Synthesis and functionalization of meso-aryl-substituted corroles

Article abstract of DOI:10.1021/jo005661t

The Rothemund condensation reaction of pyrrole and aldehydes is an extensively used route to meso-tetraarylporphyrins, but simple modifications of the reaction conditions allow the formation of different macrocycles other than the expected porphyrin. In the presence of an excess of pyrrole, this modified Rothemund approach leads to the synthesis of meso-triaryl-substituted corroles. This methodology allows the preparation of a wide range of substituted corroles starting from commercially available products. Higher yields have been obtained in the case of benzaldehydes bearing electron-withdrawing substituents, while the reaction fails in the presence of 2,6-disubstituted benzaldehydes. Although if not isolated, some experimental evidences indicate that the linear 5,10,15-triphenylbilane 4 is the precursor of the final corrole ring. Reaction of 5,10,15-triphenylcorrole 2 with an excess of NBS leads to the complete bromination of the macrocycle. Spectroscopic characterization seems to indicate the formation of the porphodimethene-like structure 5, where the macrocyclic aromatic conjugation is interrupted at the 10 position. Metalation of this compound with cobalt acetate and PPh₃ affords the corresponding complex. The X-ray crystal structure of triphenylphosphine [2,3,7,8,12,13,17,18-octabromo-5,10,15-tris(4-nitropheriyl)corrolato]-cobalt( III) 8 confirms the ability of corrole ring to retain an almost planar conformation when fully substituted at the peripheral positions.

Full text of DOI:10.1021/jo005661t

550
J . Org. Chem. 2001, 66, 550-556
Syn th esis a n d F u n ction a liza tion of m eso-Ar yl-Su bstitu ted Cor r oles
Roberto Paolesse,*,† Sara Nardis,† Francesco Sagone,† and Richard G. Khoury‡
Dipartimento di Scienze e Tecnologie Chimiche and INFM, Universita` di Roma Tor Vergata,
00133 Roma, Italy, and Lab. Chimie Supramoleculaire ISIS, Universite Louis Pasteur,
67000 Strasbourg, France
roberto.paolesse@uniroma2.it
Received September 26, 2000
The Rothemund condensation reaction of pyrrole and aldehydes is an extensively used route to
meso-tetraarylporphyrins, but simple modifications of the reaction conditions allow the formation
of different macrocycles other than the expected porphyrin. In the presence of an excess of pyrrole,
this modified Rothemund approach leads to the synthesis of meso-triaryl-substituted corroles. This
methodology allows the preparation of a wide range of substituted corroles starting from
commercially available products. Higher yields have been obtained in the case of benzaldehydes
bearing electron-withdrawing substituents, while the reaction fails in the presence of 2,6-
disubstituted benzaldehydes. Although if not isolated, some experimental evidences indicate that
the linear 5,10,15-triphenylbilane 4 is the precursor of the final corrole ring. Reaction of 5,10,15-
triphenylcorrole 2 with an excess of NBS leads to the complete bromination of the macrocycle.
Spectroscopic characterization seems to indicate the formation of the porphodimethene-like structure
5, where the macrocyclic aromatic conjugation is interrupted at the 10 position. Metalation of this
compound with cobalt acetate and PPh3 affords the corresponding complex. The X-ray crystal
structure of triphenylphosphine [2,3,7,8,12,13,17,18-octabromo-5,10,15-tris(4-nitrophenyl)corrolato]-
cobalt(III) 8 confirms the ability of corrole ring to retain an almost planar conformation when fully
substituted at the peripheral positions.
In tr od u ction
compounds has greatly increased because the structural
modifications can induce new and interesting properties
to the resulting macrocycles,5 the synthetic pathways
developed for these porphyrin analogues are generally
related to â-alkyl substitution patterns.
This limitation has been true also for corrole, a
porphyrin analogue that has recently undergone a sig-
nificant renaissance in interest.6 Corroles have demon-
strated the ability to coordinate a wide range of metals
and exhibits a coordinative behavior that is quite peculiar
and different from that of porphyrins.7
Synthetic porphyrins have been a useful tool to fully
understand the complex reactions performed by these
macrocycles in biological systems.1 Furthermore, the
attempt to mimic the peculiar chemistry operated by
natural porphyrins opened the way to their possible
exploitation in a wide range of practical applications.2
While â-octaalkylporphyrins represent the more direct
counterpart to natural porphyrins, meso-tetraarylpor-
phyrins, with meso-tetraphenylporphyrin (H2TPP) as the
prototypical example, have been the most often used
macrocycles in these studies. The popularity of H2TPP
is probably due to the facile preparation from com-
mercially available products and to the possibility to
introduce a wide range of functional groups at the
â-pyrrolic or at the meso-aryl positions.3 For this reason,
H2TPP has been the ideal platform used to build more
complex systems, providing a fruitful and significant
contribution to porphyrin studies.
We have been interested in the chemistry of this
compound for a certain time and have studied the
possibility to develop a synthetic route to meso-aryl-
substituted corroles.8 In 1994, we reported the first
example of a meso-aryl-substituted corrole from a one-
pot reaction,9 however this reaction was of limited
synthetic interest and in fact it afforded the fully
substituted corrole as triphenylphosphino-Co(III) complex
1. Both Cobalt and â-substituted pyrroles were necessary
to drive the tetramerization of the starting pyrrole to the
However, a similar possibility is rarely found with
regards to porphyrin analogues, macrocycles having some
modification in the molecular skeleton with respect to
that of the parent porphyrin.4 While the interest on these
(4) The Porphyrin Handbook; Kadish, K. M., Smith, K. M., Guilard,
R., Eds.; Academic Press: New York, 2000; Vol. 2.
(5) Sessler J . L.; Weghorn, S. J . Expanded, Contracted & Isomeric
Porphyrins; Tetrahedron Org. Chem. Series; Pergamon: Oxford, 1997;
Vol. 15.
(6) Paolesse, R. The Porphyrin Handbook; Kadish, K. M., Smith, K.
M., Guilard, R., Eds.; Academic Press: New York, 2000; Vol. 2, p 201.
(7) Erben, C.; Will, S.; Kadish, K. M. The Porphyrin Handbook;
Kadish, K. M., Smith, K. M., Guilard, R., Eds.; Academic Press: New
York, 2000; Vol. 2, p 235.
(8) Paolesse, R.; Licoccia, S.; Fanciullo, M.; Morgante, E.; Boschi,
T. Inorg. Chim. Acta 1993, 203, 107.
(9) Paolesse, R.; Licoccia, S.; Bandoli, G.; Dolmella, A.; Boschi, T.
Inorg. Chem. 1994, 33, 1171.
* To whom correspondence should be addressed. Phone:
39.6.72594752. Fax: 39.6.72594328.
† Universita` di Roma Tor Vergata.
‡ Universite Louis Pasteur.
(1) The Porphyrins; Dolphin, D. Ed.; Academic Press: New York,
1978; Vol. VII, Parts A and B.
(2) The Porphyrin Handbook; Kadish, K. M., Smith, K. M., Guilard,
R., Eds.; Academic Press: New York, 2000; Vol. 6.
(3) J aquinod, L. The Porphyrin Handbook; Kadish, K. M., Smith,
K. M., Guilard, R. Eds.; Academic Press: New York, 2000; Vol. 1, p
201.
10.1021/jo005661t CCC: $20.00 © 2001 American Chemical Society
Published on Web 12/28/2000

Synthesis of meso-Aryl-Substituted Corroles
J . Org. Chem., Vol. 66, No. 2, 2001 551
Ta ble 1. Com p a r ison of Rea ction Yield s of Differ en t
corrole. In the other cases, the expected porphyrin was
the macrocycle obtained.
m eso-Tr ia r ylcor r oles
aryl aldehyde
benzaldehyde
5,10,15-TAC yield (%)
6
4-NO2 benzaldehyde
3-NO2 benzaldehyde
2-NO2 benzaldehyde
4-Br benzaldehyde
3-Br benzaldehyde
2-Cl benzaldehyde
22
15
8
5
9
9
4-CH3 benzaldehyde
4-OCH3 benzaldehyde
2,5-(OCH3)2 benzaldehyde
F5-benzaldehyde
6
7
5
4
4-pyridinecarboxaldehyde
2-furalaldehyde
mesitaldehyde
2,6-(OCH3)2 benzaldehyde
2,6-Cl2 benzaldehyde
9
trace
no
no
no
We did not attempt to direct the reaction between
pyrrole and benzaldehyde to obtain 2 because the com-
petitive formation of H2TPP seemed to be an overwhelm-
ing obstacle to the formation of 2. The seminal papers of
and aromatic aldehydes has never been reported, taking
into account the wide application of this reaction for the
synthesis of meso-tetraarylporphyrins.3 However, it should
be noted that the formation of corrole is significantly
influenced by the molar ratio between pyrrole and
aldehydes and only in the presence of excess pyrrole is
it possible to obtain corrole in appreciable yields.12 In fact,
we obtained the highest yields of corrole using a 3:1
pyrrole/benzaldehyde molar ratio in refluxing acetic acid.
In these conditions the expected H2TPP was obtained in
comparable yields, while other chromophores, such as
N-confused porphyrin and sapphyrin, were observed by
TLC in very low amounts and were not separated.
When the reaction was carried out in the presence of
lower pyrrole/benzaldehyde molar ratios the formation
of corrole significantly decreased and H2TPP was the
major product of the reaction. A higher excess of pyrrole
led to a drastic drop in the formation of macrocyclic
products and instead the formation of polypyrroles which
made purification of the reaction mixtures very difficult.
We obtained highest yields of 2 with pyrrole concentra-
tion in the 0.25 M range and consequently benzaldehyde
0.08 M. Both higher and lower concentrations (up to 10-4
M) of reagents induced significant decreases of the
reaction yields and from this point of view these results
are quite similar to those observed in the case of the
synthesis of H2TPPs,14 indicating that a similar mecha-
nism is probably present in these reactions, as discussed
below.
Furuta10 and Latos-Grazynsky11 groups, about the for-
mation of other macrocycles such as N-confused porphy-
rins and sapphyrins from the direct condensation of
pyrrole and aromatic aldehydes, aroused the suspicion
that also corrole can be obtained by this way. This
hypothesis was correct and in 1999 we reported the
synthesis of 2 from pyrrole and benzaldehyde by a
modified Rothemund reaction.12 That same year Gross
and co-workers reported the synthesis of 5,10,15-tris-
(pentafluoro)phenylcorrole by reaction of pyrrole and
pentafluorobenzaldehyde in the absence of solvent.13
These approaches allow for a novel approach to the
preparation of meso-aryl-substituted corroles and can also
be of tremendous importance for the complete develop-
ment of the chemistry of corrole, opening the way for its
potential applications in different fields.
A critical step of this synthetic route is the chromato-
graphic separation of 2 from H2TPP. This separation can
be performed either on silica gel (CH2Cl2 as eluent) or
on neutral alumina (Brockman grade III, CH2Cl2/hexane
1:1 as eluent), but it is necessary to take particular care
because the relative Rf values are quite similar. This
problem limits also the possibility to carry out the
reaction in larger scale.
Herein we study the generality of the Rothemund
synthesis of corroles by reaction between pyrrole and
aldehydes as well as the functionalization of the resulting
macrocycles via bromination. We report also the first
X-ray characterization of a fully brominated triarylcorrole
as a triphenylphosphinocobalt complex.
An important point to clarify of the Rothemund syn-
thesis of corroles should be its versatility to the prepara-
tion of various tri-arylcorroles. For this reason we studied
the reaction of pyrrole with different aromatic aldehydes,
using the reaction conditions giving the highest yields
of 2; the results obtained are reported in Table 1.
It is important to mention that the formation of corrole
is general for monosubstituted aldehydes whatever be the
Resu lts a n d Discu ssion
It is quite surprising that until now the formation of
corrole from the Rothemund reaction between pyrrole
(10) Furuta, H.; Asano, T.; Ogawa, T. J . Am. Chem. Soc. 1994, 116,
767.
(11) Chmielewski, P. J .; Latos-Grazynski, L.; Rachlewicz K.; Glow-
iak, T. Angew. Chem., Int. Ed. Engl. 1994, 33, 779.
(12) Paolesse, R.; J aquinod, L.; Nurco, D. J .; Mini, S.; Sagone, F.;
Boschi, T.; Smith, K. M. Chem. Commun. 1999, 1307.
(13) Gross, Z.; Galili, N.; Saltsman, I. Angew. Chem., Int. Ed. Engl.
1999, 38, 1427.
(14) Adler, A. D.; Longo, F. R.; Finarelli, J . D.; Goldmacher, J .;
Assour, J .; Korsakoff, L. J . Org. Chem. 1968, 5, 669.

552 J . Org. Chem., Vol. 66, No. 2, 2001
Paolesse et al.
Sch em e 1
substituent position on the aromatic aldehyde. This result
is very important, demonstrating that the one pot reac-
tion between pyrrole and aldehydes can allow the prepa-
ration of a wide range of substituted tri-arylcorroles.
Higher yields have been obtained in the case of
aldehydes bearing electron-withdrawing groups. In the
case of 4-NO2 benzaldehyde we obtained a quite unex-
pected 22% yield of 3 without formation of the corre-
sponding porphyrin. This is of particular importance
because it allowed us to obtain the corresponding corrole
in a gram scale along with a facile separation procedure.
In fact the absence of porphyrin among reaction products
avoid the problem of corrole-porphyrin separation, the
most tedious purification step of this synthetic approach.
The higher yields obtained in the case of benzaldehydes
bearing electron-withdrawing groups are not an obvious
result. It is not possible to discuss in detail this point
because we are in the presence of a multistep reaction
and we have no experimental indications about the
reaction mechanism. However, it is quite reasonable to
hypothesise the linear tetrapyrrole 4 as the precursor of
the final corrole ring (Scheme 1). This hypothesis is
supported by the presence of 4 as one of the reaction
products of the acidic condensation of pyrrole and benz-
aldehyde.15
The linear tetrapyrrole can lead to the porphyrinogen
and then to the porphyrin, by attack of an aldehyde unit,
or to the corrole ring by a cyclization due to a pyrrole-
pyrrole link formation.
Our group has previously observed a similar pathway
in the case of the acidic cyclization of a,c-biladienes,16 and
4 has recently been converted to 2 by oxidative cycliza-
tion.17 Furthermore when the reaction was carried out
in the presence of an oxidant, such as chloranil, we
observed a significant reduction of the reaction time with
no substantial differences in the corrole yield; this result
again supports the presence of 4 as reaction intermedi-
ate.18
In the case of electron-poor aldehydes, their higher
reactivity induces their facile reaction with pyrrole and
consequently their rapid disappearance from the reaction
mixture. In this case, the linear tetrapyrrole cannot react
with residual aldehyde units to form the porphyrinogen,
explaining the absence of porphyrin as reaction product
and the higher yields of corrole.
Heteroaromatic aldehydes have also been tested. For
example, 5,10,15-tris(4-pyridyl)corrole has been obtained
in 9% yield, while in the case of furfural aldehyde only
traces of corrole have been obtained.
The spectral characteristics of these corroles are in
good agreement with those observed for the parent
macrocycle 2.12 Their electronic spectra show a Soret-like
band around 415-420 nm and broad Q-bands in the
515-650 nm region; due to the lower symmetry of the
1
corrole ring in the H NMR spectrum there is the spread
of the â-pyrrolic protons resonances. In the case of
5,10,15-tris(4-methoxyphenyl)corrole and 5,10,15-tris(2,5-
dimethoxyphenyl)corrole we experienced a significant
1
line broadening, that prevented to obtain a resolved H
NMR spectrum. This feature is not unusual in the case
of corrole free-bases19 and can be avoided by obtaining
diamagnetic metal complexes of these macrocycles.
On the other hand, the reaction failed in the case of
2,6-disubstituted benzaldehydes, such as 2,6-dichloro or
2,6-dimethoxybenzaldehyde. This result parallels the
(15) Ka, J .-W.; Lee, C.-H. Tetrahedron Lett. 2000, 41, 4609.
(16) Licoccia, S.; Di Vona, M. L.; Paolesse, R. J . Org. Chem. 1998,
63, 3, 3190.
(17) Ka, J .-W.; Cho, W.-S.; Lee, C.-H. Tetrahedron Lett. 2000, 41,
8121.
(18) We gratefully acknowledge a referee for this helpful suggestion.
(19) Broadhurst, M. J .; Grigg, R.; J ohnson, A. W. J . Chem. Soc.,
Perkin Trans. 1 1972, 1124.

Synthesis of meso-Aryl-Substituted Corroles
J . Org. Chem., Vol. 66, No. 2, 2001 553
showed no resonances for both the â-pyrrolic protons,
confirming the complete substitution of these positions,
and the internal NH protons; an unexpected downfield
signal around 13 δ was instead present. Furthermore,
in the visible spectrum in CHCl3 (Figure 1), the typical
corrole Soret-like band was substituted by a broad
1
absorption around 440 nm. The signal at 13 δ in H NMR
spectrum disappeared by addition of a drop of NEt3,
whereas the visible spectrum showed the changes re-
ported in Figure 1 upon addition of NEt3 or TFA.
These results seem to indicate that the conjugation
pathway of the corrole macro-ring has been interrupted
and they are reasonably consistent with the formation
of the species 5, where a tautomeric shift of an internal
proton causes the interruption of the macrocyclic aro-
matic pathway, with the formation of a porphodimethene-
like structure.24
F igu r e 1. Visible spectra of 5 (solid line), 5 + TFA (dashed
line), and 5 + NEt3 (pointed line), in CH2Cl2.
lower efficiency of the Adler’s method for the preparation
of 2,6-disubstituted H2TPPs to respect the classical
Rothemund approach.20 For this reason we attempted the
reaction of pyrrole and 2,6-disubstituted aromatic alde-
hyde using collidine as solvent and copper acetate as
templating agent, but the results obtained were discour-
aging. The reaction products were the corresponding
copper porphyrin, in low yields, and only traces of the
desired corroles as supported by the FAB-mass spectrum.
The failure of the reaction in the case of 2,6-disubsti-
tuted benzaldehydes can be ascribed to steric factors that
prevent the formation of the linear tetrapyrrolic inter-
mediate. In the case of 2,5-dimethoxybenzaldehyde, in
fact, the reaction was successful and it was possible to
obtain the corresponding corrole in reasonable yield.
The use of this modified Rothemund approach can open
the way to the synthesis of a wide range of meso-
substituted corroles, which can serve as an ideal platform
to obtaining more complex systems based on these
macrocycles. To explore this possibility we decided to
study the bromination of 2, because in the case of H2TPP
this reaction is one of the most widely used to obtain
modified porphyrins.21
This structure can explicate the molecular peak ob-
tained in the FAB mass spectrum and the signal around
13 δ present in the 1H NMR spectrum, that can be in
fact related to the downfield shifts of the inner NH
protons observed in analogous porphodimethenes.24 The
attempt of the corrole ring to reduce the sterical strain
induced by the NH protons and by the peripheral bromo
substituents with the displacement of one inner proton
can reasonably explicate the formation of 5. TFA induced
the protonation of the macrocycle, with the concomitant
re-aromatization of the distorted macrocycle, as evi-
denced by the appearance in the visible spectrum of a
red shifted Soret-like band at 480 nm (Scheme 2).
The peculiarity of corrole to accommodate a similar
porphodimethene-like structure has been previously
observed in the case of the Vilsmeier formylation of octa-
alkylcorroles, with formation of the 10-dimethylamino-
methene derivative 6.22
Due to the higher reactivity of corroles with respect to
the corresponding porphyrins toward electrophilic re-
agents,22 we reacted at room temperature 2 with an
excess of NBS in CHCl3.
The progress of the reaction was monitored by TLC
until no further evidence of starting material was present.
Upon work up a green product was obtained in good
yields. FAB mass spectrum of this compound showed a
molecular peak corresponding to the fully brominated
corrole, confirming the peculiarity of corrole chemistry.
In Fact, H2TPP is tetrabrominated in these conditions
at the antipodal positions.23 The use of lower amounts of
NBS gave a mixture of partially substituted corroles due
to the lower symmetry of the corrole ring; the separation
of this mixture by chromatography has not be attempted.
Spectroscopic characterization of the reaction product
yielded however unexpected data. The 1H NMR spectrum
When the bromination reaction was carried out on 3,
we obtained the corresponding fully brominated corrole,
which showed similar spectroscopic results.
Unfortunately, we were unable to obtain crystals of 5
suitable for X-ray analysis. To gain further insights on
its structure, we decided to react 5 with cobalt (II) acetate
(20) Rothemund, P. J . Am. Chem. Soc. 1939, 61, 2912.
(21) J aquinod, L.; Khoury, R. G.; Shea, K. M.; Smith, K. M.
Tetrahedron 1999, 55, 13151.
(22) Paolesse, R.; J aquinod, L.; Senge, M. O.; Smith, K. M. J . Org.
Chem. 1997, 62, 6193.
(23) Crossley, M. J .; Burn, P. L.; Chew, S. S.; Cuttance, F. B.;
Newsom, I. A. Chem. Commun. 1991, 1564.
(24) Buchler, J . W.; Puppe, L. Liebigs Ann. Chem. 1974, 1046.

554 J . Org. Chem., Vol. 66, No. 2, 2001
Paolesse et al.
Sch em e 2
1
feature was confirmed by the H NMR spectrum, where
the signals of the axial PPh3 showed chemical shifts
similar to that of the corresponding cobalt corrole com-
plexes.7 In this case, it seems that the corrole is able to
retain a planar conformation as observed in the case of
the fully substituted corrole 1.9
We obtained in a similar way the complex 8, which
showed similar spectroscopic features. In this case X-ray
characterization performed on a single-crystal allowed
its unambiguous structural identification.25 The molec-
ular structure is shown in Figure 2.
F igu r e 2. Molecular structure of 8: (a) top view; (b) side view.
Hydrogen atoms have been omitted for clarity.
and triphenylphosphine to obtain the corresponding
complex. As observed in the case of 6, the coordination
of Co (III) ions induced the re-aromatization of the
macrocycle, with the formation of the corresponding
triphenylphosphino cobalt complex 7.
The cobalt atom exhibits a pentacoordinated geometry
and lies 0.37 Å out of the mean corrole plane. Based on
the 23 core atoms the corrole is fairly planar with a mean
plane deviation of 0.169 Å. This deviation is higher than
that observed in the case of the corresponding Co-
TPCPPh3 complex,12 but it is comparable with that
observed in the case of 1, confirming the ability of cobalt
corroles to retain an almost planar conformation when
fully substituted. This behavior is completely different
from the corresponding porphyrin complexes, where
severe deviations from the planarity has been observed.26
The Co-N bond lengths average is 1.92 Å [Co-N(21)
1.93, Co-N(22) 1.92, Co-N(23) 1.88, Co-N(24) 1.94].
These bond lengths are slightly longer than those meas-
ured in the similar complex CoTPCPPh3, where we
obtained lengths in the range 1.86-1.89 Å.
(25) The authors have deposited atomic coordinates and a full
structure description for 8 with the Cambridge Crystallographic Data
Centre. The structure can be obtained, upon request, from the Director,
Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge,
CB2 1EZ, UK.
(26) Senge, M. O. The Porphyrin Handbook; Kadish, K. M., Smith,
K. M., Guilard, R. Eds.; Academic Press: New York, 2000; Vol. 1, p
239.
Spectral characterization of 7 gave features similar to
that of a corresponding triphenylphosphino cobalt (III)
corrole.9 Quite surprisingly the visible spectrum showed
a Soret band without significant red shift, indicating the
absence in this case of deviations from the planarity. This

Synthesis of meso-Aryl-Substituted Corroles
J . Org. Chem., Vol. 66, No. 2, 2001 555
compound in 5% yield (0.26 g). UV-vis: λmax 418 (ꢀ 140 000),
577 (ꢀ 24 000), 617 (ꢀ 20 400), 650 (ꢀ 17 000) nm. 1H NMR
(CDCl3) δ -2.78 (br s, 3H), 7.9-8.15 (m, 6H), 8.25 (d, 4H, J )
4 Hz), 8.6 (m, 4H), 8.85 (d, 4H, J ) 4 Hz), 9.00 (d, 2H, J ) 4
Hz). LRMS (FAB): m/z 763 (M+). Anal. Calcd for C37H23
Br3N4: C, 58.22; H, 3.04; N, 7.34. Found: C, 58.28; H, 3.11;
N, 7.49.
5,10,15-Tr is(3-br om op h en yl)cor r ole. This corrole was
prepared, as described for 2, starting from 3-bromobenzalde-
hyde (3.73 g, 20.2 mmol) and pyrrole (4.2 mL, 60.5 mmol) in
9% yield (0.46 g,). UV-vis: λmax 419 nm (ꢀ 125 800), 513 nm
(ꢀ 14 000), 575 (ꢀ 16 800), 615 (ꢀ 13 500), 648 (ꢀ 12 300) nm.
1H NMR (CDCl3) δ -2.82 (s, 3H), 7.87 (m, 2H), 8.15 (m, 4H),
8.28 (d, 2H, J ) 8 Hz), 8.35 (d, 2H, J ) 8 Hz), 8.51 (s, 2H),
8.56, 8.62, 8.88, 9.04 (each d, 2H, J ) 4 Hz). LRMS (FAB):
m/z 763 (M+). Anal. Calcd for C37H23 Br3N4: C, 58.22; H, 3.04;
N, 7.34. Found: C, 58.04; H, 2.96; N, 7.52.
In conclusion we have accomplished, in reasonable
yields, the synthesis of different tri-aryl corroles by a
modified Rothemund reaction. The possibility to intro-
duce substituents at the peripheral positions of the
corrole has been demonstrated possible by the bromina-
tion reaction. Also, the behavior showed by corrole is
quite different from that of the corresponding meso-
tetraarylporphyrins, confirming that corrole chemistry
is peculiar and somewhat unpredictable. Further studies
are now in progress in this area and our results will be
reported in due course.
Exp er im en ta l Section
General conditions are as previously reported16 with the
exception that electronic absorption spectra were measured
using a Cary 50 spectrophotometer in CH2Cl2 solutions.
5,10,15-Tr ip h en ylcor r ole (2). Pyrrole (4.2 mL, 60.5 mmol)
and benzaldehyde (2.0 mL, 20 mmol) were dissolved in 250
mL of acetic acid. The solution was refluxed under stirring
and the course of the reaction was monitored by UV-vis
spectroscopy; after 3 h, the reaction mixture was cooled at room
temperature and 250 mL of distilled water were added. The
precipitate (H2TPP almost pure) was filtered off and NaCl was
added to the filtrate. The resulting precipitate was filtered off
and the crude solid was dissolved in CH2Cl2 and chromato-
graphed on silica gel (CH2Cl2 eluent). The first red-brown band
was H2TPP, the second green fraction contained the desired
product. This fraction was collected, the solvent evaporated
under vacuum and the residue was crystallized from CH2Cl2/
hexane to give green-violet microcrystals of 5,10,15-triphenyl-
corrole (0.21 g, 6% yield).UV-vis: λmax 415 (ꢀ 110 000), 567 (ꢀ
16 000), 615 (ꢀ 12 500), 648 (ꢀ 10 500). 1H NMR (CDCl3) δ
-2.91 (s, 3H), 7.72-7.82 (m, 9H), 8.16 (d, 2H, J ) 5 Hz), 8.39
(d, 4H, J ) 7 Hz), 8.53, 8.58, 8.86, 8.93 (each d, 2H, J ) 4 Hz).
LRMS (FAB): m/z 527 (M+). Anal. Calcd for C37H26N4: C,
84.39; H, 4.98; N, 10.64. Found: C, 84.24; H, 4.92; N, 10.49.
5,10,15-Tr is(4-n itr op h en yl)cor r ole (3). This corrole was
prepared, as described for 2, starting from 4-nitrobenzaldehyde
(3.05 g, 20.2 mmol) and pyrrole (4.2 mL, 60.5 mmol), but in
this case the formation of the corresponding porphyrin was
not observed. For this reason the residue was chromato-
graphed through a short column of silica gel eluting with
CH2Cl2. The resulting corrole was crystallized from CH2Cl2/
hexane (1:2) (0.98 g, 22% yield). UV-vis: λmax 425 nm (ꢀ
5,10,15-Tr is(2-ch lor op h en yl)cor r ole. This corrole was
prepared, as described for 2, from chlorobenzaldheyde (2.27
mL, 20.2 mmol) and pyrrole (4.2 mL, 60.5 mmol), (0.38 g, 9%
yield). UV-vis: λmax 418 nm (ꢀ 132 000), 562 (ꢀ 12 700), 604
1
(ꢀ 7600) nm. H NMR (CDCl3) δ -1.55 (s, 3H), 7.70 (m, 7H),
8.15 (m, 4H), 8.38 (d, 2H, J ) 4 Hz), 8.42 (d, 2H, J ) 4 Hz),
8.58 (d, 2H, J ) 4 Hz), 8.78 (m, 1H), 8.95 (d, 2H, J ) 4 Hz).
LRMS (FAB): m/z 630 (M+). Anal. Calcd for C37H23 Cl3N4: C,
70.54; H, 3.68; N, 8.89. Found: C, 70.33; H, 3.76; N, 8.61.
5,10,15-Tr is(4-m eth ylp h en yl)cor r ole. This corrole was
prepared, as described for 2, from p-tolualdehyde (2.38 mL,
20.2 mmol) and pyrrole (4.2 mL, 60.5 mmol), (yield: 0.23 g,
6%). UV-vis: λmax 418 (ꢀ 135 800), 518 (ꢀ 10 000), 566 (ꢀ
1
21 100), 617 (ꢀ 16 300), 648 (ꢀ 12 500) nm. H NMR (CDCl3) δ
-2.74 (br s, 3H), 2.72 (s, 6H), 2.74 (s, 3H), 7.55 (d, 2H, J ) 7.5
Hz), 7.65 (d, 4H, J ) 7.5 Hz), 8.09 (d, 2H, J ) 7.5 Hz), 8.28 (d,
4H, J ) 7.5 Hz), 8.58, 8.62, 8.88, 8.98 (each d, 2H, J ) 4 Hz).
LRMS (FAB): m/z 569 (M+). Anal. Calcd for C40H32N4: C,
84.48; H, 5.67; N, 9.85. Found: C, 84.19; H, 5.52; N, 9.58.
5,10,15-Tr is(4-m eth oxyp h en yl)cor r ole. This corrole was
prepared, as described for 2, from 4-methoxybenzaldheyde
(2.47 mL, 20.2 mmol) and pyrrole (4.2 mL, 60.5 mmol), (0.29
g, 7% yield). UV-vis: λmax 418 (ꢀ 121 000), 576 (ꢀ 29 800), 618
(ꢀ 26 400), 648 (ꢀ 19 800) nm. LRMS (FAB): m/z 617 (M+).
Anal. Calcd for C40H32N4O3: C, 77.90; H, 5.23; N, 9.08.
Found: C, 80.05; H, 5.38; N, 8.89.
5,10,15-Tr is(2,5-d im eth oxyp h en yl)cor r ole. This corrole
was prepared, as described for 2, from 2,5-dimethoxybenz-
aldheyde (3.35 g, 20.2 mmol) and pyrrole (4.2 mL, 60.5 mmol).
In this case the chromatographic separation was carried out
on neutral alumina (Brockmann, Grade III elution with
CH2Cl2), to give 0.24 g of the title compound (5% yield) after
crystallization from CH2Cl2/hexane. UV-vis: λmax 421 (ꢀ
122 500), 515 (ꢀ 22 600), 575 (ꢀ 22 300), 647 (ꢀ 19 300) nm.
LRMS (FAB): m/z 707 (M+). Anal. Calcd for C43H38N4O6: C,
73.07; H, 5.42; N, 7.93. Found: C, 72.88; H, 5.61; N, 7.78.
5,10,15-Tr is(4-p yr id yl)cor r ole. This corrole was prepared,
as described for 2, from 2,5-dimethoxybenzaldheyde (1.93 mL,
20.2 mmol) and pyrrole (4.2 mL, 60.5 mmol). In this case the
chromatographic separation was carried out on neutral alu-
mina (Brockmann, Grade III elution with CH2Cl2), to give 0.32
g of the title compound (9% yield) after crystallization from
CH2Cl2/hexane. UV-vis: λmax 417 (ꢀ 119 600), 576 (ꢀ 23 000),
616 (ꢀ 18 400), 651 (ꢀ 17 300) nm. 1H NMR (CDCl3) δ -1.76
(br s, 3H), 8.16 (d, 2H, J ) 5 Hz), 8.30 (d, 4H, J ) 5 Hz), 8.63
(d, 2H, J ) 5 Hz), 8.68 (d, 2H, J ) 5 Hz), 8.95 (d, 2H, J ) 5
Hz), 9.02 (d, 2H, J ) 5 Hz), 9.06 (d, 4H, J ) 5 Hz), 9.12 (d,
2H, J ) 5 Hz). LRMS (FAB): m/z 530 (M+). Anal. Calcd for
1
64 500), 595 nm (ꢀ 12 600). H NMR (CDCl3) δ -2.82 (s, 3H),
8.40 (t, 3H, J ) 9 Hz), 8.54 (d, 4H, J ) 8 Hz), 8.58 (d, 2H, J )
4 Hz), 8.66 (m, 2 H), 8.71 (d, 4H, J ) 8 Hz), 8.83 (s, 1H), 8.90
(d, 2H, J ) 4 Hz), 9.11 (d, 2H, J ) 4 Hz). LRMS (FAB): m/z
662 (M+). Anal. Calcd for C37H23N7O6: C, 67.17; H, 3.50; N,
14.82. Found: C, 67.34; H, 3.42; N, 14.49.
5,10,15-Tr is(2-n itr op h en yl)cor r ole. This corrole was pre-
pared, as described for 2, from 2-nitrobenzaldehyde (3.05 g,
20.2 mmol) and pyrrole (4.2 mL, 60.5 mmol), (0.36 g, 8% yield).
UV-vis: λmax 421 nm (ꢀ 135 000), 518 nm (ꢀ 10 500), 575 nm
(ꢀ 15 300), 612 nm (ꢀ 11 200), 648 nm (ꢀ 7000). 1H NMR (CDCl3)
δ -2.51 (br s, 3H), 7.90 (m, 2H), 7.96 (m, 4H), 8.35 (m, 6H),
8.63 (m, 4H), 8.95 (m, 4H). LRMS (FAB): m/z 622 (M+). Anal.
Calcd for C37H23N7O6: C, 67.17; H, 3.50; N, 14.82. Found: C,
67.02; H, 3.82; N, 14.99.
5,10,15-Tr is(3-n itr op h en yl)cor r ole. This corrole was pre-
pared as described for 2, from 3-nitrobenzaldehyde (3.04 g, 20.2
mmol) and pyrrole (4.2 mL, 60.5 mmol), (0.67 g, 15% yield).
UV-vis: λmax 421 nm (ꢀ 137 800), 515 nm (ꢀ 9500), 578 nm (ꢀ
22 200), 615 nm (ꢀ 13 800), 645 nm (ꢀ 9200). 1H NMR (CDCl3)
δ -2.80 (s, 3H), 8.05 (m, 3H), 8.55 (m, 9H), 8.85 (d, 2H, J ) 4
Hz), 9.08 (d, 4H, J ) 4 Hz), 9.25 (d, 2H, J ) 4 Hz). LRMS
(FAB): m/z 622 (M+). Anal. Calcd for C37H23N7O6: C, 67.17;
H, 3.50; N, 14.82. Found: C, 67.21; H, 3.42; N, 14.54.
5,10,15-Tr is(4-br om op h en yl)cor r ole. This corrole was
prepared as described for 2, starting from 4-bromobenzalde-
hyde (3.73 g, 20.2 mmol) and pyrrole (4.2 mL, 60.5 mmol).
Crystallization from CH2Cl2/ hexane(1:2) afforded the title
C
34H23N7: C, 77.11; H, 4.38; N, 18.51. Found: C, 76.92; H,
4.52; N, 18.19.
5,10,15-Tr is(p en ta fu or op h en yl)cor r ole. This corrole was
obtained following the procedure described above for 2, starting
from pyrrole (4.2 mL, 60.5 mmol) and pentafluorobenzaldehyde
(3.96 g, 20.2 mmol) in 4% yield. Spectral properties were in
accord with data previously reported13 and with an authentic
sample prepared according to the literature.13
2,3,7,8,12,13,17,18-Oct a b r om o-5,10,15-t r ip h e n ylcor -
r ole (5). 2 (35 mg, 0.07 mmol) was dissolved in 50 mL of

556 J . Org. Chem., Vol. 66, No. 2, 2001
Paolesse et al.
anhydrous CHCl3 and NBS (189 mg, 1.06 mmol) was added
to the solution. The mixture was stirred at room temperature,
under nitrogen, for 12 h; it was then treated with a saturated
aqueous solution of NaHCO3, dried on anhydrous Na2SO4 and
the solvent was removed under vacuum. The residue was
chromatographed on neutral alumina (Brockmann, Grade III
elution with CH2Cl2/hexane 1:1). The first green fraction was
collected, the solvent evaporated under vacuum and the
resulting residue was crystallized from CH2Cl2/hexane (1:2)
to give the title compound (47 mg, 58% yield). UV-vis: λmax
438 (ꢀ 32 700), 665 (ꢀ 6700) nm.1H NMR (CDCl3) 6.98 (d, 2H,
J ) 8 Hz), 7.38-7.52 (m, 14H), 13.12 (s, 2H). LRMS (FAB):
m/z 1158 (M+). Anal. Calcd for C37H18 Br8N4: C, 38.38; H, 1.57;
N, 4.84. Found: C, 38.24; H, 1.78; N, 4.52.
2,3,7,8,12,13,17,18-Octa br om o-5,10,15-tr is(4-n itr op h en -
yl)cor r ole. 3 (122.3 mg, 0.19 mmol), was dissolved in anhy-
drous CHCl3 (100 mL) and treated with NBS (527 mg, 2.96
mmol); the solution was refluxed for 2 h under nitrogen. The
mixture was washed with a saturated solution of NaHCO3,
dried over anhydrous Na2SO4 and the solvent was evaporated
under vacuum. The resulting residue was dissolved in CH2Cl2
and chromatographed on silica gel. The green fraction was
Tr ip h en ylp h osp h in e [2,3,7,8,12,13,17,18-Oct a b r om o-
5,10,15-tr is(4-n itr op h en yl)cor r ola to]coba lt(III) (8). This
compound was obtained, as described for 7, from 2,3,7,8,12,
13,17,18-octabromo-5,10,15-tris(4-nitro)phenylcorrole (50 mg,
0.04 mmol), (55 mg, 78% yield). UV-vis: λmax 386 (ꢀ 32 000),
1
566 (ꢀ 13 900) nm. H NMR (CDCl3) δ 8.46 (d, 4H, J ) 8 Hz),
8.41 (d, 2H, J ) 8 Hz), 7.70 (br d, 2H), 7.60 (m, 4H), 7.16 (m,
3H), 6.95 (m, 6H), 4.76 (m, 6H). LRMS (FAB): m/z 1349 (M+
- PPh3). Anal. Calcd for C55H27Br8CoN7O6P: C, 41.01; H, 1.69;
N, 6.09. Found: C, 41.08; H, 1.75; N, 6.02.
Cr ysta llogr a p h ic Da ta . Single crystals of 8, [C55H27Br8-
CoN7O6P ‚ 10(H2O)], were grown from CH2Cl2/hexane. Crystals
were placed in oil and a single purple crystal of dimensions
0.15 × 0.12 × 0.12 mm was selected, mounted on a glass fiber,
and placed in low-temperature N2 stream. The unit cell was
orthorhombic with a space group of Pna2(1). Cell dimensions:
a ) 21.870(4) Å, b ) 15.120(3) Å, c ) 31.050(6) Å, R ) â ) γ
) 90°, V ) 10267(4) Å3 and Z ) 8 (FW ) 1770.19, F ) 2.29
g‚cm-3). Reflections were collected from 1.85° e θ e 25.88°
for a total of 12 675 of which 5522 were unique (Rint ) 0.089)
having I > 4σ(I); number of parameters ) 1489. Final R factors
were R1 ) 0.09 (based on observed data, wR2 ) 0.213 (based
on all data), GOF ) 1.39, maximal residual electron density
) 0.944 eÅ-3. X-ray diffraction data for 8 was collected on a
Nonius KappaCCD diffractometer with a graphite monochro-
matized Mo KR radiation (λ ) 0.710 71 Å). æ scans, at 173 K.
The structure solution of 8 was determined using direct
methods and refined (based on F2 using all independent data)
by full matrix least-squares methods (SHELXTL 97). Hydro-
gen atoms were included at calculated positions by using a
riding model.
collected, evaporated to dryness and crystallized from
a
CH2Cl2/hexane (1:2) to give the title product (44 mg, 18% yield).
UV-vis: λmax 422 (ꢀ 33 900), 633 (ꢀ 10 100) nm. 1H NMR
(CDCl3) δ 5.95-6.12 (br m, 3H), 7.56-8.35 (m, 10H) 13.25 (s,
2H). LRMS (FAB): m/z 1293 (M+). Anal. Calcd for C37H15
Br8N7O6: C, 34.38; H, 1.17; N, 7.58. Found: C, 34.24; H, 1.04;
N, 7.49.
Tr ip h en ylp h osp h in e [2,3,7,8,12,13,17,18-Oct a b r om o-
5,10,15-tr ip h en ylcor r ola to]coba lt(III) (7). 5 (50 mg, 0.043
mmol) was dissolved in 50 mL of MeOH and treated with an
excess of cobalt(II) acetate (50 mg, 0.20 mmol) and triphenyl-
phosphine (50 mg, 0.19 mmol). The mixture was refluxed
under stirring and the course of the reaction was monitored
by UV-visible spectroscopy; after 1 h the solvent was evapo-
rated under vacuum, and the resulting residue was chromato-
graphed on silica gel (CH2Cl2 eluent).
Ack n ow led gm en t . This work was supported
by MURST and CNR (MADESS II project n. 97.
01380.PF48).
The first red fraction was collected and crystallized from
CH2Cl2/hexane (1:2) (53 mg, 83% yield). UV-vis: λmax 390 (ꢀ
Su p p or tin g In for m a tion Ava ila ble: X-ray crystallo-
graphic data of 8 and 1H NMR spectra of 2, 3, 5, and 8. This
material is available free of charge via the Internet at
http://pubs.acs.org.
1
29 000), 568 (ꢀ 12 000) nm. H NMR (CDCl3) δ 7.72 (m, 5H),
7.56 (m, 10H), 7.08 (m, 3 H), 6.72 (m, 6H), 4.81 (m, 6H). LRMS
(FAB): m/z 1214 (M+ - PPh3). Anal. Calcd for C55H30Br8-
CoN4P: C, 44.76; H, 2.05; N, 3.80. Found: C, 44.74; H, 1.99;
N, 3.73.
J O005661T