Facile dearomatisation of porphyrins using palladium-catalysed hydrazination: The 5,15-diiminoporphodimethenes and their redox products

Article abstract of DOI:10.1016/j.tet.2013.11.006

The synthesis, electronic absorption and ¹H NMR spectra of a suite of novel porphyrinoids derived from meso-bromoporphyrins by palladium-catalysed aminations using ethyl and tert-butylcarbazates are reported. Instead of the expected carbazate-substituted porphyrins, a facile oxidative dearomatisation of the porphyrin ring occurs in high yield, especially for the nickel(II) complexes, resulting in high yields of 5,15- diiminoporphodimethenes (DIPDs). The analogous zinc(II) and free base DIPDs were also characterised, the former by X-ray crystallography. The oxidation and reduction reactions of DIPDs and their precursor carbazate porphyrins were studied. Density Functional Theory (DFT) was used to calculate the optimised geometries and frontier molecular orbitals of DIPD Ni8c and bis(azocarboxylate) 19c, and Time Dependent DFT calculations allowed the prediction of electronic absorption spectra, whose characteristics corresponded well with those of the observed solution spectra. In the latter case, the calculated low-energy absorptions were unlike those of a typical porphyrin, due to the near-degeneracy of the highest filled frontier orbitals, and the wide energy separation between the unfilled orbitals. This feature was present in the observed spectrum.

Full text of DOI:10.1016/j.tet.2013.11.006

Tetrahedron 70 (2014) 517e532
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Facile dearomatisation of porphyrins using palladium-catalysed
hydrazination: the 5,15-diiminoporphodimethenes and their redox
products
Louisa J. Esdaile a, Grace L. Simpkins a, Oliver B. Locos a y, Llew Rintoul a,
,
,
,
*
Martin P. Duriska b z, Peter Turner c, John C. McMurtrie a, Dennis P. Arnold a
a School of Chemistry, Physics and Mechanical Engineering, Queensland University of Technology, GPO Box 2434, Brisbane 4001, Australia
b School of Chemistry, Monash University, Victoria 3800, Australia
c Crystal Structure Analysis Facility, School of Chemistry, University of Sydney, NSW 2006, Australia
a r t i c l e i n f o
a b s t r a c t
The synthesis, electronic absorption and 1H NMR spectra of a suite of novel porphyrinoids derived from
meso-bromoporphyrins by palladium-catalysed aminations using ethyl and tert-butylcarbazates are re-
ported. Instead of the expected carbazate-substituted porphyrins, a facile oxidative dearomatisation of
the porphyrin ring occurs in high yield, especially for the nickel(II) complexes, resulting in high yields of
5,15-diiminoporphodimethenes (DIPDs). The analogous zinc(II) and free base DIPDs were also charac-
terised, the former by X-ray crystallography. The oxidation and reduction reactions of DIPDs and their
precursor carbazate porphyrins were studied. Density Functional Theory (DFT) was used to calculate the
optimised geometries and frontier molecular orbitals of DIPD Ni8c and bis(azocarboxylate) 19c, and Time
Dependent DFT calculations allowed the prediction of electronic absorption spectra, whose character-
istics corresponded well with those of the observed solution spectra. In the latter case, the calculated
low-energy absorptions were unlike those of a typical porphyrin, due to the near-degeneracy of the
highest filled frontier orbitals, and the wide energy separation between the unfilled orbitals. This feature
was present in the observed spectrum.
Article history:
Received 18 August 2013
Received in revised form 21 October 2013
Accepted 4 November 2013
Available online 8 November 2013
Keywords:
Porphyrinoid
Diiminoporphodimethene
Crystal structure
Modelling
Density functional theory
Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction
and air is a high-yielding route to azoporphyrins Ni2/Zn22, pro-
vided that the opposite 15-meso carbon is substituted with, for
This work arose from our early attempts to prepare the then-
unknown ‘azoporphyrins’, the porphyrin analogues of the azo-
benzenes. Anderson and co-workers predicted in 2002 that
a diporphyrin covalently linked in the meso-position through an
azo bridge would display strong conjugation between the macro-
example, a phenyl group (Scheme 1).2 If the 15-position is unsub-
stituted, a remarkable head-to-tail coupling occurs, giving as the
major products partially dearomatised systems, such as 3a and
3b.2,5,6 Our original proposal to prepare the azoporphyrins, how-
ever, involved the Pd-catalysed coupling of alkoxycarbonyl-
protected hydrazines (i.e., carbazates) with haloporphyrins, to be
followed by deprotection, a second coupling, then a (presumed)
facile dehydrogenative oxidation to the target 2 (Scheme 2). This
multi-step route was based on the fact that typical pathways to
azoporphyrins by diazonium coupling or partial reduction of
nitrosoporphyrins had failed.6
Moreover, mild methods for the synthesis of azobenzenes have
been reported using this strategy. In 1999, Bridger and co-workers
synthesised tert-butoxycarbonyl (Boc)-protected aryl hydrazines
using palladium catalysis.7 Carbazates are ambidentate nucleo-
philes, and the amidation product predominated over the isomeric
amination product. A Boc-protected aryl hydrazine dimer was
formed as a side-product of the Pd-catalysed coupling of aryl bro-
mides with tert-butyl carbazate. Cho and co-workers then reported
cycle
p-systems, conferring electronic properties different from
those of the known ethene- and ethyne-linked dyads.1 Sub-
sequently, we reported the successful synthesis of the first azo-
porphyrins, and then expanded the range of these derivatives into
functionalised versions, as well as an azo/butadiyne-linked tet-
rad.2e4 It turned out that the oxidative homo-coupling of primary
aminoporphyrins, such as Ni1b or Zn1b using catalytic copper(II)
* Corresponding author. Tel.: þ61 (0)7 3138 2482; e-mail address: d.arnold@qut.
edu.au (D.P. Arnold).
y
Present address: ChemCentre, Chemistry Precinct, Bentley, PO Box 1250, Perth
6152, Australia.
z
Present address: Bruker Biosciences Pty Ltd,
1A/28 Albert St, Preston 3072, Australia.
0040-4020/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2013.11.006

518
L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
Scheme 1.
Scheme 2.
the synthesis of azobenzenes via the coupling of Boc-protected aryl
hydrazines with aryl halides.8 The reaction gave the azobenzene in
a two-step procedure. The N0-Boc diarylhydrazines were syn-
thesised using Pd(OAc)2, P(t-Bu)3 and Cs2CO3 in toluene at 110 ꢀC in
good to excellent yields with both electron-withdrawing groups
and electron-donating groups attached. These were then directly
oxidised to the corresponding azobenzenes with N-bromosuccini-
mide (NBS)/pyridine in dichloromethane (DCM) at room temper-
ature.8 Cho and co-workers then reported a copper mediated one-
pot synthesis of azobenzenes from bis-Boc aryl hydrazines and aryl
iodides.9
calculation on two candidate structures from the suite of com-
pounds we prepared. We report their geometry optimisations, their
predicted frontier orbital properties, and comparisons of experi-
mental electronic absorption spectra with the predicted electronic
transitions. From other recent results, the dearomatisation of cer-
tain meso-aminoporphyrins appears to be a more general reaction,
which is leading to new lines of research.15
2. Results and discussion
2.1. Palladium-catalysed couplings of carbazates with nick-
The amination and amidation of haloporphyrins has developed
into a useful reaction for preparing a range of nitrogen-substituted
porphyrinoids (NSPs). Various amino- and amido-porphyrins, in-
el(II) meso-bromoporphyrins
As a starting point for this work, we chose 5-bromo-10,20-
diphenylporphyrinatonickel(II) (5-bromoNiDPP) Ni5 as the sub-
strate. Its reaction with tert-butyl carbazate using the conditions of
Takanami et al.10 did not proceed cleanly to a mono-coupling
product (Scheme 3).
cluding both meso- and b-substituted cases, have been reported in
recent years using metal catalysis, including the significant papers
of Takanami et al.,10 and Zhang and co-workers,11 amongst several
others.12 In particular, this type of reaction has led to many exam-
ples of porphyrinyl tertiary amines being tested as dyes for solar
cells.13 Although an example of coupling of a carbamate was re-
ported by Zhang’s group,11b the carbazates have not been in-
vestigated by others. Herein, we report the details of our studies of
the reactions of 5-bromoporphyrins and 5,15-dibromoporphyrins
with carbazates under Pd-catalysis. Although we abandoned this
route to azoporphyrins because of our success with the copper-
catalysed amine coupling, nevertheless this chemistry has un-
covered the facile dearomatisation of the carbazate coupling
products, which leads to the previously-unknown diiminopor-
phodimethene (DIPD) macrocycle 4. Our early results were pub-
lished in preliminary form in 2004,14 and now we report all the
synthetic details and our studies of the redox chemistry of this suite
of carbazate-derived porphyrinoids. Moreover, we have performed
Density Functional Theory (DFT) and Time Dependent DFT
TLC showed two products (or three from larger-scale reactions)
were formed during this reaction and all were isolated. It was found
that only products arising from attack by the amino-nitrogen of the
carbazate were isolated from this reaction, and the possible ami-
dation product, i.e., 6, was not. The least polar compound was found
to be the bright green mono-meso-substituted azocarboxylate
Ni7a. The formation of the oxidised azocarboxylate product was
unexpected, as the reaction conditions were thought to be re-
ducing. These reactions were carried out in an argon atmosphere,
and care was taken to exclude air from the reaction vessel. This
result was promising, as it suggested that if hydrazine-linked por-
phyrin dimers could be prepared, they might spontaneously oxidise
to give an azo-linked dimer.
This selectivity of the formation of the amino-product is pre-
sumably due to the steric crowding of the meso-position on the

L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
519
Scheme 3.
porphyrin macrocycle. Buchwald and co-workers found that
amino- and amido-substituted products formed in equal amounts
from ortho-substituted iodoarenes, whereas only the amido prod-
ucts formed from para-substituted iodoarenes.16 The steric
crowding of the meso-position of a porphyrin is comparable to that
of an ortho-substituted iodoarene, and presumably the crowding of
the meso-position and the use of the bulky 2,20-bis(diphenylphos-
phino)-1,10-binaphthyl (BINAP) ligand led to the selectivity of the
formation of the terminal amino-substituted products.
More interestingly, the major product of this reaction was the
unprecedented blue-green di-meso-substituted diiminoporphodi-
methene (DIPD) Ni8a, which was isolated in a crude yield of 72%. It
was characterised using 1H NMR, high resolution MS, UV/vis
spectroscopy as well as X-ray crystallography, and was reported in
our preliminary communication.14 This product was very un-
expected, as the starting material was mono-brominated and
therefore only mono-meso-substituted products were envisaged.
The loss of aromaticity also was a major surprise at that time, as
a variety of mono- and di-aminated porphyrins had been syn-
thesised previously, with no reports of porphodimethene prod-
ucts.10,11 When the reaction of NiDPPBr with tert-butyl carbazate
was repeated on a larger-scale, after the bromoporphyrin starting
material was completely consumed, a minor red product was noted
on the TLC plate, eluting between the two green compounds. Upon
isolation, this red compound was found to be the mono-meso-
substituted carbazate porphyrin Ni9a, characterised using 1H NMR,
high resolution MS, IR, UV/vis spectroscopy and elemental analysis.
This compound was found to oxidise spontaneously to the azo-
carboxylate Ni7a upon exposure to air over a period of hours, and
when exposed to acid was found to decompose to non-porphyrinic
material.
We were able to prepare the carbazate porphyrin Ni9a se-
lectively by reducing the reaction time from 25 h to 13 h. At 13 h
the carbazate porphyrin was isolated in 84% recrystallised yield,
even when a ten-fold excess of the carbazate was present. When
the same ratios of starting materials were allowed to react for
24 h, the amount of Ni9a was reduced to 19%, and the yield of
DIPD Ni8a increased from 5% to 54%. This increase in yield of the
DIPD may be attributed to air seeping into the reaction vessel
over the longer reaction time or during sampling for TLC anal-
ysis, despite our precautions. Moreover, the mono-meso-
substituted azocarboxylate and carbazate porphyrins were then
selectively synthesised from the coupling of 1 equiv of carbazate
with a bromoporphyrin. Azocarboxylate Ni7a was synthesised in
a 63% crude yield (25% recrystallised yield, due to the low-
yielding recrystallisation) upon leaving the reaction mixture
open to the air. However, the application of the oxidant 2,3-
dichloro-5,6-dicyanobenzoquinone (DDQ) directly to the
reaction mixture resulted in the decomposition of the porphyrin
products.
To try to improve the ease of purification, separation and char-
acterisation of the products obtained from the palladium-catalysed
reactions, tert-butyl carbazate was replaced by ethyl carbazate. It
was hoped that by replacing the tert-butyl group with an ethyl
group, the solubility of the products in organic solvents might be
lowered, thus improving the low-yielding recrystallisations. In
addition, the 1H NMR signal for the tert-butyl singlet of the DIPD
resonated close to the signal for any water that was present in the
solvent. This made the differentiation between similar compounds
or the determination of the number of carbazate-type substituents
on a compound difficult.
Ethyl carbazate was found to react within the same amount of
time as the tert-butyl analogue, and gave the analogous three
products: the azocarboxylate Ni7b and carbazate Ni9b and the
DIPD Ni8b. It was found that the change in the group did result in
easier separation by chromatography. This was due to the products’
higher retention factors, so the requirement for a less polar solvent
for the elution of the product meant that polar impurities no longer
co-eluted with the DIPD product. However, it was found that
changing to ethyl carbazate unfortunately did not significantly re-
duce the solubilities of the products.
To prepare the DIPD selectively, 5,15-dibromoporphyrin Ni10
was subjected to the same conditions, using excess ethyl carbazate
(Scheme 4). However, the yield for Ni8b was found to vary just as
much when either the mono- or di-bromoporphyrin were
employed, but the dibromoporphyrin was easier to purify, so it was
henceforth used for the synthesis of DIPDs.
To synthesise the mono-substituted carbazate- and azocarbox-
ylate porphyrins, 1 equiv of ethyl carbazate was combined with
NiDPPBr. The carbazate porphyrin was either isolated or allowed to
oxidise in air to the azocarboxylate. The azocarboxylate porphyrin
Ni7b was isolated in up to a 63% yield using this method, and was
characterised using 1H NMR, MS, IR and UV/vis spectroscopy. The
redox properties of these azocarboxylate and carbazate porphyrins,
as well as those of the DIPDs, will be discussed in Section 2.3.
To avoid the dearomatisation reactions entirely, NiDPPBr Ni5
was replaced by 5-bromo-10,15,20-triphenylporphyrinatonickel(II)
(NiTriPPBr, 11). It was found that when 1 equiv of the carbazate was
employed, the reaction proceeded sluggishly, and the products
degraded over the longer reaction time required. When ten
equivalents of carbazate were used, the reaction proceeded more
rapidly (within 22 h) and upon allowing the reaction mixture to stir
open to the atmosphere, the desired azocarboxylate 12b was iso-
lated in 67% crude yield (the carbazate porphyrin 13b was not
isolated under these work-up conditions). The azocarboxylate 12b
was characterised using 1H NMR, high resolution MS and UV/vis

520
L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
Scheme 4.
spectroscopy. In order to study the redox behaviour of this suite of
compounds, we needed the carbazate porphyrin, and we found that
it could be prepared in almost pure form by treating bromopor-
phyrin 11 with ten equivalents of tert-butyl carbazate in the pres-
ence of the standard palladium catalysts, but then isolating the
products as rapidly as possible to avoid oxidation. The carbazate
porphyrin 13a was characterised using 1H NMR, high resolution MS
and UV/vis spectroscopy, and only 6% of the azocarboxylate 12a was
isolated under these conditions.
Thus the synthesis of the three major types of products formed
from the coupling of alkyl carbazates with nickel meso-bromopor-
phyrins had been optimised to synthesise each product selectively.
The broader applicability of these couplings was then studied using
the zinc(II) and free base bromoporphyrins.
was repeated on several occasions and the yields were difficult to
reproduce. The repetition gave either no stable products or the
desired DIPD highly contaminated and in low yields. The dibro-
minated starting material H2DPPBr2 H210 was also subjected to the
coupling reaction, in order to favour the formation of the DIPD
product, and ethyl rather than tert-butyl carbazate was used to aid
in the ease of separation. The desired product was formed only in
low yield and the purification was not found to be any easier.
In order to synthesise the zinc and free base mono-substituted
azocarboxylate and carbazate porphyrins, 1 equiv of ethyl carbazate
was combined with either mono-brominated ZnDPPBr Zn5 or
H2DPPBr H25. Remarkably, both of these reactions predominately
resulted in the formation of the dearomatised DIPD products Zn8b
and H28b, in low yields. This result is interesting, as these condi-
tions were thought to be the most conducive to the formation of
mono-substituted products. A large amount of unreacted starting
material was also recovered, along with other side-products in
trace amounts, which could not be characterised. It was concluded
from these experiments that the nickel(II) porphyrins formed ei-
ther mono- or di-substituted products, depending on the reaction
conditions, whereas both the zinc and free base porphyrins formed
the disubstituted DIPD products preferably, regardless of the
conditions.
To avoid this reaction, the couplings with carbazates of a free
base bromo substrate with a blocked 15-meso position, namely 5-
bromo-15-phenyl-10,20-bis(3,5-di-tert-butylphenyl)porphyrin
(H2DAPPBr, 14), was attempted (Scheme 5). In the only successful
reactions, the catalyst Pd(dppe)2 [dppe¼1,2-bis(diphenyl
phosphino)ethane] was used instead of the Pd(OAc)2/BINAP sys-
tem. The major product was the green azocarboxylate 15 (29%),
but two other porphyrinoid free bases were also isolated, one blue
and the other violet in solution. The 1H NMR spectrum of the blue
compound suggested strongly that it is the novel iminophlorin 16.
The details of the 1H NMR spectroscopic results are presented in
the Supplementary data. The violet compound appeared to be
a tautomeric iminophlorin, presumably 17, but its low yield pre-
cluded its full characterisation.
2.2. Palladium-catalysed couplings of carbazates with zinc(II)
and free base meso-bromoporphyrins and the mechanism of
the double coupling
The synthesis of the zinc(II) DIPD from 5,15-dibromo-10,20-
diphenylporphyrinatozinc(II) Zn10 was found to proceed cleanly
to yield Zn8b in up to 82% recrystallised yield. DIPD Zn8b was
characterised using 1H NMR, high resolution MS, UV/vis spectros-
copy and X-ray crystallography (see Section 2.4). The ruby-red zinc
DIPDs were found to give the highest yielding recrystallisations of
the series, as long as coordinating solvents (e.g., methanol or pyr-
idine) were avoided during the recrystallisation. Unusually for most
porphyrin synthetic research, the self-aggregation tendencies of
Zn(II) porphyrinoids were actually advantageous in this series. Al-
though the zinc complex did give the highest purified yield of all
the DIPDs on one occasion, this preparation was found to be quite
unreliable. On several occasions the reaction was repeated with no
major products observed, and numerous side-products were iso-
lated, which either degraded or were isolated in insufficient
quantities to identify their structures. We do not have a satisfactory
explanation for this lack of reproducibility.
To expand the variety of metals that could be coordinated to the
DIPDs, the free base H2DPPBr H25 was coupled with excess tert-
butyl carbazate (Scheme 3). The reaction yielded the dark orange
free base DIPD H28a in a recrystallised yield of 37%, which was
characterised using 1H NMR, high resolution MS and UV/vis spec-
troscopy. It was found that the coupling with free base porphyrin
was the most unreliable of the three that were trialled. The reaction
In summary, the synthesis of nickel DIPDs was the most re-
producible and gave the highest crude yields, the zinc series was
less repeatable but produced the highest recrystallised yields, and
the free base was the lowest yielding and the least repeatable.
Takanami et al. reported that nickel(II)-coordinated porphyrin
starting materials resulted in the highest yields for amino- and

L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
521
Scheme 5.
amido-porphyrins in Pd-catalysed couplings, followed by zinc(II),
and finally free base, which was found to result in a complex
mixture.10 A similar result was also observed by Shen et al. for the
nickel-catalysed amination of porphyrins.12d Thus for the carbazate
nucleophiles reported here, our findings agree with those for the
amines and amides reported by Takanami et al.10
Some control experiments were performed to try to understand
better the formation of a disubstituted product from a mono-
substituted starting material. The first experiment was to react
unsubstituted H2DPP 18 with excess t-butyl carbazate using the
Pd(OAc)2/rac-BINAP catalytic system. No change was observed after
25 h, the amount of time normally required for the full consump-
tion of the brominated starting materials. Thus the bis-substitution
requires the initial substitution to occur in order to allow the sec-
ond carbazate molecule to attack the opposite free meso-position.
excess carbazate, using the normal catalytic system. The reaction
mixtures were monitored by TLC and it was found that both yielded
the DIPD products, Ni8a and Ni8b, repectively. These mixtures
were reacted for two weeks, and although the starting materials
were not entirely consumed, no more product was being formed in
either mixture, and side-products were starting to form. During
these experiments, we observed by TLC that both the mono-
azocarboxylate and carbazate porphyrins were found in both re-
action mixtures. These results did not clarify, which of the azo-
carboxylate and carbazate porphyrins lead to the formation of the
DIPDs. However, the azocarboxylate porphyrin Ni7a was found to
give the DIPD product in a higher yield (47%) and with fewer side-
products than the carbazate porphyrin Ni9b (20%). This observation
led us to propose a mechanism for the formation of the diimino-
porphodimethenes, which is shown in Scheme 6. This mechanism
Scheme 6.
The synthesis of the DIPD product was then trialled from the cou-
pling of NiDPPBr2 Ni10 with excess ethyl carbazate in the absence of
both palladium and ligand, to see if a simple nucleophilic sub-
stitution reaction was responsible. The starting material was
completely consumed after four weeks, however the long reaction
time meant that many side-products also formed. The DIPD prod-
uct Ni8b was indeed isolated, but only in <10% crude yield. The
DIPD Ni8b was also isolated from the coupling of a ten-fold excess
of ethyl carbazate with Ni10, in the presence of base and Pd(OAc)2
only (i.e., ligand-free). The starting material had been consumed
within 18 h, and Ni8b was isolated in a 14% crude yield.
requires that after the formation of the carbazate porphyrin, the
latter is oxidised (perhaps due to the oxygen in the solvent, as the
solvent was not degassed for these reactions) to the azocarboxylate
porphyrin. The electron-withdrawing azocarboxylate group then
activates the opposite, unsubstituted meso-position, allowing the
nucleophilic attack of a second carbazate. This results in a phlorin-
type intermediate with one exocyclic imine bond, analogous to the
species 16/17 possibly prepared from H2DAPPBr, which is then
oxidised to give the stable DIPD. There have been some recent
examples of formal nucleophilic substitution of hydrogen by N-
nucleophiles, including some in the absence of any metal cata-
lysts.17 So the reactions in this scheme have precedents, and it is
clear that strongly electron-withdrawing substituents, such as
azocarboxylate encourage this reaction.
In order to investigate another pathway to the DIPDs, mono-
meso-substituted azocarboxylate Ni7a, and mono-meso-
substituted carbazate porphyrin Ni9b were separately treated with

522
L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
2.3. Redox reactions of the carbazate coupling products
The reverse process, reduction of the bis(azocarboxylate) 19b
was then attempted, by treatment with a ten-fold excess of NaBH4
(5 equiv per azocarboxylate moiety). After stirring at room tem-
perature for 15 min, the azocarboxylate was fully consumed, and
the major product was the DIPD Ni8b, while a minor polar product
was also observed by TLC. The mixture was stirred at room tem-
perature for 1 day and then refluxed for 1 day, as no further re-
duction appeared to be occurring at the lower temperature. Upon
separation of the resulting mixture, it was found that Ni8b
remained as the only porphyrinic material, and it was isolated in
a 59% crude yield. It is interesting to note that the fully reduced
bis(carbazate) was not observed in this reaction, which shows the
remarkable stability of the DIPDs. In the monofunctionalised series,
the carbazate/azocarboxylate redox pair can be reliably inter-
converted, while in the bis series, the DIPD/azocarboxylate pair can
be realised from both sides, but the bis(carbazate) is elusive.
The study of the redox properties of the bis(carbazate) por-
phyrins and their metal complexes is rather complicated, due to the
tendency of these compounds to oxidise partially to the DIPDs. We
thought that the oxidation of a DIPD might proceed rather simply to
a bis(azocarboxylate). However, the reduction of a bis(azocarbox-
ylate) could result in either the bis(carbazate) porphyrin or the
DIPD. The result of the reduction could of course be a mixture of the
two products, which would be difficult to separate. Therefore, we
chose to examine first the mono-meso-substituted carbazate por-
phyrin. The oxidation of the carbazate porphyrin Ni13a should re-
sult in the corresponding azocarboxylate Ni12a and the reduction
of Ni12a should re-form Ni13a. In fact, the oxidation of Ni13a
proceeded as expected to give the Ni12a in a 95% recrystallised
yield. Upon the addition of 1 equiv of the oxidant DDQ, the reaction
proceeded rapidly, with an immediate colour change from red to
green.
2.4. X-ray crystal structure of the zinc(II) 5,15-
diiminoporphodimethene Zn8b
To test the reductive process, Ni12a was treated with NaBH4
in THF/methanol at room temperature and the mixture was
monitored by TLC. After stirring the mixture at room tempera-
ture for 40 min, the Ni12a had converted to the carbazate Ni13a
quantitatively. To check the stability of the carbazate, the mixture
was then heated to reflux and the carbazate porphyrin slowly
began to convert to the primary amine Ni1b and to the meso-
unsubstituted NiTriPP. The mixture was then refluxed overnight,
after which the carbazate was fully consumed and the major
product was found to be NiTriPP, isolated in 48% yield. A small
amount of the impure aminoporphyrin Ni1b was also isolated.
The cleavage to the aminoporphyrin and the complete loss of the
nitrogen substituent are rather difficult to explain, and deserve
further investigation. We reported in 2006 the amination of
NiTriPPBr to Ni1b using unsubstituted hydrazine as the nucleo-
philic coupling partner, so NeN cleavage is perhaps not un-
expected, but reductive cleavage of the primary amino group
from a porphyrin has not been previously reported.18 Harper
observed, however that in the NaBH4/Pd/C reduction of the free
base nitroporphyrin H2TriPPNO2 to H2TriPPNH2, if the reaction is
not quenched within 10 min, extensive over-reduction to H2TriPP
It should be noted that we have drawn the relative stereo-
chemistry about the two C]N double bonds in the DIPDs as giving
the whole molecule C2 symmetry (as a planar projection). Only one
isomer was ever observed in the 1H NMR spectra of Ni8a and Zn8b,
for example, and the crystal structure we reported in our pre-
liminary communication, namely Ni(II) complex Ni8a, indeed
exhibited this disposition of the substituents.14 The exocyclic imine
ꢂ
bonds have C]N bond lengths of 1.299 and 1.297 A, whereas the
ꢂ
N
iminoeNamido bonds both have bond length of 1.349 A. The average
bond lengths of the dipyrromethane units are similar to those in
ꢂ
a fully conjugated porphyrin, with a mean bond length of 1.38(9) A,
with the exception of those flanking the imine bond. The average
C
mesoeCa bond length for the carbons adjacent to the imine bond
ꢂ
ꢂ
was found to be 1.45(4) A. This increase, in the vicinity of 0.06(5) A,
has previously been reported for the dioxoporphyrins and the
porphodimethenes.21,22 There exists strong out-of-plane distortion
of the macrocycle, similar to that previously reported for other
nickel porphodimethenes.23 This distortion is illustrated as a linear
display in Fig. 1. This shows that there is a small amount of saddle-
becomes a problem.19 In the analogous 5,10-diarylporphyrin se-
20
type distortion on the b-carbons labelled C3 and C13 (where the
ꢀ ꢁ
ries, Basic used very short reaction times to avoid this reaction.
imino-substituents are attached to C10 and C20). The average NieN4
Although it is hard to imagine an actual synthetic use for this
reaction, it is a possible method for clean removal of an NH2 or
NO2 substituent from a meso-position.
ꢂ
distance was found to be 1.891 A, which allows the central nickel
Turning now to the disubstituted series, DIPD Ni8b was treated
with an excess of DDQ, and an instantaneous colour change was
observed (Scheme 7). The mixture changed from the blue-green of
the DIPD to the bright green of the azocarboxylate. The fully aro-
matic bis(azocarboxylate) 19b was isolated in 71% recrystallised
yield and was characterised by 1H NMR, MS, IR, UV/vis spectroscopy
and elemental analysis. The tert-butyl DIPD Ni8a was also oxidised
in this manner and gave the tert-butyl bis(azocarboxylate) 19a in
a lower yield of 36%, due to difficulties in recrystallisation. We also
tried this oxidation with iodine in DCM and silver(I) in the form of
AgPF6 and AgOCOCF3 in DCM, but little or no conversion was
observed.
ꢂ
Fig. 1. Displacements (A) of atoms from the mean C20N4 plane in crystals of nickel(II)
diiminoporphodimethene Ni8a. The X-ray crystal structure was reported in Ref. 14.
Scheme 7.

L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
523
Fig. 2. Structure of zinc(II) diiminoporphodimethene Zn8b determined by single-crystal X-ray analysis. Disordered water and dioxane solvent molecules are omitted for clarity.
atom to have almost exact square planar geometry. Osuka and co-
workers reported similar NieN4 bond lengths for their quinodi-
methene product.22
from the C20N4 mean plane. Anderson’s zinc(II) tetracyanopor-
phodimethene had a severely non-planar saddle shape, which was
attributed to the steric repulsion between the b-hydrogens and the
Single crystals of the Zn(II) analogue Zn8b that were suitable for
crystallographic analysis were grown from the slow diffusion of
water into a solution in dioxane and pyridine. In the crystal struc-
ture, the metal ion is coordinated in N4 core of the porphyrin, with
a molecule of pyridine also coordinated to the zinc to give a five-
coordinate complex with approximate square-based pyramidal
geometry. The crystals also contain one dioxane and two water
molecules for each complex giving the crystal formula
[Zn8b(pyridine)]$dioxane$2H2O. Face-on and edge-on views of the
structure are shown in Fig. 2. The complex has crystallographic C2
symmetry (C2 axis passes through the zinc atom and is coincident
with the ZneNpyridine bond). The Cmeso]Nimino bond lengths of
nitrile substituents.21 To compare further the degree of modifica-
tion of the porphyrin macrocycle geometry with a similarly dear-
omatised compound, Fig. 4 shows the bond distances and out-of-
plane distortions of our DIPD Zn8b and porphodimethene 20.
ꢂ
ꢂ
1.303(2) A and the NiminoeNamino bond lengths are 1.368(2) A. The
ꢂ
average bond lengths of the dipyrromethene units were 1.39(3) A,
again with the exception of the carbons adjacent to the imine
carbon, with an average length of 1.46(8) A, the difference being
ꢂ
ꢂ
Fig. 3. Displacements (A) of atoms from the mean C20N4 plane in crystals of zinc(II)
diiminoporphodimethene Zn8b.
similar to the nickel analogue. The central zinc atom projects from
ꢂ
the C20N4 mean plane by 0.61 A, and the average ZneN4 distance is
ꢂ
ꢂ
24
2.07(8) A. The ZneNpyridine bond length was found to be 2.085(2) A,
2.5. Spectroscopic properties of the 5,15-
diiminoporphodimethenes
ꢂ
which is shorter than those in simple porphyrins (2.14e2.20 A)
and those found by Senge and Smith for the dioxoporphyrins
22
ꢂ
ꢂ
(2.12e2.13 A), and approximately equal to the 2.084 A reported
for a tetracyanoporphodimethene 20.21 Shorter ZneNpyridine bond
lengths have been attributed to stronger pyridine coordination due
to the higher positive charge on the core of the macrocycle.21
2.5.1. Electronic absorption spectra. To examine the effect of the
carbazate substituent, the absorption spectrum of the carbazate
porphyrin 13a was compared to that of unsubstituted NiTriPP. The
lone pairs of the nitrogen substituents are expected to influence the
electronic spectrum of the carbazate porphyrin. Fig. 5 shows that
NiTriPP possesses a Soret band with a maximum at 409 nm,
whereas 13a has its maximum at 422 nm. The Q-bands of NiTriPP
are found at 522 and 549 nm, whereas for 13a they are located at
489, 535 and 576 (shoulder) nm, indicating a significant electronic
effect of the substituent. The lowest-energy Q-band for the primary
amine Ni1b lies at 599 nm. Its structural and chemical properties
indicate there is a strong overlap of the N lone pair with the por-
phyrin electrons,18 so the more modest red-shift occasioned by the
carbazate group clearly shows less interaction between macrocycle
and substituent.
A linear display of the out-of-plane distortion is displayed in
Fig. 3, which shows that the macrocycle displays a hybrid distortion
between ruffled and saddled. The maximum out-of-plane distor-
Once oxidised to the mono-azocarboxylate 12a, the resulting
effects on the absorption spectrumda dramatic red-shifting of the
Soret and the Q-bandsdcan be attributed as an extension of the
porphyrin conjugation due to the azo group. The Soret band shifts
to 446 nm and the Q-band to 616 nm, a remarkable red-shift
tion is on the
tributed to a twist of the phenyl ring towards the distorted
carbon, the C6 mean plane of the phenyl ring lying at a 61.2ꢀ angle
b-pyrrolic C3 and C13. This deformation may be at-
b
-

524
L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
Fig. 4. Comparisons of the molecular dimensions (in blue) and out-of-plane atom displacements (red) for Zn8b (left) and the tetracyanoporphodimethene complex 20 (right)
reported by Blake et al.21
intensity, as expected when the conjugation of a porphyrin is
interrupted at opposite meso-positions, for example, 20.21 There is
also a strong absorption centred at 334 nm, which has an extinction
coefficient of approximately two-thirds of that of the Soret band.
This absorption is not observed in typical porphodimethenes that
lack the N-substituent. A nickel(II)OEP (H2OEP¼2,3,7,8,12,13,17,18-
octaethylporphyrin) imino-oxo-porphodimethene was reported
by Balch and co-workers to have an absorption band at 326 nm (as
well as at 456, 536 and 660 nm),26 whereas only two absorptions
(445 and 556 nm) were reported for the zinc analogue.27 The 5,15-
dioxoporphyrins also display a relatively strong absorption in the
region between 310 and 330 nm.28 Most porphyrinoids, including
porphodimethenes with exocyclic C]C bonds,21,29 do display
a weak absorption band in this region, however its intensity (both
relative to the Soret band and absolute) is not usually as high.
Fig. 5. Electronic absorption spectra (in dichloromethane) of NiTriPP (bold solid line),
nickel(II) carbazate 13a (solid line), nickel(II) mono(azocarboxylate) 12a (dashed line)
and bis(azocarboxylate) 19b (double line).
compared with 13a. While the Soret band of the azocarboxylate
maintains its intensity, relative to the carbazate porphyrin, it has
broadened significantly and the Q-bands of the azocarboxylate
have doubled in intensity, relative to the Soret band (Fig. 5).
Anderson noted similar electronic properties in the meso-arylazo-
substituted porphyrins, the only comparable examples in the lit-
erature.1 The broadening and red-shifting of the bands were more
dramatic for the latter compounds, however, as they were further
conjugated to aryl groups.
In order to examine the electronic effects when either one or
two azocarboxylate groups are substituted on a porphyrin, the
spectrum of bis(azocarboxylate) 19b is also shown in Fig. 5. The
effect of the second azocarboxylate group is to increase the in-
tensity of the Q-bands in relation to the Soret band, and to red-
shift both the Soret and Q-bands. The Soret exhibits only a mi-
nor shift from 446 nm to 452 nm (a 300 cmꢁ1 shift), while the Q-
bands have a greater shift from 616 nm to 653 nm (920 cmꢁ1). The
Q-band of 19b is also very broad. The theoretical calculations
described in Section 2.6 shed some light on this unusual spectrum.
The azocarboxylate substituent apparently possesses a similar
electronic influence to the nitro group, for which a Q-band red-
shift from 5-nitro- to 5,15-dinitroNiDPP of 562e583 nm
(640 cmꢁ1) is found; however the second nitro group has a more
profound effect on the Soret band, with a 980 cmꢁ1 shift (from 408
to 425 nm) observed.25
Fig. 6. Electronic absorption spectra (in dichloromethane) of nickel(II) diiminopor-
phodimethene Ni8a (bold solid line), zinc(II) diiminoporphodimethene Zn8b (solid
line), and free base diiminoporphodimethene H28a (dashed line).
The UV/vis spectrum of the DIPD Ni8a shown in Fig. 6 can be
compared with a fully oxidised bis(azocarboxylate) 19a in Fig. 5
(the spectra were also compared in our communication).14 This
comparison dramatically illustrates the effect of the reduction in
the conjugation of the porphyrin macrocycle, especially in the loss
of intensity of the Soret band. Although the Q-bands have ap-
proximately the same intensity in both compounds, the Soret band
of aromatised 19a has regained the typical intensity of a normal
porphyrin. The intensity of the band at 334 nm is also reduced once
the macrocyclic aromaticity is restored, adding to the argument
that this band should be attributed to the exocyclic imine bonds
(see Section 2.6 below).
The spectra in Fig. 6 show that the zinc and free base DIPDs
display blue-shifted spectra in comparison to the nickel analogue.
The blue-shift is the most dramatic for the free base H28a, for which
the Soret band is reduced in intensity as well. In fact the spectrum
The UV/vis spectrum of the DIPD Ni8a is shown below in Fig. 6,
together with the zinc and free base analogues Zn8b and H28a. The
Soret band is reduced and the Q-bands have increased in relative

L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
525
of H28a is very different from even its Ni(II) analogue Ni8a, and so
product 19b. The
b
-protons of the DIPD appear as four sharp dou-
unlike a typical free base porphyrin, that is, it unrecognisable as
such. This radical change in the spectrum is dramatic for a putative
minor change of central substituent, but similar observations have
been reported for the dioxoporphyrins. For example, the free base
OEP dioxoporphyrin absorbs at 315, 408 and 492 nm, with ex-
tinction coefficient ratio 2.2:4.7:1.0,28 whereas its Ni(II) complex
absorbs very differently at 337, 458, 546 and 676 (weak) nm
(7.7:23:4.3:1.0).30 The Zn(II) analogue Zn8b is also unusual, having
characteristics of both the others, but the absorption band attrib-
uted above to the exocyclic imine substituent is remarkably con-
sistent for all three DIPDs. Modelling of the transitions by Density
Functional Theory is described in Section 2.6.
blets between 6 and 7 ppm (Fig. 8).
2.5.2. The 1H NMR spectra. The porphyrin regions of the 1H NMR
spectra of the tert-butylazocarboxylate and -carbazate porphyrins
Ni7a and Ni9a are illustrated in Fig. 7. The effect of the electron-
withdrawing azocarboxylate substituent can be seen in the broad
Fig. 8. Porphyrin proton region of the 400 MHz 1H NMR spectra of nickel(II) bis(a-
zocarboxylate) 19b (top spectrum) and nickel(II) mono(azocarboxylate) Ni7b (bottom)
in CDCl3.
spread of b-H peaks of >1 ppm. The most downfield b-H doublet,
Porphodimethenes and dioxoporphyrins naturally also show
this effect.5,29,31 The amido protons give rise to the slightly broader
singlet at 8.78 ppm, which was confirmed using D2O exchange. We
initially thought that DIPDs might exhibit only one pair of doublets
due to the 3- and 7-Hs, is remarkably shifted further downfield
than the singlet meso-peak. This unusual peak inversion was not
observed in the arylazo-substituted porphyrins, however it was
observed in acetylene- and imine-linked porphyrin dimers.1 Con-
versely, carbazate Ni9a possesses a broader doublet, resonating at
arising from the
the imino nitrogens. The 2D-NOESY experiment revealed cross-
peaks between only one of the -peaks (at 6.85 ppm) and the NH
b-protons, depending on the rate of inversion of
9.44 ppm, for the 3,7-
stituent. The broadening may reflect rotational restriction of the
bulky substituent. The 3,7- -protons signal for the azocarboxylate
b-protons adjacent to the carbazate sub-
b
peak. This indicates that, on the 1H NMR timescale, the amido
proton lies in closer proximity to one side of the macrocycle than
the other, rather than lying orthogonal to the macrocycle, and im-
plies a slow rate of inversion of the imino nitrogens. Thus in solu-
tion the average structure has C2 symmetry, which is more or less
maintained in the solid state (Section 2.4).
b
Ni7a are much sharper, presumably due to the removal of the in-
terfering NH atoms upon oxidation.
The zinc DIPD Zn8b has a 1H NMR spectrum very similar to that
of the nickel analogue Ni8b, and the spectrum of the free base H28b
is also similar in the pyrrole proton region. Notably, the inner NH
protons resonate at 12.42 ppm, whereas in an aromatic free base
porphyrin their signals are typically located at approximately
ꢁ3 ppm. In the absence of macrocyclic conjugation, these protons
resemble those of simple pyrroles. This effect has also been noted
for porphodimethenes and dioxoporphyrins.21,29,32,33 Two of the
signals for the b-pyrrole protons in the spectra of H28a and H28b
are broader than the others; this is due to small couplings to the
pyrrolic NH in the same ring. As the macrocycle is no longer aro-
matic, the inner NHs are not undergoing tautomerism, allowing
observation of this coupling. When H28a was treated with D2O, not
only does the signal at 12.42 ppm disappear, but the coupling is
Fig. 7. Porphyrin proton region of the 400 MHz 1H NMR spectra of nickel(II) mono(-
azocarboxylate) Ni7a (top spectrum) and nickel(II) carbazate Ni9a (bottom) in CDCl3.
The triphenylporphyrin carbazate 13a has a 1H NMR spectrum
typical of an unsymmetrically substituted 5,15-disubstituted por-
phyrin, and the NH signals were assigned as a result of the different
rates of deuterium exchange with D2O: a slightly broad peak at
7.67 ppm was found to disappear rapidly, so this peak was assigned
as the amino NH, whereas the broader peak at 7.25 ppm took longer
to diminish and was therefore assigned as the amido NH peak.
The 1H NMR spectra of the mono- and bis(azocarboxylate)s (e.g.,
Ni7b, 19b) make an interesting comparison (Fig. 8). The spectrum
lost, and all four
b-pyrrole signals appear as sharp doublets
(Supplementary data, pages S7, S8).
2.6. Density Functional Theory modelling of nickel(II) 5,15-
diiminoporphodimethene and its oxidation product
Because of the unique nature of the DIPDs, and the fact that we
had the solid state structures in hand, it seemed worthwhile to
attempt to model the new macrocyclic system using Density
Functional Theory (DFT) calculations, and to see whether the cal-
culations could predict the geometry and the electronic absorption
spectrum. To this end, the model structures Ni8c and 19c were
subjected to theoretical calculations using the DFT model. Elec-
tronic structure and time-dependent calculations were sub-
sequently performed on the resulting optimised geometries using
the same parameters. The comprehensive results for both com-
pounds are provided in the Supplementary data of the optimised
geometries, the frontier molecular orbital surfaces (HOMOꢁ1,
HOMO, LUMO, LUMOþ1), and tables of the most intense excited
state transitions in the visible to near-IR regions.
19b is typical for a symmetrically substituted NiDPP. The
b-peaks
occupy a smaller range and the most downfield -peak is upfield,
b
compared with that of Ni7b, despite the presence of two strongly
electron-withdrawing groups. As also suggested by the electronic
spectra (Fig. 5) and the DFT calculations (section 2.6 below), this
reflects the reduction in the aromaticity/ring current of this
strongly axially-asymmetric porphyrin. Remarkably, this effect is
also manifest in an upfield shift of the signals for the ortho-protons
of the 10,20-phenyl groups. Because of the complete elimination of
the macrocyclic ring current, the
b-protons of the nickel DIPD Ni8b
are shifted substantially upfield when compared to its oxidation

526
L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
The optimised molecular geometry for Ni8c closely resembles
that found in the solid state for Ni8a. The macrocycle is strongly
ruffled, and the overall symmetry is close to C2. The Cmeso
azocarboxylate substituents results in the interesting situation of
the HOMOꢁ1 and HOMO being almost degenerate. The occupied
frontier orbitals (deriving from the Gouterman orbitals of a D4h
metalloporphyrin) comprise the heavily modified ‘a2u-derived’
HOMOꢁ1 and the ‘a1u-derived’ HOMO (Fig. 10), the former lying
only 0.076 eV lower in energy. In contrast, the unoccupied frontier
orbitals differ in energy by 0.86 eV, the LUMO having an interesting
form. It has almost no electron density on the transverse 10,20-
meso carbons, and large coefficients on the 5,15 carbons and the azo
substituents, whereas the LUMOþ1 has the opposite bias, but
smaller coefficients on the azo groups.
When the transitions to electronically-excited states are exam-
ined (Table 1), an interesting pattern emerges. While the lowest-
energy Q-band is mostly contributed (86%) by the HOMOeLUMO
transition, the accompanying bands at 624 and 592 nm have a very
mixed composition, as expected from the near-degeneracy of the
occupied frontier orbitals. There are significant contributions from
much deeper levels to the LUMO and LUMOþ1. This situation is
actually the reverse of the classical D4h situation, for which the
filled a2u/a1u HOMOꢁ1/HOMO are close in energy, and the unfilled
frontier orbitals are the eg degenerate pair. This leads to the ‘stan-
dard’ situation described by Gouterman that explains the allowed/
forbidden intensity pattern of the B and Q bands. In the present
case with 19c, the unusual large energy difference between the
LUMO and LUMOþ1 leads to the prediction of three ‘Q bands’ at
671, 624, and 592 nm. The ‘Electron Density Difference Map’
(EDDM), representing the difference between the electron distri-
butions before and after the lowest-energy transition is shown in
Fig. 11. The extent of charge transfer from the pyrrole rings of the
core to the meso carbons and their attached electron-accepting
substituents is dramatically illustrated in this plot.
]
NeNHeC(]O)eOeC moieties are almost planar, implying exten-
sive ‘amide resonance’. On the other hand, the optimised geometry
for 19c has C1 symmetry, and the N]N units are disposed towards
the same side of the macrocycle, yet the C]O units point away from
each other. Other conformers were explored, and several lie in close
energetic proximity. In the global minimum-energy conformer, one
of the CmesoeN]NeC(]O)eOeC moieties is nearly planar, but the
other has a torsion angle of 50ꢀ along the N]NeC]O unit. Not-
withstanding this fact, the torsion angles CaeCmesoeN]N, which
are indicative of the delocalisation of the azo group with the por-
phyrin
p
system, are 24ꢀ for the substituent with the out-of-plane
ester carbonyl, and 20ꢀ for the ‘fully conjugated’ substituent.
We carried out the TD-DFT transition modelling using the global
minimum structure; the one-electron transitions for Ni8c are tab-
ulated in the Supplementary data. As far as we know, this is the first
report of the electronic structure calculations on a porphyrinoid
like this with interrupted macrocyclic conjugation. The calculated
gas-phase one-electron transitions are compared with the experi-
mental spectra of the analogues Ni8b and 19b measured in DCM in
Fig. 9. As expected for a comparison of gas-phase with solution
spectra, quantitative agreement is not observed, and also the cal-
culated transitions are systematically higher in energy. This is often
observed in TD-DFT calculations on mononuclear porphyrinoids
using these methods.34 Notably, three key features are reproduced
by the calculations, namely (i) the red-shift when the aromaticity is
restored in 19; (ii) the broad asymmetric Q-bands of 19b com-
prising three transitions with similar oscillator strengths at 670,
624 and 592 nm; and (iii) the increase in relative intensity of the Q-
like band versus the Soret-like band for the DIPD, not unexpected
for an interrupted porphyrinoid structure.
Ni8b exhibits quite a broad tail on the red edge of the lowest-
energy visible band, but there are no transitions with oscillator
strength f >0.05 predicted in this region for the optimised structure
(Fig. 9, top). This feature, as well as the rather strong absorption in
the 500 nm region, in which no significant transitions are calcu-
lated, may well be the result of populations of conformers rather
different from the single one used in the TD-DFT calculations.
Further discussions of these are unwarranted in the scope of this
initial study.
3. Conclusion
The 5,15-diiminoporphodimethenes reported here are the first
compounds of their type to be characterised. These porphyrinoids
with reduced conjugation and exocyclic imino-bonds are closely-
related to the oxoporphyrins and porphodimethenes. ‘Interrupted
porphyrins’, such as these have interesting properties: oxopor-
phyrins are involved in the degradation of biological porphyrins,
including the hemes, and 5,15-porphodimethenes, with their
marked electronic asymmetry across the macrocycle, resemble the
5,15-dialkynylporphyrins, which exhibit non-linear optical prop-
erties. Quinoidal/cumulenic structures possessing porphodime-
thene character are thought to contribute to the excited states of
the latter class of compound.21
As mentioned above, there are two reports of the closely-related
5-oxo-15-iminoporphodimethenes, namely the zinc and nickel OEP
derivatives (H2OEP¼2,3,7,8,12,13,17,18-octaethylporphyrin). The
former was reported by Fuhrhop and co-workers in 1970, and was
prepared by treatment of 5-aminoZnOEP with Fe3þ salts in the
presence of air.27 This porphodimethene was found to hydrolyse
easily to the dioxoporphyrin when left in chloroform or during
chromatography. Fuhrhop rationalised the formation of this species
as the oxidation of the aminoporphyrin to the radical-cation, which
was stabilised by delocalisation from the amino N lone pair,
forming an iminophlorin radical, which was then quenched by
Fig. 9. Comparison of experimental absorption spectra in DCM (solid lines) with
transitions calculated by TD-DFT in the gas-phase for Ni8b/Ni8c (top; vertical bars and
filled markers) and 19b/19c (bottom; vertical bars and unfilled markers).
The most interesting feature derived from the calculations is the
multi-component Q-band of 19c. The strong transverse electronic
asymmetry across the 5,15-positions due to the conjugated

L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
527
Fig. 10. DFT calculated frontier orbital surfaces for Ni bis(azocarboxylate) 19c.
Table 1
One-electron excitations for Ni bis(azocarboxylate) 19c calculated by TD-DFT (300e700 nm, fꢂ0.01, transition contribution ꢂ10%)
Excited state
Orbital composition
Excitation
eV
Oscillator strength
nm
cmꢁ1
f
1
2
HOMO
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
/
LUMO
LUMO
LUMO
LUMO
LUMO
LUMO
LUMOþ1
LUMOþ3
LUMOþ1
LUMOþ2
LUMO
86
34
43
25
14
31
12
18
15
39
47
36
19
33
32
25
10
28
19
59
14
70
35
12
20
33
15
5
1.849
1.986
671
624
14,911
16,016
0.0712
0.1050
HOMOꢁ5
HOMOꢁ1
HOMOꢁ5
HOMOꢁ2
HOMOꢁ1
HOMO
4
2.096
592
16,901
0.0974
11
HOMOꢁ19
2.754
450
22,213
0.0560
HOMO
HOMO
13
16
HOMOꢁ7
HOMOꢁ1
HOMOꢁ7
HOMOꢁ4
HOMOꢁ1
HOMOꢁ3
HOMOꢁ1
HOMO
2.836
2.918
437
425
22,870
23,532
0.1241
0.1215
LUMOþ1
LUMO
LUMOþ1
LUMOþ1
LUMOþ1
LUMO
LUMOþ1
LUMOþ2
LUMOþ1
LUMOþ1
LUMOþ2
LUMO
17
3.028
409
24,424
0.5140
HOMO
18
HOMOꢁ3
3.040
408
24,522
0.2498
HOMO
19
27
HOMOꢁ1
HOMOꢁ12
HOMOꢁ9
HOMOꢁ2
HOMOꢁ13
HOMO
HOMOꢁ15
HOMOꢁ13
HOMO
3.095
3.371
401
368
24,963
27,192
0.1629
0.2748
LUMO
LUMOþ2
38
39
LUMO
3.856
3.868
322
321
31,104
31,195
0.2276
0.2247
LUMOþ4
LUMO
LUMO
LUMOþ4
13
47

528
L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
(under argon): tetrahydrofuran (THF) from sodium wire and ben-
zophenone; DCM, 1,2-dichloroethane (DCE) and MeCN from CaH2;
pyrrole was distilled without additives. Preparative column chro-
matography was performed on silica gel (40e63 mm), which was
purchased from Qindao Haiyang Chemical Co. Ltd. China. Analytical
TLC was performed on Merck Silica Gel 60 F254 TLC plates. The
following porphyrin starting materials were prepared by literature
procedures: H2DPP,38 H2DPPBr,25 H2DPPBr2,39 NiDPPBr,25 ZnDPP
Br,39 NiDPPBr2,25 ZnDPPBr2,39 H2TriPP,40 H2TriPPBr,41 NiTriPPBr,41
H2DAPPBr.41
1H NMR experiments were recorded on a Bruker Avance 400
spectrometer at 400.155 MHz. All samples were recorded in CDCl3
and 0.1% pyridine-d5 was added to the solutions of zinc(II) por-
phyrins, unless otherwise stated. DQF COSY and gradient NOESY
experiments were acquired using standard pulse sequences with
2048 points in f2 dimension and 256 points in f1 dimension. The
gradient NOESYexperiments were performed with a mixing time of
1.5 s. All coupling constants are recorded in Hertz. UV/Vis spectra
were recorded on a Varian Cary 3 or Varian Cary 50 UVevisible
spectrometer, and were recorded in DCM, unless otherwise stated.
Accurate Electrospray Ionisation (ESI) and Laser Desorption
Ionisation (LDI) mass spectra were recorded at the School of
Chemistry, Monash University, Melbourne. ESI measurements were
performed using an Agilent 1100 Series LC attached to an Agilent
G1969A LC-TOF system with reference mass correction using NaI
clusters. An eluent of 50:50 DCM/MeOH with 0.1% formic acid was
employed using a flow rate of 0.3 mL/min. Solvent aspiration was
achieved by nitrogen gas flowing at 8 L/min. The source tempera-
ture was set to 350 ꢀC and the capillary voltage to 4.0 kV. Liquid
secondary ion mass spectrometry (LSIMS) mass spectra were
recorded at the Organic Mass Spectrometry Facility, University of
Tasmania, Hobart. LSIMS measurements were collected using m-
nitrobenzyl alcohol as proton donor and CsI/TEG mixture for ex-
ternal reference, on a Kratos Concept ISQ double focussing Mag-
netic/Electrostatic Mass Spectrometer. IR spectra were collected
using a Nicolet 870 Nexus Fourier Transform Infrared (FTIR) system
equipped with a diamond internal reflection element. Samples for
elemental microanalysis were dried at 80 ꢀC under vacuum for 12 h
and analyses were performed by the Microanalytical Service,
School of Molecular and Microbial Sciences, The University of
Queensland.
Fig. 11. Electron density difference map for the lowest-energy electronic transition
(86% HOMOeLUMO) of 19c calculated by TD-DFT. Blue: electron-poor region after
transition; orange: electron-rich region after transition.
aerial oxygen.27 More recently, Balch and co-workers synthesised
the nickel analogue, by treatment of 5-aminoNiOEP with FeCl3 in
ethanol-stabilised chloroform.26 The chemistry of porphyrinoids
bearing double bonds at one or more meso-positions has been
reviewed by Lee et al.35
The behaviour of alkyl carbazates in favouring the formation of
DIPDs readily from the initial amination products is so far unique.
We have tried other apparently similar nucleophiles in Pd-
catalysed aminations of bromoporphyrins, with no success: tosyl-
hydrazine, semicarbazide, thiosemicarbazide and 1,2-bis(ethoxycar
bonyl)hydrazine. With hydrazine or 1,1-diphenylhydrazine, the
NeN bond of the nucleophile is cleaved; in the former case, this led
to the 1ꢀ aminoporphyrin, the 2ꢀ bis(porphyrinyl)amine, and the
hydroxyporphyrin.18 In the latter case, remarkably the only amino-
product was the 2ꢀ amine.36 Another interesting point is that there
has so far been no mention of 5,15-bis(secondary amino)porphyrins
undergoing dearomatisation to DIPDs: several authors have re-
ported double aminations of 5,15-dihaloporphyrins,11c,12a,12f but no
one has reported oxidation chemistry analogous to what we have
found for the carbazates, even for 5,15-bis(4-methoxyanilino)por-
phyrin, which might be expected to be a prime candidate for such
an oxidation.12f This contrasts with the facile dearomatisation
found when aminoporphyrins undergo the head-to-tail oxidative
coupling exemplified by structures 3a and 3b.2,6 This type of
dearomatisation has been reported to afford oligomeric analogues
of 3a under the appropriate conditions.15
4.2. Palladium-catalysed reactions of carbazates and carba-
mates with meso-bromoporphyrins
Brominated porphyrin, amine or hydrazine source
(0.5e10 equiv), Pd(OAc)2 (7 mol %), rac-BINAP (20 mol %) and
Cs2CO3 (7 equiv) were placed into a dry, argon-filled Schlenk flask
and were dried under vacuum for 30 min. Freshly distilled, dry THF
(5 cm3 per 20 mg porphyrin) was added to the vessel and the re-
action was stirred under argon at 68 ꢀC. The reaction was monitored
by TLC [DCM/n-hexane (1:1)] until the bromoporphyrin starting
material was either fully consumed or perceived to be no longer
being consumed. The reaction was quenched with water and stirred
for 1 h open to air. The mixture was extracted into DCM, and the
solvent was then removed by rotary evaporation. The products
were separated using column chromatography. This procedure was
utilised for the palladium-catalysed amination of porphyrins in all
cases, unless otherwise stated.
In a subsequent report, we will describe the properties of the
unique 5,10-diimino ‘corner porphyrin’ analogues and compare
them with the present 5,15-disubstituted series.37
4. Experimental section
4.1. General
4.2.1. 5-(tert-Butoxycarbonylazo)-10,20-diphenylporphyrinatonickel
(II) (Ni7a) and N,N0-bis(tert-butoxycarbonylamino)-5,15-diimino-
10,20-diphenylporphodimethenatonickel(II) (Ni8a). NiDPPBr (Ni5,
39.4 mg, 0.0668 mmol), tert-butyl carbazate (10 equiv, 88.6 mg,
0.668 mmol), Pd(OAc)2 (1.0 mg, 0.0045 mmol), rac-BINAP (8.3 mg,
0.0013 mmol), Cs2CO3 (155 mg, 0.476 mmol) and dry THF (10 cm3)
All solvents were Analytical Reagent grade, unless otherwise
stated. Triethylamine (TEA) and pyridine were stored over KOH;
toluene and diethyl ether were stored over freshly pressed sodium
wire. The following solvents were freshly distilled prior to use

L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
529
were stirred for 27 h at 68 ꢀC. The products were separated using
column chromatography [DCM then DCM/EtOAc (19:1)]. The first
eluting compound after the orange starting material was the bright
green Ni7a (15 mg, 35% crude yield), which resisted recrystallisation.
The final product eluted upon the addition of MeOH and was the
blue-green Ni8a (37 mg, 71% crude yield). DIPD Ni8a was recrys-
tallised from a mixture of MeOH and H2O (13 mg, 25%). Ni7a: 1H
(CO); UV/vis: lmax/nm (ε/103 Mꢁ1 cmꢁ1) 335 (33.4), 438 (48.7), 605
(23.0).
4.2.5. N,N0-Bis(ethoxycarbonylamino)-5,15-diimino-10,20-
diphenylporphodimethenatonickel(II) (Ni8b) synthesised without
palladium. NiDPPBr2 (Ni10, 20.0 mg, 0.0295 mmol), ethyl carbazate
(10 equiv, 30.7 mg, 0.295 mmol), Cs2CO3 (68.2 mg, 0.209 mmol) and
dry THF (5 cm3) were stirred at 68 ꢀC for 4 weeks. The blue-green
product was purified using column chromatography [DCM then
DCM/MeOH (100:1)] and was isolated in 9% yield (1.9 mg).
NMR
d
: 9.68 (2H, d, J 4.9,
b
H), 9.67 (1H, s, meso H), 8.99 (2H, d, J 4.9,
H), 7.98e7.96 (4H, m,
b
H), 8.83 (2H, d, J 4.9,
b
H), 8.67 (2H, d, J 4.6, b
phenyl H), 7.74e7.68 (6H, m, phenyl H), 1.81 (9H, s, CH3); m/z (HRESI)
647.1706 ([MþH]þ, C37H28N6NiO2 calcd 646.1627); nmax (cmꢁ1) 3330
(NH), 1740 (CO); UV/vis: lmax/nm (ε/103 Mꢁ1 cmꢁ1) 334 (31.7), 437
4.2.6. 5-(tert-Butoxycarbonylazo)-10,15,20-triphenylporphyrinatonic
kel(II) (12a) and 5-[2-(ethoxycarbonyl)hydrazo]-10,15,20-
triphenylporphyrinatonickel(II) (13a). NiTriPPBr (11, 50 mg,
0.074 mmol), tert-butyl carbazate (10 equiv, 101 mg, 0.74 mmol),
Pd(OAc)2 (1.3 mg, 0.006 mmol), rac-BINAP (9.3 mg, 0.015 mmol),
Cs2CO3 (172 mg, 0.528 mmol) and dry THF (12.5 cm3) were stirred
for 17.5 h at 68 ꢀC. The products were purified using column
chromatography [DCM/n-hexane (1:1) then DCM]. The first eluting
compound after the starting material was the bright green azo-
carboxylate 12a, which was recrystallised by the evaporation of
a DCM/pentane mixture (3.2 mg, 6%). The second major product
eluted upon changing to DCM and was the red carbazate porphyrin
13a, which was recrystallised by the evaporation of a DCM/pentane
(47.7), 536 (12.0), 606 (23.4). Ni8a: 1H NMR
d
: 8.72 (2H, br s, N H),
7.49e7.42 (10H, m, phenyl H), 6.96 (2H, d, J 4.6, H), 6.85 (2H, d, J 4.2,
H), 6.69 (2H, d, J 4.9, H), 6.68 (2H, d, J 4.4, b H), 1.47 (18H, s, CH3);
b
b
b
m/z (HRESI) 777.2439 ([MþH]þ, C42H38N8NiO4 calcd 776.2369); UV/
vis: lmax/nm (ε/103 Mꢁ1 cmꢁ1) 334 (31.7), 437 (47.7), 606 (23.4).
Crystals were grown by the slow diffusion of water into a solution of
Ni8a in dioxane. CCDC-268791 contains the crystallographic data for
this compound.
4.2.2. 5-[2-(tert-Butoxycarbonyl)hydrazo]-10,20-diphenylporphyrin
atonickel(II) (Ni9a). NiDPPBr (Ni5, 20.0 mg, 0.0334 mmol), tert-
butyl carbazate (10 equiv, 111 mg, 0.836 mmol), Pd(OAc)2 (1.3 mg,
0.0056 mmol), rac-BINAP (10.4 mg, 0.0167 mmol), Cs2CO3 (193 mg,
0.594 mmol) and dry THF (12.5 cm3) were stirred for 13.5 h at 68 ꢀC.
The product was purified using column chromatography (DCM)
and the red product Ni9a was recrystallised by the slow evapora-
mixture (35 mg, 65%). 12a: 1H NMR
(2H, d, J 5.1, H), 8.48 (2H, d, J 4.9,
d
: 9.54 (2H, d, J 5.1,
H), 8.40 (2H, d, J 4.9, b
b
H), 8.68
H),
b
b
7.86e7.84 (6H, m, phenyl H), 7.63e7.56 (9H, m, phenyl H), 1.71 (9H,
s, CH3); m/z (HR-ESI) 745.1833 ([MþNa]þ, C43H32N6NaNiO2 calcd
tion of DCM (46 mg, 84%). 1H NMR
d, J 5.1, H), 8.99 (2H, d, J 4.8, H), 8.77 (2H, d, J 4.8,
J 5.0, H), 8.01e7.98 (4H, m, phenyl H), 7.74 (1H, br s, amine N H),
d: 9.60 (1H, s, meso H), 9.44 (2H,
745.1838); nmax (cmꢁ1
)
1740 (CO); UV/vis: lmax/nm
b
b
b
H), 8.76 (2H, d,
(ε/103 Mꢁ1 cmꢁ1) 446 (200.8), 616 (25.3); Found: C, 70.72; H, 4.69;
b
N,11.11%; (Calcd C, 70.81; H, 4.06; N,12.09).13a: 1H NMR (CD2Cl2)
d:
7.72e7.66 (6H, m, phenyl H), 7.17 (1H, br s, amide N H), 1.54 (9H, t, J
9.34 (2H, d, J 4.9,
8.54 (2H, d, J 5.0,
b
b
H), 8.64 (2H, d, J 5.0,
H), 7.93e7.88 (6H, m, phenyl H), 7.67 (1H, br s,
b H), 8.56 (2H, d, J 5.0, b H),
7.1, CH3); m/z (LSIMS) 648.1788 (Mþ, C37H30N6NiO2 calcd 648.1784);
nmax (cmꢁ1
)
3354 (NH), 1718 (CO); UV/vis: lmax/nm
amine N H) 7.65e7.59 (9H, m, phenyl H), 7.25 (1H, br s, amide N H),
1.45 (9H, s, CH3), amine and amide N H peaks assigned using D2O
exchange experiments; m/z (HR-ESI) 747.1985 ([MþNa]þ,
(ε/103 Mꢁ1 cmꢁ1) 415 (199.8), 528 (13.6), 582 sh. (3.0); Found: C,
68.27; H, 4.77; N, 12.81%; (Calcd C, 68.44; H, 4.66; N, 12.94).
C
43H34N6NaNiO2 calcd 747.1994); UV/vis: lmax/nm (ε/103 Mꢁ1
4.2.3. 5-(Ethoxycarbonylazo)-10,20-diphenylporphyrinatonickel(II)
(Ni7b). NiDPPBr (Ni5, 20.0 mg, 0.0334 mmol), ethyl carbazate
(1 equiv, 3.5 mg, 0.0334 mmol), Pd(OAc)2 (0.5 mg, 0.002 mmol),
rac-BINAP (4.2 mg, 0.0067 mmol), Cs2CO3 (77.3 mg, 0.237 mmol)
and dry THF (5 cm3) were stirred for 18 h at 68 ꢀC. The quenched
reaction mixture was allowed to stir at room temperature, open to
air, overnight. The product was purified using column chromatog-
raphy (DCM, 13 mg, 63% crude yield) and the bright green Ni7b was
cmꢁ1) 422 (202), 489 (10.4), 535 (12.6) 576 sh (3.8).
4.2.7. 5-(Ethoxycarbonylazo)-10,15,20-triphenylporphyrinatonickel
(II) (12b). NiTriPPBr (Ni5, 20.0 mg, 0.0297 mmol), ethyl carbazate
(10 equiv, 30.9 mg, 0.297 mmol), Pd(OAc)2 (0.5 mg, 0.002 mmol), rac-
BINAP (3.7 mg, 0.0059 mmol), Cs2CO3 (68.6 mg, 0.211 mmol) and dry
THF (5 cm3) were stirred for 22 h at 68 ꢀC. The quenched reaction
mixture was allowed to stir at room temperature, open to air, over-
night. The product was purified using column chromatography
(DCM). The resulting product 12b was bright green (14 mg, 67% crude
yield) and resisted recrystallisation. 1H NMR showed that the sample
recrystallised from THF/H2O (5.2 mg, 25%). 1H NMR
5.1, H), 9.50 (1H, s, meso H), 8.87 (2H, d, J 4.7, H), 8.78 (2H, d, J 5.1,
H), 8.59 (2H, d, J 4.7, H), 7.94e7.91 (4H, m, phenyl H), 7.70e7.68
d: 9.67 (2H, d, J
b
b
b
b
(6H, m, phenyl H), 4.69 (2H, q, J 7.1, CH2), 1.61 (3H, t, J 7.1, CH3); m/z
(MALDI-TOF-MS) 619.14 ([MþH]þ, C35H25N6NiO2 calcd 619.14);
nmax (cmꢁ1) 1733 (CO); UV/vis: lmax/nm (ε/103 Mꢁ1 cmꢁ1) 443
(79.5), 534 (5.4), 612 (10.7).
was not pure enough for elemental analysis. 1H NMR
5.1, H), 8.75 (2H, d, J 5.1, H), 8.54 (2H, d, J 4.9, H), 8.46 (2H, d, J 5.1,
H), 7.94e7.92 (6H, m, phenyl H), 7.71e7.65 (9H, m, phenyl H), 4.68
(2H, q, J 7.1, CH2), 1.60 (3H, t, J 7.1, CH3); m/z (LSIMS) 694.1650 (Mþ,
41H28N6NiO2 calcd 694.1627); UV/vis: lmax/nm 449, 621.
d: 9.36 (2H, d, J
b
b
b
b
C
4.2.4. N,N0-Bis(ethoxycarbonylamino)-5,15-diimino-10,20-
diphenylporphodimethenatonickel(II)
(Ni8b). NiDPPBr2
(Ni10,
4.2.8. N,N0-Bis(ethoxycarbonylamino)-5,15-diimino-10,20-
diphenylporphodimethenatozinc(II) (Zn8b). ZnDPPBr2 (Zn10,
20.0 mg, 0.0295 mmol), ethyl carbazate (2.2 equiv, 6.75 mg,
0.065 mmol), Pd(OAc)2 (0.5 mg, 0.002 mmol), rac-BINAP (3.7 mg,
0.0059 mmol), Cs2CO3 (68.2 mg, 0.209 mmol) and dry THF (5 cm3)
were stirred at 68 ꢀC for 26 h. The product was purified using
column chromatography [DCM then DCM/EtOAc (19:1)]. The
resulting product Ni8b was dark green (13 mg, 61% crude yield) and
20.0 mg, 0.0293 mmol), ethyl carbazate (10 equiv, 30.5 mg,
0.293 mmol), Pd(OAc)2 (0.5 mg, 0.002 mmol), rac-BINAP (3.6 mg,
0.0059 mmol), Cs2CO3 (67.7 mg, 0.208 mmol) and dry THF (5 cm3)
were stirred at 68 ꢀC for 17.5 h. The product was purified by column
chromatography [DCM then DCM/MeOH (100:1)], and the resulting
red product Zn8b was recrystallised from DCM/pentane (17 mg,
was recrystallised from dioxane/H2O at 4 ꢀC. 1H NMR
s, N H), 7.52e7.41 (10H, m, phenyl H), 6.96 (2H, d, J 4.9,
(2H, d, J 4.4, H), 6.71 (2H, d, J 4.9, H), 6.68 (2H, d, J 4.4,
d: 8.80 (2H, br
b
H), 6.84
H), 4.33
82%). 1H NMR
7.25 (2H, H, found from COSY and NOESY experiments), 6.88 (2H,
d, J 3.9, H), 6.77 (2H, d, J 4.4, H), 6.67 (2H, d, J 3.9, H), 4.27 (4H,
q, J 7.1, CH2), 1.31 (6H, t, J 7.2, CH3); m/z (MALDI-TOF-MS) 727.1553
d: 9.07 (2H, br s, N H), 7.50e7.45 (10H, m, phenyl H),
b
b
b
b
(4H, q, J 7.1, CH2), 1.34 (6H, t, J 7.1, CH3); m/z (MALDI-TOF-MS) 721.15
b
b
b
([MþH]þ, C38H31N8NiO4 calcd 721.18); nmax (cmꢁ1) 3327 (NH), 1728

530
L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
([MþH]þ, C38H31N8O4Zn calcd 727.1760); nmax (cmꢁ1) 3330 (NH),
1730 (CO); UV/vis: lmax/nm (ε/103 Mꢁ1 cmꢁ1) 328 (35.8), 415 (36.7),
437 (53.4), 552 (33.3). Crystals of [Zn8b(pyridine)]$dioxane$2H2O
suitable for X-ray structure determination were grown by the slow
diffusion of water into a solution of Zn8b in dioxane and pyridine.
(1.1 equiv, 11 mg, 0.0049 mmol) was added. The mixture was
shaken and then passed through a short column [DCM/n-hexane
(1:1)]. The bright green azocarboxylate 12b was recrystallised by
the evaporation of a mixture of DCM/pentane, to give a dark green
powder (30 mg, 95% from two crops), having 1H NMR data identical
to those of the product described in Section 4.2.6.
4.2.9. N,N0-Bis(tert-butoxycarbonylamino)-5,15-diimino-10,20-
diphenylporphodimethene (H28a). H2DPPBr (H25, 22.0 mg,
0.0407 mmol), tert-butyl carbazate (10 equiv, 53.9 mg, 0.407 mmol),
Pd(OAc)2 (0.6 mg, 0.003 mmol), rac-BINAP (5.0 mg, 0.0081 mmol),
Cs2CO3 (94.1 mg, 0.289 mmol) and dry THF (6.5 cm3) were stirred for
14.5 h at 68 ꢀC. The mixture was purified using column chroma-
tography [DCM then DCM/MeOH (100:1)]. The orange product H28a
eluted upon the addition of MeOH and was purified from hot/cold
4.3.2. The reduction of 5-(tert-butoxycarbonylazo)-10,15,20-
triphenylporphyrinatonickel(II) (12a). Azocarboxylate porphyrin
12a (26 mg, 0.035 mmol) was dissolved in THF/MeOH (4 cm3/8
drops) and NaBH4 (10 equiv, 13 mg, 0.35 mmol) was added. The
mixture was stirred at room temperature for 40 min, after which
the only significant product was the carbazate porphyrin 13a. The
mixture was then refluxed for 20 h and the major product was
NiTriPP, which was recrystallised from DCM/pentane (10 mg, 48%).
A small amount of the aminoporphyrin Ni1b was also isolated, and
was recrystallised by the evaporation of a mixture of DCM/MeOH
(1 mg, 5%). Both products were identified by comparison of their 1H
NMR spectra with those of authentic samples.18,42
MeOH/H2O recrystallisation (8.8 mg, 37%). 1H NMR
inner NH), 9.01 (2H, br s, NH), 7.50e7.45 (10H, m, phenyl H), 7.25 (2H,
d: 12.42 (2H, br s,
d, J 4.7,
dd,
b H), 6.85 (2H, br dd, b H), 6.77 (2H, d, J 4.7, b H), 6.44 (2H, br
b
H), 1.55 (18H, s, CH3); m/z (ESI) 721.3260 ([MþH]þ, C42H41N8O4
calcd 721.3251); nmax (cmꢁ1) 3330 (NH), 1720 (CO); UV/vis: lmax/nm
(ε/103 Mꢁ1 cmꢁ1) 316 (35.7), 394 (38.2), 489 (32.2).
4.3.3. The oxidation of N,N0-bis(tert-butoxycarbonylamino)-5,15-
diimino-10,20-diphenylporphodimethenatonickel(II) (Ni8a) to give
5,15-bis(t-butoxycarbonylazo)-10,20-diphenylporphyrinatonickel(II)
(19a). The DIPD (Ni8a 28 mg, 0.035 mmol) was dissolved in DCM
(4 cm3), DDQ (2.1 equiv, 17 mg, 0.075 mmol) was added and the
mixture was shaken briefly. The mixture was separated using col-
umn chromatography (DCM) to give the bright green bis(azo-
carboxylate) 19a (21 mg, 77%), which was recrystallised from DCM/
4.2.10. N,N0-Bis(ethoxycarbonylamino)-5,15-diimino-10,20-
diphenylporphodimethene (H28b). H2DPPBr2 (H25, 20.0 mg,
0.0322 mmol), ethyl carbazate (10 equiv, 33.5 mg, 0.322 mmol),
Pd(OAc)2 (0.5 mg, 0.002 mmol), rac-BINAP (4.0 mg, 0.0064 mmol),
Cs2CO3 (74.4 mg, 0.228 mmol) and dry THF (5 cm3) were stirred at
68 ꢀC for 20 h. The product was purified by column chromatography
[DCM/TEA (200:1) then DCM/MeOH/TEA (200:1:1), two columns],
yielding impure dark orange product H28b (13 mg, 54%). 1H NMR
12.47 (2H, br s, inner N H), 9.08 (2H, br s, N H), 7.50e7.45 (10H, m,
phenyl H), 6.85 (2H, d, J 4.9, H), 6.84 (2H, dd, J 4.0,1.8, H), 6.55 (2H,
d:
pentane to give a dark green powder (12 mg, 36%). 1H NMR
(4H, d, J 5.0, H), 8.69 (4H, d, J 5.0, H), 7.93e7.91 (4H, m, phenyl H),
d: 9.50
b
b
b
b
7.73e7.67 (6H, m, phenyl H),1.79 (18H, s, CH3); m/z (LSIMS) 774.2223
(Mþ, C42H36N8NiO4 calcd 774.2213); nmax (cmꢁ1) 1740 (CO); UV/vis:
lmax (ε/103 Mꢁ1 cmꢁ1) 321 (22.8), 449 (128), 645 (21.5).
d, J 4.9,
(6H, t,
b H), 6.54 (2H, dd, J 4.0, 1.8, b H), 4.34 (4H, q, J 7.1, CH2), 1.35
J
7.1, CH3); m/z (MALDI-TOF-MS) 665.2613 ([MþH]þ,
C
38H33N8O4 calcd 665.2625); nmax (cmꢁ1) 3330 (NH), 1720 (CO).
4.3.4. The oxidation of N,N0-bis(ethoxycarbonylamino)-5,15-diimino-
10,20-diphenylporphodimethenatonickel(II) (Ni8b) to give 5,15-
bis(ethoxycarbonylazo)-10,20-diphenylporphrinatonickel(II)
(19b). The DIPD Ni8b (15 mg, 0.021 mmol) was dissolved in DCM
(2 cm3), DDQ (2.1 equiv, 10 mg, 0.045 mmol) was added and the
mixture was shaken briefly. The mixture was separated using col-
umn chromatography (DCM) to give the bright green bis(azo-
carboxylate) 19b (13 mg, 86%), which was recrystallised from DCM/
4.2.11. 5-(tert-Butoxycarbonylazo)-10,20-bis(3,5-di-tert-butyl-
phenyl)-15-phenylporphyrin (15). H2DAPPBr (14, 50 mg,
0.059 mmol), tert-butyl carbazate (10 equiv, 63.1 mg, 0.590 mmol),
Pd(dppe)2 (7 mol %, 3.8 mg, 0.004), Cs2CO3 (7.1 equiv, 137.4 mg,
0.420 mmol) and dry THF (10 cm3) were stirred at 68 ꢀC for 23 h and
the major product seen by TLC was the green azocarboxylate 15. This
product was separated using column chromatography [DCM]. Two
other compounds were isolated by a second column [toluene: ethyl
acetate 1%], these being blue 16 and violet 17. The compounds were
recrystallised from dioxane/water (15: 15.3 mg, 29% yield,16: 5.8 mg,
pentane to give the fluffy blue product (11 mg, 71%). 1H NMR
(4H, d, J 5.0, H), 8.63 (4H, d, J 5.0, H), 7.88e7.86 (4H, m, phenyl
d: 9.46
b
b
H), 7.71e7.64 (6H, m, phenyl H), 4.69 (4H, q, J 7.1, CH2), 1.59 (6H, t, J
17: 1.0 mg). 15: 1H NMR
4.9 Hz, H), 8.65 (4H, br s,
d
: 9.84 (2H, d, J 4.9 Hz,
b H), 8.88 (2H, d, J
7.1, CH3); m/z (MALDI-TOF-MS) 719.12 ([MþHþ], C38H29N8NiO4
b
b
H), 8.14 (2H, d, phenyl H), 8.01 (4H, m,
calcd 719.17); nmax (cmꢁ1) 1741 (CO); UV/vis: lmax (ε/103 Mꢁ1 cmꢁ1
)
phenyl H), 7.80 (2H, m, phenyl H), 7.73 (3H, m, phenyl H), 1.55 (9H, s,
330 (24.7), 452 (132), 653 (23.6); Found: C, 63.43; H, 4.04; N,
15.50%; (Calcd C, 63.45; H, 3.92; N, 15.58).
t
But H), 1.53 (36H, s, Bu H), ꢁ1.29 (2H, s, inner H); m/z (ESIMS)
891.5314 ([MþH]þ, C59H67N6O2 calcd 891.5326); UV/vis: lmax/nm
(ε/103 Mꢁ1 cmꢁ1) 435 (65.1), 602 (8.7), 677 (5.2), 824 (2.4). 16: 1H
4.3.5. The reduction of 5,15-bis(ethoxycarbonylazo)-10,20-diphe-
nylporphrinatonickel(II) (19b). Bis(azocarboxylate) 19b (5.0 mg,
0.0070 mmol) was dissolved in THF/MeOH (2 cm3/4 drops) and
NaBH4 (10 equiv, 2.6 mg, 0.070 mmol) was added. The mixture was
stirred at room temperature for 1 day and then refluxed for 23 h.
The mixture was purified by column chromatography [DCM/EtOAc
(9:1) then (1:1)] to give DIPD Ni8b (3.0 mg, 59%), which was
identified by comparison of its 1H NMR spectrum with that of an
authentic sample prepared as in Section 4.2.4.
NMR
d: 12.54 (1H, s, N H), 10.51 (1H, s, N H), 10.28 (1H, s, N H), 8.45
(1H, s, N H), 7.73 (2H, d, o-Ar H), 7.65 (1H, t, p-Ar H), 7.54 (1H, t, p-Ar
H), 7.45e7.37 (5H, m, Ph H), 7.34 (2H, d, o-Ar H), 6.92 (1H, dd,
6.86 (1H, d, J 5.5 Hz, H), 6.78 (1H, d, J 4.5 Hz, H), 6.71 (1H, dd,
6.52 (1H, d, J 4.5 Hz, H), 6.51 (1H, d, J 4.5 Hz,
4.5 Hz, H), 6.03 (1H, br d, J 5.5 Hz,
H), 1.46 (9H, s, tBu H),1.36 (18H,
b
H),
b
b
b
H),
b
b H), 6.29 (1H, d, J
b
b
t
t
s, Bu H), 1.35 (18H, s, Bu H); UV/vis: lmax/nm (intensity ratio) 370
(1.9), 580 (1.0). 16 and 17: m/z (ESIMS) 16: 925.5382, 17: 925.5385
([MþHþCH3OH]þ, C60H73N6O3 calcd 925.5744).
4.4. Crystal structure determination of [Zn8b(pyridine)]$
dioxane$2H2O
4.3. Redox reactions of the coupling products
4.3.1. The oxidation of 5-[2-(ethoxycarbonyl)hydrazo]-10,15,20-
triphenylporphyrinatonickel(II) (13b). Carbazate porphyrin 13b
(32 mg, 0.044 mmol) was dissolved in DCM (1.5 cm3) and DDQ
Data for [Zn8b(pyridine)]$dioxane$2H2O were collected at ap-
proximately 150 K using double diamond monochromated syn-
ꢂ
chrotron radiation (0.48595 A) using
u and j scans at the

L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
531
ChemMatCARS beamline at the Advanced Photon Source. Data in-
tegration and reduction were undertaken with SAINT and XPREP.42a
Subsequent computations were carried out using the WinGX-32
graphical user interface.42b Structures were solved by direct
methods using SIR97.42c Data were refined and extended with
SHELXL-2013.42d Non-hydrogen atoms were refined anisotropically.
The complex has crystallographic twofold symmetry (the twofold
axis is coincident with the ZneNpyridine bond and the dioxane mol-
eculehascrystallographicinversionsymmetry). Thewatermolecules
occupy general positions in the unit cell. Carbon-bound hydrogen
atoms were included in idealised positions and refined using a riding
model. The nitrogen-bound hydrogen atom H(3) was located in the
difference map and refined without restraints. As oxygen bound
hydrogen atoms (associated with the water of crystallization) could
not be located in the difference Fourier map they were not modelled.
Crystal data and refinement details for [Zn8b(pyridine)]$diox-
Ni8c, and tables of TD-DFT calculated excited states for Ni8c. The
optimised geometries are provided as .mol files. Supplementary
data related to this article can be found at http://dx.doi.org/10.1016/
j.tet.2013.11.006.
References and notes
1. Screen, T. E. O.; Blake, I. M.; Rees, L. H.; Clegg, W.; Borwick, S. J.; Anderson, H. L. J.
Chem. Soc., Perkin Trans. 1 2002, 320.
2. Esdaile, L. J.; Jensen, P.; McMurtrie, J. C.; Arnold, D. P. Angew. Chem., Int. Ed. 2007,
46, 2090.
ꢀ ꢁ
3. Basic, B.; McMurtrie, J. C.; Arnold, D. P. J. Porphyrins Phthalocyanines 2010, 14,
482.
ꢀ ꢁ
4. Basic, B.; McMurtrie, J. C.; Arnold, D. P. Eur. J. Org. Chem. 2010, 4329.
5. Shen, D. M.; Lin, C.; Chen, X. G.; Chen, Q. Y. Synlett 2009, 945.
6. Smith, M. J. D. Phil. Thesis, University of Oxford, 2004.
7. Wang, Z.; Skerlj, R. T.; Bridger, G. J. Tetrahedron Lett. 1999, 40, 3543.
8. Lim, Y. K.; Lee, K. S.; Cho, C. G. Org. Lett. 2003, 5, 979.
9. Kim, K.-Y.; Shin, J.-T.; Lee, K.-S.; Cho, C.-G. Tetrahedron Lett. 2004, 45, 117.
10. Takanami, T.; Hayashi, M.; Hino, F.; Suda, K. Tetrahedron Lett. 2003, 44, 7353.
11. Chen, Y.; Gao, G.-Y.; Zhang, X. P. Tetrahedron Lett. 2005, 46, 4965; (b) Gao, G.-Y.;
Chen, Y.; Zhang, P. Org. Lett. 2004, 6, 1837; (c) Chen, Y.; Zhang, X. P. J. Org. Chem.
2003, 68, 4432.
ane$2H2O:
C47H47N9O8Zn,
M¼931.10,
monoclinic,
P2/c,
ꢂ
ꢂ
ꢂ
b
a¼12.4001(8) A, b¼12.8989(8) A, c¼13.5031(9) A,
¼96.163(3)ꢀ,
3
ꢂ
V¼2147.3(2) A , Dc¼1.440
g
cmꢁ3
,
Z¼2, crystal size
0.090ꢃ0.080ꢃ0.015 mm, purple plate, 150(2) K,
l(Mo-K
a
)¼
12. (a) Akita, M.; Hiroto, S.; Shinokobu, H. Angew. Chem., Int. Ed. 2012, 51, 2894; (b)
Ladomenou, K.; Lazarides, T.; Panda, M. K.; Charalambidis, G.; Daphnomili, D.;
0.71073, 2qmax¼44.9ꢀ, hkl range ꢁ19 to 19, e20 to 18, e21 to 19,
ꢁ
Coutsolelos, A. G. Inorg. Chem. 2012, 51, 10548; (c) Abada, Z.; Ferrie, L.; Akagah,
N¼33,020, Nind¼7943 (Rmerge¼0.060), Nobs¼5558 (I>2
s(I)),
ꢃ
B.; Lormier, A. T.; Figadere, B. Tetrahedron Lett. 2012, 53, 6961; (d) Pereira, A. M.
Nvar¼299, residuals R1(F, 2
s
)¼0.051, wR2(F2, all)¼0.125, GoF(all)¼
V. M.; Neves, M. G. P. M. S.; Cavaleiro, J. A. S.; Jeandon, C.; Gisselbrecht, J.-P.;
Choua, S.; Ruppert, R. Org. Lett. 2011, 13, 4742; (e) Sakamoto, R.; Nishikawa, M.;
Yamamura, T.; Kume, S.; Nishihara, H. Chem. Commun. 2010, 2028; (f) Shen, D.-
M.; Liu, C.; Chen, X.-G.; Chen, Q.-Y. J. Org. Chem. 2009, 74, 206; (g) Liu, C.; Shen,
D.-M.; Chen, Q.-Y. J. Org. Chem. 2007, 72, 2732.
ꢂꢁ3
1.351, Drmin,max¼ꢁ0.79, 1.15 e A . The refinement residuals are
defined as R1¼SjjFojꢁjFcjj/SjFoj for Fo>2 (Fo) and wR2¼
s
{S[w(Fo2ꢁF2c)2]/S[w(F2c)2]}1/2 where w¼1/[
s
2(Fo2)þ(0.04P)2þ0.5P],
P¼(Fo2þ2F2c)/3. The CIF file has been deposited with the Cambridge
Structural Database (CCDC reference number 953602) and can be
obtained free of charge via www.ccdc.cam.ac.uk/conts/retrie-
ving.html (or from the Cambridge Crystallographic Data Centre, 12
Union Road, Cambridge CB2 1EZ, UK; fax: (þ44) 1223-336-033; or
deposit@ccdc.cam.ac.uk).
13. (a) Warnan, J.; Favereau, L.; Meslin, F.; Severac, M.; Blart, E.; Pellegrin, Y.; Jac-
quemin, D.; Odobel, F. ChemSusChem 2012, 5, 1568; (b) Liu, Y.; Lin, H.; Dy, J. T.;
Tamaki, K.; Nakazaki, J.; Nakayama, D.; Uchida, S.; Kubo, T.; Segawa, H. Chem.
Commun. 2011, 4010; (c) Lee, C.-W.; Lu, H.-P.; Reddy, N. M.; Lee, H.-W.; Diau, E.
W.-G.; Yeh, C.-Y. Dyes Pigm. 2011, 91, 317; (d) Mathew, S.; Iijima, H.; Toude, Y.;
Umeyama, T.; Matano, Y.; Ito, S.; Tkachenko, N. V.; Lemmetyinen, H.; Imahori,
H. J. Phys. Chem. C 2011, 115, 14415; (e) Imahori, H.; Matsubara, Y.; Iijima, H.;
Umeyama, T.; Matano, Y.; Ito, S.; Niemi, K.; Tkachenko, N. V.; Lemmetyinen, H. J.
Phys. Chem. C 2010, 114, 10656; (f) Lee, C.-W.; Lu, H.-P.; Lan, C.-M.; Huang, Y.-L.;
Liang, Y.-R.; Yen, W.-N.; Liu, Y.-C.; Lin, Y.-S.; Diau, E. W.-G.; Yeh, C.-Y. Chem.dEur.
J. 2009, 15, 1403.
4.5. Computational methods
14. Esdaile, L. J.; McMurtrie, J. C.; Turner, P.; Arnold, D. P. Tetrahedron Lett. 2005, 46,
6931.
The initial geometries of the porphyrin 19c and porphodime-
thene Ni8c were first calculated at the semi-empirical level using
the PM3 method. The geometry was then further optimised by
performing a restricted closed-shell calculation with Gaussian 0943
using the DFT model with Becke’s three-parameter hybrid func-
tional, Lee, Yang and Parr’s correction for correlation interactions
and a 6-31G(d,p) basis set.44 Time-dependent DFTcalculations were
subsequently performed on the resulting optimised geometries at
the same level. The electron density difference map of the HOMO to
LUMO transition of 19c was constructed with the aid of the Gaus-
15. Yamashita, K.-i.; Takeuichi, S.; Asano, M. S.; Sugiura, K.-i. Book of Abstracts. In-
ternational Conference on Porphyrins and Phthalocyanines; Society of Por-
phyrins and Phthalocyanines: Jeju, South Korea, Jul 1e6, 2012; p 583.
16. Kwong, F. Y.; Klapars, A.; Buchwald, S. L. Org. Lett. 2002, 4, 581.
17. (a) Yamashita, K.-i.; Kataoka, K. .; Asano, M. S.; Sugiura, K.-i. Org. Lett. 2012, 14,
190; (b) Balaban, M. C.; Gillot-Chappaz, C.; Canard, G.; Fuhr, O.; Roussel, C.;
Balaban, T. S. Tetrahedron 2009, 65, 3733.
18. Esdaile, L. J.; Senge, M. O.; Arnold, D. P. Chem. Commun. 2006, 4192.
19. Harper, S. R.; Arnold, D. P. Unpublished results.
ꢀ ꢁ
20. Basic, B. Ph.D. Thesis, Queensland University of Technology, 2010.
21. Blake, I. M.; Anderson, H. L.; Beljonne, D.; Bredas, J.-L.; Clegg, W. J. Am. Chem.
Soc. 1998, 120, 10764.
22. Senge, M. O.; Smith, K. M. Z. Naturforsch., B: Chem. Sci. 1993, 48, 991.
23. Kamo, M.; Tsuda, A.; Nakamura, Y.; Aratani, N.; Furukawa, K.; Kato, T.; Osuka, A.
Org. Lett. 2003, 5, 2079.
3
sum program43 and displayed using an isovalue of 0.0004 e/A .
ꢂ
Acknowledgements
24. Senge, M. O. In The Porphyrin Handbook; Kadish, K. M., Smith, K. M., Guilard, R.,
Eds.; Academic: London, 2000; Vol. 10, pp 1e218.
We thank Queensland University of Technology (QUT) for
a post-graduate scholarship for L.E., the Australian Research Council
for financial assistance (DP0663774), and the High Performance
Computing team, Queensland University of Technology (QUT) for
their support of the Gaussian 09 program. Use of the Chem-
MatCARS Sector 15 at the Advanced Photon Source was supported
by the Australian Synchrotron Research Program, which is funded
by the Commonwealth of Australia under the Major National Re-
search Facilities Program. ChemMatCARS Sector 15 is also sup-
ported by the National Science Foundation/Department of Energy.
The Advanced Photon Source is supported by the U.S. Department
of Energy, Basic Energy Sciences.
25. Arnold, D. P.; Bott, R. C.; Eldridge, H.; Elms, F. M.; Smith, G.; Zojaji, M. Aust. J.
Chem. 1997, 50, 495.
26. Sprutta, N.; Rath, S. P.; Olmstead, M. M.; Balch, A. L. Inorg. Chem. 2005, 44, 1452.
27. Fuhrhop, J.-H. J. Chem. Soc., Chem. Commun. 1970, 781.
28. Barnett, G. H.; Evans, B.; Smith, K. M. Tetrahedron 1975, 31, 2711.
29. Smith, M. J.; Clegg, W.; Nguyen, K. A.; Rogers, J. E.; Pachter, R.; Fleitz, P. A.;
Anderson, H. L. Chem. Commun. 2005, 2433.
30. Balch, A. L.; Mazzanti, M.; Olmstead, M. M. Inorg. Chem. 1993, 32, 4737.
31. Blake, I. M.; Rees, L. H.; Claridge, T. D. W.; Anderson, H. L. Angew. Chem., Int. Ed.
2000, 39, 1818.
32. McCallien, D. W. J.; Sanders, J. K. M. J. Am. Chem. Soc. 1995, 117, 6611.
33. Fuhrhop, J.-H.; Baumgartner, E.; Bauer, H. J. Am. Chem. Soc. 1981, 103, 5854.
34. Stillman, M. J. In Handbook of Porphyrin Science; Kadish, K. M., Smith, K. M.,
Guilard, R., Eds.; World Scientific: Singapore, 2010; Vol. 14, pp 461e524.
35. Lee, C.-H.; Roznyatovskiy, V.; Hong, S.-J.; Sessler, J. L. In Handbook of Porphyrin
Science; Kadish, K. M., Smith, K. M., Guilard, R., Eds.; World Scientific: Singa-
pore, 2010; Vol. 13, pp 197e252.
36. Schill, F.; Arnold, D.P. Unpublished results.
37. Basic, B.; Locos, O. B.; Rintoul, L.; Kan, J.; Jiang, J.; Arnold, D. P. In preparation.
Supplementary data
ꢀ ꢁ
38. Manka, J. S.; Lawrence, D. S. Tetrahedron Lett. 1989, 30, 6989.
39. DiMagno, S. G.; Lin, V. S. Y.; Therien, M. J. J. Org. Chem. 1993, 58, 5983.
40. Senge, M. O.; Feng, X. J. Chem. Soc., Perkin Trans. 1 2000, 3615.
These data include partial 1H NMR spectra for all previously
unreported compounds, calculated orbital surfaces for compounds

532
L.J. Esdaile et al. / Tetrahedron 70 (2014) 517e532
41. Hartnell, R.D.;Edwards, A. J.;Arnold, D.R.J.PorphyrinsPhthalocyanines2002, 6, 695.
42. (a) Bruker. SMART, SAINT and XPREP; Bruker, Analytical X-ray Instruments:
Madison, Wisconsin, USA, 1995; (b) Farrugia, L. J. J. Appl. Crystallogr. 1999, 32,
837; (c) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G. L.; Giocavazzo, C.;
Guagliardi, A.; Moliterni, A. G. C.; Polidori, G.; Spagna, S. J. Appl. Crystallogr.
1999, 32, 115; (d) Sheldrick, G. M. Acta Crystallogr. 2008, D64, 112.
E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.;
Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J.
M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.;
Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.;
Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.;
€
Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, O.; Foresman,
43. Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.;
Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.;
Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.;
Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers,
J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision A.1; Gaussian:
Wallingford CT, 2009.
44. (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648; (b) Lee, C.; Yang, W.; Parr, R. G.
Phys. Rev. B 1988, 37, 785; (c) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980,
58, 1200; (d) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys.
Chem. 1994, 98, 11623.