Locked π-expanded chlorins in two steps from simple tetraarylporphyrins

Article abstract of DOI:10.1021/ol052103y

(Chemical Equation Presented) Upon tandem Reformatsky reaction, easily accessible porphyrinic ketones give "locked" chlorinic diester. Both ketones and diesters, as bases or palladium complexes, efficiently generate singlet dioxygen, as demonstrated by tr

Full text of DOI:10.1021/ol052103y

ORGANIC
LETTERS
2
005
Vol. 7, No. 23
257-5260
Locked
π-Expanded Chlorins in Two
5
Steps from Simple Tetraarylporphyrins
Julien Fouchet, Christophe Jeandon, Romain Ruppert, and Henry J. Callot*
U.M.R. 7177 du C.N.R.S., Facult e´ de Chimie, UniVersit e´ Louis Pasteur,
1, Rue Blaise Pascal, F-67000 Strasbourg, France
callot@chimie.u-strasbg.fr
Received September 1, 2005
ABSTRACT
Upon tandem Reformatsky reaction, easily accessible porphyrinic ketones give “locked” chlorinic diester. Both ketones and diesters, as
bases or palladium complexes, efficiently generate singlet dioxygen, as demonstrated by trapping with cholesterol.
Whether as photosensitizers for photodynamic therapy or as
building blocks for models of photosynthetic systems,
chlorins (â,â′-dihydroporphyrins) are often chosen since their
(locked â positions), (e) photochemical stability above 600
nm, and (f) functional groups that are not conjugated to the
chromophore and that can be easily modified in order to tune
the hydro- or lipophilicity of the photosensitizer.
chromophores show intense long wavelength absorptions
1
(
g600 nm). These compounds are accessible by modifica-
tions of natural products and chlorophylls or by total
synthesis. Whatever the route selected, the stability of the
products vs. â,â′-dehydrogenation or -elimination to form
more stable porphyrins is crucial. In recent years several
routes to “locked” chlorins (general structure 1) have been
developed, the most recent and efficient one by Gryko and
2
-6
Galezowski.
The prerequisites for an optimal photosen-
sitizer are (a) efficient generation of singlet oxygen, (b)
significant absorption at wavelengths g600 nm, (c) easy
access (limited number of good yield steps, simple reaction
conditions, minimal number of chromatographic separations,
easy scaling up, cheap reagents), (d) chemical stability
Figure 1.
In this letter we present a shorter and more economical
route to such chlorins, as well as tests demonstrating that
our chromophores and some of their metal complexes act
as stable and efficient photosensitizers to produce singlet
oxygen.
(1) (a) Bonnett, R. Chem. Soc. ReV. 1995, 19. (b) Sternberg, E. D.;
Dolphin, D.; Br u¨ ckner, C. Tetrahedron 1998, 54, 4151. (c) Ali, H.; van
Lier, J. E. Chem. ReV. 1999, 99, 2379. (d) Pandey, R. K.; Zheng, G. In The
Porphyrin Handbook; Kadish, K. M., Smith, K. M., Guilard, R., Eds.;
Academic Press: Boston, MA, 2000; Vol. 6, p 157.
(
2) Gryko, D. T.; Galezowski, M. Org. Lett. 2005, 7, 1749.
During the study of the benzoylation of nickel tet-
raarylporphyrins, we showed that Ni-2 (as well as the p-OEt
(3) Battersby, A. R.; Dutton, C. J.; Fookes, C. J. R. J. Chem. Soc., Perkin
Trans. 1 1968, 1569.
1
0.1021/ol052103y CCC: $30.25
© 2005 American Chemical Society
Published on Web 10/19/2005

7
analogue), instead of yielding the expected benzoylporphy-
To perform the next step, the Reformatsky reaction was
selected, but instead of hydroxyester Ni-5, diester Ni-6, a
product of a double Reformatsky reaction, was obtained in
48% and 40% yield (ethyl and methyl esters, respectively).
Independently, reaction of ketone Ni-4 with the anion of ethyl
acetate at -78 °C gave hydroxyester Ni-5 in nearly quantita-
tive yield.
The structure of diesters Ni-5 and Ni-6 was unambiguously
established from their UV-visible and NMR data. The UV-
visible spectrum of Ni-6 (R ) Me) shows two major
absorption bands: a Soret band at 450 nm and an intense
absorption at 668 nm as one would expect for a chlorin.
However, due to the extended conjugated system, both bands
show a bathochromic shift relative to simple meso-tetraaryl-
chlorins whose corresponding bands are located in the 615-
rin, gave, after oxidation under basic conditions, corrole Ni-3
up to 24%) and ketone Ni-4 (up to 35%). To synthesize
precursors of corrole Ni-3 and widen the scope of the ring
contraction, we wished to (a) optimize the yield of ketone
Ni-4 at the expense of corrole Ni-3 and other byproducts
and (b) introduce an acetic acid side chain where the carbonyl
group was initially located in ketone Ni-4.
(
8
Scheme 1
6
20 nm range. Similar chlorins bearing an additional
9
,10
unsaturated ring have been described and also showed a
bathochromic shift (670 nm for the longest wavelength band
of Ni-7 ).
1
0
The NMR data for Ni-6 (R ) Me) are in full agreement
with the proposed structure: the two acetic chains show the
expected singlet for the methyl groups and an AB system
for the diastereotopic methylenic protons. On the other hand,
hydroxyester Ni-5 (R ) Et), while showing an almost
identical UV-visible spectrum, gave unequivocal NMR data
for the side chain (two pairs of diastereotopic protons) as
well as an exchangeable signal for the hydroxyl proton. For
Ni-5 as well as for all diesters, the signals for methyl and
ethyl protons are shifted to high field as one would expect
for groups placed between two aromatic rings.
Scheme 2
The first target was reached when we found that ketone
Ni-4 could be produced as the only product of low polarity
in 80% yield from the corresponding porphyrin under
acylating, then oxidizing conditions. The presence of acid
allowed the isomerization of the initial tertiary alcohol
produced by the acylation + cyclization sequence, thus
ensuring the full transformation of the product to ketone Ni-4
via the pyrrolic secondary alcohol. Under the same conditions
the corresponding palladium complex Pd-2 gave ketone Pd-4
in 79% yield.
(4) (a) Strachan, J.-P.; O’Shea, D. F.; Balasubramanian, T.; Lindsey, J.
S. J. Org. Chem. 2000, 65, 3160. (b) Balasubramanian, T.; Strachan, J.-P.;
Boyle, P. D.; Lindsey, J. S. J. Org. Chem. 2000, 65, 7919. (c) Taniguchi,
M.; Kim, H.; Ra, D.; Schwartz, J. K.; Kirmaier, C.; Hindin, E.; Diers, J.
R.; Prathapan, S.; Bocian, D. F.; Holten, D.; Lindsey, J. S. J. Org. Chem.
2
002, 67, 7329. (d) Taniguchi, M.; Ra, D.; Mo, G.; Balasubramanian, T.;
Lindsey, J. S. J. Org. Chem. 2001, 66, 7342. (e) Taniguchi, M.; Kim, M.
N.; Ra, D.; Lindsey, J. S. J. Org. Chem. 2005, 70, 275.
(5) (a) Montforts, F.-P. Angew. Chem., Int. Ed. Engl. 1981, 20, 778. (b)
Schmidt, W.; Montforts, F.-P. Synlett 1997, 903. (c) Montforts, F.-P.; Kutzki,
O. Angew. Chem., Int. Ed. 2000, 39, 599.
(6) Van der Haas, R. N. S.; Tang, S.; Arenas Hernandez, B.; de Jong,
R. L. P.; Erkelens, C.; Liu, Y.; Gast, P.; Smijs, T. G. M.; Schuitmaker, H.
J.; Lugtenburg, J. Eur. J. Org. Chem. 2005, 3813.
(
7) Jeandon, C.; Ruppert, R.; Callot, H. J. Chem. Commun. 2004, 1090.
Examples of tandem Reformatsky reactions forming gem
C-C bonds are scarce and are, to our knowledge, only
(8) A full account of these results and the last developments of the
1
1
mechanistic hypotheses is in preparation.
5258
Org. Lett., Vol. 7, No. 23, 2005

known for substrates whose reactive positions are at the
carboxylic acid oxidation level (nitriles, acid chlorides, ortho
esters, or formates).
In our case, we suggest that a hydroxyester (as a zinc
alcoholate) was formed initially, but that the loss of the
oxygenated group was facilitated by both the zinc ion and
the stability of the resulting cation, which in turn reacted
with excess reagent. We checked that, under acidic conditions
Although sterically hindered by two neighboring aryl rings,
the two ester groups of Pd-6 could be easily hydrolyzed by
sodium hydroxyde in n-propanol into the dicarboxylic acid.
In turn this diacid could be esterified back to Pd-6 in acidic
methanol.
All new compounds possessing long wavelength absorp-
tions (see examples in Figure 2) were tested under illumina-
(
TsOH in CH
corresponding cation (λmax ) 407 nm (ꢀ ) 33 000), 499
21 500), 720 (10 000), 792 (9 800), 875 (10 600). Such long
2 2
Cl ), hydroxyester Ni-5 produced easily the
(
wavelength absorptions, extending to the near-infrared
region, are known for cations generated on a carbon directly
bound to the porphyrin ring.12
We also checked the reactivity of ketones lacking the
9
phenyl ring on the additional six-membered ring but
obtained only untractable mixtures under Reformatsky
conditions.
To extend the scope of the reaction to other metal
complexes and the corresponding bases, we first subjected
ketone Pd-4 to the Reformatsky conditions and obtained
diester Pd-6 in 24% yield. To obtain the corresponding base
we could not start with Ni-6, whose attempted acid-catalyzed
demetalation led to extensive decomposition. We demetalated
Figure 2. UV-visible spectra of Pd-4 and Pd-6.
2 4
ketone Ni-4 in TFA-H SO (95% yield) and remetalated the
resulting base H
2
-4 with zinc(II) (Scheme 3). This complex
tion (λ > 510 nm, to avoid excitation in the Soret band
region) in order to check their ability to generate singlet
oxygen.
Scheme 3
Many spectroscopic or chemical methods of singlet oxygen
detection are available. Among them, the oxidation of
13
cholesterol is well documented, used as detector for singlet
1
4
oxygen in biological systems, and very easy to carry out
(see the Supporting Information for experimental details).
Scheme 4
The results are collected in Table 1. For comparison, the
2
corresponding meso-tetraarylporphyrin base H -2 was tested
under the same experimental conditions. As expected, ketone
(11) Tandem Reformatsky condensation forming gem C-C bonds
starting with alkyl formates and orthoformates, see: (a) Reformatsky, S.
Ber. 1895, 28, 3263. (b) Gerzon, K.; Flynn, E. H.; Sigal, M. V., Jr.; Wiley,
P. F.; Monahan, R.; Quark, U. C. J. Am. Chem. Soc. 1956, 78, 6396. (c)
Gawronski, J. K. Tetrahedron Lett. 1984, 25, 2605. Starting with nitriles
and acid chlorides: (d) El Alami, N.; Belaud, C.; Villi e´ ras, J. J. Organomet.
Chem. 1987, 319, 303. General reviews on Reformatsky reaction: (e)
Rathke, M. W. Org. React. 1975, 22, 423. (f) F u¨ rstner, A. Synthesis 1989,
was then subjected to the Reformatsky conditions and gave
diester Zn-6 in 20% yield. Acid treatment of Zn-6 gave base
H -6.
2
5
71. (g) Ocampo, R.; Dolbier, W. R., Jr. Tetrahedron, 2004, 60, 9325.
(12) Xu, Y.; Jaquinod, L.; Wickramasinghe, A.; Smith, K. M. Tetrahe-
dron Lett. 2003, 44, 7753.
(
9) Ishkov, Y. V. Russ. J. Org. Chem. 2001, 37, 288.
(13) Beckwith, A. L. J.; Davies, A. G.; Davison, I. G. E.; Maccoll, A.;
Mruzek, M. H. J. Chem. Soc., Perkin Trans. 2 1989, 815.
(14) Girotti, A. W.; Korytowski, W. Methods Enzymol. 2000, 319, 85.
(10) Faustino, M. A.; Neves, M. G. P. M. S.; Vicente, M. G. H.; Silva,
A. M. S.; Cavaleiro, J. M. S. Tetrahedron Lett. 1995, 36, 5977.
Org. Lett., Vol. 7, No. 23, 2005
5259

turnovers in 3 h). Ketone Pd-4 was recovered unchanged
even after long-term irradiation (10 h and more than 100
turnovers).
Table 1. Cholesterol Photochemical Oxidationsa
longest wavelength
turnover no.
(after 3 h of
irradiation)
This reaction sequence represents by far the shortest and
most efficient route to locked chlorins that are good
photosensitizers since (a) ketones Ni-4 and Pd-4 were
prepared in 80% and 79% in a one-pot sequence from a
nickel or palladium tetra-p-tert-butylphenylporphyrin, itself
made in a one-pot sequence from pyrrole and p-tert-
butylbenzaldehyde in 23% yield, and transformed into Ni-6
and Pd-6 in one step, via a novel tandem Reformatsky
reaction, (b) the overall yields of diesters M-6 from shelf
chemicals are 8.9% (M ) Ni) and 4.5% (M ) Pd), (c)
“locked” diester Pd-6 efficiently catalyzes the oxidation of
cholesterol by singlet oxygen, (d) all new compounds show
strong absorption above 650 nm, and (e) the presence of
functional groups that can be transformed and are indepen-
dent of the chromophore opens a possibility to tune the
hydrophilic/lipophilic properties of the sensitizers, since the
chlorinic chromophores proved to withstand acidic or basic
conditions, as well as photooxygenation. In addition, the very
photosensitizer
PS)
band (λmax and
ꢀ in C6H6)
(
H2-2
Ni-6 (R ) Me)
Zn-6
H2-4
H2-6
648 (3400)
668 (14000)
670 (25000)
724 (26000)
686 (27000)
686 (40000)
648 (40000)
9
0
0
27
2.5
35
38
Pd-4
Pd-6
a
-3
Experimental conditions: 20 mL of benzene, cholesterol (2 × 10
-
5
mol/L), PS (10 mol/L), 3 h of irradiation (250 W slide projector bulb, λ
510 nm), cholesterol/PS ) 200.
>
Ni-4 and chlorin Ni-6 (R ) Me) were photochemically
inactive, but both were recovered intact after irradiation. No
cholesterol oxidation products were detected in the presence
of zinc complexes Zn-4 and Zn-6 (R ) Me), but each one
was converted cleanly in one single compound. Full char-
acterization of both compounds is still under way, but mass
spectroscopic analysis showed that these oxidized compounds
are the zinc complexes of products resulting from the
2
accessible ketonic precursors H -4 and Pd-4 are also
excellent photosensitizers.
Acknowledgment. We thank P. Wehrung for mass
spectra.
oxidative cleavage of the tetrapyrrole ring. Both bases H
2
-4
and H -6 and palladium complexes Pd-4 and Pd-6 were
2
Supporting Information Available: Experimental pro-
cedures for the preparation of Ni-4, Ni-5, Ni-6, Pd-4, Pd-6,
H -4, H -6, Zn-4, and Zn-6; full experimental details for
2 2
the photosensitized oxidation of cholesterol. This material
is available free of charge via the Internet at http://pubs.acs.org.
found to be active and stable under irradiation conditions,
as shown by the fact that their electronic spectrum taken
after irradiation was superimposable with the starting spec-
trum. H
when compared to porphyrin base H
9 turnovers in 3 h) and entries 4, 6, and 7 (from 27 to 38
2
-6, Pd-4, and Pd-6 are excellent photosensitizers
2
-2: compare entry 1
(
OL052103Y
5260
Org. Lett., Vol. 7, No. 23, 2005