Experimental and theoretical characterization of 5,10- diminoporphodimethenes: Dearomatized porphyrinoids from palladium-catalyzed hydrazinations of 5,10-diarylporphyrins

Article abstract of DOI:10.1002/cplu.201300433

A new dearomatized porphyrinoid, 5,10-diiminoporphodimethene (5,10-DIPD), has been prepared by palladium-catalyzed hydrazination of 5,10-dibromo-15,20- bis(3,5-di-tert-butylphenyl)porphyrin and its nickel(II) complex, by using ethyl and 4-methoxybenzyl carbazates. The oxidative dearomatization of the porphyrin ring occurs in high yield. Further oxidation with 2,3-dichloro-5,6- dicyanobenzoquinone forms the corresponding 5,10-bis(azocarboxylates), thereby restoring the porphyrin aromaticity. The UV/visible spectra of the Ni ^II DIPDs exhibit remarkable redshifts of the lowest-energy bands to 780 nm, and differential pulse voltammetry reveals a contracted electrochemical HOMO-LUMO gap of 1.44 V. Density functional theory (DFT) was used to calculate the optimized geometries and frontier molecular orbitals of model 5,10-DIPD Ni7c and 5,10-bis(azocarboxylate) Ni8c. The conformations of the carbamate groups and the configurations of the C=N-Z unit were considered in conjunction with the NOESY spectra, to generate the global minimum geometry and two other structures with slightly higher energies. In the absence of solution data regarding conformations, ten possible local minimum conformations were considered for Ni8c. Partition of the porphyrin macrocycle into tri- and monopyrrole fragments in Ni7c and the inclusion of terminal conjugating functional groups generate unique frontier molecular orbital distributions and a HOMO-LUMO transition with a strong element of charge transfer from the monopyrrole ring. Time-dependent DFT calculations were performed for the three lowest-energy structures of Ni7c and Ni8c, and weighting according to their energies allowed the prediction of the electronic spectra. The calculations reproduce the lower-energy regions of the spectra and the overall forms of the spectra with high accuracy, but agreement is not as good in the Soret region below 450 nm. A new dearomatized porphyrinoid, 5,10-diiminoporphodimethene, has been prepared and characterized spectroscopically and by theoretical calculations (see figure). The strongly redshifted HOMO-LUMO absorption band has the character of a charge transfer from the unique pyrrole ring to the tripyrrole fragment. Copyright

Full text of DOI:10.1002/cplu.201300433

CHEMPLUSCHEM
FULL PAPERS
DOI: 10.1002/cplu.201300433
Experimental and Theoretical Characterization of 5,10-
Diminoporphodimethenes: Dearomatized Porphyrinoids
from Palladium-Catalyzed Hydrazinations of 5,10-
Diarylporphyrins
[a]
[a]
[b]
[b]
ˇ ´ [a]
Bruno Basic, Oliver B. Locos, Llew Rintoul, Jinglan Kan, Jianzhuang Jiang, and
Dennis P. Arnold*[a]
A new dearomatized porphyrinoid, 5,10-diiminoporphodime-
thene (5,10-DIPD), has been prepared by palladium-catalyzed
hydrazination of 5,10-dibromo-15,20-bis(3,5-di-tert-butylphe-
nyl)porphyrin and its nickel(II) complex, by using ethyl and 4-
methoxybenzyl carbazates. The oxidative dearomatization of
the porphyrin ring occurs in high yield. Further oxidation with
2,3-dichloro-5,6-dicyanobenzoquinone forms the correspond-
ing 5,10-bis(azocarboxylates), thereby restoring the porphyrin
aromaticity. The UV/visible spectra of the NiII DIPDs exhibit re-
markable redshifts of the lowest-energy bands to 780 nm, and
differential pulse voltammetry reveals a contracted electro-
chemical HOMO–LUMO gap of 1.44 V. Density functional
theory (DFT) was used to calculate the optimized geometries
and frontier molecular orbitals of model 5,10-DIPD Ni7c and
5,10-bis(azocarboxylate) Ni8c. The conformations of the carba-
mate groups and the configurations of the C=NÀZ unit were
considered in conjunction with the NOESY spectra, to generate
the global minimum geometry and two other structures with
slightly higher energies. In the absence of solution data re-
garding conformations, ten possible local minimum conforma-
tions were considered for Ni8c. Partition of the porphyrin mac-
rocycle into tri- and monopyrrole fragments in Ni7c and the in-
clusion of terminal conjugating functional groups generate
unique frontier molecular orbital distributions and a HOMO–
LUMO transition with a strong element of charge transfer from
the monopyrrole ring. Time-dependent DFT calculations were
performed for the three lowest-energy structures of Ni7c and
Ni8c, and weighting according to their energies allowed the
prediction of the electronic spectra. The calculations reproduce
the lower-energy regions of the spectra and the overall forms
of the spectra with high accuracy, but agreement is not as
good in the Soret region below 450 nm.
Introduction
used recently by Chen and co-
workers.[4] Numerous other com-
pounds having exocyclic oxo or
alkylidene substituents have been
characterized,[5] including dimeric
species reported by Anderson’s
group.[6] Very few of these have
the interrupting substituents in
the adjacent 5,10-positions, apart
from a free base dioxo example
(isolated as the monohydrate)[5e]
and some modified tetrapyrroles
The porphyrin nucleus represents one of the most studied of
all non-benzenoid aromatic systems. Imaginative synthetic
chemists have developed a plethora of related macrocycles
that can be thought of as derivatives or analogues of the tetra-
pyrrolic framework of the porphyrins.[1] Porphyrinoids in which
the 18/22 p-electron aromaticity is interrupted by atoms
doubly bonded to peripheral (mostly meso) carbon(s) are an in-
teresting class of tetrapyrroles.[2] The typical allowed/forbidden
(B/Q band) nature of the major electronic transitions is lost,
and strong polarization along the new electronic axes is pres-
ent. Such compounds have been known for many years. 5,15-
Dioxoporphyrins (1)[3] (Scheme 1) were initially prepared using
thallium(III) salts, but the less toxic iron(III) chloride has been
with exocyclic double bonds in
three or four meso positions.[7]
COVER
Some core-modified and expand-
ˇ ´
[a] Dr. B. Basic, Dr. O. B. Locos, Dr. L. Rintoul, Assoc. Prof. D. P. Arnold
ed porphyrinoids with two exocyclic double bonds in adjacent
meso sites have also been reported.[8] The only porphyrinoid
with this substitution pattern that has been characterized in
School of Chemistry, Physics and Mechanical Engineering
Queensland University of Technology
GPO Box 2434, Brisbane 4001 (Australia)
Fax: (+61)7-3138-1804
1
detail (by X-ray crystal structure, H NMR and UV/visible spec-
troscopy) is “NiII(cis-OEPO2)” 2, that is, the nickel(II) complex of
5,10-dioxo-2,3,7,8,12,13,17,18-octaethylporphyrin.[9] Its crystallo-
graphic analysis was complicated by disorder in the arrange-
ments of the oxo substituents in the crystals (not uncommon
for OEP-based structures), as well as cocrystallization with the
trans isomer and the mono-oxy radical.[9a] However, it was clear
E-mail: d.arnold@qut.edu.au
[b] J. Kan, Prof. J. Jiang
Department of Chemistry
University of Science and Technology Beijing
Beijing 100083 (P. R. China)
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cplu.201300433.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 813 – 824 813

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
novel azoporphyrins (the porphyrin analogues of azoben-
zenes),[15] we found that a free meso position always led to pre-
dominant head-to-tail dimerization, rather than the desired (at
the time) head-to-head azoporphyrin formation. This was con-
firmed by Chen and co-workers.[4]
Herein, we now report the carbazate coupling and in situ
partial oxidation chemistry for the 5,10-diarylporphyrin (“corner
porphyrin”) system, which has led to the first examples of 5,10-
DIPDs (prototype structure 4). As noted above, porphyrinoids
with exocyclic double bonds of any kind in the 5,10-position
are very rare, so the facile formation of the DIPDs represents
an unprecedented opportunity to study this little-known type
of macrocycle. So in addition to normal spectral characteriza-
tion and voltammetry, we subjected the structure to theoreti-
cal quantum chemical calculations using density functional
theory (DFT), and simulated the electronic spectrum by time-
dependent DFT (TD-DFT). This represents the first published at-
tempt to apply these methods to this “3+1” tetrapyrrolic mac-
rocycle, and the results can now be compared with those for
the more familiar 5,15-dearomatized porphyrinoids.[11a–c] The
experimental and theoretical results reveal profoundly different
electronic structures for the two systems.
Scheme 1. Structures of dearomatized porphyrinoids 1–4.
Results and Discussion
that the electronic absorption spectra of the 5,15- and 5,10-
dioxo species differ significantly.[9b]
Synthesis and reactions
The development of palladium-catalyzed aminations of 5,15-
diarylporphyrins[10] led us to use hydrazine derivatives in place
of simple amines.[11] In the case of carbazates (RO2CÀNHÀNH2),
the initial amination products undergo facile in situ partial oxi-
dation to a novel class of dearomatized porphyrinoids, which
we described as carbamate-substituted 5,15-diiminoporphodi-
methenes (DIPDs), exemplified by structure 3.[11a–c] These are
formally bis(hydrazides) of 5,15-dioxoporphyrins, but have not
yet been prepared from the latter compounds. This spontane-
ous dearomatization reaction is one manifestation of the inter-
esting properties of the nitrogen-substituted porphyrinoids,
that is, tetrapyrroles with amino-type nitrogen atoms attached
at one or more peripheral carbon atoms. The lone pairs of
electrons on the attached nitrogen atoms are strongly conju-
gated with the macrocyclic p system, thus raising the energy
of the highest occupied molecular orbital (HOMO). The meso-
meric electron donation is apparent from the significant effects
of the nitrogen atom on oxidation potentials and electronic
spectra,[11c,d] and the near or complete coplanarity of the NH/
NH2 groups found in X-ray crystal structures of porphyrinyl pri-
mary and secondary amines.[11c,d,12] One graphic example of the
novel effects of the attached nitrogen was discovered by Rup-
pert and co-workers,
The starting porphyrin for this work was 5,10-bis(3,5-di-tert-bu-
tylphenyl)porphyrin (H2DAP), which was brominated by the
known method to form H2DAPBr2 H25.[16] Insertion of NiII gave
the complex Ni5 (Scheme 2). We began the hydrazinations by
using the same combination of palladium precatalysts and
phosphine ligand as we had used for the 5,15-disubstituted
porphyrins.[11a–c] In turn, we derived these from the report of
Takanami et al.[10a] on aminations of haloporphyrins. The 2,2’-
bis(diphenylphosphino)-1,1’-binaphthyl (BINAP) ligand gave
reasonable success in preparing the 5,15-DIPDs, but initial ef-
forts to aminate 5,10-dibromoporphyrin 5 (free base, nickel(II),
or zinc(II) complexes) with alkyl carbazates were inconsistent
and low-yielding. Immediate success was forthcoming when
the bidentate diphosphine bis(2-diphenylphosphinophenyl)
ether (DPEphos) was used instead of BINAP. Because one can
readily separate the catalysts, ligands, and bases from porphy-
rin-based coupling products by chromatography, we also in-
creased the amounts of Pd precursor/ligand to 20/55 mol%
from the previous 7/20 mol%. Under these conditions, the re-
actions described in Scheme 2 were performed successfully.
The starting dibromoporphyrins were usually consumed within
two hours of heating in THF at its boiling point. In other work,
we have found that this catalyst system is active enough to
allow amination reactions of sensitive substrates to be com-
pleted in acceptable times (ca. 24 h) even at ambient tempera-
ture.[17]
who demonstrated the conjugation of the N lone pair in
bis[10,15,20-triphenylporphyrinatonickel(II)-5-yl]amine by using
voltammetry combined with electronic and EPR spectrosco-
py.[13] Moreover, there is a strong tendency for head-to-tail oli-
gomerization of meso-aminoporphyrins, followed by loss of
the porphyrin macrocyclic aromaticity, as shown by several re-
search groups.[4,11a,c,14] In the course of our research into the
The nickel complex Ni7a was prepared as in Scheme 2 in
yields of purified product of at least 60%. The initial product
of the double coupling of ethyl carbazate with a dibromopor-
phyrin is presumed to be the bis(carbazate), in this case Ni6a.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 813 – 824 814

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
termination. As its spectroscopic
properties were essentially iden-
tical to those of the ethyl ver-
sion, no further experiments
were pursued with it.
We also prepared the free
base DIPD H27a, and in this case
it seemed that the auto-oxida-
tion was quite fast, as the com-
plicated chromatographic behav-
ior exhibited by the nickel com-
plex was not evident (Scheme 3).
However, one side product was
separated and characterized,
namely the mono-azocarboxy-
late H29, isolated in 5% yield. Al-
though various other minor side
products were also observed in
the purification of Ni7a, the
analogous NiII mono-azocarboxy-
late was not able to be charac-
Scheme 2. Synthesis of the nickel(II) 5,10-diiminoporphodimethenes and 5,10-bis(azocarboxylates). Reagents:
1) ethyl or 4-methoxybenzyl carbazate, Cs2CO3, Pd(OAc)2, bis(2-diphenylphosphinophenyl) ether (DPEphos) in THF;
2) air; 3) 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ), CH2Cl2.
Indeed, TLC analyses of the reaction mixture after the dibromo
starting material is fully consumed show two major spots: the
less polar one is green and the more polar one is purple. Isola-
tion by column chromatography was at first puzzling, as the
second band when isolated appeared to be the same com-
terized. The yield of the desired H27a as the crude product
was good, but the target compound was difficult to purify,
and the sample remained contaminated with some ethyl car-
bazate and other porphyrinoids (see the Supporting Informa-
tion). Note that these compounds are not easy to handle, as
they are very soluble in almost all organic solvents, and at-
tempts to produce high-quality crystals proved to be frustrat-
ing. The preparation of the original 5,10-H2DAP nucleus is tedi-
ous and yields are low, so preparations on scales above about
20 mg that might make purification of end products more ef-
fective are difficult.
1
pound as the first, according to their H NMR spectra in CDCl3.
This behavior results from the fact that the carbazate is auto-
oxidized readily in solution, particularly in chlorinated solvents,
and apparently even more rapidly on silica gel. Thus, when the
second fraction is re-chromatographed, the green less-polar
substance is increasingly seen to be present as the leading
edge of the purplish band. To solve these problems, the proce-
dure in the Experimental Section was adopted, which involved
a residence time of some hours on the second silica gel
column, leading to isolation of the pure green DIPD Ni7a. This
situation was known to us for the 5,15-DIPD analogues, and in
the procedure developed by Esdaile for these compounds, the
amination solution was stirred
More vigorous oxidation of 7 with DDQ produced the 5,10-
bis(azocarboxylates) Ni8a, Ni8b, and H28a (Schemes 2 and 3).
The reaction with the 4-methoxybenzyl version was not clean,
and numerous products were formed along with the desired
Ni8b, which was not obtained pure. Presumably side reactions
of the electron-rich 4-methoxybenzyl unit were interfering.
overnight under air to complete
the auto-oxidation.[11c] In the
present case, this procedure was
less successful, as the conver-
sion to DIPD seemed to be
slower, but was assisted by
chlorinated solvents and/or
silica gel. In the end, the result
was satisfactory and Ni7a and
the 4-methoxybenzyl analogue
Ni7b were purified successfully
and fully characterized by elec-
1
tronic absorption and H, COSY,
NOESY, and 13C NMR spectrosco-
py. The latter derivative was pre-
pared as part of our many at-
tempts to achieve single crystals
suitable for X-ray structure de-
Scheme 3. Synthesis of the free base 5,10-diiminoporphodimethene and 5,10-bis(azocarboxylate). Reagents:
1) ethyl carbazate, Cs2CO3, Pd(OAc)2, DPEphos in THF; 2) DDQ, CH2Cl2.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 813 – 824 815

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
1
Figure 1. Comparison of the downfield sections of the H NMR spectra of 5,10-di(ethoxycarbonylazo)diiminoporphodimethene Ni7a (bottom) and diethyl
bis(azocarboxylate) Ni8a (top). Proton assignments shown; 7.26 ppm=residual CHCl3.
However, the ethyl analogue Ni8a was well-behaved and it
was isolated as the pure product in 60% yield. The free base
H28a was also characterized, although again its purification
was imperfect. The successful isolation of pure Ni8a was im-
portant, as it gave us a 5,10-disubstituted porphyrin that pos-
sesses highly polar substituents in the “corner” disposition on
a fully aromatic ring, for comparison with the electronic struc-
ture of DIPD Ni7a. This is helpful for interpreting the electronic
absorption spectra and the theoretical calculations (see below).
Scheme 4. Proposed major conformation of Ni7 in solution.
1H NMR spectra
1
The effects on the H NMR spectra of the replacement of the
porphyrin aromaticity with dipyrrolic cross-conjugation are fa-
miliar from studies of dioxoporphyrins[3] and 5,15-DIPDs.[11a–c]
The signals for the b-pyrrolic protons experience a strong up-
field shift into the region 6.5–7 ppm, while the ortho protons
of the aryl substituents also feel the loss of the ring current
and resonate closer to the CHCl3 signal than those in typical
porphyrins. In the present 5,10-disubstituted case, similar ef-
fects apply, as shown in Figure 1, in which the downfield por-
tions of the spectra of Ni7a and Ni8a are compared. The signal
for the imino-NHs is observed at 8.93 ppm as a broad singlet
that undergoes fast H/D exchange with D2O. The two-dimen-
sional NOESY experiments showed conclusively a conformation-
al preference for a solution structure with the geometry of the
imino substituents as drawn in Scheme 4. As shown in Fig-
ure S1 (in the Supporting Information) there is a strong dipolar
coupling between the carbazate NHs and the neighboring 7,8-
b-pyrrole protons, which is possible only if the NHs point “in-
wards” as shown in Scheme 4. There are some very minor im-
purities visible in the spectra that may result from other rotam-
ers, but these make up less than 5% of the product. The 4-me-
thoxybenzyl carbamate analogue Ni7b displays completely
analogous NMR spectra, including the corresponding NOESY
correlation. The azocarboxylate substituents in the bis(azocar-
boxylate) Ni8a exert a strong electron-withdrawing effect on
the neighboring b-pyrrolic protons, and the 7,8- and 3,12-pro-
tons resonate at 9.71 and 9.36 ppm, respectively. The former
shift is typical of nickel(II) porphyrins with ÀN=NÀ substituents,
and is exceeded in our experience only by those for dinickel
azoporphyrin dimers and tetramer.[15]
The free base analogues H27a and H28a behave similarly,
and the effects of the interruption of the diamagnetic ring cur-
rent are clear from the chemical shifts of the inner NHs. In the
former, two broad exchangeable signals are observed at 12.46
and 11.28 ppm. In addition, small couplings are observed be-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 813 – 824 816

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
tween the NHs and their companion b-pyrrolic protons,
namely 2.0 and 2.4 Hz, respectively. These signals were as-
signed using two-dimensional spectra as the tripyrrole NH and
monopyrrole NH, respectively. Moreover, the signals show the
existence of tautomerism involving the two possible sites for
the former protons. The two NH signals each resolve into pairs
of almost equal broad singlets when the temperature is low-
ered to 228 K in CDCl3 solution (Figure S2). As for the NiII com-
plex Ni7a, dipolar coupling is observed between the carba-
mate NHs (9.22 ppm) and the 7,8-b-pyrrole protons, which sug-
gests the prevalence of “inwards-pointing” geometries for the
hydrazone units (Scheme 4). In the spectrum of the corre-
sponding bis(azocarboxylate) porphyrin H28a, the NH reso-
nates at À0.36 ppm. This position is relatively far downfield
compared with those for typical porphyrins, including 5,10-
H2DAPBr2 (À2.71 ppm) and 5,10-H2DAP itself (À3.27 ppm). The
inner NHs of the mono-azocarboxylate H29 resonate at an in-
termediate position, namely À1.15 ppm, as a very broad
signal. Such large downfield shifts arising from the two ÀN=NÀ
CO2Et substituents could be attributed to a significant diminu-
tion in ring current (perhaps owing to severe out-of-plane dis-
tortion of the macrocycle), or a decrease in pKa of the NH pro-
tons. The former explanation is less plausible, as the 7,8-b-pyr-
rolic protons in the same three compounds appear at 9.94,
9.68, and 8.99 ppm, respectively, an order quite in line with the
chemical shifts of the corresponding protons in the NiII com-
plexes (see above).
UV/visible/near-IR absorption spectra
The effects of interruption of the porphyrin aromaticity on the
electronic absorption spectra of tetrapyrrolic macrocycles are
profound. The electronic asymmetry of either the 5,15- or 5,10-
dioxo- or diiminoporphodimethenes removes the near-degen-
eracy of the transitions from the ground to the first and
second excited electronic states. Thus, the classical Gouter-
man B/Soret (strongly allowed) versus Q (almost forbidden)
model is entirely inappropriate. Instead, there may be signifi-
cant redshifts of the lowest-energy band, together with
marked loss of intensity and/or broadening of the “Soret-like”
bands in the blue region of the visible spectrum. In the pres-
ent case, the spectrum of 5,10-DIPD Ni7a features bands in the
visible region at 410, 702, and 776 nm, with a broad region of
absorption spanning the 500–700 nm region. The spectra of
Ni7a, Ni8a, and the dibromo precursor Ni5 recorded in toluene
are compared in Figure 2a. The spectra were also recorded in
a range of solvents of differing polarities, in view of the
charge-transfer nature of some of the major electronic transi-
tions calculated by TD-DFT (see below), but no pattern of sol-
vatochromism was apparent.
Figure 2. a) Electronic absorption spectra of 5,10-DIPD Ni7a (c), 5,10-bi-
s(azocarboxylate) Ni8a (g), both in toluene, and NiDAPBr2 Ni5 in CH2Cl2
(a; intensity divided by 2.5). b) Electronic absorption spectra of 5,10-DIPD
Ni7a (c) and 5,15-DIPD 3 (M=Ni, R=Et; g) in toluene; the intensities
are normalized to the maximum of the lowest-energy bands.
bands of Ni8a are in marked contrast to the quite narrow and
very typical bands of the dibromo precursor Ni5. In fact, in Fig-
ure 2a the vertical scale of the spectrum of Ni5 has been divid-
ed by 2.5 to fit conveniently on the same plot. As was found
for the analogous 5,15-DIPDs,[11b] the lowest-energy band of
free base H27a is significantly blueshifted (>100 nm) com-
pared with that for the NiII complex Ni7a. In addition, the spec-
trum of rearomatized H28a is not very different from that of
H27a. Such a dramatic effect of nickel(II) complexation has also
been reported for the related 5,15-dioxoporphyrins, namely
5,15-H2OEP(=O)2 (lowest-energy band maximum at 492 nm)
and 5,15-NiOEP(=O)2 (676 nm).[9b]
The final comparison of interest is between the new 5,10-
DIPDs and the previously reported 5,15-isomers. For complete-
ness, we have re-recorded the spectrum of 3 (M=Ni, R=Et) in
toluene, and this spectrum and that of Ni7a are compared in
Figure 2b. The intensities have been normalized to the extinc-
tion coefficient of the lowest-energy band, so the energy shifts
and relative intensities of the bands can be compared more
easily. Here the differences in the electronic structures and
symmetry are apparent: the “3+1” partition of a tetrapyrrolic
A somewhat more typical porphyrin spectrum is restored by
oxidation to the bis(azocarboxylate) Ni8a, the near-IR band
being replaced by a broad, weaker band peaking near 600 nm.
However, the Soret-like band is very broad, presumably reflect-
ing the strong polarity and low symmetry of this derivative,
and the presence of several conformations of similar energies
(see Theoretical Calculations section). The unusually broad
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 813 – 824 817

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
macrocycle produces a large shift to the red compared with
the “2+2” division (776 vs. 609 nm). The spectra do resemble
each other in overall form, in that the ratios of extinction coef-
ficients of the Soret-like bands and Q-like bands are quite simi-
lar and there is the same broad “infill” across the region be-
tween the two major bands. The influence of the extension of
the pyrrolic conjugation by the imino termini is also expected
to differ between the two DIPD systems. These points were ad-
dressed by theoretical calculations of electronic structures and
predictions of electronic absorption spectra (see below).
those of the others. The dibromo compound Ni5 has the larg-
est voltammetric gap, congruent with its blueshifted Q band
(see Figure 2a).
Theoretical calculations of electronic structures and absorp-
tion spectra
In our previous paper on the 5,15-DIPDs, we reported gas-
phase TD-DFT calculations of the electronic spectral transitions
for the methyl ester analogue of 3 (M=Ni).[11b] We have now
performed similar work on model structures Ni7c and Ni8c:
full geometry optimizations, frontier molecular orbital eigenval-
ues, and electronic transitions involving the 32 lowest singlet
excited states. The models differ from the actual structures
only in the deletion of the peripheral tert-butyl groups and the
substitution of methyl for ethyl on the carbamates. Details of
the computational methods are provided in the Experimental
Section. In our detailed theoretical study of a series of dizinc(II)
diporphyrins covalently linked by ethane, ethene, ethyne,
imine, and azo bridges,[18] we found better agreement with ex-
perimental spectra by including the solvent in the calculations
by means of the polarizable continuum model, rather than re-
lying on gas-phase structures. So in the present work we have
recorded the experimental spectra of Ni7a, Ni8a, and the 5,15-
analogue 3 (M=Ni and R=Et) in toluene and performed all
the calculations on the model structures using the same
medium.
Electrochemistry
We have not located any voltammetric data for 5,10-dearomat-
ized porphyrinoids in the literature, so we subjected the series
of nickel(II) complexes 5,10-NiDAP, Ni5, Ni7a, and Ni8a to
cyclic voltammetry and differential pulse voltammetry (DPV) in
dichloromethane (CH2Cl2). Unfortunately, the compounds did
not behave tractably in the former measurements, and despite
their intrinsically high solubilities, they caused problems of
electrode coating and/or aggregation. The DPV results also
were not completely satisfactory, but some of the redox waves
were defined, as listed in Table 1. The voltammograms are pre-
sented in Figure S3. A rigorous comparison among the four
Table 1. Differential pulse voltammetry results.[a]
Because the dispositions of the imino substituents in Ni7a
and Ni7b are so clear in the solution NMR spectrum, only
structures for Ni7c in which the imino NHs are disposed as in
Scheme 4 were subjected to optimizations with respect to ro-
tamers about the amide bonds. Other orientations possessed
much higher energies, so were ignored. Only these three were
included in the calculated absorption spectrum to be dis-
cussed below. The global minimum geometry, having
C2 symmetry and syn arrangements of both N=NHÀC=O units,
is shown in Figure 3a. Two other rotamers, having respectively
anti,syn and anti,anti geometries, were calculated to lie 0.31
and 0.85 kJmolÀ1 higher in energy. In all conformers, the mac-
rocycle is strongly distorted in the ruffled sense;[19] for exam-
ple, in the global minimum conformer, the hydrazone carbon
atoms C5 and C10 lie Æ1.035 ꢁ out of the C20N4 mean plane,
and the aryl-attached C15 and C20 are out of that plane by
Æ0.755 ꢁ.
Compound
E0red2 [V][b]
À1.23
E0red1 [V][b]
E0ox1 [V][b]
DE [V][c]
NiDAP
Ni5
Ni7a
Ni8a
À0.78
À1.06
À0.43
À0.37
+1.09
+1.15
+1.01
+1.35
1.87
2.21
1.44
1.72
[d]
–
À0.79
[d]
–
[a] In
a
0.1 molLÀ1 solution of [Bu4N][ClO4] in CH2Cl2 containing
0.5 mmolLÀ1 of sample; referenced to ferrocenium/ferrocene (Fc+/Fc)
+
1
couple (E = (Fc /Fc)=0.501 V vs. SCE). [b] Estimated as the average of
2
peak potentials of anodic and cathodic scans. [c] DE=E0ox1ÀE0
.
red1
[d] Poorly defined waves, not measured.
compounds is not possible, as in no case were both the first
oxidation and first reduction waves free of complications. The
first reductions are easiest for the DIPD Ni7a and bis(azocar-
boxylate) Ni8a, whereas the bromo substituents (NiDAP cf.
Ni5) are strongly electron-donating with respect to the addi-
tion of an electron to the macrocycle. On the other hand, the
first oxidation potential is less sensitive to the structural
changes, although clearly the hardest to oxidize is the bis(azo-
carboxylate) Ni8a, which is expected considering the strongly
electron-withdrawing nature of the substituents. The compari-
son of redox potentials with the energies of the Q-band
maxima is often worthwhile for checking internal consistency,
although the two estimates of the HOMO–LUMO energy gaps
are almost never directly related. The voltammetric gaps DE=
On the other hand, for the bis(azocarboxylate) Ni8c, we
have no evidence for preferred solution conformations, so
both CmesoÀN=N and =NÀC=O single bond rotamers were ex-
plored in the geometry optimization. Thus ten local minimum
structures were considered, and their geometric parameters
are described in Table S1. These were labeled in accordance
with the relative arrangements of the monopyrrole ring, the
N=N vectors, and the C=O vectors, when the foremost is point-
ing down the page. The results show that there are three
structures close to the global minimum energy, which account
for about 93% of the population. Only these three were in-
cluded in the calculated absorption spectrum to be discussed
below. The out-of-plane distortion is again of the ruffled type,
E0ox1ÀE0
for the four compounds are compared with the
red1
spectroscopic data in Figure S4. The two sets of data do trend
in the same directions, and certainly the HOMO–LUMO gap of
the dearomatized macrocycle Ni7a is markedly smaller than
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 813 – 824 818

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
Figure 4. Frontier molecular orbitals calculated by DFT for the optimized ge-
ometry of the model 5,10-DIPD Ni7c.
Figure 3. Dispositions of the substituents in the calculated minimum-energy
conformations of a) Ni7c viewed along the C2 axis and b) Ni8c viewed per-
pendicular to the C2 axis.
as is commonly found for NiII porphyrins in any case, but be-
cause of the bulky substituents at C5 and C10, these structures
are even more distorted than simpler tetraaryl or -alkyl NiII por-
phyrins (Figure 3b). Considering all ten structures, the out-of-
plane distances of C5 and C10 span a range of 0.734 to
0.774 ꢁ, whereas for C15 and C20 the range is 0.650 to
0.682 ꢁ. The dispositions of the CmesoÀN=NÀC=O moieties
clearly influence the total energies significantly, as the ten
structures span an energy range of 32.2 kJmolÀ1.
Figure 5. Frontier molecular orbitals calculated by DFT for the lowest-energy
conformer of the model 5,10-bis(azocarboxylate) Ni8c.
The four frontier orbitals involved in the low-energy elec-
tronic transitions for the lowest-energy conformers of Ni7c and
Ni8c are pictured in Figures 4 and 5, respectively. The unusual
tripyrrole+monopyrrole framework of the 5,10-DIPD macrocy-
cle in Ni7c results in a very interesting composition of the
HOMO, LUMO, and LUMO+1 orbitals, and although Ni8c is
a “normal” porphyrin, it possesses unusual electronic distor-
tions owing to the 5,10-substitution. As this porphyrinoid sub-
stitution pattern and aromaticity interruption have not previ-
ously been subjected to theoretical examination, we will de-
scribe the orbitals and the electronic transitions in some detail.
The orbital eigenvalues for the three lowest-energy conformers
of Ni7c and Ni8c are listed in Table S2. The HOMO of Ni7c is
concentrated on the imino substituents and the monopyrrole
ring, whereas both the LUMO and LUMO+1 have almost no
electron density on the latter. In the HOMO there are nodes
across the C5ÀC6 and C9ÀC10 bonds, which implies linear con-
jugation extending from the C=O of one hydrazone unit right
through the tripyrrole to the other terminal hydrazone, thus
isolating the unique pyrrole ring. Notably, the LUMO and
LUMO+1 have almost zero contribution from this pyrrole unit.
In stark contrast, as expected, the porphyrin Ni8c has a more
conventional pattern in its frontier orbitals, although remarka-
bly, the HOMO has no contribution from p orbitals of the con-
jugated substituent. It is recognizable as akin to the a1u frontier
orbital of a D4h metalloporphyrin, being concentrated on the
pyrrole carbon atoms. The HOMOÀ1, which lies quite close in
energy, is similarly reminiscent of the a2u orbital, concentrated
on the meso carbon atoms and the inner N atoms. The isola-
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 813 – 824 819

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
tion of the monopyrrole ring is still apparent in the LUMO and
LUMO+1.
unusually large LUMO to LUMO+1 gap, as listed in Table S2.
This results in two changes relative to the other two conform-
ers, DUUD and DUUU, which are very similar to each other in
orbital energies and hence transition energies and oscillator
strengths. Conformer UDDU has a smaller HOMO–LUMO gap,
a redshifted absorption band, the transition of which is more
purely “HOMO–LUMO” in composition (89 vs. 78/79%), and
a much higher oscillator strength (0.0567 vs. 0.0108 and
0.00550, respectively). It can be seen in Figure 7b that the
form of the experimental spectrum is fairly well reproduced in
the region >450 nm, with predicted transitions even coincid-
ing almost exactly with the Soret maximum, an eventuality not
normally arising from TD-DFT calculations on porphyrins. How-
ever, as for Ni7c, this correspondence does not extend further
into the higher-energy region.
The characteristics of the lowest-energy transition for both
structures can be illustrated using electron density difference
maps; these are shown for the lowest-energy conformers of
Ni7c and Ni8c in Figure 6. The strongly redshifted and rather
Finally, the comparison with the 5,15-DIPD 3 (M=Ni, R=
Me), the experimental and predicted spectra of which are over-
laid in Figure 7c, shows how “electronic cleavage” of porphyri-
noid aromaticity in the 2+2 manner leads to a smaller redshift
of the HOMO–LUMO transition than does the 3+1 separation.
This is not surprising when the more extended conjugation in
the 5,10-isomer revealed by the frontier orbitals shown in
Figure 6 is considered.
Figure 6. Electron density difference maps for the ground state to first excit-
ed state electronic transitions for the DFT-optimized global minimum-energy
structures of a) model 5,10-DIPD Ni7c: 92% HOMO–LUMO; b) model 5,10-bi-
s(azocarboxylate) Ni8c: 79% HOMO–LUMO. Blue: electron-poor region after
transition; orange: electron-rich region after transition.
intense (f ca. 0.15) lowest-energy absorption band, largely
composed of the HOMO–LUMO transition, results in a marked
cross-ring redistribution of electrons, away from the isolated
pyrrole ring and the hydrazone groups on the 5,10-positions,
towards the tripyrrole region. For Ni8c, on the other hand, the
much weaker transition (f<10% of that for Ni7c) is more or
less the same, but in the reverse sense, from the tripyrrole
region towards the monopyrrole/azocarboxylate side of the
ring.
Conclusion
The palladium-catalyzed coupling of hydrazine derivatives (in
this case, carbazates) with 5,10-dibromoporphyrins is a simple
route to a new class of 5,10-dearomatized porphyrinoids,
namely diiminoporphodimethenes (DIPDs), with attached car-
bamates. These derivatives were prepared as both NiII com-
plexes and free bases, although the former are, as is typical for
many porphyrins, more stable and more easily purified. The
DIPDs can be further transformed by DDQ oxidation to rearo-
matize the macrocycle, forming novel 5,10-bis(azocarboxy-
lates). The key compounds were characterized spectroscopical-
ly, electrochemically, and by theoretical calculations of geome-
tries, frontier molecular orbitals, and electronic transitions by
using DFT and TD-DFT methods. So far, we have not been able
to obtain single crystals of sufficient quality to enable X-ray
structure determinations of any of the compounds in this
study. The calculations afforded excellent predictions of the ab-
sorption spectra in the low-energy regions, but the corre-
spondence of theory and experiment at higher energies was
less impressive. The consideration of conformers close in
energy to the global minimum was important for the bis(azo-
carboxylate) Ni8c, as one of these exhibited a redshifted transi-
tion that was also apparent in the experimental spectrum of
Ni8a. The frontier orbitals of Ni7c revealed conjugation across
the whole O=CÀN=NÀtripyrroleÀN=NÀC=O fragment of the
DIPD molecule, which results in a lowest-energy absorption
band near 780 nm. In contrast, the analogous 5,15-DIPD exhib-
its its lowest-energy absorption band near 600 nm. In view of
the significant cross-ring charge-transfer nature of the HOMO–
LUMO transition for 5,10-DIPDs, the nonlinear optical behavior
may be of theoretical interest, but practical application would
be somewhat problematic in view of the relative difficulty of
Turning now to the predicted electronic absorption spectra,
the three lowest-energy conformers of both Ni7c and Ni8c, as
well as the global minimum-energy C2 geometry of 3 (M=Ni,
R=Me), were simulated using TD-DFT calculations; the former
two were Boltzmann-weighted to show the respective intensi-
ty contributions. The experimental spectra recorded in toluene
and the most intense calculated transitions in the visible/near-
IR region are compared in Figure 7. The transitions making up
the predicted spectra are collected in Table S3 (Ni7c), Table S4
(Ni8c), and Table S5 (3). The calculated near-IR transitions for
the Ni7c conformers are almost indistinguishable, and the pre-
dicted peak maximum near 12800 cmÀ1 (780 nm) corresponds
very closely in energy with the experimental value for Ni7a
(776 nm, see Figure 7a). The overall form of the spectrum is re-
produced in the >500 nm range, but the calculations simulate
the Soret region very poorly; the predicted splitting into bands
near 430 and 380 nm is not present in the real spectrum, al-
though the latter does exhibit a very broad envelope across
this region.
It was noted above that the experimental spectrum of Ni8a
is much broader than that of typical nickel(II) porphyrins, even
when 5,10-disubstituted (cf. Ni5). The inclusion of the three
major conformers goes some way to explaining this result. The
broad tail on the red edge of the spectrum is apparently
a result of the contribution of the third-lowest-energy confor-
mer, which we defined as UDDU (Table S1), which exhibits an
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 813 – 824 820

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
preparing and purifying the pre-
cursor porphyrins. If this hin-
drance could be overcome, our
results imply that these systems
are worthy of deeper investiga-
tion.
Experimental Section
General procedures
Chemicals used were of Analytical
Reagent grade and purchased
from Sigma–Aldrich. The following
solvents were distilled under argon
prior to use: CH2Cl2 from CaH2, tet-
rahydrofuran (THF) from sodium
wire and benzophenone, and pyr-
role without additives. Triethyla-
mine (TEA) and pyridine were
stored over KOH. Solvents were
evaporated under reduced pres-
sure at 408C. Analytical TLC was
performed using Merck Silica
Gel 60 F254 plates. For preparative
column chromatography, Merck
silica gel (230–400 mesh) was
used. 1H NMR spectra were mea-
sured on a Bruker Avance instru-
ment operating at 400 MHz. Unless
otherwise stated, samples were
prepared in deuterated chloro-
form. The chemical shifts are re-
ported in ppm referenced against
the residual chloroform (CHCl3)
peak at 7.26 ppm. Coupling con-
stants are given in hertz. UV/visible
spectra were recorded in CH2Cl2
for all samples, and methanol,
hexane, or toluene for solvato-
chromism studies, on
a Varian
Cary 50 instrument. Liquid secon-
dary ion (LSI) and electrospray ioni-
zation (ESI) mass spectra were re-
corded at the Organic Mass Spec-
trometry Facility, University of Tas-
mania, Hobart. LSIMS data were
obtained on a Kratos ISQ double
focusing magnetic sector mass
spectrometer using a direct inser-
tion probe and fitted with an
LSIMS ion source, using 10 kV
cesium ions as the primary beam
and m-nitrobenzyl alcohol as the
liquid matrix. ESI spectra were ob-
tained on a Thermo Electron LTQ
Orbitrap mass spectrometer at
Monash University, Melbourne.
Samples were dissolved in CH2Cl2
and then diluted 1:100 in metha-
nol before being infused into the
mass spectrometer at a flow rate
of 50 mLminÀ1. A voltage of 4500 V
Figure 7. Comparison of experimental and calculated electronic absorption spectra (in toluene, proportioned ac-
cording to the Boltzmann populations). a) DIPD Ni7a (c) and TD-DFT calculated transitions for the model Ni7c
~
*
*
(vertical bars, symbols: conformer syn,syn ( ); syn,anti ( ); anti,anti ( )); b) bis(azocarboxylate) Ni8a (c) and
~
*
TD-DFT calculated transitions for the model Ni8c (vertical bars, symbols: conformer DUUD ( ); DUUU ( ); UDDU
*
(
)); c) 5,15-DIPD 3 (M=Ni, R=Et) (c) and TD-DFT calculated transitions for the model with R=Me (vertical
*
bars, ( )). See text and Table S1 for explanation of the vertical scales and conformer coding.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 813 – 824 821

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
was applied at the tip. Isotopic modeling was calculated by Mas-
sLynx V3.5 software by Micromass Limited and online using Isoto-
pident on the Expert Protein Analysis System (ExPASY) server. IR
spectra were recorded using a Nicolet 870 Nexus Fourier transform
infrared (FTIR) system equipped with an attenuated total internal
reflectance (ATR) accessory with a single-bounce diamond cell.
jected to column chromatography using CH2Cl2/ethyl acetate/TEA
(200:100:1) as eluent. Two major bands were collected. The first
green band was the target compound Ni6a. The second band,
which began as purple, slowly developed a green leading edge,
and when isolated this fraction proved to be a mixture of Ni7a
and a nickel(II) porphyrin, presumably the bis(carbazate) Ni6a (see
Results and Discussion). This fraction was re-chromatographed as
follows. The sample was dissolved in CH2Cl2, this solution was ap-
plied to the column, and some less-polar contaminants were
eluted with CH2Cl2, during which the major component(s) were
largely immobile. Then the eluent was changed to CH2Cl2/ethyl
acetate/TEA (6:1:0.01), which eluted a concentrated narrow band
containing the pure green product Ni7a. The compound was diffi-
cult to crystallize without severe losses, so generally it was left as
a dark green residue after evaporation, washing with pentane, and
Electrochemical measurements were performed with a BAS CV-
50W voltammetric analyzer. The cell comprised inlets for a glassy
carbon disk working electrode of 3.0 mm diameter and a silver-
wire counter electrode. The reference electrode was Ag/Ag+
(0.01 molLÀ1), which was connected to the solution by a Luggin ca-
pillary, the tip of which was placed close to the working electrode.
It was corrected for junction potentials by being referenced inter-
+
+
1
nally to the ferrocenium/ferrocene (Fc /Fc) couple [E = (Fc /Fc)=
2
0.501 V vs. SCE]. Typically, a 0.1 molLÀ1 solution of [Bu4N][ClO4] in
CH2Cl2 containing 0.5 mmolLÀ1 of sample was purged with nitro-
gen for 5 min, then the voltammogram was recorded at ambient
temperature. Porphyrin starting material H2DAP was prepared as
previously described.[15b]
1
drying under high vacuum (14 mg, 67%). H NMR: d=8.93 (bs, 2H;
3
3
NHcarbazate), 7.52 (t, J=1.8 Hz, 2H; p-Haryl), 7.24 (d, J=1.8 Hz, 4H; o-
Haryl), 7.11 (s, 2H; 7,8-H), 7.09 (d, 3J=4.8 Hz, 2H; 3,12-H), 6.94 (d,
3J=4.8 Hz, 2H; 2,13-H), 6.39 (s, 2H; 17,18-H), 4.38 (q, 3J=7.1 Hz,
4H; CH2), 1.39 (t, 3J=7.1 Hz, 6H; CH3), 1.35 ppm (s, 36H; tBu-H);
13C NMR: d=162.1, 150.2, 145.9, 143.6, 141.7, 137.5, 135.2, 128.4,
126.1, 124.9, 124.3, 122.9, 113.8, 62.4, 34.8, 31.4, 14.5 ppm; UV/Vis
(CH2Cl2): lmax (log e)=348 (4.48), 410 (4.76), 702 (3.97), 776
(4.27) nm; MS (ESI): m/z calcd for C54H62N8NiO4 +H+: 945.4326
[M+H+]; found: 945.4338.
Synthesis
5,10-Dibromo-15,20-bis(3,5-di-tert-butylphenyl)porphyrin
5,10-Bis(3,5-di-tert-butylphenyl)porphyrin (H2DAP;
(H25):
100 mg,
146 mmol) was dissolved in chloroform (5 mL) and stirred in an ice
bath. N-Bromosuccinimide (65 mg, 2.5 equiv, 363 mmol) and pyri-
dine (294 mL, 3.63 mmol) were added. The solution was stirred for
1 h at room temperature then the reaction was quenched with
acetone. The solvent was removed under vacuum. The residue was
suspended in methanol (5 mL), collected on a cotton wool plug in
a Pasteur pipette, and then washed with additional methanol to
remove succinimide. The collected solid was eluted with CH2Cl2,
which was removed under vacuum. After confirmation of its struc-
5,10-N,N’-Bis(ethoxycarbonylazo)-15,20-bis(3,5-di-tert-butylphenyl)-
porphyrinatonickel(II) (Ni8a): Compound Ni7a (15 mg, 16 mmol)
was dissolved in CH2Cl2 (1.5 mL) and stirred. DDQ (13 mg, 3 equiv,
55 mmol) was added to the solution. There was an immediate color
change from green to brown. TLC showed the starting material
was consumed within 10 min, so the reaction mixture was applied
directly to a chromatography column and the product was eluted
with CH2Cl2. A single greenish-brown compound was isolated,
which was recrystallized from CH2Cl2/MeOH to give very fine fi-
brous, dark brown crystals (8 mg, 58%). 1H NMR: d=9.71 (s, 2H;
7,8-H), 9.36 (d, 3J (H,H)=5.0 Hz, 2H; 3,12-H), 8.72 (d, 3J (H,H)=
1
ture and purity by H NMR spectroscopy, the product was subject-
ed to nickel(II) insertion. 1H NMR: d=9.68 (s, 2H; 7,8-H), 9.61 (d,
3
3J=4.6 Hz, 2H; 3,12-H), 8.89 (d, J=4.6 Hz, 2H; 2,13-H), 8.81 (s, 2H;
3
3
3
17,18-H), 8.00 (d, J=1.7 Hz, 4H; o-Haryl), 7.81 (t, J=1.7 Hz, 2H; p-
Haryl), 1.53 (s, 36H; tBu-H), À2.71 ppm (bs, 2H; inner NH). The data
agree with those previously reported.[16]
5.0 Hz, 2H; 2,13-H), 8.42 (s, 2H; 17,18-H), 7.75 (d, J (H,H)=1.8 Hz,
3
3
4H; o-Haryl), 7.73 (t, J (H,H)=1.8 Hz, 2H; p-Haryl), 4.72 (q, J (H,H)=
7.0 Hz, 4H; CH2), 1.63 (t, 3J (H,H)=7.0 Hz, 6H; CH3), 1.47 ppm (s,
36H; tBu-H); 13C NMR: d=161.7, 149.4, 143.8, 141.6, 141.1, 140.5,
138.1, 136.1, 135.5, 133.2, 132.4, 128.1, 127.2, 125.7, 121.9, 64.4,
35.0, 31.6, 14.4 ppm; UV/Vis (CH2Cl2): lmax (log e)=345 (4.40), 415
(4.63), 488 (4.89), 605 (4.05) nm; MS (ESI): m/z calcd for
C54H60N8NiO4 +H+: 943.4169 [M+H+]; found: 943.4178.
5,10-Dibromo-15,20-bis(3,5-di-tert-butylphenyl)porphyrinatonicke-
l(II) (Ni5): Compound H25 (80 mg, 95 mmol) was dissolved in tolu-
ene (4 mL), then [Ni(acac)2] (80 mg, 0.31 mmol; acac=acetylaceto-
nate) was added. The solution was stirred and heated at reflux.
After 3 h the starting material was consumed. The toluene was re-
moved and the residue purified by column chromatography using
CH2Cl2 as eluent. The product was recrystallized from CH2Cl2/MeOH
N,N’-Bis(4-methoxybenzyloxycarbonylamino)-15,20-bis(3,5-di-tert-
butylphenyl)-5,10-diiminoporphodimethenatonickel(II) (Ni7b): This
analogue was prepared as for Ni7a above, by substituting ethyl
carbazate with 4-methoxybenzyl carbazate. The crude product,
containing the target compound and partially oxidized com-
pounds, was kept in CHCl3 solution under air for 4 days, during
which the brown/purple solution became green. After evaporating
the solvent, the residue was subjected to chromatography using
CHCl3/ethyl acetate (1:9) to elute the major olive-green band. The
solvents were removed under vacuum and the residue was triturat-
ed with pentane and dried under high vacuum to leave a dark
green powder. 1H NMR: d=8.92 (bs, 2H; NHcarbazate), 7.48 (t, 3J=
1.8 Hz, 2H; p-Haryl), 7.36 (d, 3J=8.8 Hz, 4H; Ar-H), 7.19 (d, 3J=
1
to give tiny red-purple crystals (77 mg, 90%). H NMR: d=9.53 (s,
3
3
2H; 7,8-H), 9.46 (d, J=4.8 Hz, 2H; 3,12-H), 8.80 (d, J=4.8 Hz, 2H;
2,13-H), 8.72 (s, 2H; 17,18-H), 7.82 (bs, 4H; o-Haryl), 7.74 (bs, 2H; p-
Haryl), 1.49 ppm (s, 36H, tBu-H); UV/Vis (CH2Cl2): lmax (log e)=421
(5.36), 531 (4.24) nm; MS (ESI): m/z calcd for C48H50Br2N6Ni:
901.1769 [M+]; found: 901.1795.
N,N’-Bis(ethoxycarbonylamino)-15,20-bis(3,5-di-tert-butylphenyl)-
5,10-diiminoporphodimethenatonickel(II) (Ni7a): Compound Ni5
(20 mg, 22 mmol) together with ethyl carbazate (23 mg, 10 equiv,
0.22 mmol), Cs2CO3 (50 mg, 7 equiv, 0.16 mmol), Pd(OAc)2 (1 mg,
20 mol%, 4.5 mmol), and DPEphos (10 mg, 55 mol%, 17.6 mmol)
were placed in a Schlenk flask, dried under vacuum, then the flask
was filled with argon and freshly distilled THF (5 mL) was added.
TLC analysis showed that even after a few minutes, some of the
desired product was apparent. The mixture was stirred at 678C for
2 h under Ar. The solvent was removed and the residue was sub-
3
1.8 Hz, 4H; o-Haryl), 7.03 (d, J=4.8 Hz, 2H; 3,12-H), 6.99 (s, 2H; 7,8-
3
3
H), 6.89 (d, J=4.8 Hz, 2H; 2,13-H), 6.89 (d, J=8.8 Hz, 4H; Ar-H),
6.35 (s, 2H; 17,18-H), 5.23 (s, 2H; CH2), 3.80 (s, 3H; OCH3),
1.31 ppm (s, 36H; tBu-H); 13C NMR: d=162.0, 159.8, 150.2, 145.9,
132.6, 141.6, 137.4, 135.1, 130.4, 128.4, 127.7, 126.0, 124.9, 124.3,
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 813 – 824 822

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
122.9, 113.9, 67.7, 55.3, 34.8, 31.4 ppm; IR: n˜ =3318 (NH), 1726,
1751 (CO) cmÀ1; UV/Vis (toluene): lmax (log e)=411 (4.73), 707
(4.10), 776 (4.40) nm; MS (ESI): m/z calcd for C66H71N8NiO6 +H+:
1129.4845 [M+H+]; found: 1129.4874.
(1.06) nm; MS (ESI): m/z calcd for C54H62N8O4 +H+: 887.4972
[M+H+]; found: 887.4965.
Theoretical calculations
N,N’-Bis(ethoxycarbonylamino)-15,20-bis(3,5-di-tert-butylphenyl)-
5,10-diiminoporphodimethene (H27a) and 5-N-(ethoxycarbonyla-
zo)-10,15-bis(3,5-di-tert-butylphenyl)porphyrin (H29): Porphyrin H25
(20 mg, 24 mmol) together with ethyl carbazate (25 mg, 10 equiv,
0.236 mmol), Cs2CO3 (49 mg, 7 equiv, 0.165 mmol), Pd(OAc)2 (ca.
1 mg, 20 mol%, 4.5 mmol), and DPEphos (7 mg, 55 mol%,
12.9 mmol) were placed in a Schlenk flask and dried under high
vacuum. The flask was charged with argon and freshly distilled
THF (5 mL) was added. The solution was stirred at 678C for 2 h.
The solvent was removed and the residue washed with water and
extracted with CH2Cl2. The solvent was removed and the residue
was purified by column chromatography using CH2Cl2/TEA (100:1)
as eluent. The first light green band comprised compound H29,
which was recrystallized from CH2Cl2/MeOH to give a sticky green
All DFT calculations were performed using the Gaussian 09 suite of
programs.[20] The initial geometries of compounds Ni7c and Ni8c
were calculated semiempirically using the PM3 method. Further
optimizations were performed using the Becke three-parameter
Lee–Yang–Parr (B3LYP) hybrid functional[21] at the 6-31G(d,p) level.
The solvent effect of toluene was included using the integral equa-
tion formalism variant of the polarizable continuum model
(IEFPCM). Ten potential conformers of Ni8c were identified by vari-
ous rotations about the C5ÀN, C10ÀN, and CNs of their respective
azocarboxylate groups while maintaining the substituent in a more
or less planar configuration. For Ni7c the geometries were limited
to conformers in which the NH of the carbamate group pointed to-
wards the pyrrole that connected the two substituents, to corre-
spond with the NMR results. Three conformers fitting this criterion
were found by rotation about the NHÀC=O bonds. A local mini-
mum was found for the ten conformers of Ni8c and three con-
formers of Ni7c as confirmed by the absence of imaginary frequen-
cies in the subsequent vibrational frequency calculations. Electronic
transitions for the three lowest-energy conformers were simulated
using TD-DFT also at the 6-31G(d,p) level. Frontier molecular orbi-
tals were displayed using an isovalue of 0.02 electronsꢁÀ3. Electron
density difference maps for the HOMO to LUMO transition were
constructed with the aid of the GaussSum program[22] and dis-
1
3
powder (1 mg, 5%). H NMR: d=10.03 (s, 1H; 20-H), 10.02 (d, J=
3
3
4.8 Hz, 1H; 3-H), 9.90 (d, J=4.8 Hz, 1H; 7-H), 9.31 (d, J=4.8 Hz,
3
3
1H; 2-H), 9.11 (d, J=4.8 Hz, 1H; 17-H), 8.88 (d, J=4.8 Hz, 1H; 18-
3
3
H), 8.87 (d, J=4.8 Hz, 1H; 8-H), 8.75 (d, J=4.8 Hz, 1H; 13-H), 8.66
(d, 3J=4.8 Hz, 1H; 12-H), 8.02 (d, 3J=1.7 Hz, 2H; o-Haryl), 8.00 (d,
3J=1.7 Hz, 2H; o-Haryl), 7.82 (m, 2H; p-Haryl), 4.80 (q, J=7.2 Hz, 2H;
3
CH2), 1.70 (t, 3J=7.2 Hz, 3H; CH3), 1.55 (s, 18H; tBu-H), 1.54 (s,
18H; tBu-H), À1.15 ppm (vbr, 2H; inner NH); UV/Vis (CH2Cl2): lmax
(log e)=422 (4.84), 442 (4.90), 603 (4.14), 679 (3.85) nm; MS (ESI):
m/z calcd for C51H58N6O2 +H+: 787.4700 [M+H+]; found: 787.4688.
The eluent was changed to CH2Cl2/TEA/methanol (100:1:2) and
a second blue fraction was isolated, which comprised mainly com-
pound H27a. The residue after evaporation of solvents was crystal-
lized from CH2Cl2/MeOH to give a blue-green sticky powder
(20 mg), but NMR spectra showed it to be still contaminated with
played using an isovalue of 0.0004 electronsꢁÀ3
.
Acknowledgements
1
ethyl carbazate and other unknown impurities. H NMR: d=12.46
We thank the Australian Research Council for financial support
through Discovery Grant DP0663774, the High Performance Com-
puting team, Queensland University of Technology (QUT) for their
support of the Gaussian 09 program, and Prof. K.-i. Sugiura for
collaboration on the corner porphyrin synthesis. B.B. and O.B.L.
thank QUT for Postgraduate Scholarships. J.J. thanks NSFC,
China, for financial support.
(bs, 1H; inner NH on tripyrrole rings), 11.28 (bs, 1H; inner NH on
3
monopyrrole ring), 9.22 (bs, 2H; NHcarbazate), 7.51 (t, J=1.6 Hz, 2H;
3
3
p-Haryl), 7.40 (d, J=4.6 Hz, 2H; 3,12-H), 7.27 (d, J=1.6 Hz, 4H; o-
3
3
Haryl), 7.05 (bd, J=2.0 Hz, 2H; 7,8-H), 6.85 (d, J=4.6 Hz, 2H; 2,13-
3
3
H), 6.28 (bd, J=2.4 Hz, 2H; 17,18-H), 4.39 (q, J=7.2 Hz, 4H; CH2),
1.41 (t, 3J=7.2 Hz, 6H; CH3), 1.36 ppm (s, 36H; tBu-H); UV/Vis
(CH2Cl2): lmax (intensity ratios)=316 (2.08), 393 (3.59), 589 (1.00),
633 (1.13) nm; MS (ESI): m/z calcd for C54H64N8O4 +H+: 889.5129
[M+H+]; found: 889.5119.
Keywords: amination
· density functional calculations ·
diiminoporphodimethenes · macrocycles · porphyrinoids
5,10-N,N’-Bis(ethoxycarbonylazo)-15,20-bis(3,5-di-tert-butylphenyl)-
porphyrin (H28a): DIPD H27a (25 mg, 28 mmol) was dissolved in
CH2Cl2 (2.5 mL) and stirred. DDQ (22 mg, 3.5 equiv, 98 mmol) was
added to the solution. There was an immediate color change from
blue to brownish green. TLC showed a major product and the
starting material. The reaction was monitored for 2 h but there
was no change with respect to the initial TLC. The reaction mixture
was purified by column chromatography (CH2Cl2) to give three
fractions. The first was compound H29, the second was mainly
compound H28a, and the third was the starting material H27a.
Compound H28a was recrystallized from CH2Cl2/MeOH to give
a dark green powder (6 mg), the NMR spectrum of which showed
it to be contaminated with H27a and another ethyl ester-contain-
ing porphyrinoid. 1H NMR: d=9.94 (s, 2H; 7,8-H), 9.63 (d, 3J=
4.8 Hz, 2H; 3,12-H), 8.80 (d, 3J=4.8 Hz, 2H; 2,13-H), 8.55 (s, 2H;
[1] For compilations of references for various modified porphyrinoids, see:
a) J. M. Lim, M.-C. Yoon, K. S. Kim, J.-Y. Shin, D. Kim in Handbook of Por-
phyrin Science, Vol. 1 (Eds.: K. M. Kadish, K. M. Smith, R. Guilard), World
Scientific Publishing, Singapore, 2010, pp. 507–558; b) J. Waluk in
Handbook of Porphyrin Science, Vol. 7 (Eds.: K. M. Kadish, K. M. Smith, R.
Guilard), World Scientific Publishing, Singapore, 2010, pp. 359–435;
c) M. Toganoh, H. Furuta in Handbook of Porphyrin Science, Vol. 2 (Eds.:
K. M. Kadish, K. M. Smith, R. Guilard), World Scientific Publishing, Singa-
´
pore, 2010, pp. 295–367; d) M. Pawlicki, L. Latos-Graz˙ynski in Handbook
of Porphyrin Science, Vol. 2 (Eds.: K. M. Kadish, K. M. Smith, R. Guilard),
World Scientific Publishing, Singapore, 2010, pp. 103–192; e) T. D. Lash
in Handbook of Porphyrin Science, Vol. 16 (Eds.: K. M. Kadish, K. M. Smith,
R. Guilard), World Scientific Publishing, Singapore, 2012, pp. 1–329; f) J.
Mack, N. Kobayashi, Z. Shen in Handbook of Porphyrin Science, Vol. 23
(Eds.: K. M. Kadish, K. M. Smith, R. Guilard), World Scientific Publishing,
Singapore, 2012, pp. 281–372.
3
3
17,18-H), 7.96 (d, J=1.8 Hz, 4H; o-Haryl), 7.83 (t, J=1.8 Hz, 2H; p-
Haryl), 4.80 (q, 3J=7.0 Hz, 4H; CH2), 1.70 (t, 3J=7.0 Hz, 6H; CH3),
1.53 (s, 36H; tBu-H), À0.36 ppm (bs, 2H; inner NH); UV/Vis (CH2Cl2):
lmax (intensity ratios)=317 (2.43), 396 (3.80), 605 (1.00), 657
[2] C.-H. Lee, V. Roznyatovskiy, S.-J. Hong, J. L. Sessler in Handbook of Por-
phyrin Science, Vol. 13 (Eds.: K. M. Kadish, K. M. Smith, R. Guilard), World
Scientific Publishing, Singapore, 2011, pp. 197–252.
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 813 – 824 823

CHEMPLUSCHEM
FULL PAPERS
www.chempluschem.org
[3] a) G. H. Barnett, B. Evans, K. M. Smith, Tetrahedron 1975, 31, 2711–2717;
[12] S. R. Harper, J. C. McMurtrie, D. P. Arnold, unpublished results.
b) D. Lahaye, K. Muthukumaran, C.-H. Hung, D. Gryko, J. S. RebouÅas, I.
[13] A. M. V. M. Pereira, M. G. P. M. S. Neves, J. A. S. Cavaleiro, C. Jeandon, J.-P.
Gisselbrecht, S. Choua, R. Ruppert, Org. Lett. 2011, 13, 4742–4745.
[14] a) M. J. Smith, DPhil Thesis, University of Oxford, Oxford (United King-
dom), 2004; b) K.-i. Yamashita, S. Takeuchi, M. S. Asano, K.-i. Sugiura,
Book of Abstracts, 7th International Conference on Porphyrins and Phtha-
locyanines (ICCP7) 2012, 583.
´
´
Spasojevic, I. Batinic-Haberle, J. S. Lindsey, Bioorg. Med. Chem. 2007, 15,
7066–7086; c) J.-H. Fuhrhop, E. Baumgartner, H. Bauer, J. Am. Chem.
Soc. 1981, 103, 5854–5861; d) M. O. Senge, K. M. Smith, Z. Naturforsch.
B: J. Chem. Sci. 1993, 48, 991–999; e) D. W. J. McCallien, J. K. M. Sanders,
J. Am. Chem. Soc. 1995, 117, 6611–6612; f) Z. Clyde-Watson, N. Bampos,
J. K. M. Sanders, New J. Chem. 1998, 22, 1135–1138.
[15] a) L. J. Esdaile, P. Jensen, J. C. McMurtrie, D. P. Arnold, Angew. Chem.
2007, 119, 2136–2139; Angew. Chem. Int. Ed. 2007, 46, 2090–2093;
[4] D.-M. Shen, C. Liu, X.-G. Chen, Q.-Y. Chen, Synlett 2009, 6, 945–948.
[5] See, for example: a) M. J. Smith, W. Clegg, K. A. Nguyen, J. E. Rogers, R.
Pachter, P. A. Fleitz, H. L. Anderson, Chem. Commun. 2005, 2433–2435;
b) I. M. Blake, H. L. Anderson, D. Beljonne, J.-L. Brꢂdas, W. Clegg, J. Am.
Chem. Soc. 1998, 120, 10764–10765; c) N. L. Bill, M. Ishida, S. Bꢃhring,
J. M. Lim, S. Lee, C. M. Davis, V. M. Lynch, K. A. Nielson, J. O. Jeppesen, K.
Ohkubo, S. Fukuzumi, D. Kim, J. L. Sessler, J. Am. Chem. Soc. 2013, 135,
10852–10862; d) M. Kamo, A. Tsuda, Y. Nakamura, N. Aratani, K. Furuka-
wa, T. Kato, A. Osuka, Org. Lett. 2003, 5, 2079–2082; e) J.-H. Fuhrhop, S.
Besecke, J. Subramanian, C. Mengersen, D. Riesner, J. Am. Chem. Soc.
1975, 97, 7141–7152.
ˇ ´
b) B. Basic, J. C. McMurtrie, D. P. Arnold, J. Porphyrins Phthalocyanines
ˇ ´
2010, 14, 481–493; c) B. Basic, J. C. McMurtrie, D. P. Arnold, Eur. J. Org.
Chem. 2010, 4381–4392.
[16] C. Maeda, P. Kim, S. Cho, J. K. Park, J. M. Lim, D. Kim, J. Vura-Weis, M. R.
Wasielewski, H. Shinokubo, A. Osuka, Chem. Eur. J. 2010, 16, 5052–
5061.
[17] M. S. Goh, M. C. Pfrunder, J. C. McMurtrie, D. P. Arnold, manuscript in
preparation.
[18] L. Rintoul, S. R. Harper, D. P. Arnold, Phys. Chem. Chem. Phys. 2013, 15,
18951–18964.
[6] a) I. M. Blake, A. Krivokapic, M. Katterle, H. L. Anderson, Chem. Commun.
2002, 1662–1663; b) I. M. Blake, L. H. Rees, T. D. W. Claridge, H. L. An-
derson, Angew. Chem. 2000, 112, 1888–1891; Angew. Chem. Int. Ed.
2000, 39, 1818–1821.
[7] a) J. P. Hill, I. J. Hewitt, C. E. Anson, A. K. Powell, A. L. McCarty, P. A. Karr,
M. E. Zandler, F. D’Souza, J. Org. Chem. 2004, 69, 5861–5869; b) A. J.
Golder, L. R. Milgrom, K. B. Nolan, D. C. Povey, J. Chem. Soc. Chem.
Commun. 1989, 1751–1753; c) T. G. Traylor, K. B. Nolan, R. Hildreth, J.
Am. Chem. Soc. 1983, 105, 6149–6151.
[8] a) S.-D. Jeong, J. L. Sessler, V. Lynch, C.-H. Lee, J. Am. Chem. Soc. 2008,
130, 390–391; b) S.-D. Jeong, S.-J. Hong, K. J. Park, S. Ham, J. L. Sessler,
V. Lynch, C.-H. Lee, J. Org. Chem. 2007, 72, 6232–6240; c) E.-K. Sim, S.-D.
Jeong, D.-W. Yoon, S.-J. Hong, Y. Kang, C.-H. Lee, Org. Lett. 2006, 8,
3355–3358.
[9] a) A. L. Balch, B. C. Noll, S. L. Phillips, S. M. Reid, E. P. Zovinka, Inorg.
Chem. 1993, 32, 4730–4736; b) A. L. Balch, M. M. Olmstead, S. L. Phillips,
Inorg. Chem. 1993, 32, 3931–3936.
[10] a) T. Takanami, M. Hayashi, F. Hino, K. Suda, Tetrahedron Lett. 2003, 44,
7353–7357; b) Y. Chen, X. P. Zhang, J. Org. Chem. 2003, 68, 4432–4438.
[11] a) L. J. Esdaile, J. C. McMurtrie, P. Turner, D. P. Arnold, Tetrahedron Lett.
2005, 46, 6931–6935; b) L. J. Esdaile, G. L. Simpkins, O. B. Locos, L. Rin-
toul, M. P. Duriska, P. Turner, J. C. McMurtrie, D. P. Arnold, Tetrahedron
2014, 70, 517–532; c) L. J. Esdaile, PhD Thesis, Queensland University of
Technology, Brisbane (Australia), 2007; d) L. J. Esdaile, M. O. Senge, D. P.
Arnold, Chem. Commun. 2006, 4192–4194.
[19] M. O. Senge, Chem. Commun. 2006, 243–256.
[20] Gaussian 09, Revision A.1, M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E.
Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Men-
nucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian,
A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara,
K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O.
Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta, F. Ogliaro,
M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Ko-
bayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyen-
gar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B.
Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann,
O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,
K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg,
S. Dapprich, A. D. Daniels, ꢄ. Farkas, J. B. Foresman, J. V. Ortiz, J. Cio-
slowski, D. J. Fox, Gaussian, Inc., Wallingford CT, 2009.
[21] a) A. D. Becke, J. Chem. Phys. 1993, 98, 5648–5652; b) C. Lee, W. Yang,
R. G. Parr, Phys. Rev. B 1988, 37, 785–789; c) S. H. Vosko, L. Wilk, M.
Nusair, Can. J. Phys. 1980, 58, 1200–1211; d) P. J. Stephens, F. J. Devlin,
C. F. Chabalowski, M. J. Frisch, J. Phys. Chem. 1994, 98, 11623–11627.
[22] N. M. O’Boyle, A. L. Tenderholt, K. M. Langner, J. Comput. Chem. 2008,
29, 839–845.
Received: December 23, 2013
Published online on February 20, 2014
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPlusChem 2014, 79, 813 – 824 824