, this new methodology allows access to heterobimetallic complexes such as the Fe^III-Ni^II containing carbon monoxide dehydrogenase active site analogue. The isolated species were investigated by various techniques, including ion mobility spectrometry, DFT calculations, and UV/Vis spectroscopy. This allowed us to probe the influence of interplane distance on Soret band splitting.

Full text of DOI:10.1002/chem.202002394

Full Paper
doi.org/10.1002/chem.202002394
Chemistry—A European Journal
&
Porphyrin Chemistry |Hot Paper|
A Synthetic Strategy for Cofacial Porphyrin-Based Homo- and
Heterobimetallic Complexes
Christoph Schissler,[a] Erik K. Schneider,[b] Benjamin Felker,[a] Patrick Weis,[b] Martin Nieger,[c]
Manfred M. Kappes,*[b, d] and Stefan Bräse*[a, e]
Abstract: We present a straightforward and generally appli-
cable synthesis route for cofacially linked homo- and hetero-
bimetallic porphyrin complexes. The protocol allows the syn-
thesis of unsymmetrical aryl-based meso-meso as well as b-
meso-linked porphyrins. Our method significantly increases
the overall yield for the published compound known as o-
phenylene-bisporphyrin (OBBP) by a factor of 6.8. Besides
the synthesis of 16 novel homobimetallic complexes con-
taining MnIII, FeIII, NiII, CuII, ZnII, and PdII, we achieved the first
single-crystal X-ray structure of an unsymmetrical cofacial
benzene-linked porphyrin dimer containing both planar-
chiral enantiomers of a NiII2 complex. Additionally, this new
methodology allows access to heterobimetallic complexes
such as the FeIII-NiII containing carbon monoxide dehydro-
genase active site analogue. The isolated species were inves-
tigated by various techniques, including ion mobility spec-
trometry, DFT calculations, and UV/Vis spectroscopy. This al-
lowed us to probe the influence of interplane distance on
Soret band splitting.
Introduction
tion. Correspondingly, strongly interacting metal sites are the
basis for the unique catalytic activity of many multinuclear
metalloproteins such as hemocyanin,[1,2] hemerythrin,[3] super-
oxide dismutase,[4] carbon monoxide dehydrogenase,[5] and cy-
tochrome c oxidase.[6] To gain further understanding of the
unique coordination chemistry of such complexes, one ideally
needs a rigid molecular system that allows one to tune the dis-
tances between the different metal centers without also influ-
encing them by significantly changing the ligand field. It then
becomes possible to vary the interactions between metal cen-
ters, particularly those that give rise to cooperative catalytic ef-
fects, by slight variations of the rigid framework. One system
of interest in this context is covalently linked dimeric porphyrin
metal complexes that can be tuned by systematic structural
variation and whose cooperative properties can be compared
with those of the corresponding constituent monomers.
In nature, the catalytically active sites of metalloenzymes are
often rigidly fixed structures in an adaptive protein matrix
which defines a spatial arrangement of the metal-containing li-
gands relative to each other. In such systems, proximity of sev-
eral metal cations is typically required to achieve catalytic func-
[a] C. Schissler, B. Felker, S. Bräse
Institute of Organic Chemistry, Karlsruhe Institute of Technology (KIT)
Fritz-Haber-Weg 6, 76131 Karlsruhe (Germany)
E-mail: braese@kit.edu
[b] E. K. Schneider, P. Weis, M. M. Kappes
Institute of Physical Chemistry, Karlsruhe Institute of Technology (KIT)
Fritz-Haber-Weg 2, 76131 Karlsruhe (Germany)
E-mail: manfred.kappes@kit.edu
[c] M. Nieger
Department of Chemistry, University of Helsinki
P.O. Box 55, 00014 Helsinki (Finland)
Porphyrins offer the opportunity to coordinate numerous
different metal ions without changing ligands and can be syn-
thesized in a tailor-made way.[7,8] Stepwise syntheses are of es-
pecially great interest toward rationally accessing artificially
made heterobimetallic active site analogues, for example,
those of the above-mentioned enzymes. This makes dimeric
porphyrins a perfect choice not only to model and understand
elementary enzymatic reactivity, but also to study cooperative
magnetic, catalytic, and optical properties of two spatially
close-lying metal ions. In principle, three different bisporphyrin
topologies are possible: coplanar, tilted, and cofacial. Cofacial
orientation provides closer metal–metal separations and is
therefore of greatest interest for cooperative effects.
[d] M. M. Kappes
Institute for Nanotechnology, Karlsruhe Institute of Technology (KIT)
Hermann-von-Helmholtz-Platz 1
76344 Eggenstein-Leopoldshafen (Germany)
[e] S. Bräse
Institute for Biological and Chemical Systems— Functional Molecular
Systems, (IBCS-FMS), Karlsruhe Institute of Technology (KIT)
Hermann-von-Helmholtz-Platz 1
76344 Eggenstein-Leopoldshafen (Germany)
Supporting information and the ORCID identification number(s) for the
author(s) of this article can be found under:
https://doi.org/10.1002/chem.202002394.
ꢀ 2021 The Authors. Chemistry - A European Journal published by Wiley-
VCH GmbH. This is an open access article under the terms of the Creative
Commons Attribution Non-Commercial License, which permits use, distribu-
tion and reproduction in any medium, provided the original work is proper-
ly cited and is not used for commercial purposes.
Over the last fifty years, multiple synthetic routes have ad-
dressed cofacial bisporphyrins, starting with the synthesis of
doubly urea and amide bridged meso-connected sandwich
Chem. Eur. J. 2021, 27, 3047 –3054
3047 ꢀ 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202002394
Chemistry—A European Journal
porphyrin dimers by Collman and Chang.[9] Shortly thereafter,
Collman et al. demonstrated electrocatalytic four-electron re-
duction of oxygen to water using cofacial binuclear cobalt por-
phyrins.[10] Inspired by this first bimetallic catalysis based on
porphyrins, Reed and colleagues developed a novel synthetic
route to imidazolate- and oxo-bridged metalloporphyrins in
which only one linker moiety and one bridging ligand were
needed.[11] Collman et al. then first synthesized dimeric b-linked
face-to-face amide-bridged porphyrin dimers.[12]
Figure 1. Targeted structures: 1: o-phenylene-bisporphyrin (OBBP), 2: o-
phenylene-ethoxycarbonyl-bisporphyrin (EOBBP), 3: o-phenylene-b-meso-bis-
porphyrin (BMOBBP). The molecules comprise porphyrin-based cofacial
ligand systems which differ regarding the number and type of residues at
meso-position and the connecting atom to the phenyl backbone.
At about the same time, Chang and Abdalmuhdi began to
focus on cofacial porphyrin dimers connected via rigid linker
moieties and prepared 8-anthryldiporphyrin by condensing
two dipyrromethane derivatives at
a linking anthracene
moiety.[13] Later bisphenylenyldiporphyrin[14] and dibenzofura-
nyldiporphyrin[15] became available based on the same ap-
proach.
and the corresponding aldehydes. The ligands differ regarding
the number and type of residues in the meso-position as well
as the position in which the porphyrin is connected to the
phenyl backbone. The first two compounds require an asym-
metric porphyrin as a starting point for the synthesis of the di-
meric porphyrin-based ligands. Starting with a simple conden-
sation reaction of pyrrole and paraformaldehyde dipyrrome-
thane, 5 can be synthesized in up to 65% yield (Scheme 1).
This can be used to build up the porphyrin cores 6 and 7 with
the corresponding residues at the 5- and 15-positions depend-
ing on the aldehyde used. The next step is crucial for the dis-
tance between the porphyrin planes. Either the third meso-po-
sition is substituted by nucleophilic aromatic substitution with
phenyllithium to obtain 8 or it remains unsubstituted. Mono-
bromination with NBS under differing conditions then leads to
bromo-porphyrin precursors 9 and 10. On the way to ligand 3,
straightforward b-bromination of commercially available tetra-
phenylporphyrin (TPP) as the first step has to be conducted
to afford 12. The synthetic route via mono-brominated
porphyrins as precursors enables incorporation of the benzene
linker moiety without a second mixed-condensation as de-
Parallel to the approaches using linker moieties that intrinsi-
cally favored cofacial orientations, Kobuke et al. developed a
synthetic route to o-phenylene-bisporphyrins. Interestingly,
these also showed a cofacial arrangement due to strong p-
stacking interactions which can compensate for the 608 bite
angle of the o-benzene linker.[16] Osuka et al. were able to con-
duct X-ray structure analyses of single crystals of 1,2-phenyl-
ene-bridged alkyl-porphyrin dimers, which proved the cofacial
arrangement with an average plane separation of 3.43 Š.[17]
Fletcher and Therien then established a new route toward co-
facial porphyrin dimers via [2+2+2] cycloaddition of 1,6-hepta-
diyne with ethynyl-linked porphyrins.[18] They were able to syn-
thesize meso-meso, meso-b, and b-b connected zinc complexes
of cofacial porphyrin dimers by this method.[19] Through a
novel Suzuki–Miyaura cross-coupling methodology toward co-
facial bisporphyrins anchored by xanthene and dibenzofuran,
starting with boronated porphyrin monomers, Nocera et al. fa-
cilitated the route developed earlier, to achieve porphyrin
plane distances around 4 Š.[20] Since this publication in 2003,
to our knowledge, there have been no further groundbreaking
synthetic method developments on the way to spatially close
cofacial porphyrins.
The pioneering syntheses described above suffer from quite
modest overall yields. For example, the highest published yield
for the overall synthesis of benzene-linked symmetrical por-
phyrin dimers is at the 0.65%[16] level. Furthermore, the proce-
dures developed, for example, by Fletcher and Therien,[19] are
not readily generalizable to the complexation of a wide range
of different transition metal ions— in particular, heterobimetal-
lic combinations. Therefore, in this study we aimed at develop-
ing a new high-yielding, tolerant, and robust synthetic meth-
odology, which is suitable for both homo- and heterobimetallic
cofacial bisporphyrin complexes and which allows for tunable
metal–metal distances.
Results and Discussion
Scheme 1. Synthesis of brominated precursors: a) formaldehyde, InCl3,
NaOH, 558C, 3 h, 65%; b) benzaldehyde, TFA, DDQ, 4 h, 63%; c) ethyl-4-for-
mylbenzoate, TFA, DDQ, 18 h, 53%, Ar: C6H4COOEt; d) PhLi, DDQ, THF,
08C!RT, 30 min, 95%; e) NBS, CH2Cl2, 08C!RT, 3 h, 97%; f) NBS, pyridine,
CH2Cl2, 08C, 20 min, 60%; g) NBS, CH2Cl2, D, 8 h, 54%.
Homobimetallic complexes
We have developed a facile route to the three different cofacial
ligands shown in Figure 1 (1, 2, and 3) starting with pyrrole
Chem. Eur. J. 2021, 27, 3047 –3054
www.chemeurj.org
3048 ꢀ 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202002394
Chemistry—A European Journal
scribed,[16,17,21] but instead with the simple Suzuki–Miyaura
cross-coupling reaction using 2-formylphenylboronic acid as
the coupling counterpart as shown in Scheme 2.
because pyrrole and benzaldehyde prefer the condensation to
TPP, excluding the aldehyde-porphyrins. Using an excess of
pyrrole and benzaldehyde to achieve full conversion of the
starting material is desirable, as TPP as a side product can
easily be removed as the first purple fraction of the subse-
quent flash column chromatography on silica gel. Because the
formation of the sterically less hindered TPP is observed before
the bisporphyrin, even more sterically hindered porphyrin tri-
mers could not be obtained by the described methodology.
The final condensations lead to the above shown cofacial
porphyrin dimers. Our novel synthetic route to the known
compound 1 increases its overall yield from 0.65%[16] or
0.30%[22] to 4.4% (i.e., by at least a factor of 6.8). The new
compounds 2 and 3 could be synthesized in 1.9% and 1.3%
yields, respectively. Whereas the final condensation of 1 and 2
rank within regular yields, for 3 the yield drops significantly.
Besides steric issues faced during synthesis, the workup is ag-
gravated due to higher basicity originating from the closer
stacked porphyrin planes, which led to decreases in the isolat-
ed yield.
In addition to that, asymmetric cofacial porphyrin dimers
with one free meso-position have become available. Further-
more, bromination of one of the pyrrole carbon atoms enables
the introduction of the backbone at the b-position which is
not possible by conventional condensation reactions.
This robust procedure provides monomeric precursors in
good overall yields (13: 26%, 14: 12%, 15: 28%). For com-
pound 14 single crystals were obtained, which were suitable
for X-ray crystallography (Figure 2). The molecular structure
confirms the aldehyde functionality in the o-position of the an-
ticipated linking moiety and clarifies that there should be
enough space for the subsequent condensation reaction to
obtain two cofacial porphyrin subunits linked together by an
o-substituted phenyl bridge (Scheme 2).
The aldehyde functionalities of 13, 14, and 15 now take part
in the concluding mixed condensation with pyrrole and benz-
aldehyde in a ratio of (1:4:3). As catalyst BF3·OEt2 or TFA were
used, depending on the starting material. Crucial is the consec-
utive addition of pyrrole and benzaldehyde until the conver-
sion of the aldehyde-porphyrins 13, 14, and 15, as monitored
by TLC, slows down. We achieve an increased yield by not
sticking to the 1:4:3 ratio as the molecular structure suggested
Note that in a similar 1,2-biphenylene-bridged porphyrin
dimer, Osuka et al. showed by X-ray crystallography that the
molecule has a near-parallel, cofacial porphyrin ring arrange-
ment with a dihedral (twist) angle of 6.68 (see also schematic
structure in the Supporting Information defining the dihedral
angle).[17] In the absence of crystal structures, it is unclear
whether 1, 2, and 3 have the same topology in the solid state.
DFT calculations (on isolated molecules; see below) identify
similarly cofacial benzene-linked porphyrin planes with dihe-
dral angles between the porphyrins by way of the linker ben-
zene hinge of 10.08 (OBBP), 0.18 (EOBBP), and 2.28 (BMOBBP),
respectively.
The synthesized ligands 1, 2, and 3 were subsequently
doubly metallated with six different transition metals
(Scheme 3), and products systematically investigated by ion
mobility spectrometry and DFT calculations. In total, 18 cofacial
homobimetallic complexes could be synthesized and charac-
terized by this new method, out of which 16 compounds were
unpublished. Additionally, we were able to grow single crystals
of compound 24, shown in Figure 3, which represents the
first unsymmetrical cofacial benzene-linked metalloporphyrin
dimer.
Scheme 2. Synthesis of the monomeric porphyrins covalently linked to the
phenyl backbone (13–15): a) 2-formylphenylboronic acid, Pd(PPh3)4, K3PO4,
THF, 808C, 6–23 h, 13: 68%, 14: 57%, 15: 52%. Syntheses of cofacial por-
phyrin dimers (1–3): b) pyrrole (14.2 equiv), benzaldehyde (9.15 equiv), TFA
(1.08 equiv), RT, 115 h, 17%; c) pyrrole (6.07 equiv), benzaldehyde
(3.93 equiv), BF3OEt2 (3.47 equiv), RT, 21.5 h, 16%; d) pyrrole (14.1 equiv),
benzaldehyde (9.13 equiv), TFA (3.21 equiv), RT, 46 h, 4.7%.
Interestingly, X-ray diffraction data of the single crystal show
two crystallographically independent but identical molecules
in the asymmetric unit. The respective chiral space group is
P21 (“Sohncke space group”) but crystallized as a twin contain-
ing both enantiomers. Compound 24 was refined as an inver-
sion twin with BASF=0.38(2) (Hooft’s y-parameter) y=
0.39(1).[23] Therefore, the abundance ratio between the two
enantiomers in the measured crystal was approximately 62:38.
The unit cell of 24 contains two crystallographically independ-
ent molecules with identical chirality in an asymmetric unit
(see least-squares fit (L.S.-fit) in the Supporting Information,
which resembles two “Pac-Man” characters biting each other).
The missing phenyl ring opposite to the backbone of one of
the two porphyrins enables this curious packing, which dis-
ables direct dispersion interactions between the intramolecular
Figure 2. Single-crystal X-ray structure of 14. For clarity, only the hydrogen
atoms bound to nitrogen atoms are displayed.
Chem. Eur. J. 2021, 27, 3047 –3054
www.chemeurj.org
3049 ꢀ 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202002394
Chemistry—A European Journal
chiral due to intrinsic planar chirality based on the unsymmet-
rical linkage to the linker benzene moiety.
Heterobimetallic complexes
As mentioned in the introduction, the demand for tailor-made
heterobimetallic complexes as artificial active site analogues of
enzymes is huge. Our synthetic approach provides for a new
facile route to an artificially synthesized carbon monoxide de-
hydrogenase active site analogue containing NiII and FeIII-cat-
ions. For this, the following synthetic route was developed as a
proof-of-principle reaction to achieve a porphyrin-based transi-
Scheme 3. Porphyrin-based homobimetallic complexes: a) (16) *MnIII: MnCl2,
DMF, 1508C, 2 h, 86%; (17) **FeIII: FeBr2, HCl, DMF, 1408C, 1 h, 93%; (18) NiII:
Ni(acac)2, DMF, 1008C, 19.5 h, 86%; (19) CuII: Cu(OAc)2, CHCl3/MeOH, 808C,
2 h, 93%; (20) ZnII: Zn(OAc)2, CHCl3/MeOH, RT, 1 h, 72%; (21) PdII: PdCl2,
DMF, 1008C, 19.5 h, 69%; b) (22) *MnIII: MnCl2, DMF, 1508C, 22 h, 73%; (23)
**FeIII: FeBr2, HCl, DMF, 1408C, 2 h, 93%; (24) NiII: Ni(acac)2, DMF, 1008C, 2 h,
96%; (25) CuII: Cu(OAc)2, CHCl3/MeOH, 808C, 1 h, 90%; (26) ZnII: Zn(OAc)2,
CHCl3/MeOH, 808C, 1 h, 99%; (27) PdII: PdCl2, DMF, 808C, 20 h, 98%; c) (28)
*MnIII: MnCl2, DMF, 1308C, 15.5 h, 79%; (29) **FeIII: FeCl2, DMF, 1508C, 14 h,
87%; (30) NiII: Ni(acac)2, DMF, 1508C, 5 h, 76%; (31) CuII: Cu(OAc)2, DMF,
608C, 4 h, 93%; (32) ZnII: Zn(OAc)2, CHCl3/MeOH, RT, 1 h, 95%; (33) PdII:
PdCl2, DMF, 1008C, 3 h, 99%. *Was obtained with two coordinated chlorides
at the manganese centers. **Was obtained as the m-oxido complex.
tion-metal-containing heterobimetallic complexes. After
a
Suzuki–Miyaura cross-coupling reaction, the monomeric por-
phyrin 13 was treated with Ni(acac)2 at 1508C for 4 h, to
obtain the NiII complex 34 in 72% yield (Scheme 4). Complexa-
tion of the NiII cation can be proven by missing NH protons
and slightly increased coupling constants of the b protons ad-
jacent to the 2-formyl-benzene residue from 3J=4.8 to 3J=
5.0 Hz, as observed in the corresponding 1H NMR spectra.
Through a subsequently performed condensation reaction sim-
ilar to that shown in Scheme 2, the NiII-H2-o-phenylene-bispor-
phyrin 35 can then be obtained in 15% yield. The ring current
of the NiII-containing porphyrin affects the free-base porphyrin
in shifting the NH protons to À3.82 ppm, which is typical for
cofacially linked porphyrin dimers. The remaining free-base
porphyrin in complex 35 can undergo a reaction with FeCl2 in
DMF at 1508C for 4 h to incorporate FeIII as the second cation
in 95% yield. The overall yield to obtain the first cofacial por-
phyrin-based heterobimetallic complex 36 is 2.2% beginning
with pyrrole and benzaldehyde as starting material and proves
the robustness of the developed synthetic methodology.
Based on the results shown, one can expect this protocol to
be suitable for many other analogous homo- and heterobime-
tallic complexes which can find multiple applications in fields
ranging from magnetism/spintronics, catalysis, and optical sen-
Figure 3. Crystal structure of NiII EOBBP 24. The X-ray diffraction of the
2
single crystal shows two crystallographically independent but identical mol-
ecules. The missing phenyl ring opposite to the backbone of one porphyrin
subunit per porphyrin dimer enables intermolecular p-stacking. See the Sup-
porting Information (Section 6. Crystallographic Data) for a more detailed
description. All hydrogen atoms are omitted for clarity.
porphyrin planes (in contrast to the system reported by Osuka
et al.,[17] which has fewer sterically demanding peripheral sub-
stituents and can therefore collapse into a cofacial structure
with a significantly strained 1,2-biphenylene link in the crystal-
line solid). An interdigitated arrangement such as in 24 can
also overcome the difficulties of growing single crystals of
enantiomeric mixtures, which was often problematic in the
past. Note that the interdigitated structure in solid state very
likely does not correspond to the situation in liquid (or gas)
phase, as shown later in the structure determination part. Due
to that, we cannot make statements regarding the metal–
metal distances in solution. Additionally, the chiral conformers
are only distinguishable in solid state because the low inter-
conversion barrier leads to averaging dihedral angles between
the porphyrin planes. Nonetheless, all derivatives of 3 are
Scheme 4. Synthesis of an artificial carbon monoxide dehydrogenase active
site analogue (36) through porphyrin building on the monomeric NiII-con-
taining porphyrin (34) and subsequent insertion of FeIII in the second por-
phyrin ring: a) Ni(acac)2, 1508C, 4 h, 71%; b) pyrrole, benzaldehyde, TFA,
DDQ, RT, 16 h, 15%; c) FeCl2, 1508C, 4 h, 95%.
Chem. Eur. J. 2021, 27, 3047 –3054
www.chemeurj.org
3050 ꢀ 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202002394
Chemistry—A European Journal
sors. Furthermore, the impact of cooperative interactions be-
tween (different) spatially proximate metal ions is of funda-
mental interest, e.g., for enzymatic reactivity in vivo.
rings. IMS is a gas-phase method to determine the collision
cross-section (CCS) of an ion, which can be easily combined
with MS.[24] It relies on determination of the drift time of an ion
in an inert collision gas (typically helium or nitrogen) guided
by an electrical field. With the recent development of several
instrumental variants which can provide greatly improved IMS
resolution, the method has gained in importance, for example,
in studies of proteins,[25] polysaccharides,[26] and fullerenes,[27]
and can now provide an additional useful identification param-
eter in proteomics.[28] In our measurements, we used a high-
resolution variant of IMS, trapped ion mobility spectrometry
(TIMS) coupled with a ToF-mass spectrometer (timsTOFꢁ,
Bruker). The operational mode of TIMS has been described
elsewhere.[29] Details regarding the IMS measurements and cal-
culations can be found in the Supporting Information. In brief,
with this method, it is possible to differentiate between iso-
mers or conformers that differ in CCS by less than 0.5%. Be-
sides isomer separation, measured CCS can be used to validate
structure predictions based on quantum chemical calculations
(and collision gas scattering trajectory calculations) and thus to
obtain a first-order structural assignment. In detail, the meth-
odology we applied is as follows: For each porphyrin complex
we performed a DFT-based geometry optimization (Turbomole
package,[30] BP-86 functional,[31,32] def2-SVP basis set,[33] Grimme
D3-BJ dispersion correction[34,35]). The partial charges were cal-
culated by the Mulliken algorithm and the CCS calculations
were conducted with the IMoS package.[36,37] The coordinates
of the ZnII2 dimers are given in Supporting Information Sec-
tion 5.2. The optimization was performed for isolated cations,
without counterions or solvation effects. Based on these opti-
mized structures we calculated CCS with the trajectory
method, which are listed in Table 1. As is evident, all measured
CCS values for the covalently linked porphyrin dimers prepared
in this study differ by only a few percent from the value of
[(H2TPP)2 +H]+. Finally, we compared its calculated CCS with
the measured CCS. The structures of noncovalently linked TPP
dimers are published.[38] We thus determined a scaling factor
of 0.94. Note that the calculated CCS depends on both geome-
try and charge distribution as well as the assumed Lennard–
Jones parameters of the constituting atoms. The interplay be-
tween these factors is rather delicate, and deviations of several
Structure determination and UV/Vis spectra
While the above-mentioned complexes have all been well
1
characterized by MS, 13C NMR, H NMR, UV/Vis (Figure 4) and IR
spectroscopy, their 3D structure could not be established by
the standard structure determination method, X-ray diffraction,
due to crystallization problems (except for 24, see above). Fur-
thermore, we expect that in solid phase, packing effects and
intermolecular interactions significantly disrupt the rather deli-
cate balance between p-stacking and van der Waals attraction
of porphyrinic moieties on the one hand and Coulomb repul-
sion of the positively charged metal centers on the other. As a
consequence, we applied a combination of quantum chemical
calculations and ion mobility spectrometry (IMS) to gain access
to at least some structural parameters of corresponding isolat-
ed monocations such as the average distance of the porphyrin
Table 1. Experimental TIMSCCSN2 values of the different porphyrin dimers.
Central atoms
TIMSCCSN2 [Š2][a]
OBBP
EOBBP
BMOBBP
TPP dimer
2H2 +H
MnIII2 +Cl
FeIII2 +O
NiII2
339.7Æ0.2
345.8Æ0.8
338.1Æ0.4
337.5Æ0.4
336.1Æ0.4
334.9Æ1.2
335.5Æ0.1
338.0Æ0.1
365.9Æ1.2
370.3Æ0.1
362.7Æ1.3
362.4Æ0.6
363.8Æ1.1
360.5Æ0.7
361.8Æ1.9
350.3Æ0.4
357.8Æ0.1
347.8Æ0.4
351.0Æ0.7
350.8Æ0.1
352.2Æ0.3
348.7Æ1.1
354.2Æ0.5
CuII2
ZnII2
PdII2
NiIIH2 +H
NiIIFeIII
Figure 4. UV/Vis spectra of the various ZnII dimers. Both OBBP and BMOBBP
2
show a second absorption band in the Soret region that is redshifted by
roughly 20 nm relative to the band of highest intensity. Additionally, the
most intense band of BMOBBP is redshifted by about 7 nm relative to OBBP.
We attribute these shifts (and splittings) to different distances between the
ZnII cations.
335.5Æ1.2
[a] Values are the mean Æ standard deviation of n=2–4 independent
measurements.
Chem. Eur. J. 2021, 27, 3047 –3054
www.chemeurj.org
3051 ꢀ 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202002394
Chemistry—A European Journal
percent between experimental and absolute theoretical CCS
are not unusual. A common strategy enabling a better com-
parison with theory is to calibrate against a well-studied inter-
nal standard with a similar structure such as [(H2TPP)2 +H]+
(Figure 5). After applying this scaling factor to the theoretical
CCS of [ZnII2-OBBP]+, [ZnII2-EOBBP]+ and [ZnII2-BMOBBP]+, we
reached a deviation between theory and experiment of below
2% (see Table 2 for details). We therefore conclude that our
DFT geometries are quite close to the actual structures for all
dimers studied, as (OBBP<BMOBBP<EOBBP). The complexes
with divalent metal centers (NiII, CuII, ZnII, PdII) show the same
CCS, indicating that the nature of the complexed metal does
not significantly influence three-dimensional structures. The
CCS of the trivalent MnIII dimers, on the other hand, are sys-
tematically larger due to the additional ClÀ ion located be-
tween the respective metal atoms, which pushes the mono-
mers a bit further apart. The structure of FeIII dimers binding
oxygen is analogous, as previously reported.[15]
Based on the DFT-optimized structures of the ZnII dimers
(with trends validated by TIMS measurements), we obtain
metal-to-metal distances that decrease from EOBBP (3.276 Š)
to OBBP (3.243 Š) to BMOBBP (3.207 Š); see Figure 6. Because
these correspond to the metal–metal distances of gaseous
ions at 0 K, the absolute numbers do not necessarily describe
the intramolecular metal center separations in solution or
solid, but we assume that even for these condensed phase en-
vironments the trend toward decreasing metal center separa-
tion in going from EOBBP to BMOBBP will remain the same. In-
terestingly, the sequence correlates with the increasing promi-
nence of a shoulder (near 435 nm) in the Soret band region of
the UV/Vis spectrum. Furthermore, the UV/Vis spectrum of
Figure 6. The DFT-calculated structures of the ZnII dimers are shown in
2
front and side views. The values on the right show the calculated distances
of the two ZnII cations in each dimer. For clarity, only the hydrogen atoms
bound to nitrogen atoms are displayed. The linking phenyl ring is drawn in
orange for clarity as well. The coordinates are included in x,y,z format in Sec-
tion 5.2 of the Supporting Information.
[ZnII2-BMOBBP] shows a broadening and slight asymmetry of
the Soret band, which might indicate an additional absorption
band. We tentatively assign this to enhanced spatial p-interac-
tion and hence stronger coupling between the chromophores
as the intermetal distance decreases. In a 2018 study, Jäger
et al. investigated the changes in the Q-band absorption
region of gaseous dimeric porphyrin ions, which were induced
by structural differences.[39] In the case of isolated ZnII2 dimers,
analogous shifts of Q-band peak maxima by up to 8 nm were
reported.
This is consistent with a previous related study by Takai et al.
who showed, that increasing distances between the two
monomers of a porphyrin dimer also increases the reorganiza-
tion energies associated with electron transfer between the
chromophores, proving that the distance can change the prop-
erties of porphyrin dimers and presumably higher oligomers as
well.[40] Furthermore, Bolze et al. showed that covalently linked
dimers of PdII porphyrins and monomers of PdII porphyrins can
differ in their Q-band region absorption quite drastically (the
Soret region was not shown).[41]
The slight decrease of metal center distance between ZnII2-
EOBBP and ZnII2-OBBP can be attributed to enhanced attractive
interaction between proximal phenyl rings of the porphyrin
dimers (whereas EOBBP lacks some corresponding phenyls).
Compared with ZnII2-OBBP, the phenyls of the two porphyrin
rings of ZnII2-BMOBBP can be better intercalated. Thus the sep-
aration of the metals can be decreased further.
Figure 5. The DFT-calculated structure of [(H2TPP)2 +H]+ is shown in front
and side views. For clarity, only the hydrogen atoms bound to nitrogen
atoms are displayed.
Table 2. Experimental CCS and their scaled, theoretical counterparts are
listed along with their corresponding deviation.[a]
Measured CCS [Š2]
Scaled CCS [Š2]
Deviation [%]
[(H2TPP)2 +H]+
[ZnII2-OBBP]+
[ZnII2-EOBBP]+
[ZnII2-BMOBBP]+
354.2
334.9
360.5
352.2
354.2
329.7
362.1
345.6
calibrant
1.6
À0.4
1.9
Conclusions
[a] As can be seen, the difference between experiment and theory is
below 2%, and therefore in good agreement.
We present a straightforward new synthesis route for both
homo- and heterobimetallic porphyrin complexes. The proto-
Chem. Eur. J. 2021, 27, 3047 –3054
www.chemeurj.org
3052 ꢀ 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202002394
Chemistry—A European Journal
col allows the synthesis of aryl-based meso-meso as well as b-
meso-linked porphyrins with an optional phenyl residue on
one of the porphyrin subunits. Where comparisons with litera-
ture can be drawn, the developed methodology generally pro-
vides higher yields, for example, by a factor of 6.8 for the well-
known symmetric porphyrin dimer 1. Based on this improved
accessibility we synthesized 18 different homobimetallic spe-
cies containing the transition metals MnIII, FeIII, NiII, CuII, ZnII,
and PdII. In the case of NiII and FeIII we were able to prepare
heterobimetallic species. The isolated species were character-
Keywords: active site analogues · bimetallic complexes ·
carbon monoxide dehydrogenase · ion mobility spectrometry ·
porphyrin dimers
[1] J. Brown, L. Powers, B. Kincaid, J. Larrabee, T. G. Spiro, J. Am. Chem. Soc.
1980, 102, 4210–4216.
[2] K. Tatsumi, R. Hoffmann, J. Am. Chem. Soc. 1981, 103, 3328–3341.
[3] I. M. Klotz, T. A. Klotz, H. A. Fiess, Arch. Biochem. Biophys. 1957, 68, 284–
299.
[4] A. Desideri, M. Falconi, F. Polticelli, M. Bolognesi, K. Djinovic, G. Rotilio,
J. Mol. Biol. 1992, 223, 337–342.
1
ized by H NMR, 13C NMR, UV/Vis, IR, MS, and high-resolution
IMS measurements, which were contrasted with DFT calcula-
tions. For a set of three different bimetallic ZnII complexes, we
could show that metal–metal distances differ systematically
with ligand (due to steric effects) and were able to correlate
this trend with Soret band shifts and splitting. Furthermore, an
unsymmetrical NiII2 complex could be crystallized in the form
of a twin containing both planar-chiral enantiomers. Thus we
were able to obtain the first single-crystal X-ray structure of an
unsymmetrical cofacial benzene-linked porphyrin dimer.
[5] H. Dobbek, V. Svetlitchnyi, L. Gremer, R. Huber, O. Meyer, Science 2001,
293, 1281–1285.
[6] J. P. Collman, N. K. Devaraj, R. A. DecrØau, Y. Yang, Y.-L. Yan, W. Ebina,
T. A. Eberspacher, C. E. Chidsey, Science 2007, 315, 1565–1568.
[7] A. Treibs, Liebig Ann. Chem. 1969, 728, 115–143.
[8] M. O. Senge, Chem. Commun. 2011, 47, 1943–1960.
[9] C. Chang, J. Heterocycl. Chem. 1977, 14, 1285–1288.
[10] J. P. Collman, P. Denisevich, Y. Konai, M. Marrocco, C. Koval, F. C. Anson,
J. Am. Chem. Soc. 1980, 102, 6027–6036.
[11] J. T. Landrum, D. Grimmett, K. J. Haller, W. R. Scheidt, C. A. Reed, J. Am.
Chem. Soc. 1981, 103, 2640–2650.
[12] J. P. Collman, C. S. Bencosme, C. E. Barnes, B. D. Miller, J. Am. Chem. Soc.
1983, 105, 2704–2710.
[13] C. Chang, I. Abdalmuhdi, J. Org. Chem. 1983, 48, 5388–5390.
[14] C. K. Chang, I. Abdalmuhdi, Angew. Chem. Int. Ed. 1984, 23, 164–
165.
[15] Y. Deng, C. J. Chang, D. G. Nocera, J. Am. Chem. Soc. 2000, 122, 410–
411.
[16] H. Meier, Y. Kobuke, S.-i. Kugimiya, Chem. Commun. 1989, 923–924.
[17] A. Osuka, S. Nakajima, T. Nagata, K. Maruyama, K. Toriumi, Angew. Chem.
Int. Ed. 1991, 30, 582–584.
[18] J. T. Fletcher, M. J. Therien, J. Am. Chem. Soc. 2000, 122, 12393–12394.
[19] J. T. Fletcher, M. J. Therien, J. Am. Chem. Soc. 2002, 124, 4298–4311.
[20] L. L. Chng, C. J. Chang, D. G. Nocera, J. Org. Chem. 2003, 68, 4075–
4078.
Experimental Section
The synthetic procedures for all synthesized compounds are avail-
able in the Supporting Information. Deposition Number 1988039
(compound 24) contains the supplementary crystallographic data
for this paper. These data are provided free of charge by the joint
Cambridge Crystallographic Data Centre and Fachinformationszen-
trum Karlsruhe Access Structures service: www.ccdc.cam.ac.uk/
structures. Due to the poor quality of the data for 14, these were
not deposited with The Cambridge Crystallographic Data Centre.
[21] Y. Shimazaki, H. Takesue, T. Chishiro, F. Tani, Y. Naruta, Chem. Lett. 2001,
30, 538–539.
[22] H. S. Cho, D. H. Jeong, M.-C. Yoon, Y. H. Kim, Y.-R. Kim, D. Kim, S. C.
Jeoung, S. K. Kim, N. Aratani, H. Shinmori, J. Phys. Chem. A 2001, 105,
4200–4210.
[23] R. W. Hooft, L. H. Straver, A. L. Spek, J. Appl. Crystallogr. 2008, 41, 96–
103.
[24] E. A. Mason, E. W. McDaniel, NASA STI/Recon Tech. Rep. A 1988, 89, 576.
[25] M. E. Ridgeway, J. A. Silveira, J. E. Meier, M. A. Park, Analyst 2015, 140,
6964–6972.
[26] Y. Pu, M. E. Ridgeway, R. S. Glaskin, M. A. Park, C. E. Costello, C. Lin, Anal.
Chem. 2016, 88, 3440–3443.
[27] P. Weis, F. Hennrich, R. Fischer, E. K. Schneider, M. Neumaier, M. M.
Kappes, PCCP 2019, 21, 18877–18892.
[28] D. Helm, J. P. Vissers, C. J. Hughes, H. Hahne, B. Ruprecht, F. Pachl, A.
Grzyb, K. Richardson, J. Wildgoose, S. K. Maier, Mol. Cell. Proteomics
2014, 13, 3709–3715.
[29] K. Michelmann, J. A. Silveira, M. E. Ridgeway, M. A. Park, J. Am. Chem.
Mass. Spec. 2015, 26, 14–24.
[30] F. Furche, R. Ahlrichs, C. Hättig, W. Klopper, M. Sierka, F. T. Weigend, Mol.
Sci. 2014, 4, 91–100.
Acknowledgements
We thank S. Jaschik, R. Wurst, S. Palloks, P. Kern, and S. Marsch-
ner (Karlsruhe Institute of Technology (KIT)) for their help in
conducting the synthesis of the precursors, and G. Niedner-
Schatteburg (TU Kaiserslautern) for stimulating discussions. C.S.
gratefully acknowledges the Fonds der Chemischen Industrie
(FCI) and the graduate program of the federal state of Baden-
Württemberg (LGF) for financial support. S.B. also thanks the
Collaborative Research Center TRR 88 “3MET” and the Deut-
sche Forschungsgemeinschaft (DFG, German Research Founda-
tion) for support through project B2. Additionally, S.B. acknowl-
edges support under Germany’s Excellence Strategy via the Ex-
cellence Cluster 3D Matter Made to Order (EXC-2082/1-
390761711). E.K.S., P.W., and M.M.K. gratefully acknowledge
support by the DFG as administered by the Collaborative Re-
search Center TRR 88 “3MET” through project C6. M.M.K.
thanks KIT for funding the TIMS-TOFMS used in this study.
Open access funding enabled and organized by Projekt DEAL.
[31] J. P. Perdew, Phys. Rev. B 1986, 33, 8822.
[32] A. D. Becke, Phys. Rev. A 1988, 38, 3098.
[33] F. Weigend, R. Ahlrichs, PCCP 2005, 7, 3297–3305.
[34] S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys. 2010, 132,
154104.
[35] S. Grimme, S. Ehrlich, L. Goerigk, J. Comput. Chem. 2011, 32, 1456–
1465.
[36] C. Larriba, C. J. Hogan, J. Comput. Phys. 2013, 251, 344–363.
[37] C. Larriba, C. J. Hogan, J. Phys. Chem. A 2013, 117, 3887–3901.
[38] K. Brendle, U. Schwarz, P. Jäger, P. Weis, M. Kappes, J. Phys. Chem. A
2016, 120, 8716–8724.
Conflict of interest
The authors declare no conflict of interest.
Chem. Eur. J. 2021, 27, 3047 –3054
www.chemeurj.org
3053 ꢀ 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202002394
Chemistry—A European Journal
[39] P. Jäger, K. Brendle, E. Schneider, S. Kohaut, M. K. Armbruster, K. Fink, P.
Weis, M. M. Kappes, J. Phys. Chem. A 2018, 122, 2974–2982.
[40] A. Takai, C. P. Gros, J. M. Barbe, R. Guilard, S. Fukuzumi, Chem. Eur. J.
2009, 15, 3110–3122.
[41] F. Bolze, C. P. Gros, P. D. Harvey, R. Guilard, J. Porphyrins Phthalocyanines
2001, 5, 569–574.
Manuscript received: May 14, 2020
Revised manuscript received: November 16, 2020
Version of record online: January 18, 2021
Chem. Eur. J. 2021, 27, 3047 –3054
www.chemeurj.org
3054 ꢀ 2021 The Authors. Chemistry - A European Journal published by Wiley-VCH GmbH