A survey of acid catalysis and oxidation conditions in the two-step, one-flask synthesis of meso-substituted corroles via dipyrromethanedicarbinols and pyrrole

Article abstract of DOI:10.1021/jo0496493

The reaction of dipyrromethanedicarbinols with pyrrole leading to meso-substituted corroles was investigated to determine whether mild acid catalysts [Dy(OTf)₃, Yb(OTf)₃, Sc(OTf)₃, and InCl₃] known to provide po

Full text of DOI:10.1021/jo0496493

A Su r vey of Acid Ca ta lysis a n d Oxid a tion Con d ition s in th e
Tw o-Step , On e-F la sk Syn th esis of Meso-Su bstitu ted Cor r oles via
Dip yr r om eth a n ed ica r bin ols a n d P yr r ole
G. Richard Geier, III,* J effrey Forris Beecham Chick, J ennifer B. Callinan,
Christopher G. Reid, and Waldemar P. Auguscinski
Department of Chemistry, Colgate University, 13 Oak Drive, Hamilton, New York 13346
ggeier@mail.colgate.edu
Received March 2, 2004
The reaction of dipyrromethanedicarbinols with pyrrole leading to meso-substituted corroles was
investigated to determine whether mild acid catalysts [Dy(OTf) , Yb(OTf) , Sc(OTf) , and InCl ]
3 3 3 3
known to provide porphyrins from dipyrromethanecarbinol species while suppressing undesired
reversibility (resulting in scrambling) are applicable to reactions affording corrole, and to explore
the requirements of the oxidation step. We examined a model reaction leading to meso-
triphenylcorrole (TPC) to survey the effect of acid catalyst, acid concentration, ratio of pyrrole to
dipyrromethanedicarbinol, oxidant, oxidant quantity, and reaction time on the yield of TPC (by
UV-vis) in reactions performed at room temperature in CH Cl . Key to this survey was a
2 2
modification of the well-known spectrophotometric method for monitoring reactions leading to
porphyrin. The survey revealed that TPC could be prepared via a subset of the mild acid catalysts
3 3
[Dy(OTf) and Yb(OTf) ], and a preparative-scale reaction afforded an isolated yield of TPC of 49%,
devoid of porphyrin. Suppression of reversible processes was further demonstrated by the synthesis
of three corroles bearing different meso substituents in defined locations in isolated yields ranging
from 50% to 80%. The reaction conditions were applicable to a dipyrromethanedicarbinol bearing
electron-withdrawing pentafluorophenyl substituentssprovided that the reaction time of the
condensation step was increased. We identified circumstances under which DDQ can cause severe
interference with the detection and isolation of some corroles, we found that the yield and purity
of the corrole depend on judicious selection of oxidation conditions, and we assessed the sensitivity
toward light of dilute solutions of the corroles prepared in this study.
In tr od u ction
interest in corrole’s unique properties. The core structure
of corrole is distinguished from that of porphyrin by the
absence of one of the four meso-positions, resulting in a
direct bipyrrole linkage. This deviation leads to an overall
contraction of the macrocycle core, and results in a free
base structure containing three “pyrrole-type” hydrogens.
These structural differences endow corrole with unique
properties (e.g., stabilization of metal ions in unusually
In recent years, studies of porphyrinic macrocycles with
expanded, contracted, isomeric, and heteroatom modified
core structures have intensified. New molecules have
1
2
been discovered such as N-confused porphyrin, and well-
known molecules have enjoyed a resurgence in interest.
Corrole is an important example of the latter category.
The relationship of corrole to the corrin macrocycle of
vitamin B12 led to past interest, and recent progress in
synthetic methodology has helped to fuel the current
4
high oxidation states relative to porphyrin ). The inter-
esting core structure of corrole also presents a synthetic
3
challenge. While the syntheses of â-substituted corroles
3
date back to the historic efforts to prepare vitamin B12,
(1) (a) Sessler, J . L.; Seidel, D. Angew. Chem., Int. Ed. 2003, 42,
such syntheses involved inconvenient stepwise proce-
dures. Unlike the case with porphyrin, syntheses of meso-
substituted corroles from readily available starting ma-
terials did not exist prior to the late 1990s.5
5
134-5175. (b) Sessler, J . L.; Gebauer, A.; Vogel, E. In The Porphyrin
Handbook; Kadish, K. M., Smith, K., Guilard, R., Eds.; Academic
Press: San Diego, CA 2000; Vol. 2, pp 1-54. (c) Sessler, J . L.; Gebauer,
A.; Weghorn, S. J . In The Porphyrin Handbook; Kadish, K. M., Smith,
K., Guilard, R., Eds.; Academic Press: San Diego, CA 2000; Vol. 2, pp
,6
5
5-124. (d) Latos-Grazynski, L. In The Porphyrin Handbook; Kadish,
K. M., Smith, K., Guilard, R., Eds.; Academic Press: San Diego, CA
000; Vol. 2, pp 361-416. (e) Sessler, J . L.; Weghorn, S. Expanded,
Contracted, and Isomeric Porphyrins; Elsevier: Oxford, UK, 1997.
2) (a) Chmielewski, P. J .; Latos-Grazynski, L.; Rachlewicz, K.;
2
(4) (a) Vogel, E.; Will, S.; Schulze-Tilling, A.; Neumann, L.; Lex, J .;
Bill, E.; Trautwein, A. X.; Wieghardt, K. Angew. Chem., Int. Ed. Engl.
1994, 33, 731-735. (b) Gross, Z. J . Biol. Inorg. Chem. 2001, 6, 733-
738. (c) Erben, C.; Will, S.; Kadish, K. M. In The Porphyrin Handbook;
Kadish, K. M., Smith, K., Guilard, R., Eds.; Academic Press: San
Diego, CA 2000; Vol. 2, pp 233-300.
(5) (a) Gross, Z.; Galili, N.; Saltsman, I. Angew. Chem., Int. Ed. 1999,
38, 1427-1429. (b) Gross, Z.; Galili, N.; Simkhovich, L.; Saltsman, I.;
Botoshansky, M.; Blaser, D.; Boese, D.; Goldberg, I. Org. Lett. 1999,
1, 599-602.
(
Glowiak, T. Angew. Chem., Int. Ed. Engl. 1994, 33, 779-781. (b)
Furuta, H.; Asano, T.; Ogawa, T. J . Am. Chem. Soc. 1994, 116, 767-
7
68.
(
3) (a) J ohnson, A. W.; Kay, I. T. J . Chem. Soc. 1965, 1620-1629.
b) Paolesse, R. In The Porphyrin Handbook; Kadish, K. M., Smith,
K., Guilard, R., Eds.; Academic Press: San Diego, CA 2000; Vol. 2, pp
(
2
01-232.
1
0.1021/jo0496493 CCC: $27.50 © 2004 American Chemical Society
Published on Web 05/18/2004
J . Org. Chem. 2004, 69, 4159-4169
4159

Geier et al.
SCHEME 1. Dip yr r om eth a n ed ica r bin ol + P yr r ole
Rou te to Meso-Su bstitu ted Cor r ole
The recent progress in the preparation of meso-
substituted corroles has largely mirrored the major
developments in the syntheses of meso-substituted
7
porphyrinssalbeit with a sharply compressed time frame.
5
6
The initial independent reports by Gross and Paolesse
of the syntheses of meso-substituted corroles bearing
three identical meso substituents (A -corroles) from reac-
tion of pyrrole with an arylaldehyde parallel the Rothe-
approaches are similar to the preparation of porphyrins
bearing up to four different substituents in defined
locations.19
3
8
9
Given our involvement in previous studies of reactions
of dipyrromethanecarbinols leading to meso-substituted
mund and Adler syntheses of tetraarylporphyrins. More
recently, Paolesse and co-workers and Gryko and co-
worker11 have reported two-step, one-flask conditions
that are similar to the Lindsey method. Successful
syntheses of A -corroles were closely followed by synthe-
1
0
2
0
porphyrins sespecially the identification of mild acid
1
2
catalysts that suppress undesired reversibility leading
2
1
to scrambling sa dipyrromethanedicarbinol + pyrrole
3
route to meso-substituted corroles was attractive to us
ses of corroles bearing two different substituents in an
alternating pattern (ABA-corroles) from a dipyrromethane
(Scheme 1). In a practical sense, this route could allow
the tuning of macrocycle properties by the incorporation
of a broad range of meso substituents in defined locations,
and it would avoid the formation of porphyrin byproduct
thereby simplifying the purification of the corrole. In a
fundamental sense, we were interested in comparing the
requirements of dipyrromethanedicarbinol reactions lead-
ing to corroles to those of similar reactions leading to
porphyrins. It is that latter point that is a primary focus
of this paper.
precursor.1
1,13-15
This approach is akin to the stepwise
-porphyrins.16 Close on the heels
of those studies were reports from Guilard, Gryko, and
2 2
preparation of trans-A B
co-workers1 and Gryko and co-workers describing the
synthesis of corroles bearing up to three different meso
substituents in defined locations via the reaction of a
dipyrromethanedicarbinol with excess pyrrole, and a
report from Collman and co-worker involving reaction of
a dipyrromethanedicarbinol with 2,2′-bipyrrole.18 Both
7a
17b
Since the initiation of our study, Guilard, Gryko, and
co-workers have published the results of their related
(6) (a) Paolesse, R.; J aquinod, J .; Nurco, D. J .; Mini, S.; Sagone, F.;
17
work. The study reported herein is complementary to
Boschi, T.; Smith, K. M. Chem. Commun. 1999, 1307-1308. (b)
Paolesse, R.; Nardis, S.; Sagone, F.; Khoury, R. G. J . Org. Chem. 2001,
their reports. Our study was aimed at examining reaction
conditions with close analogy to those developed for the
preparation of porphyrins from dipyrromethanecarbinol
species.21 A model reaction of a dipyrromethanedicarbinol
with pyrrole leading to meso-triphenylcorrole (TPC) was
investigated so that our results could be compared to the
6
6, 550-556.
(
7) Gryko, D. T. Eur. J . Org. Chem. 2002, 1735-1743.
(
8) (a) Rothemund, P. J . Am. Chem. Soc. 1936, 58, 625-627. (b)
Rothemund, P. J . Am. Chem. Soc. 1939, 61, 2912-2915.
(
9) Adler, A. D.; Longo, F. R.; Finarelli, J . D.; Goldmacher, J .; Assour,
J .; Korsakoff, L. J . Org. Chem. 1967, 32, 476.
10) Paolesse, R.; Marini, A.; Nardis, S.; Froiio, A.; Mandoj, F.;
(
rich data available from studies of the synthesis of meso-
Nurco, D. J .; Prodi, L.; Montalti, M.; Smith, K. M. J . Porphyrins
Phthalocyanines 2003, 7, 25-36.
tetraphenylporphyrin (TPP).2
0,22
In complementary fash-
(
11) Gryko, D. T.; Koszarna, B. Org. Biomol. Chem. 2003, 1, 350-
ion, Guilard, Gryko, and co-workers primarily investi-
gated reactions in neat pyrrole with TFA catalysis, and
their model reaction utilized a dipyrromethanedicarbinol
3
57.
12) Lindsey, J . S.; Schreiman, I. C.; Hsu, H. C.; Kearney, P. C.;
Marguerettaz, A. M. J . Org. Chem. 1987, 52, 827-836.
(
(
13) (a) Gryko, D. T. Chem. Commun. 2000, 2243-2244. (b) Gryko,
D. T.; J adach, K. J . Org. Chem. 2001, 66, 4267-4275. (c) Gryko, D.
T.; Piechota, K. J . Porphyrins Phthalocyanines 2002, 6, 81-97.
(
(18) Decreau, R. A.; Collman, J . P. Tetrahedron Lett. 2003, 44, 3323-
3327.
(19) Rao, P. D.; Dhanalekshmi, S.; Littler, B. J .; Lindsey, J . S. J .
Org. Chem. 2000, 65, 7323-7344.
(20) Geier, G. R., III; Littler, B. J .; Lindsey, J . S. J . Chem. Soc.,
Perkin Trans. 2 2001, 712-718.
(21) Geier, G. R., III; Callinan, J . B.; Rao, D. P.; Lindsey, J . S. J .
Porphyrins Phthalocyanines 2001, 5, 810-823.
(22) (a) Geier, G. R., III; Lindsey, J . S. J . Chem. Soc., Perkin Trans.
2 2001, 677-686. (b) Geier, G. R., III; Lindsey, J . S. J . Chem. Soc.,
Perkin Trans. 2 2001, 687-700.
14) Brinas, R. P.; Br u¨ ckner, C. Synlett 2001, 442-444.
15) Asokan, C. V.; Smeets, S.; Dehaen, W. Tetrahedron Lett. 2001,
(
4
2, 4483-4485.
(
16) (a) Lee, C.-H.; Lindsey, J . S. Tetrahedron 1994, 50, 11427-
1
6
1440. (b) Littler, B., J .; Ciringh, Z.; Lindsey, J . S. J . Org. Chem. 1999,
4, 2864-2872.
(17) (a) Guilard, R.; Gryko, D. T.; Canard, G.; Barbe, J .-M.; Ko-
szarna, B.; Brandes, S.; Tasior, M. Org. Lett. 2002, 4, 4491-4494. (b)
Gryko, D. T.; Tasior, M.; Koszarna, B. J . Porphyrins Phthalocyanines
2
003, 7, 239-248.
4
160 J . Org. Chem., Vol. 69, No. 12, 2004

Synthesis of Meso-Substituted Corroles
bearing a strongly electron-withdrawing pentafluoro-
phenyl substituent.17a They do report an attempt to
SCHEME 2. Th e Mod el Rea ction Lea d in g to TP C
3 2 2
employ the mild acid Yb(OTf) in CH Cl , but they
obtained a poor yield of corrole (2%) and they did not
report investigating such conditions further.
In this paper we describe studies of the dipyr-
romethanedicarbinol + pyrrole route to meso-substituted
corroles. A primary aim was to determine whether mild
3 3 3 3
conditions [Yb(OTf) , Dy(OTf) , Sc(OTf) , or InCl cataly-
sis in CH Cl ] of proven success in reactions of dipyr-
2
2
romethanecarbinol species leading to porphyrin are
amenable to syntheses of corroles. Toward that end, we
systematically investigated the effects of key reaction
parameters (acid catalyst, acid catalyst concentration,
ratio of pyrrole to dipyrromethanedicarbinol, and reaction
time) on a model reaction leading to TPC formation. The
yield of TPC in analytical-scale reactions was assessed
by a modification of the UV-vis spectrophotometric
method commonly employed in porphyrin syntheses.12
We also examined the effects of important reaction
parameters in the oxidation step (oxidant and quantity
of oxidant) as it has been proposed that macrocycle
formation occurs during this step,6
b,10,11,14,17,23
and the
literature has accumulated a variety of oxidation methods
without systematic comparison. Results from analytical-
scale experiments were confirmed by representative
preparative-scale syntheses. Undesired reversibility (lead-
ing to scrambling or formation of porphyrin) was assessed
by UV-vis and TLC analysis for the presence of TPP.24
The suppression of scrambling was further confirmed by
laser desorption mass spectrometry (LD-MS) and TLC
analysis of three corroles bearing different substituents
in defined locations prepared under one of the best
reaction conditions identified. The applicability of this
condition to the preparation of corroles bearing electron-
withdrawing substituents was investigated via a reac-
tion leading to 5,10,15-tris(pentafluorophenyl)corrole
required for systematic, analytical-scale reactions (Sup-
porting Information).
1
. Deter m in a tion of th e Yield of Cor r ole in An a -
lytica l-Sca le Rea ction s. UV-vis spectrophotometric
analysis was investigated for the rapid determination of
the yield of TPC from an aliquot of a crude reaction
mixture due to the success of UV-vis monitoring of
reactions leading to porphyrin,12 and reports of spectro-
6
,10,18
photometric monitoring of reactions affording corrole
although no details of the methodology as applied to
(
corrole have been published to the best of our knowledge).
Despite the encouragement, we expected that UV-vis
monitoring of TPC would be challenging. The molar
extinction coefficient of the Soret band of TPC is about
(TpFPC). Finally, the sensitivity toward light of the
corroles prepared in this study was qualitatively com-
pared by monitoring changes in UV-vis spectra as a
function of time.
6
one fifth that of TPP, so there is heightened potential
2
5
for interference from other absorbing species. TPC is
known to be both basic and acidic, and the UV-vis
spectra of the various species differ.5 Reaction aliquots
containing residual acid catalyst or excess weak base
used to neutralize the acid could give rise to inconsistent
spectra.
a,6a
Resu lts a n d Discu ssion
In vestiga tion of th e Mod el Rea ction Lea d in g to
m eso-Tr ip h en ylcor r ole (2a ). The studies of the model
reaction of dipyrromethanedicarbinol 1a -OH (prepared
To address these concerns, a series of experiments were
freshly via NaBH
dipyrromethane 1a ) with pyrrole leading to TPC 2a
Scheme 2) involved the development of a spectrophoto-
4
reduction of 1,9-bis-benzoyl-5-phenyl-
6
performed that utilized solutions of authentic TPC doped
with individual and combinations of species that would
be present in crude, oxidized reaction mixtures (acid
catalyst, DDQ, pyrrole, triethylamine, and oligopyrryl-
methenes). DDQ was found to severely interfere with the
detection of TPC by causing a near total loss of the Soret
band (Figure 1b) and strong retention of TPC on silica
TLC. The latter observation is consistent with a report
of TPC retention by a silica pad employed prior to HPLC
analysis of DDQ oxidized reaction mixtures of pyrrole +
(
metric method to quantify TPC in analytical-scale reac-
tions, acid screening experiments, an examination of the
ratio of pyrrole to dipyrromethanedicarbinol, reaction
time course experiments, oxidant screening experiments,
and a preparative-scale synthesis of TPC. Key to these
studies was our observation that 1a could be purified
without chromatographysaffording the gram quantities
(
23) Ka, J .-W.; Cho, W.-S.; Lee, C.-H. Tetrahedron Lett. 2000, 41,
121-8125.
24) TPP is expected to only arise from reaction of a dipyr-
(25) The porphyrin byproduct produced in some reported reactions
leading to corrole would interfere with spectral monitoring of corrole
given the similar λmax of the Soret bands of porphyrin and corrole and
the greater molar extinction coefficient of the porphyrin. In our reaction
route, porphyrin can only arise as the result of undesired reversibility.
Thus, spectral monitoring of corrole generated from reactions devoid
of scrambling will not suffer from interference due to porphyrin.
8
(
romethanedicarbinol with pyrrole under reversible conditions (by
analogy to reactions of 5-phenyldipyrromethane). Geier, G. R., III;
Littler, B. J .; Lindsey, J . S. J . Chem. Soc., Perkin Trans. 2 2001, 701-
7
11.
J . Org. Chem, Vol. 69, No. 12, 2004 4161

Geier et al.
It has been reported that DDQ produces an adduct with
2
7
tetraphenylporphyrin upon 3-5 days of exposure.
A
DDQ/TPP (2:1) adduct was identified, and the spectral
evidence was consistent with the interaction of a lone pair
of a pyrrolenine nitrogen of the TPP with an empty π*
orbital of a -CN group of DDQ.27 Nevertheless, the
rapid and severe interference of DDQ in the detection
of triphenylcorrole is interesting given that DDQ has
been employed in syntheses of a variety of cor-
roles.5
,11,13,14,15,17,18,23,28,29
The presence of other absorbing
species in the sample matrix (e.g., oligopyrrylmethenes)
interfered with detection of TPC. We found that passing
an aliquot of the oxidized reaction mixture through a
Pasteur pipet filled with silica followed by elution with
2 2
Cl
removed interfering species.30 Excess pyrrole
CH
>50-fold relative to TPC) partly disrupted the interac-
(
tion of TPC with DDQ according to analysis by TLC, but
the UV-vis spectrum was still sharply altered. The Soret
band of TPC was broadened and its intensity was
attenuated by the combined presence of pyrrole and
DDQ, and the interference persisted after dithionite wash
(Figure 1d) and filtration through a silica pipet. Although
we were unable to identify a suitable additive that could
counteract the interference brought on by pyrrole + DDQ,
we found that the effect was constant at >6-fold excess
pyrrole relative to TPC, and that the absorbance of the
broadened Soret band (∼420 nm) varied linearly with the
concentration of TPC. Thus, the level of TPC in the
reaction mixtures could be reproducibly quantified de-
spite the altered appearance of the spectra. The optimum
methodology to emerge from these studies is as follows:
oxidation of an aliquot of a reaction mixture with a
solution of DDQ, washing the oxidized mixture with aq
dithionite, passing an aliquot of the washed mixture
through a Pasteur pipet containing silica gel, and record-
ing the UV-vis spectrum of the eluent. This protocol
provided spectra with the appearance of that shown in
Figure 1d.
2
. Acid -Scr een in g Exp er im en ts. Acid-screening ex-
periments were performed to identify suitable reaction
conditions for the condensation step. These experiments
utilized 1a -OH (2.5 mM) in CH Cl at room temperature
2 2
in keeping with conditions previously identified for the
analogous reaction providing porphyrin.21 Four mild acids
[
Dy(OTf)
3 3 3 3
, Yb(OTf) , Sc(OTf) , and InCl ] were investi-
3
1
gated at concentrations of 0.32, 1.0, 3.2, 10, and 32 mM.
F IGURE 1. Comparison of UV-vis spectra of TPC under
conditions that mimic aspects of the spectrophotometric
analysis of crude reaction mixtures. An aliquot (200 µL) of a
TFA catalysis was also investigated as a control for
reversible processes, as previous studies of analogous
solution of TPC in CH
2
Cl
2
(0.625 mM) was treated as follows:
reactions leading to porphyrin detected scrambling under
(
(
(
A) diluted in toluene (300 µL), (B) diluted in a DDQ solution
10 mM) in toluene (300 µL), (C) diluted in a DDQ solution
10 mM) in toluene (300 µL) followed by washing with an aq
20,21
elevated concentration of TFA in CH
2
Cl
2
.
Pyrrole was
employed at concentrations of 125, 250, and 500 mM,
corresponding to a ratio of pyrrole to carbinol unit of 25,
2 2 4
solution of 5% Na S O and 5% NaOH, and (D) addition of
5
0, and 100, respectively. The yield of TPC was deter-
pyrrole (100 equiv relative to TPC), followed by dilution with
a DDQ solution (10 mM) in toluene (300 µL) and washing with
2 2 4
an aq solution of 5% Na S O and 5% NaOH. A portion (50
mined spectrophotometrically at reaction times of 15 min
µL) of each solution was then diluted in CH
2
Cl
2
/EtOH (3:1, 3
(27) Mohajer, D.; Dehghani, H. J . Chem. Soc., Perkin Trans. 2 2000,
1
99-205.
28) Wasbotten, I. H.; Wondimagegn, T.; Ghosh, A. J . Am. Chem.
Soc. 2002, 124, 8104-8116.
mL) and UV-vis spectra were recorded.
(
benzaldehyde.26 DDQ interference was reversible, as
vigorous shaking of the mixture in an aq basic solution
of sodium dithionite led to the reappearance of the Soret
band (Figure 1c), and the typical retention on silica TLC.
(
29) Collman, J . P.; Decreau, R. A. Tetrahedron Lett. 2003, 44, 1207-
210.
(30) Geier, G. R., III; Lindsey, J . S. J . Org. Chem. 1999, 64, 1596-
603.
31) The quantity of acid is reported in concentration units of
molarity for convenience of comparison of reaction conditions; however,
these four acids are poorly soluble in CH Cl
a1n,2d6 the reaction mixtures
were heterogeneous as reported previously.
1
1
(
(26) Geier, G. R., III; Ciringh, Y.; Li, F.; Haynes, D. M.; Lindsey, J .
2
2
2
S. Org. Lett. 2000, 2, 1745-1748.
4
162 J . Org. Chem., Vol. 69, No. 12, 2004

Synthesis of Meso-Substituted Corroles
TABLE 1. Con d en sa tion Con d ition s P r ovid in g Good
Yield s of TP C Devoid of TP P Byp r od u ct for Ea ch
Con cen tr a tion of P yr r ole In vestiga ted u n d er Ca ta lysis
a
by Dy(OTf)3, Yb(OTf)3, or TF A
[
acid],
[pyrrole],
mM
% yield
of TPCc
b
entry
acid
mM
1
2
3
4
5
6
7
8
9
Dy(OTf)3
Dy(OTf)3
Dy(OTf)3
Yb(OTf)3
Yb(OTf)3
Yb(OTf)3
TFA
1.0
1.0
3.2
0.32
0.32
0.32
0.32
0.32
0.32
125
250
500
125
250
500
125
250
500
54
63
65
40
58
60
43
56
66
TFA
TFA
a
The reactions were performed with 1a -OH (2.5 mM) on a 5-10
F IGURE 2. Yield of TPC as a function of the ratio of pyrrole
to dipyrromethanedicarbinol for reactions of 1a -OH (2.5 mM)
with Dy(OTf) (1.0 mM) in CH Cl at room temperature.
3 2 2
Reactions were monitored spectrophotometrically for the yield
of TPC.
mL scale in CH2Cl2 at room temperature. The reactions were
monitored at 15 min and 1 h. The lowest acid concentration for
each concentration of pyrrole that provided a near-maximal yield
of TPC. Higher quantities of Dy(OTf)3 and Yb(OTf)3 commonly
provided yields of corrole similar to those reported here. c The
highest yield (UV-vis) at either of the two time points is reported.
b
(1.0 mM). The yield of TPC was monitored spectropho-
tometrically at 15 min and 1 h. The maximum yield of
TPC (∼65%) was obtained at a ratio of pyrrole to carbinol
units ranging from 50:1 to 244:1 (Figure 2). This result
suggests that optimal formation of the tetrapyrrane
intermediate requires a fairly large excess of pyrrole, and
that our initial selection of conditions for the acid-
screening experiments was appropriate. The decline in
the yield of TPC at the highest quantities of pyrrole could
stem from interference with the acid catalyst by the
weakly basic pyrrole or by the formation of oligomers
containing multiple bipyrrole units.1 Low levels of TPP
(<3%) were detected at the lowest ratios of pyrrole to
carbinol units investigated (e2.5:1). This observation
suggests that even with this mild acid, reversible pro-
cesses can occur if there is insufficient pyrrole to rapidly
react with the free carbinol units. Such reversibility is
not observed in the stoichiometric reaction of a dipyr-
romethanedicarbinol and a dipyrromethane leading to
porphyrins bearing similar substituents.21
and 1 h. The presence of TPP (indicating undesired
reversible processes) was assessed by TLC and by UV-
vis.
The best reaction conditions identified from the acid-
screening experiments are summarized in Table 1 (see
the Supporting Information for plots of the yield of TPC
as a function of acid concentration). The highest yield of
TPC with no detected TPP was provided by Dy(OTf)
3
, Yb-
7b
(
OTf) , and low concentrations of TFA (e1.0 mM). TFA
3
concentrations of >3.2 mM provided no detectable TPC
and instead gave low levels of TPP (1-3%) indicative of
undesired reversible processes. This observation is con-
sistent with previous reports pertaining to dipyrrometh-
anecarbinol reactions leading to porphyrin.20 InCl
and
provided detectable tetraphenylporphyrin (3-
0%) at all acid quantities investigated. The level of
3
Sc(OTf)
3
1
detected TPP declined with an increasing ratio of pyrrole
to carbinol units; nevertheless, it was surprising to obtain
any TPP from these reactions given the mild nature of
the catalysts and the large excess of pyrrole. The origin
of this result is not clear, but the observation suggests
that each acid should be investigated when adapting
reaction conditions devised for porphyrin syntheses to the
synthesis of other macrocycles. The highest spectral
yields of TPC obtained under the best conditions (∼65%)
are in line with the 65% yield of TPC obtained by Lee
and co-workers upon DDQ oxidation of a linear tri-
aryltetrapyrrane.23 The yield of TPC generally increased
upon an increase in ratio of pyrrole to carbinol units from
4. Rea ction Tim e Cou r se Exp er im en ts. Reactions
were performed under three of the best reaction condi-
tions identified from the acid-screening experiments
(Table 1, entries 2, 5, and 9), and the yield of TPC was
monitored spectrophotometrically from 1 min to 24 h
(Figure 3). The maximum yields of TPC (∼65%) were in
good agreement with the earlier results of the screening
experiments. The reaction under TFA catalysis reached
a maximum yield of TPC within 1 min of the addition of
the TFA. The yield of TPC increased more gradually
under catalysis by Dy(OTf) and Yb(OTf) , reaching a
3
3
maximum yield by 30 and 15 min, respectively. In all
cases, the yield of TPC remained fairly constant through-
out the reaction until reaction times >7 h. Trace levels
of TPP (<1%) were generally detected at only the longest
time pointsafter the yield of TPC was observed to
decline. Thus, a trace level of undesired reversibility can
occur at long reaction times, but it can be readily avoided.
In general, the reaction time course of the model reaction
was similar to analogous reactions leading to porphy-
rins.21
2
5:1 to 50:1 (45-55% and 55-65%, respectively) whereas
little additional improvement was observed at the 100:1
ratio. The yield of TPC was fairly insensitive to the
quantity of Dy(OTf)
3 3
and Yb(OTf) . This observation is
consistent with a report pertaining to dipyrrometh-
2
1
anecarbinol reactions leading to porphyrin. Catalysis
by TFA in acetonitrile solvent19 was also investigated,
and found to provide poorer results in terms of the yield
and purity of the isolated TPC (Supporting Information).
3
. Ra tio of P yr r ole to Dip yr r om eth a n ed ica r bin ol.
Analytical-scale reactions employing ratios of pyrrole to
a -OH ranging from 2:1 to 5760:1 were performed in
CH Cl at room temperature under catalysis by Dy(OTf)
5. Oxid a n t Scr een in g Exp er im en ts. In the synthe-
sis of porphyrin it is common to assume that the por-
phyrinogen produced during the condensation step is
rapidly and quantitatively oxidized to porphyrin upon
1
2
2
3
J . Org. Chem, Vol. 69, No. 12, 2004 4163

Geier et al.
Thus, we investigated the impact of oxidant and oxidant
quantity on the yield and ease of isolation of TPC.
i. DDQ Oxid a tion . Aliquots from the condensation of
1
a -OH (2.5 mM) with pyrrole (250 mM) catalyzed by Dy-
(
OTf) (1.0 mM) in CH Cl were treated with 0.5 to 80
3
2
2
equiv of DDQ relative to 1a -OH. The yield of TPC was
determined spectrophotometrically after washing with aq
dithionite, and the complexity of the mixtures was
assessed by TLC. TPC was observed in all cases, and the
optimal quantity (∼60%) was obtained from 3 to 4 equiv
of DDQ. A gradual decline in yield of TPC was observed
with higher equivalents of DDQ (e.g., 52% yield of corrole
with 8 equiv of DDQ compared to 43% yield with 80
equiv). The oxidation conditions employed in our protocol
for spectrophotometric monitoring of the yield of TPC (6
equiv of DDQ) compare favorably to this result.
ii. p-Ch lor a n il Oxid a tion . Aliquots from the conden-
sation of 1a -OH (2.5 mM) with pyrrole (250 mM) cata-
lyzed by Dy(OTf) (1.0 mM) in CH Cl and quenched by
3 2 2
the addition of triethylamine (5 equiv relative to the acid)
were treated with 0.25 to 16 equiv of p-chloranil relative
to 1a -OH. The samples were heated to reflux, and
monitored from 30 min to 2 h by TLC and UV-vis.
Oxidation with 2 equiv of p-chloranil for 45-60 min was
found to provide the best result. [Note that this quantity
of oxidant is lower than the stoichiometric quantity of 3
equiv.] Reactions employing >2 equiv of p-chloranil
provided a brown pigment (R
desired green TPC pigment (R
CH Cl /hexanes (1:1)]. Lee and co-workers reported
f
0.50) in addition to the
f
0.40) [TLC analysis, silica,
2
2
observing a similar pigment that they provisionally
assigned as a bilatriene derivative.23 The intensity of the
brown pigment increased with the quantity of p-chloranil
and with the time of oxidation (e.g., treatment with g4
equiv of p-chloranil for 1.5 h provided almost exclusively
the brown pigment at the expense of TPC). The brown
pigment fell below the limits of detection by TLC in
reactions employing e2 equiv of p-chloranil. However,
the yield of TPC was also impacted by the use of low
quantities of p-chloranil (e.g., the yield of TPC from 0.25,
F IGURE 3. Yield of TPC as a function of condensation time
for reactions of dipyrromethanedicarbinol 1a -OH (2.5 mM)
with excess pyrrole in CH
following conditions: (A) pyrrole (250 mM) and Dy(OTf)
mM), (B) pyrrole (250 mM) and Yb(OTf) (0.32 mM), and (C)
2
Cl
2
at room temperature under the
0
.50, 1.0, and 2.0 equiv of p-chloranil was 2%, 7%, 18%,
3
(1.0
and 50%, respectively). None of the reactions produced
detectable TPP, indicating that the oxidation was rapid
relative to reversible processes.
3
pyrrole (500 mM) and TFA (0.32 mM). Reactions were moni-
tored spectrophotometrically. Note the log scale for time.
iii. Com p a r ison of DDQ a n d p-Ch lor a n il Oxid a -
tion s in Sem ip r ep a r a tive-Sca le Rea ction s. Aliquots
addition of DDQ. As a result, studies of porphyrin
syntheses have largely focused on the condensation step.
In the case of corrole synthesis, the conditions of the
oxidation step might be more exacting. For example, Lee
and co-workers found that the yield of TPC obtained upon
the oxidation of a purified triaryltetrapyrrane was im-
(
50-100 mL) from the condensation of 1a -OH (2.5 mM)
with pyrrole (250 mM) catalyzed by Dy(OTf) (1.0 mM)
in CH Cl were treated with DDQ or p-chloranil to
compare isolated yields of TPC and ease of isolation in
3
2
2
3
3
the absence of a dithionite wash (Table 2). The low
isolated yield of TPC from the DDQ oxidation is consis-
tent with interference by DDQ. The p-chloranil reactions
provided TPC that was readily isolated in yields that
were similar to the spectral yields obtained from the
analytical-scale reactions involving p-chloranil oxidation.
pacted by the choice of solvent, oxidant, and inorganic
additive.23 The ease of corrole isolation might also depend
on the choice of oxidation conditions in light of our
observation that DDQ can interfere with the detection
and TLC retention of TPC. Control experiments involving
p-chloranil showed no interference with TPC.32 However,
it was not clear whether the weaker oxidant would
provide a yield of corrole comparable to DDQ oxidation.23
2 2
TLC analysis [silica, CH Cl /hexanes (1:1)] under over-
loaded conditions revealed a faint tan band consistent
with the previously observed brown pigment under all
conditionssincluding DDQ oxidation and conditions in-
(
32) p-Chloranil lacks the cyano groups that are reportedly impor-
(33) The aq dithionite wash of crude reaction mixtures on a
2
7
tant in the interaction between DDQ and porphyrin, and are
potentially important in the corresponding interaction of DDQ with
corrole.
semipreparative scale is irksome as the mixtures are prone to forma-
tion of an emulsion and the organic/aqueous interface is challenging
to locate.
4
164 J . Org. Chem., Vol. 69, No. 12, 2004

Synthesis of Meso-Substituted Corroles
TABLE 2. Isola ted Yield s of TP C Obta in ed fr om
silica gel to afford TPC contaminated with a low level of
the less polar brown pigment. The TPC was further
purified via a short silica column, eluting with CH Cl /
2 2
Sem ip r ep a r a tive-Sca le Rea ction s Com p a r in g Oxid a tion
a
Con d ition s In volvin g DDQ a n d p-Ch lor a n il
%
of TPC
yield
hexanes (1:1) to obtain 128 mg, 49% yield. The discrep-
ancy between the isolated yield of TPC (∼50%) and the
spectral yield (∼65%) could reflect the differences in the
oxidation conditions, losses of TPC incurred during
isolation, and/or a degree of systematic overestimation
of the yield of TPC by spectral monitoring. The isolated
yield of TPC compared favorably to the ∼50% yield of
TPC obtained from the semipreparative-scale experi-
ment.
equivb
c
entry
oxidant
DDQ
p-chloranil
p-chloranil
p-chloranil
p-chloranil
1
2
3
4
5
3
2d
18
26
54
53
1.5
1.75
1.9
2.0
a
The condensation of 1a -OH (2.5 mM) + pyrrole (250 mM) was
catalyzed by Dy(OTf)3 (1.0 mM) in CH2Cl2 at room temperature.
Aliquots (50-100 mL) of the reaction mixture were treated with
oxidant at room temperature (DDQ) or at reflux (p-chloranil) for
P r ep a r a tion of Cor r oles Bea r in g Differ en t Su b-
stitu en ts in Defin ed Loca tion s. To rigorously confirm
the absence of scrambling under reaction conditions
identified for the model reaction, we investigated the
preparation of corroles bearing two or three different
substituents in defined locations (Table 3). Analytical-
scale reactions leading to 2b and 2d were followed for
yield of corrole (UV-vis) as a function of time under the
three best conditions identified from studies of the model
reaction (Table 1, entries 2, 5, and 9). These studies
provided results very similar to those obtained from the
model reaction (Supporting Information). Preparative-
scale reactions were performed in identical fashion to the
model reaction. In each case, a trace level of a brown
pigment was produced in addition to the corrole. The
exceptionally high yield of 2b (∼80%) was reproducible,
but the origin of the high yield for that particular corrole
is unclear. The corroles were produced devoid of porphy-
rin and without detectable scrambling as evidenced by
TLC and LD-MS analysis. Thus, the reaction conditions
identified from the model reaction are suitable for the
4
5 min. TPC was isolated by filtration through a silica pad
(CH2Cl2) followed by silica gel chromatography [CH2Cl2/hexanes
b
(1:1)]. Equivalents of oxidant relative to 1a -OH originally present
c
in the aliquot. Isolated yield of TPC after drying under vacuum.
d
Retention of TPC in the presence of DDQ was observed during
the initial filtration through a silica pad.
volving lower quantities of p-chloranil. Together, these
observations suggest that ∼2 equiv of p-chloranil for 30
min to 1 h of reflux provides the best balance of yield of
TPC, ease of isolation, and minimization of impurities.
. P r ep a r a tive-Sca le Syn th esis of TP C. Dipyr-
romethanedicarbinol 1a -OH (0.50 mmol, 2.5 mM) and
pyrrole (250 mM) were treated with Dy(OTf) (1.0 mM)
in CH Cl at room temperature (Table 3). The reaction
was monitored spectrophotometrically at 15-min inter-
vals. Consistent with analytical-scale reactions, a spectral
yield of ∼65% was obtained by 45 min. Triethylamine (5
equiv. relative to acid) was added followed by p-chloranil
6
3
2
2
(1.0 mmol, 2 equiv relative to 1a -OH) and the mixture
was heated to reflux for 45 min. The TPC was isolated
from polar byproducts upon passage through a pad of
TABLE 3. P r ep a r a tive Syn th esis of Meso-Su bstitu ted Cor r oles
J . Org. Chem, Vol. 69, No. 12, 2004 4165

Geier et al.
efficient preparation of corroles bearing meso substitu-
ents in defined locations.
withdrawing groups and/or facially encumbering
groups.5
,11,13,15,29
Success in these cases is understandable
P r ep a r a tion of 5,10,15-Tr is(p en ta flu or op h en yl)-
cor r ole. Given general interest in corroles bearing
electron-withdrawing substituents, we examined the
reaction of 1e-OH with pyrrole leading to TpFPC 2e
if the mechanism by which DDQ interacts with corrole
proves to be similar to that of porphyrin.27 An electron-
poor corrole would not be expected to readily donate
electron density to a -CN group of DDQ, and it has been
shown that a facially encumbered porphyrin, tetramesi-
tylporphyrin, does not readily form an adduct with DDQ
for presumably steric reasons.27 Recently, reports have
appeared that describe exacting conditions involving
DDQ oxidation. The authors proposed that excess DDQ
(
Table 3). Authentic TpFPC was prepared by the method
5
of Gross and co-workers, and was used to confirm the
applicability of our spectrophotometric analysis for this
corrole, and to investigate potential interference of the
isolation of TpFPC by DDQ. This latter point was of
interest as DDQ has been widely used in the preparation
of corroles bearing electronegative substituents.5,11,13,15,29
DDQ interference with the isolation of TpFPC was
observed, but to a much lesser extent than was the case
with TPC (e.g., isolation of authentic TpFPC after treat-
ment with DDQ resulted in a 15% loss of TpFPC
compared to a 53% loss of TPC in side-by-side experi-
ments).
1
3c,18
can result in overoxidation of corrole,
formation of a corrole-DDQ complex
or in the
leading to poor
11,13c
isolated yields of corrole. Corroles that lack electron-
withdrawing groups and/or facially encumbering substit-
uents were initially prepared by air oxidation in an
6
“Adler-like” procedure. Reports of the use of p-chloranil
with a diverse range of substituents have appeared more
recently, but without explanation for the choice of oxidant
1
0,14,15
Analytical-scale reactions leading to TpFPC were fol-
lowed for yield of corrole (UV-vis) as a function of time
under the three best conditions identified from studies
of the model reaction (Table 1, entries 2, 5, and 9). The
reactions differed sharply from those leading to the other
corroles in that the condensation reactions were much
slower and the maximum yields of TpFPC were lower.
or without direct comparison to the use of DDQ.
Interestingly, there are a handful of reports that are
at apparent odds with the general trends in the litera-
ture. (1) Lee and co-workers reported the synthesis of
TPC in a yield of up to 65% via DDQ oxidation of a
triaryltetrapyrrane species.23 (2) Br u¨ ckner and co-work-
ers reported the synthesis of TPC in a yield of 40% via
condensation of an aldehyde with excess 5-phenyldipyr-
romethane followed by DDQ oxidation.14 (3) Guilard,
Gryko, and co-workers and Gryko and co-workers re-
ported the preparation of TPC in a yield of 22-25% via
condensation of a dipyrromethanedicarbinol 1a -OH in
neat pyrrole (25 equiv per carbinol unit) followed by
Catalysis by Dy(OTf)
3
provided a 36% yield of TpFPC by
8
h, Yb(OTf) provided a 40% yield by 8 h, and TFA
3
provided a 25% yield by 8 h (Supporting Information).
This result demonstrates the importance of monitoring
the reactionshad the condensation step of these reactions
been performed for 1 h (as was determined appropriate
for the other corroles) we would have erroneously con-
cluded that the reactions had failed.
Preparative-scale experiments were performed under
reaction conditions identical to the other corroles (with
the exception that 0.20 mmol of 1e was used due to
limited quantities). The reaction was allowed to proceed
for 10 h providing a spectral yield of TpFPC of 30% prior
to oxidation with p-chloranil (2 equiv relative to 1e-OH).
After 45 min of reflux, little corrole was detected by TLC
or UV-vis. DDQ (3 equiv) was then added at room
temperature over 30 min, followed by stirring for 30 min
at room temperature. An isolated yield of 24% (39 mg)
was obtained. The reaction was repeated under identical
condensation conditions (the reaction required 8 h to
reach a spectral yield of 34%) followed by oxidation with
2 2
dilution in CH Cl and treatment with DDQ (3 equiv per
1a -OH).17 (4) Ghosh and co-workers reported the syn-
thesis of TPC in a yield of ∼6% from pyrrole and
2
8
benzaldehyde using the Gross method involving DDQ
oxidation.5 We investigated each of these reports to
reconcile them with our observations and with the
general trends found in the literature.
We suspected that the results of Lee and co-workers
could be understood, in part, by their choice of solvent.
They obtained their best results using propionitrile, and
only a trace quantity of TPC was obtained from a reaction
2
3
performed in CH Cl in the absence of other additives.
2
2
We treated propionitrile solutions of authentic TPC with
0-10 equiv of DDQ relative to TPC, and then attempted
to pass aliquots through a silica pipet with elution with
CH Cl . In all cases, a band of TPC was observed to elute
3
equiv of DDQ. After the solution was stirred at room
2
2
temperature for 1 h, the quantity of TpFPC appeared to
be low based on TLC and UV-vis analysis. Additional
DDQ (3 equiv) was added over 30 min, and an isolated
yield of corrole of 18% (28 mg) was obtained. Although
additional work is required to optimize the oxidation step
of this reaction, it is clear from these results that mild
catalysis conditions can promote the condensation reac-
tion leading to corroles that bear electron-withdrawing
substituents.
On th e Ch oice of Oxid a n t. The observation of
interference of detection and isolation of TPC by DDQ
and the poor result of p-chloranil in the synthesis of
TpFPC prompted a reexamination of oxidation conditions
reported in the literature. Consistent with our findings,
DDQ oxidation has been generally applied with best
success in the preparation of corroles bearing electron-
rapidly through the silica pipet, although the intensity
of Soret and Q-bands were diminished in samples of TPC
treated with >3 equiv of DDQ. Thus, it appears that
propionitrile assists with the isolation of TPC in the
presence of modest levels of DDQ. In the work reported
by Br u¨ ckner and co-workers, the presence of excess
5-phenyldipyrromethane attracted our attention. Control
experiments involving solutions of TPC doped with DDQ
and varying quantities of 5-phenyldipyrromethane re-
vealed that the addition of the dipyrromethane allowed
improved recovery of TPC. Interestingly, 6 equiv of
dipyrromethane relative to DDQ were required to com-
pletely nullify the interference of DDQ with the TPCs
close to the quantity of excess of dipyrromethane em-
ployed in Br u¨ ckner’s conditions. Thus, the requirement
for excess dipyrromethane in this particular reaction may
4
166 J . Org. Chem., Vol. 69, No. 12, 2004

Synthesis of Meso-Substituted Corroles
be related to counteracting interference from DDQ in
addition to providing an optimal oligomer composition
for corrole formation. (Note that p-chloranil was used to
prepare the other corroles in their study.) Our investiga-
tion of the procedure of Guilard, Gryko, and co-workers
provided findings that were consistent with interference
of the isolation of TPC by DDQ. The condensation of 1a -
OH with pyrrole mediated by TFA was performed in
accordance with the published procedure1 (Supporting
7a
2 2
Information). After dilution in CH Cl , the mixture was
divided evenly between two flasks and subjected to
oxidation by DDQ (3 equiv) at room temperature or
p-chloranil (2 equiv) at reflux. While spectral monitoring
of the two oxidation reactions provided similar results,
DDQ oxidation afforded an isolated yield of TPC of 11%
F IGURE 4. Absorbance of the Soret band as a function of
time for dilute solutions of TPC 2a and TpFPC 2e in toluene
in the presence of ambient light. The λmax of the Soret band of
TPC shifted from 420 to 401 nm during the experiment. The
(11.2 mg) whereas p-chloranil oxidation provided an
isolated yield of 50% (50.0 mg). The difference between
our isolated yield of TPC under DDQ oxidation (11%) and
the yield reported by Guilard, Gryko, and co-workers
λmax of the Soret band of TpFPC was 414 nm.
(
22-25%) could reflect variability in the extent to which
of some corroles toward light and air has been reported
in the literature,34 but we were unable to find detailed
information on the scope and magnitude of corrole
instability. Thus, we performed a simple side-by-side
comparison of the solution stabilities of the corroles
prepared in this study. Solutions of each corrole were
prepared in toluene and UV-vis spectra were recorded
upon exposure to ambient light. Changes in the absor-
bance of the Soret band for TPC and TpFPC as a function
DDQ interfered with the initial filtration through a silica
pad. Regardless, it is clear that oxidation with p-chloranil
provided a higher isolated yield. Oxidation with p-
chloranil also provided a cleaner reaction mixture and
final product as assessed by TLC analysis. Both oxidants
provided the characteristic green spot of TPC along with
a low level of the brown pigment. Additionally, DDQ
oxidation provided an olive green band that eluted
slightly ahead of the TPCsconsistent with published
of time are summarized in Figure 4. Consistent with
3
observations pertaining to the preparation of A -corroles
literature reports,1
1,17a
the corrole bearing electron-
1
0
from pyrrole + aldehyde, and ABA-corroles via dipyr-
withdrawing substituents was found to be fairly stable.
In contrast, TPC underwent substantial decomposition
over a short period of time (e.g., ∼50% of the intensity of
the Soret band was lost by 2 h of exposure to light). Light
is required for the decomposition as an identical solution
of TPC stored in the dark did not undergo decomposition.
Corroles 2b-d showed rates of decomposition similar to
or even a little faster than that of TPC (Supporting
Information). We have not identified the products of the
decomposition, but others have identified an open chain
tetrapyrrole (biliverdin) structure in the decomposition
of a corrole bearing a combination of â and meso sub-
romethane + aldehyde,1 but not reported in corrole
3b
1
7
syntheses via dipyrromethanedicarbinols. Finally, our
attempts to replicate the report28 of the preparation of
TPC via the Gross method were not successful even upon
attempted dithionite wash of the crude reaction mixture
prior to isolation of the TPC. The origin of our lack of
success is not clear, but this observation is consistent
with the original report of Gross and co-workers.5
On the basis of our observations and reports from the
literature, the choice of oxidant appears to be critical to
the success of the reaction and/or to the ease of corrole
isolation. While we make no claims that the generally
superior yield of corroles bearing electron-withdrawing
and/or facially-encumbering groups in reactions involving
DDQ stem solely from differential interactions between
various corroles and DDQ, we suggest that it may be a
factor in at least some situations. p-Chloranil appears
to be best suited for corroles that tend to interact strongly
with DDQ, and DDQ appears to be best suited for
reactions leading to corrole that require stronger oxidiz-
ing conditions. Corroles lacking electron-withdrawing or
facially-encumbering substituents appear to fall in the
former group, and corroles with strongly electron-
withdrawing substituents appear to fall in the latter
group. For intermediate cases, the best oxidation condi-
tions are not presently clear. A comparison of p-chloranil
34
stituents. As corroles become candidates for a variety
of applications, their stability in solution will need to be
delineated. Key questions include the following: How
many electron-withdrawing groups are necessary to
impart stability? How strong of an electron-withdrawing
group is required? Does the placement of the electron-
withdrawing group (5 or 10 position) matter? Can electron-
donating groups be tolerated if other substituents are
present?
Con clu sion s
3 3
We have found that mild acids [Dy(OTf) or Yb(OTf) ]
or a low concentration of TFA can be applied to the
synthesis of meso-substituted corroles devoid of scram-
(
2 equiv. at reflux for 45 min) and DDQ (3 equiv at room
3 3
bling. The effect of Dy(OTf) or Yb(OTf) concentration
temperature for 1 h) would seem to provide a good
starting point.
St a bilit y of Cor r oles. While determining molar
extinction coefficients for the corroles prepared in this
study, we noted a gradual decrease in the absorbance
of the Soret band over time. The instability of solutions
on the yield of TPC and the reaction time course under
the best reaction conditions was similar to that of
previous studies of analogous reactions leading to por-
(34) Tardieux, C.; Gros, C. P.; Guilard, R. J . Heterocycl. Chem. 1998,
35, 965-970.
J . Org. Chem, Vol. 69, No. 12, 2004 4167

Geier et al.
pyrrole (0.25, 0.50, or 1.0 M) was prepared in CH Cl in a
phyrin. The behavior of InCl
3
and Sc(OTf)
3
deviated from
2
2
second 100-mL volumetric flask. The contents of the two
volumetric flasks were mixed in an amber bottle equipped with
a solvent pump, thereby providing a stock solution of 1a -OH
our expectations in that TPP (indicative of reversible
processes) was observed under all conditions investigated.
A dipyrromethanedicarbinol species bearing electron-
withdrawing pentafluorophenyl substituents treated simi-
larly to the model reaction also underwent reaction
leading to corrolesalbeit more slowly. Together, these
results suggest that sharp deviations from the catalytic
conditions used to prepare porphyrins from dipyr-
romethanedicarbinols are not required to obtain good
yields of corrole devoid of scramblingsalthough not all
of the mild acid catalysts work equally well. The critical
role of the choice and quantity of oxidant was identifieds
including a direct comparison of DDQ and p-chloranil.
An examination of dilute corrole solutions provided a
comparison of their instability toward light and air. As
a whole, this study adds to the growing understanding
of reaction conditions affording corrole, it demonstrates
similarity between dipyrromethanedicarbinol routes to
corroles and to porphyrins, and it provides methodology
complementary to that previously reported for the prepa-
ration of corroles bearing substituents in defined loca-
tions.
(2.5 mM) and pyrrole (125, 250, or 500 mM). Reactions were
performed at room temperature in tightly capped 20-mL vials
that were stirred with a micro stir bar. Solid acids were
weighed into all reaction vials prior to beginning the reaction
sequence for the day, and each reaction was started by the
addition of 5-10 mL of the reactant solution from the solvent
pump. Reactions involving TFA were initiated by the addition
of TFA to reaction vials already containing 5 mL of the
reactant stock solution. The reactions were monitored spec-
trophotometrically for yield of TPC at 15 min and 1 h as
described above. TLC was performed on the crude, oxidized
mixture [silica, CH
2 2
Cl /hexanes (2:1)] after washing with the
aq dithionite solution.
Ra tio of P yr r ole to Dip yr r om eth a n ed ica r bin ol Ex-
p er im en ts. Portions (5 mL) of a freshly prepared solution of
dipyrromethanedicarbinol 1a -OH (2.5 mM) in CH Cl were
2
2
transferred to 20-mL vials. Pyrrole was added by syringe to
afford concentrations of pyrrole ranging from 5.0 to 488 mM.
Solutions containing pyrrole in concentrations ranging from
1
.22 to 14 M were prepared by concentrating the 5-mL portion
of the solution of 1a -OH, transferring the concentrated solu-
tion to a volumetric flask (5 mL), adding pyrrole, and filling
2 2
to the mark with CH Cl . The contents of the volumetric flasks
were returned to 20-mL vials. Reactions were initiated by
Exp er im en ta l Section
pouring each solution of 1a -OH and pyrrole into a 20-mL vial
containing Dy(OTf) (3.0 mg, 0.0050 mmol, 1.0 mM). The
3
reactions were monitored spectrophotometrically for yield of
TPC at 15 min and 1 h as described above. TLC was performed
UV-Vis Sp ectr op h otom etr ic Deter m in a tion of th e
Yield of Cor r ole. Corrole-forming reactions were monitored
by transferring an aliquot (200 µL) of the condensation reaction
mixture by syringe to a 1.8-mL microcentrifuge tube contain-
ing a DDQ solution (300 µL, 10 mM in toluene). The mixture
was vortex mixed for 2-5 s. To the oxidized reaction mixtures
on the crude, oxidized mixture [silica, CH
2 2
Cl /hexanes (2:1)]
after washing with the aq dithionite solution.
Rea ction Tim e Cou r se Exp er im en ts. Reaction monitor-
ing as a function of time was performed as described above
for the acid-screening experiments with the exception of using
a 10-mL reaction volume. Each reaction was initiated by the
addition of the appropriate acid at room temperature (Table
1, entries 2, 5, and 9). The reactions were monitored for yield
of corrole as described above at 1 min, 4 min, 8 min, 15 min,
30 min, 1 h, 2 h , 4 h, 7 h, and 24 h. TLC was performed on
2 2 4
was added an aq solution of 5% Na S O , 5% NaOH (0.5 mL)
followed by vigorous shaking for 10 s. Samples were spun in
a microcentrifuge at 1000 g for 1-3 min to hasten the
separation of the aqueous and organic layers. A portion of the
upper organic layer (50 µL) was transferred by syringe to a
3
0
Pasteur pipet filled two-thirds full with silica gel (∼1.5 g).
The sample was eluted with three, 1-mL portions of CH Cl .
2
2
Solvent was driven off the column by using a handheld pipet
tool. The eluent was transferred to a cuvette and the UV-vis
spectrum was recorded. The yield of corrole was determined
by comparing the absorbance at 420 nm with the theoretical
absorbance calculated from the following equation:
the crude, oxidized mixtures [silica, CH
2 2
Cl /hexanes (2:1)] after
washing with the aq dithionite solution.
Deter m in a tion of th e Sta bility of Cor r oles. Solutions
of corroles 2a -e were prepared in toluene in the dark. The
concentration of each solution was adjusted to provide an
absorbance of ∼1.0. UV-vis spectra were recorded in the dark,
and again from 5 min to 24 h after the room lights were turned
on. Decomposition of corrole was inferred from changes in the
intensity of the Soret band. A control experiment involving
corrole solutions maintained in the dark was performed
similarly.
Atheoretical ) (110 000 cm M-1)(1.0 cm)(0.0025 M) ×
-1
2
5
00 µL
00 µL 3050 µL
50 µL
)(1.30)(0.306) ) 0.716
(
) (
-
1
M-1 is the reported molar extinction
where 110 000 cm
6
b
coefficient for TPC, 1.0 cm is the path length of the cuvette,
.0025 M is the theoretical concentration of corrole given the
initial concentration of dipyrromethanedicarbinol (2.5 mM),
200 µL/500 µL) ,and (50 µL/3050 µL) account for the dilution
Gen er a l P r oced u r e for th e P r ep a r a tive-Sca le Syn th e-
ses of Cor r oles, Given for 5,10,15-Tr ip h en ylcor r ole (2a ).
The reduction of 1a (215 mg, 0.500 mmol) with NaBH (946
4
mg, 25.0 mmol) in THF/methanol (40 mL, 3:1) afforded the
corresponding carbinol 1a -OH. The carbinol was dried under
vacuum for 30 min and then immediately subjected to con-
densation with pyrrole (3.47 mL, 50.0 mmol) in the presence
0
(
of the corrole upon oxidation of an aliquot in the DDQ solution
and passage through the silica pipet, respectively, 1.30 is an
2 2
empirically derived correction factor for the volume of CH Cl
eluent that is retained on the silica pipet,30 and 0.306 is an
empirically derived correction factor for the attenuation of the
absorbance of the Soret band of corrole in the presence of
pyrrole and DDQ. See the Supporting Information for a
discussion of the determination of the two correction factors.
Acid -Scr een in g Exp er im en ts. Immediately prior to the
condensation reactions, 1a (215 mg, 0.500 mmol) was reduced
to the corresponding dipyrromethanedicarbinol 1a -OH with
3 2 2
of Dy(OTf) (122 mg, 0.200 mmol) in CH Cl (200 mL) for 1.3
h at room temperature followed by the addition of triethy-
lamine (0.139 mL, 1.00 mmol) and oxidation with p-chloranil
(246 mg, 1.00 mmol) at reflux for 45 min. The entire reaction
mixture was filtered through a pad of silica gel and eluted with
CH
sample was dissolved in CH
gel (15 g), concentrated, and purified by chromatography
[silica, CH Cl /hexanes (1:1)]. The green band containing TPC
2
Cl
2
. The filtrate was concentrated to dryness. The corrole
2
2
Cl (75 mL), adsorbed onto silica
NaBH
following a literature procedure. After being dried under
vacuum for 30 min, the dicarbinol was dissolved in CH Cl
and transferred to a 100-mL volumetric flask. A solution of
4
(1.3 g, 36 mmol) in THF/methanol (40 mL, 3:1)
1
9
2
2
2
2
was collected, concentrated, and dried under vacuum affording
TPC (128 mg, 49%) that was subsequently crystallized from
4
168 J . Org. Chem., Vol. 69, No. 12, 2004

Synthesis of Meso-Substituted Corroles
CH
published values.
,15-Bis(4-m eth ylp h en yl)-10-p h en ylcor r ole (2b). The
reduction of 1b (229 mg, 0.500 mmol) followed by condensation
with pyrrole (3.47 mL, 50.0 mmol) in the presence of Dy(OTf)
122 mg, 0.200 mmol) in CH Cl (200 mL) for 1 h at room
2
Cl
2
/hexanes providing analytical data consistent with
(s, 3H), 4.08 (s, 3H), 7.28 (d, J ) 8.2 Hz, 2H), 7.62 (d, J ) 7.4
Hz, 2H), 7.71 (t, J ) 7.3 Hz, 1H), 7.81 (t, J ) 7.3 Hz, 2H), 8.07
(d, J ) 8.2 Hz, 2H), 8.25 (d, J ) 7.4 Hz, 2H), 8.34 (d, J ) 7.2
Hz, 2H), 8.56 (m, 4H), 8.85-8.91 (m, 4H); LD-MS obsd 570.1
6
5
+
+
3
(M ); HRMS (FAB) obsd 571.2515 (MH ), calcd 571.2498
+
(
2
2
(MH ) (C39
,10,15-Tr is(p en ta flu or op h en yl)cor r ole (2e). The reduc-
tion of 1e (140 mg, 0.200 mmol) followed by condensation with
pyrrole (1.39 mL, 20.0 mmol) in the presence of Dy(OTf) (48.8
mg, 0.0800 mmol) in CH Cl (80 mL) for 10 h at room
30 4
H N O).
temperature, addition of triethylamine (0.139 mL, 1.00 mmol),
oxidation with p-chloranil (246 mg, 1.00 mmol) at reflux for
5
4
5 min, and purification identical to the general procedure
3
afforded 2b (222 mg, 80%) that was subsequently crystallized
2
2
-
3
from CH
2
Cl
2
/hexanes. λabs (toluene, ꢀ × 10 ) 421 (130), 564
temperature, addition of triethylamine (056 µL, 0.40 mmol),
oxidation with DDQ (136 mg, 0.600 mmol) at room tempera-
ture for 1 h, addition of a further quantity of DDQ (136 mg,
1
(
16.4), 617 (13.4), 651 (13.7); H NMR [CD
3
OD (0.6%) in
3
5
3
CDCl ] δ 2.67 (s, 6H), 7.62 (d, J ) 7.7 Hz, 4H), 7.73 (m, 3H),
8
4
8
.16 (d, J ) 5.7 Hz, 2H), 8.24 (d, J ) 7.7 Hz, 4H), 8.52 (d, J )
0
.600 mmol) over 30 min, and purification identical to the
general procedure afforded 2e (28 mg, 18%) that was subse-
quently crystallized from CH Cl /hexanes providing analytical
.6 Hz, 2H), 8.57 (d, J ) 3.5 Hz, 2H), 8.87 (d, J ) 4.6 Hz, 2H),
+
.91 (d, J ) 4.0 Hz, 2H); LD-MS obsd 554.1 (M ); HRMS (FAB)
2
2
+
+
obsd 555.2583 (MH ), calcd 555.2549 (MH ) (C39
H
30
N
4
).
5
data consistent with published values.
5
,10-Dip h en yl-15-(4-m eth ylp h en yl)cor r ole (2c). The re-
duction of 1c (222 mg, 0.500 mmol) followed by condensation
with pyrrole (3.47 mL, 50.0 mmol) in the presence of Dy(OTf)
122 mg, 0.200 mmol) in CH Cl (200 mL) for 1.25 h at room
Ack n ow led gm en t. This research was supported by
an award from Research Corporation. The authors
thank the Merck/AAAS Undergraduate Science Re-
search Program for Summer Undergraduate Research
Fellowships to C.G.R and J .F.B.C. Partial support for
this work was provided by the National Science Foun-
dation’s Course, Curriculum, and Laboratory Improve-
ment Program under grant DUE-0088227. LD-MS and
FAB mass spectra were obtained at the Mass Spectrom-
etry Laboratory for Biotechnology at North Carolina
State University. Partial funding for the facility was
obtained from the North Carolina Biotechnology Center
and the NSF.
3
(
2
2
temperature, addition of triethylamine (0.139 mL, 1.00 mmol),
oxidation with p-chloranil (246 mg, 1.00 mmol) at reflux for
4
5 min, and purification identical to the general procedure
afforded 2c (135 mg, 50%) that was subsequently crystallized
-
3
from CH
2
Cl
2
/hexanes. λabs (toluene, ꢀ × 10 ) 421 (133), 564
1
(16.9), 617 (13.1), 651 (13.3); H NMR [CD
3
OD (0.6%) in CDCl
3
]
δ 2.68 (s, 3H), 7.62 (d, J ) 7.7 Hz, 2H), 7.73 (m, 4H), 7.81 (m,
H), 8.16 (d, J ) 5.7 Hz, 2H), 8.25 (d, J ) 7.7 Hz, 2H), 8.35 (d,
J ) 7.2 Hz, 2H), 8.53-8.57 (m, 4H), 8.87 (m, 2H), 8.92 (m,
2
+
2
(
H); LD-MS obsd 540.2 (M ); HRMS (FAB) obsd 541.2366
+
+
MH ), calcd 541.2392 (MH ) (C38
28 4
H N ).
5
-P h en yl-10-(4-m eth oxyp h en yl)-15-(4-m eth ylp h en yl)-
cor r ole (2d ). The reduction of 1d (237 mg, 0.500 mmol)
followed by condensation with pyrrole (3.47 mL, 50.0 mmol)
Su p p or tin g In for m a tion Ava ila ble: General experimen-
tal methods, synthesis of diacyldipyrromethanes 1a and 1c,
description of oxidant screening experiments, discussion of
correction factors for spectrophotometric analysis of corroles,
plots of the yield of TPC as a function of acid concentration,
investigation of TFA/acetonitrile condensation conditions, time
course experiments for reactions leading to corroles 2b,d ,e,
3 2 2
in the presence of Dy(OTf) (122 mg, 0.200 mmol) in CH Cl
(200 mL) for 1 h at room temperature, addition of triethyl-
amine (0.139 mL, 1.00 mmol), oxidation with p-chloranil (246
mg, 1.00 mmol) at reflux for 45 min, and purification identical
to the general procedure with the exception that CH
2 2
Cl /
hexanes (3:2) was used during chromatography afforded 2d
(
Cl
169 mg, 59%) that was subsequently crystallized from CH
2
-
1
data pertaining to the stability of corroles 2a -e, H NMR
-
3
2
/hexanes. λabs (toluene, ꢀ × 10 ) 421 (121), 563 (15.8), 619
spectra of diacyl dipyrromethanes 1a and 1c, H NMR spectra
1
1
(12.4), 652 (13.4); H NMR [CD
3
OD (0.6%) in CDCl
3
] δ 2.68
of corroles 2a -e, and COSY spectra of corroles 2b-d . This
material is available free of charge via the Internet at
http://pubs.acs.org.
(
35) The 1H NMR spectral results that we obtained differ slightly
14
from those previously reported for 2b. This likely stems from
differences in resolution between the spectra.
J O0496493
J . Org. Chem, Vol. 69, No. 12, 2004 4169