Benzoylation at the meso position of a zinc(II) deuteroporphyrin

Article abstract of DOI:10.1007/PL00010106

By short treatment of zinc(II) deuteroporphyrin-IX dimethyl ester with benzoic anhydride/SnCl₄, a mixture of benzoylated derivatives was obtained from which a tribenzoyl derivative with benzoyl groups at the 2, 4, and γ-meso positions could be isolated in low yield. Except for the known Vilsmeier formylation, this constitutes the first example of a Friedel-Crafts acylation taking place at the meso position of a porphyrin chelate. The expected lack of regioselectivity of this reaction, even at low conversion rates, and the difficulty to separate the products limit, of course, its synthetic value for the preparation of the particular benzoyl derivatives. This limitation cannot be overcome by starting with a completely β-substituted porphyrin or with a symmetrically substituted porphyrin, since attempts to benzoylate Zn(II) mesoporphyrin-IX dimethyl ester lead exclusively to demetallation. However, these findings add to the knowledge of the fundamental properties of porphyrin reactivity. A rationalization based on calculations at a semiempirical level, which are in acceptable agreement with the experimental results, is presented.

Full text of DOI:10.1007/PL00010106

Monatshefte fuÈr Chemie 129, 69±79 (1998)
Benzoylation at the meso Position of a Zinc(II)
Deuteroporphyrin
1
1
2;Ã
Â
Â
Jordi Vazquez , Maria M. Gonzalez , Cristina Martõ , Santi Nonell
Â2
,
and Francesc R. Trull1;y
1
 Á
Departament de Quõmica Organica, Universitat de Barcelona, E-08028 Barcelona, Catalunya, Spain
 Á
CETS Institut Quõmic de Sarria, Universitat Ramon Llull, E-08017 Barcelona, Catalunya, Spain
2
Summary. By short treatment of zinc(II) deuteroporphyrin-IX dimethyl ester with benzoic
anhydride/SnCl4, a mixture of benzoylated derivatives was obtained from which a tribenzoyl
derivative with benzoyl groups at the 2, 4, and ꢀ-meso positions could be isolated in low yield.
Except for the known Vilsmeier formylation, this constitutes the ®rst example of a Friedel-Crafts
acylation taking place at the meso position of a porphyrin chelate. The expected lack of
regioselectivity of this reaction, even at low conversion rates, and the dif®culty to separate the
products limit, of course, its synthetic value for the preparation of the particular benzoyl derivatives.
This limitation cannot be overcome by starting with a completely ꢁ-substituted porphyrin or with a
symmetrically substituted porphyrin, since attempts to benzoylate Zn(II) mesoporphyrin-IX
dimethyl ester lead exclusively to demetallation. However, these ®ndings add to the knowledge of
the fundamental properties of porphyrin reactivity. A rationalization based on calculations at a
semiempirical level, which are in acceptable agreement with the experimental results, is presented.
Keywords. Electrophilic substitution; Benzoylation; Porphyrins; Semiempirical calculations.
Zur Benzoylierung an der meso-Position eines Zink(II)-Deuteroporphyrins
Zusammenfassung. Durch kurzes Behandeln von Zink(II)-Deuteroporphyrin-IX-dimethylester mit
wurde ein Gemisch von benzoylierten Derivaten erhalten, aus welchen
È
Benzoesaureanhydrid/SnCl4
in geringer Ausbeute ein Tribenzoylderivat isoliert werden konnte, das die Benzoylgruppen in 2-,4-
È
und ꢀ-meso-Position enthalt. Abgesehen von der Vilsmeier-Formylierung stellt diese Reaktion das
erste Beispiel einer Friedl-Crafts-Acylierung an der meso-Position eines Porphyrinchelates dar.
È
Mangelnde Selektivitat der Reaktion auch bei geringem Umsatz und die Schwierigkeiten bei der
Trennung der Produkte engen natuÈrlich ihre Bedeutung fuÈr die Synthese bestimmter Benzoylderivate
È
ein. Dieser Mangel konnte auch nicht durch Verwendung von vollstandig ꢁ-substituierten oder
symmetrisch substituierten Porphyrinen behoben werden, da Versuche, Zn(II)-Mesoporphyrin-IX-
dimethylester zu benzoylieren, ausschlieûlich das entmetallierte System ergaben. Allerdings tragen
diese Resultate zum Wissen uÈber die fundamentalen Reaktionsweisen von Porphyrinen wesentlich
bei. Eine Rationalisierung der Ergebnisse kann auf der Basis von semiempirischen Rechen-
methoden, deren Resultate mit den experimentellen Daten im Einklang stehen, gegeben werden.
ÐÐÐÐ
y
Deceased on June 28, 1997 at the age of 43

Â
J. Vazquez et al.
70
Introduction
The study of the relative reactivity of the different positions in the porphyrin
periphery (i.e., ꢁ and meso positions) is an important goal of porphyrin chemistry
[1], particularly because it could set the path to differently substituted porphyrins
which are of general interest for the evaluation of the effects of these substituents
upon the chemical behaviour of the porphyrins [2]. Among those, benzoylated
porphyrins are a priori candidates as non-¯uorescent singlet molecular oxygen
1
ꢀO2ꢀ Ág†† sensitizers [3].
In principle, deuteroporphyrin-IX 1a with two free ꢁ positions should be a
suitable system to study the relative reactivity of ꢁ vs. meso positions of porphyrins
and for the subsequent preparation of their structurally related derivatives. A
number of electrophilic substitution reactions of deuteroporphyrin or its iron(III)
chloride have been investigated [1, 2, 4]. The results of these studies are basically
in agreement with the calculated reactivity indices for a metalloporphyrin towards
an electrophile: 0.29 (meso), 0.05 (ꢁ-pyrrole), 0.14 (ꢂ-pyrrole), 0.07 (N) [2].
Exceptional preferences for ꢁ-pyrrolic substitutions have been mostly rationalized
on grounds of steric hindrance of the meso carbons [1]. Substitution of the two NH

Benzoylation of a Zinc(II) Deuteroporphyrin
71
protons in a porphyrin by metal ions of low electronegativity, e.g. Mg2‡, lead to
porphyrin ligands of relatively high nucleophilicity. However, the acid lability of
magnesium porphyrins prevents their general use in electrophilic reactions.
Previously, the synthesis of the 6-benzoyl porphyrins 1e and 1f has been
achieved by treatment of the iron(II) complex of the corresponding 6-unsubstituted
precursor (1c, 1d) with benzoic anhydride using SnBr4 as the Lewis acid [4].
Iron(III) porphyrins were chosen as substrates because metal-free porphyrins did
not react. Apparently, meso substitution has never been observed, although similar
reactions have been applied to a number of porphyrins.
Interestingly enough, Fischer has prepared both the 4-benzoyl porphyrin 1g and
the symmetrical 1,4-dibenzoyl porphyrin 1h using a completely different strategy:
by coupling of the two dipyrrin halves [4]. On the other hand, no benzoylated
porphyrin of type IX has been reported so far.
Acetylation of copper and iron(III) porphyrins has only been described for ꢁ-
pyrrolic positions. The usual product is the 2,4-diacetyl derivative, but careful
control of reaction conditions led to formation of a mixture of 2- and 4-acetyl
deuteroporphyrin dimethyl esters [5, 6].
In summary, non of the Friedel-Crafts acylations reported to date leads to meso
substitution. On the contrary, the Vilsmeier formylation does so in almost
quantitative yield [7]. No rationalization has been offered for these results.
Introduction of nitrogen, sulfur, and halogen substituents into the porphyrin
periphery has also been achieved [1, 5, 8]. These reactions are regioselective and
usually take place either at the ꢁ positions (bromination, iodination) or at the meso
positions (nitration, chlorination), but not at both.
Because the photophysical properties of porphyrin derivatives of the
benzophenone type (i.e. with structures like 1i±1k) are of interest, we decided to
further investigate Fischer's benzoylation procedure upon the easily accesible
deuterohemin dimethyl ester (iron(III) deuteroporphyrin-IX dimethyl ester, Fe(III)-
1a). In addition, we extended the benzoylation reaction to zinc(II) deutero-
porphyrin-IX dimethyl ester (Zn(II)-1a) which contains a more easily removable
metal.
Results and Discussion
The product of the above reaction is a partially demetallated mixture of mainly
mono-, di-, and tribenzoylated porphyrins where benzoylation at the 2 and 4ꢁ
positions predominates. HPLC permits to isolate the 2,4,ꢀ-tribenzoyl deutero-
porphyrin-IX dimethyl ester 1k. This compound constitutes the ®rst reported meso
substituted derivative prepared by Friedel-Crafts acylation of a porphyrin chelate.
It is also interesting because of its low ¯uorescence quantum yield (ca. one third
that of the unbenzoylated precursor). Interestingly, benzoylation of a completely ꢁ-
substituted porphyrin chelate such as zinc(II) mesoporphyrin-IX dimethyl ester,
Zn(II)-1b, under analogous conditions does not take place, and the reaction leads
exclusively to demetallation. These results have been used to propose a mechanism
for the benzoylation in the deuteroporphyrin series. Quantum mechanical
calculations at a semiempirical level have been performed of the relative reactivity
towards electrophiles of deuteroporphyrin-IX, its Zn(II) complex, and the 2-

Â
J. Vazquez et al.
72
benzoyl and 2,4-dibenzoyl derivatives of the latter. The theoretical results are
compared to the experimental ones.
Benzoylation of iron(III)-deuteroporphyrin-IX
(duterohemin, Fe(III)-1a, R6 ˆ ꢁ(CH2)2-COOH)
According to Fischer [4], treatment of the iron(III) complex of the ꢁ-free
porphyrins 1c and 1d with benzoic anhydride/SnBr4 for 10±12 min at 70ꢂC,
followed by a cumbersome work-up which includes as key steps treatment with
ice-cooled concentrated H2SO4 (probably to demetallate the product) and
esteri®cation of the resulting free acid with CH2N2, affords the corresponding
metal-free monobenzoyl porphyrin dimethyl esters 1e and 1f.
The same procedure was applied to iron(III)-deuteroporphyrin-IX (deutero-
hemin, Fe(III)-1a, R6 ˆ ꢁ(CH2)2-COOH). A mixture (by TLC) containing one or
more benzoylated derivative(s) of the starting compound was obtained: the Soret
band with a shoulder was observed near 450 nm. This should be compared with
2,4-diacetyldeuterohemin which has a Soret maximum at 440 nm in pyridine [9].
Phenyl signals where found near 7.9±7.2 ppm in the 1H NMR spectrum (DMSO-d6/
NaCN), and a ketone C=O band near 1660 cmꢁ1 was observed in the IR spectrum.
However, all attempts to demetallate this product(s) were unsuccessful. These
included treatment with: 1) concentrated H2SO4 at 0ꢂC for 1 h; 2) MeOH/HCl(g)
in the presence of FeSO4 as described by Caughey et al. [5] for the one-step
conversion of deuterohemin into deuteroporphyrin-IX dimethyl ester (1a); and 3)
concentrated H2SO4/FeSO4 at 0ꢂC for 3 h. With the MeOH/HCl mixture, complete
esteri®cation of the components of the mixture took place (new singlet near
1
3.8 ppm in the H NMR, ester C=O band near 1725 cmꢁ1 in the IR).
These results indicate that iron(III) removal from the benzoylated porphyrin(s)
is more dif®cult to achieve relative to that from the unbenzoylated precursor. The
results from benzoylation reactions of Zn(II)-1a (see below) con®rm this
observation.
Lack of reactivity of deuteroporphyrin-IX dimethyl ester (1a)
under the benzoylating conditions
Upon treatment of 1a with 1) benzoyl chloride/SnCl4 in CH2Cl2 at ꢁ3ꢂC for
20 min following the procedure previously reported for the benzoylation of other
aromatic hydrocarbons [10] and 2) benzoic anhydride/SnCl4 following essentially
the procedure reported by Fischer [4], only unreacted material was recovered even
after prolonged reaction time. This con®rms previous reports on the lack of
reactivity of 1a towards acylation reactions [1, 2].
Benzoylation of the zinc(II) chelates Zn(II)-1a and Zn(II)-1b
Starting from Zn(II)-1a, benzoylation took place to some extent in all attempts
when SnCl4 was used as catalyst (i.e. presence and absence of nitrobenzene as
solvent, reaction time 15 to 30 min, 70ꢂC). In contrast, when the more reactive
catalyst but stronger Lewis acid AlCl3 was used in Friedel-Crafts benzoylations

Benzoylation of a Zinc(II) Deuteroporphyrin
73
[10], no reaction took place, and the free base 1a was recovered. Apparently, AlCl3
is too strong a Lewis acid so that the acidic medium led to a demetallation of the
substrate.
Clearly, with SnCl4, demetallation of Zn(II)-1a occurs at a lower rate, thus
allowing for the benzoylation process to take place to some extent. Indeed, with
SnCl4 (for additional details see Experimental), a complex mixture of Zn(II)- and
free base deuteroporphyrin-IX derivatives with a variable number of benzoyl groups
per porphyrin was obtained. In addition, whereas the metal free derivatives contain
one to three benzoyl groups, up to ®ve benzoyl units are found in the Zn(II)-
compounds (by MS). This con®rms that metal removal from the benzoylated
porphyrin is more dif®cult than from the unbenzoylated precursor and that this
dif®culty increases with the number of benzoyl groups present on the chelate.
From the benzoylation mixture, one pure compound (2,4,ꢀ-tribenzoyldeutero-
porphyrin-IX dimethyl ester, 1k) could be isolated. This is the ®rst example of a
Friedel-Crafts acylation of a porphyrin chelate occurring at one meso position. The
structure of 1k was established spectroscopically: high resolution mass spectrom-
etry con®rmed the tribenzoylation. The location of the three benzoyl groups was
deduced from the 1H NMR spectrum where the C2-H and C4-H signals are absent;
in addition, substitution at the ꢀ-meso position was concluded on the basis of the
strong distortion of the shape and chemical shift of the -CH2-CH2-COOMe signals
relative to both deuteroporphyrin-IX dimethyl ester (1a) and 2,4-diacetyldeutero-
porphyrin-IX dimethyl ester (11). In the UV/Vis spectrum in CH2Cl2, the Soret
band is red-shifted by 19 nm and the Q bands by 14±17 nm relative to 1a. These
shifts are only slightly lower than those found for the 2,4-diacetyl derivative 1l
(25 nm for the Soret band and 19±22 nm the Q bands). The differences between 1k
and 1l may be interpreted in terms of a larger deviation from coplanarity of the
C=O group relative to the porphyrin in 1k, probably imposed by steric hindrance
exerted by the phenyl group. The ¯uorescence maximum in benzene is also red-
shifted (by 21 nm) relative to 1a and nearly coincides with that of 1l. Interestingly,
the ¯uorescence quantum yield of 1k is ca. 0.04, i.e. one third of that of 1a (one
fourth of that of 1l). These results con®rm that both acetylated and benzoylated
porphyrins could in principle be used as singlet molecular oxygen photosensitizers
of low ¯uorescence.
In contrast, treatment of zinc(II) mesoporphyrin-IX dimethyl ester (Zn(II)-1b)
under analogous benzoylation conditions leads exclusively to demetallation, and
the free base 1b is obtained. The last compound is inert towards benzoylation, as
was the deuteroporphyrin-IX dimethyl ester (DPDME) 1a. The previous result is
important because it indicates that when all ꢁ positions of porphyrin chelates are
substituted ± as in Zn(II)-1b ± benzoylation of the meso position(s) competes with
demetallation, the latter process apparently being more ef®cient and thus leading
exclusively to loss of zinc(II) from the porphyrin chelate. This result is also
important because it helps to rationalize the outcome of the benzoylation reaction
of Zn(II)-1a: in this case, competition between demetallation and benzoylation also
occurs. However, benzoylation of (most likely) the ꢁ position(s) is now more
ef®cient relative to demetallation, leading to a ꢁ-benzoylated zinc(II) deutero-
porphyrin dimethyl ester derivative. Secondly, and perhaps most importantly, once
ꢁ-benzoylated, the porphyrin chelate is more stable towards demetallation and can

Â
J. Vazquez et al.
74
therefore be further benzoylated (in meso). The presence of the benzoyl group(s) in
ꢁ-position of the porphyrin chelate is therefore a conditio sine qua non for its
subsequent meso benzoylation. In conclusion, of the three possible processes upon
the original chelates (demetallation, ꢁ-benzoylation, and meso benzoylation, the
present and previous results indicate the following order of ef®ciency: ꢁ-
benzoylation > demetallation > meso benzoylation. In addition, once the original
chelate has been ꢁ-benzoylated, the relative order of ef®ciency between
demetallation and meso benzoylation is reversed, so that now meso benzoylation
competes effectively with demetallation.
Calculations
The above results and those of previous authors [1, 2, 4±6] indicate that deutero-
porphyrin and other free base porphyrins show only poor reactivity towards
acylating agents and, in particular, towards benzoic anhydride. On the other hand,
Fischer et al. [4] have monobenzoylated a few Fe(III) porphyrins, and the present
report described the polybenzoylation of Zn(II) deuteroporphyrin-IX.
In order to check whether these results correspond with the predictions of
quantum mechanical calculations, we have performed a semi-empirical investiga-
tion of the relative reactivity of deuteroporphyrin-IX, its Zn(II) complex, and the 2-
benzoyl and 2,4-dibenzoyl derivatives of the last compound towards electrophiles.
The regioselectivity of these reactions has also been investigated. The energy gap
between the HOMO an subHOMO of the porphyrin on the one hand and the
LUMO of the benzoyl cation on the other (ꢁ6.507 eV) has been chosen as a
criterion for relative reactivity with the porphyrin derivative showing the smallest
gap corresponding to the most reactive one. Regioselectivity, in turn, has been
evaluated from the electron density maps of the HOMO and subHOMO of each
porphyrin, with positions of higher electron density expectedly being the most
reactive. Both criteria are based upon Fukui's frontier orbital model [11].
The results of these calculations (Table 1) show that, for all the porphyrins
investigated, the HOMO and subHOMO are close in energy (ca. 0.25 eV for
deuteroporphyrin and ca. 0.07 eV for the three Zn(II)-porphyrins studied). In
Table 1. Energies (eV) of the most signi®cant molecular orbitals of deuteroporphyrin-IX, Zn(II)-
deuteroporphyrin-IX, Zn(II)-2-benzoyldeuteroporphyrin-IX, and Zn(II)-2,4-dibenzoyldeuteropor-
phyrin-IX from semi-empirical calculations
Molecular
Orbital
1a
free acida
Zn(II)-1a
free acid
Zn(II)-1f
free acid
Zn(II)-1g
free acid
LUMO
ꢁ1.596
ꢁ7.848
ꢁ9.092
ꢁ0.395
ꢁ9.425
ꢁ9.546
ꢁ1.920
ꢁ7.746
ꢁ7.829
ꢁ9.213
ꢁ9.394
ꢁ9.536
ꢁ2.012
ꢁ7.889
ꢁ7.962
ꢁ9.330
ꢁ9.492
ꢁ9.633
ꢁ2.122
ꢁ7.988
ꢁ8.045
ꢁ9.412
ꢁ9.618
ꢁ9.714
HOMO
subHOMO
HOMO-2
HOMO-3
HOMO-4
a
Engergies for the N21-H, N23-H tautomer; the ordered values for the N22-H, N24-H tautomer are ꢁ1.619,
ꢁ7.857, ꢁ8.107, ꢁ-9.322, ꢁ9.457, ꢁ9.544 eV

Benzoylation of a Zinc(II) Deuteroporphyrin
75
contrast, the rest of HOMOs have much lower in energy (1.4±1.5 eV for HOMO-2
in all cases). According to these results, both HOMO and subHOMO have been
taken into consideration in the theoretical evaluation of the reactivity towards
electrophiles of the porphyrins.
The (EHOMO ± ELUMO) ‡ (EsubHOMO ± ELUMO)) values (in eV) thus obtained
are: deuteroporphyrin, 2.938 (average of two tautomers); Zn(II)-deuteroporphyrin,
2.561; Zn(II)-2-benzoyldeuteroporphyrin, 2.837; and Zn(II)-2,4-dibenzoyldeutero-
porphyrin, 3.019. These numbers would explain why Zn(II)-deuteroporphyrin, with
the smallest energy gap, is more reactive towards benzoylation than deuteropor-
phyrin, but not why the latter does not react under conditions where Zn(II)-2,4-
dibenzoyldeurteroporphyrin, with a sightly higher energy gap, apparently does.
The regioselectivity of the benzoylation process was analyzed from the electron
density (HOMO ‡ subHOMO) maps. The following conclusions can be drawn:
1) For deuteroporphyrin, the meso positions are expected to be the most
reactive ones; the pyrrole nitrogens display an intermediate reactivity, and the ꢁ-C2
and -C4 positions show the least reactivity.
2) For zinc(II) deuteroporphyrin, the electron densities of the meso and ꢁ-C2
and -C4 positions are comparable, and therefore a similar reactivity is expected for
them. If steric effects are considered, one would expect reactions with electrophiles
to occur preferentially at the ꢁ-C2 and/or -C4 positions.
3) For zinc(II) 2-benzoyldeuteroporphyrin, perhaps surprisingly, the most
reactive position is not C4 but the meso positions and, more speci®cally, the ꢁ and
ꢀ bridges. Even though the consideration of steric effects might alter this
theoretical order of reactivity, benzoylation of the ꢁ and ꢀ bridges probably occurs
as proven by the obtention of a mixture of dibenzoylated products.
4) For zinc(II) 2,4-dibenzoyldeuteroporphyrin, the ꢀ-meso position displays the
highest electron density and therefore one should expect benzoylation to occur
preferentially at this position; however, other dibenzoyl deuteroporphyrin isomers
are expected to be benzoylated at different positions, and as a result, a mixture of
tribenzoylporphyrins has to be expected.
Experimental
Chemicals
Deuteroporphyrin-IX dimethyl ester (1a) and mesoporphyrin-IX dimethyl ester (1b) were prepared
from hemin (Aldrich) as reported previously [5]. 2,4-Diacetyldeuteroporphyrin-IX dimethyl ester
(1l) was obtained from deuterohemin as described [5]. Zinc(II) mesoporphyrin-IX dimethyl ester
(Zn(II)-1b) was prepared as previously reported [12]. Benzoic anhydride (m.p.: 42ꢂC), benzoyl
chloride (both from Aldrich), aluminum chloride (Merck), and stannic chloride (Fluka) were puri®ed
as reported [13]. All solvents and solutions were rendered oxygen-free and argon saturated by several
cycles of evacuation at room temperature under magnetic stirring followed by argon saturation.
Reactions were typically carried out in the dark and under argon.
Calculations
Quantum mechanical calculations were performed at a semiempirical level using the SPARTAN
computer program, version 4.1.1, installed on a Silicon Graphics Power Indigo 2 work station [14].

Â
J. Vazquez et al.
76
All structures were optimized at the RHF/PM3 level [15] using the following criteria: maximum
gradient 1.0 Â 10ꢁ4 a.u., maximum shift 3.0 Â 10ꢁ5 nm, maximum energy difference
4.18 Â 10ꢁ3 kJ Á molꢁ1. No restrictions were used in the optimization of the geometries.
Analytical methods
Analytical thin layer chromatography (TLC) was carried out on Merck 60 F254 silica gel plates
(0.2 mm layer). Preparative thin layer chromatography (PTLC) was carried out on home-made
Merck 60 H silica gel plates (0.5 mm layer, 20 Â 20 cm). Column chromatograpyhy was performed
on basic Al2O3 (activity I, Merck; 70±230 mesh). Analytical HPLC was carried out on a home-made
Nucleosil-120 silica column (0.46 Â 25 cm; particle size: 3 mm) with a Shimadzu LC-7A pump (for
bio liquid chromatography; 1 ml/min) using a Shimadzu SPD-M6A photodoide array detector and a
toluene : n-heptane : MeOH ˆ 40:60:1 (v/v/v) solvent system. Preparative HPLC was performed on a
Chromasil-100 (MZ-Analysentechnik Mainz) silica column (2 Â 25 cm; particle size: 5 mm) with a
Shimadzu LC-8A pump (7 ml/min) using a Shimadzu SPD-6AV UV/Vis detector (detection at
400 nm) and a toluene:n-hexane:MeOH ˆ 40:60:1 (v/v/v) solvent system.
Melting points (not corrected) were determined on a Ko¯er-Reichert micro-hot stage apparatus.
For Zn(II)-1l, it could only be obtained under rapid heating; otherwise, decomposition occured, and
the product did not melt below 350ꢂC. UV/Vis spectra were recorded on a Varian Cary 5E
instrument. Fluorescence spectra were determined with an Aminco-Bowman Series 2 luminescence
spectrometer using air-saturated benzene or CH2Cl2 solutions of the porphyrins. Emission spectra
were recorded exciting at the maximum of the Soret band. Absorbances at the excitation wavelength
were optically matched (A ꢃ 0.1; concentrations ꢃ 0.7 mM). Fluorescence quantum yields (Èf) for
the free base porphyrins were calculated from the integration of the complete corrected ¯uorescence
spectrum and comparison to mesoporphyrin-IX dimethyl ester (1b) solutions of equal absorbance at
the excitation wavelength and taking the literature value (0.12) for Èf of this compound [16]. Èf
values for the Zn(II)-porphyrin complexes were obtained similarly by comparison to Zn(II)-
1
mesoporphyrin-IX dimethyl ester (Èf ˆ 0.05) [16]. H and 13C NMR spectra were run on a Varian
Gemini XL-200 (200.6 MHz) or a Varian Unity-300 (300 MHz) spectrometer using tetramethylsilane
(TMS) as internal standard. Immediately before sample preparation, the solvent (CDCl3) was ®ltered
through a short column of basic Al2O3 (Act. I, Merck), and the ®rst few drops were discarded. IR
spectra were determined in a Fourier-Transform Nicolet spectrophotometer. Mass spectra were run
on a Hewlett-Packard 5988A instrument equipped for FAB analysis with a Capillaritron Frasor (Xe)
or on a VG QUATTRO (Fisons) instrument. Nitrobenzyl alcohol (NBA) was used as matrix. High
resolution mass measurements were carried out with an AutoSpec-Q instrument using the FAB
technique at resolution 5000, and with NBA as matrix. The cesium gun energy was 20 KV.
All spectroscopic measurements were performed at room temperature.
Zinc(II)-deuteroporphyrin-IX dimethyl ester (Zn(II)-1a; C32H32N4O4Zn)
A modi®cation of the procedure previously reported for the conversion of mesoporphyrin-IX
dimethyl ester into its zinc(II) complex was used [12]. A pinky red solid (96% yield; one spot on
TLC (SiO2, CH2Cl2:MeOH ˆ 10:1; Rf ˆ 0.85) was obtained which was spectroscopically identi®ed
as the title compound.
M.p.: 253±256ꢂC (decomposition); 1H NMR (CDCl3, ꢃ, ppm): 9.50, 9.49, 9.31, 9.28 (4 Â s,
4H, ˆ Cꢂ,ꢁ,ꢀ,ꢃH-), 8.80, 8.67 (2  s, 2H, ˆ C2;4H-), 4.20, 4.12 (2  t, 4H, -CH2CH2CO2-), 3.66, 3.65
(2 Â s, 6H, -CO2CH3), 3.60, 3.51, 3.47, 3.34 (4 Â s, 12H, C1;3;5;8-CH3), 3.12, 3.06 (2 Â t, 4H,
-CH2CH2CO2-); 13C NMR (CDCl3, ꢃ, ppm): 173.6 (-CO2CH3), 148±138 (Cꢂ,ꢁ-pyrrole), 129.1,
128.0 ( ˆ C2;4H-), 100.5, 99.75, 97.0, 95.8 ( ˆ CmesoH-), 51.6 (-CO2CH3), 37.0 (-CH2CH2CO2-),
21.7 (-CH2CH2CO2-), 13.5, 11.5, 11.4 (C1;3;5;8-CH3); IR (KBr, ꢄ, cmꢁ1): 3440, 2915, 1730 (ester

Benzoylation of a Zinc(II) Deuteroporphyrin
77
C=O); UV/Vis (CH2Cl2): ꢅmax (") ˆ 400 (344000), 530 (14600), 566 (18000) nm; ¯uorescence
‡
(benzene): ꢅexc ˆ 400 nm, ꢅem ˆ 570, 623 nm; Èf ˆ 0.045; MS (FAB): m/z ˆ 604, 602, 600 (M
68Zn, 66Zn, 64Zn).
Zinc(II)-2,4-diacetyldeuteroporphyrin-IX dimethyl ester (Zn(II)-1l; C36H36N4O6Zn)
Zn(II)-1l was prepared by brief treatment of 2,4-diacetyldeuteroporphyrin dimethyl ester (1l) with
zinc(II) acetate as reported previously for the conversion of mesoporphyrin-IX dimethyl ester into its
zinc(II) complex [9]. A greenish red solid (quantitative yield; one spot on TLC (SiO2, CH2Cl2:
MeOH ˆ 10:1; Rf ˆ 0.83) was obtained.
M.p.: 258±262ꢂC (vide supra); 1H NMR (CDCl3, ꢃ, ppm): 10.45, 10.31, 9.6, 9.5 (4 Â s,
4H, ˆ Cꢂ,ꢁ,ꢀ,ꢃH-), 4.26, 4.24 (2  t, 4H, -CH2CH2CO2-), 3.77, 376, 3.66, 3.65, 3.50, 3.46 (6  s,
18H, C1;3;5;8-CH3,-CO2CH3), 3.26, 3.23 (2 Â s, 6H, CH3CO-), 3.16, 3.14 (2 Â t, 4H,-CH2CH2CO2-);
13C NMR (CDCl3, ꢃ, ppm): 199.5, 199.0 (-COCH3), 173.4 (-CO2CH3), 146±136 (Cꢂ,ꢁ-pyrrole),
104±101 ( ˆ CmesoH-), 51.7 (-CO2CH3), 36.6 (-CH2CH2CO2-), 33.1 (-COCH3), 21.55
(-CH2CH2CO2-), 14.2 11.6 (C1;3;5;8-CH3); IR (KBr, ꢄ, cmꢁ1): 3430, 2925, 2855, 1740 (ester
C=O), 1650 (ketone C=O); UV/Vis (CH2Cl2): ꢅmax (") ˆ 424 (91000), 550 (7500), 587 (7000) nm;
¯uorescence (benzene): ꢅexc ˆ 416 nm, ꢅem ˆ 597, 646 nm; Èf ˆ 0.03; MS (FAB): m/z ˆ 688, 686,
684 (M‡ 68Zn, 66Zn, 64Zn).
Benzoylation of Zn(II)-deuteroporphyrin-IX dimethyl ester (Z(II)-1a)
Zn(II) deuteroporphyrin-IX dimethyl ester (150 mg, 0.25 mmol) was heated to 70ꢂC in the presence
of an excess of benzoic anhydride (2.7 g, 12 mmol) in a 10 ml ¯ask equipped with a re¯ux condenser
and a CaCl2 tube. The anhydride melted and eventually dissolved the Zn(II)-porphyrin.
SnCl4(2.0 ml) was then added dropwise resulting in a gradual change in colour from purple to
dark green of the contents of the ¯ask which also solidi®ed. The mixture was reacted in the dark at
70ꢂC for a total of 30 min. After cooling to room temperature, CH2Cl2 (100 ml) was added, and the
mixture was washed with water (6 Â 50 ml), ®ltered through paper, and evaporated. A dark brown
solid was obtained from which benzoic anhydride benzoic acid were removed by repeated washing
with hexane. Subsequent column chromatography of the resulting mixture over basic Al2O3 afforded
four fractions. Fraction 1, eluted with CH2Cl2, corresponded to deuteroporphyrin dimethyl ester (1a;
38 mg; one spot on TLC; SiO2, CH2Cl2:MeOH ˆ 10:1; Rf ˆ 0.85). Fraction 2, eluted with
CH2Cl2:MeOH ˆ 500:1, consisted of a mixture of metal free, benzoylated porphyrins plus 1a
(66 mg; one spot on TLC; SiO2, CH2Cl2:MeOH ˆ 10:1; Rfꢄ0.8). Fraction 3, eluted with
CH2Cl2:MeOH ˆ 100:1, corresponded to a mixture of zinc(II) deuteroporphyrins with one to ®ve
benzoyl groups (see below; 16 mg in total; one long spot on TLC; SiO2, CH2Cl2:MeOH ˆ 10:1;
Rf ˆ 0.65±0.5). Fraction 4, eluted with CH2Cl2:MeOH ˆ 10:1, was not further investigated (11 mg;
several spots on TLC; SiO2Cl2:MeOH ˆ 10:1; Rf ˆ 0.4±0.25).
Fraction 2 was dissolved in ether and washed repeatedly with 2% (w/w) aqueous HCl. The acid
retained 1a, and the ether fraction showed ®ve main peaks by HPLC with Rt in min (in brackets:
relative integration): 6.59 (6%), 7.00 ( ‡ 7.6, shoulder; 4.6%), 7.93 (34%), 11.67 (8.4%), (12.3,
shoulder ‡ ) 12.55 (45.7%). UV/Vis spectra in the mobile phase: 395, 486, 517, 555 nm (Rt ˆ 6.59);
399, 490, 526, 560 nm (Rt ˆ 7.00); 403.5, 495, 529, 564 nm (Rt ˆ 7.93); 405, 495, 527, 564 nm (Rt ˆ
11.67); and 407, 497, 528, 565 nm (Rt ˆ 12.55). Under identical conditions, 1a has Rt ˆ 5.71 min and
UV/Vis: 389, 481, 509, 555 nm. MS (FAB): m/z : 747 (53%), 643 (100%) (Rt ˆ 6.59); 747 (79%),
643 (100%) (Rt ˆ 7.00); 851 (30%), 747 (100%), 643 (75%) (Rt ˆ 7.93); 851 (69%), 747 (100%),
1
643 (48%) (Rt ˆ 11.67); and 851 (100%), 745.5 (70%) (Rt ˆ 12.55). H NMR (CDCl3, ꢃ, ppm): for
all the components, except that with Rt ˆ 12.3 (below), spectra with a larger number of peaks than
expected for pure compounds were obtained.

Â
J. Vazquez et al.
78
2,4,ꢀ-Tribenzoyldeuteroporphyrin-IX dimethyl ester (1k; C53H46N4O7)
Isolated from fraction 2 (see above) by preparative HPLC (Rt ˆ 12.3 min in analytical HPLC).
M.p.: 105±108ꢂC; H NMR (CDCl3, ꢃ, ppm): 10.20, 10.19, 10.17 (3  s, 3H, ˆ Cꢂ,ꢁ,ꢃH-), 8.2±
1
7.1 (m, 15H, ˆ C2;4H-), 4.43, 4.25 (t ‡ dt, 4H, -CH2CH2CO2-), 3.65, 3.64, 3.61, 3.59 (4  s, 12H,
C
5;8-CH3,-CH2CH3), 3.30, 3.09 (2 Â t, 4H, -CH2CH2CO2-), 3.13, 2.86 (2 Â s, 6H, C1;3-CH3), ꢁ2.99,
ꢁ3.04 (2 Â s, 2H, NH); IR (KBr, ꢄ, cmꢁ1): 3429 (C-H), 2990, 2929, 1752 (ester C=O), 1652
(benzoyl C=O), 1256, 1177, 1160, 1098, 1017; UV/Vis (CH2Cl2): ꢅmax (") ˆ 418 (138600), 512
(12700), 547 (6600), 583 (5800), 635 (2100) nm; ¯uorescence (benzene): ꢅexc ˆ 416 nm; ꢅem ˆ 644,
710 nm; Èf ˆ 0.04; high resolution MS (FAB): m/z ˆ 851.3446 (calculated: 851.3445; error: ꢁ0.2
‡
ppm), corresponds to an elemental composition of C53H47N47O7 (M ‡ 1).
Fraction 3
1H NMR (CDCl3, ꢃ, ppm): 10.28, 10.21, 10.15, 10.10, 10.08 ( ˆ CmesoH-), 8.0±6.8 (PhCO-), 4.42,
4.40 (-CH2CH2CO2-), 3.71±3.62, 2.95 (8 Â s, Cꢁ-CH3, -CO2CH3), 3.41, 3.39 (-CH2CH2CO2-), 2.56
(not assigned): 13C NMR (CDCl3, ꢃ, ppm): 199.55, 199.18 (-COPh), 173.63 (-CO2CH3), 150±138
(Cꢂ,ꢁ-pyrrole), 133.0, 129.9, 128.2 (PhCO-), 100±98 ( ˆ CmesoH-), 51.76 (-CO2CH3), 37.06, 36.89
(-CH2CH2CO2-), 21.97, 21.81 (-CH2CH2CO2-), 15.19, 11.74 (Cꢁ-CH3); IR (KBr, ꢄ, cmꢁ1): 3440
(C-H), 2925, 1737 (ester C=O), 1665 (benzoyl C=O), 1260; UV/Vis (CH2Cl2); ꢅmax (") ˆ 418
(128000), 545 (10000), 581 (7600) nm; ¯uorescence (benzene): ꢅexc ˆ 416 nm; ꢅem ˆ 595 nm; Èf ˆ
0.04; MS (FAB): m/z ˆ 915.4, 914.5, 913.6, 912.8 (100%), 810.5, 809.5, 808.6, 807.7 (100%), 706.6,
705.6, 704.6, 703.6 (25%).
In one run, this fraction (10 mg) dissolved in CH2Cl2 (100 ml) was demetallated with 3 M
aqueous HCl (50 ml) during 2 h at room temperature under argon. The solution was then neutralized
with 0.5 M aqueous NaHCO3, washed with water, and dried. The recovered sample was esteri®ed
with CH2N2 in ether for 2 min at RT, puri®ed by PTLC (SiO2, CH2Cl2:MeOH ˆ 10:1), and the
fraction with Rfꢄ0.4 was analyzed.
MS (FAB): m/z ˆ 641.6 (26%), 745.8, 746.9 (77%), 851.6 (100%), 955.7 (7.4%), 1058.9
(27.5%); UV/Vis (CH2Cl2): ꢅmax ˆ 414, 509, 545, 579, 634 nm.
Acknowledgements
This work has been supported by a grant (QFN 94-4613-CO2) from the CICYT and the Generalitat
Â
de Catalunya. We thank Dr. M. Pericas (Univ. Barcelona) for allowing us to use his work station.
Special thanks are due to Prof. S. E. Braslavsky (Max Planck Insitut fuÈr Strahlenchemie, MuÈlheim
a.d. Ruhr, Germany) for help with the HPLC separation and subsequent spectroscopic analysis. C.M.
Â
thanks the spanish Ministerio de Educacion y Cultura for a fellowship.
References
[1] Fuhrhop J-H (1979) In: Dolphin D (ed) The Porphyrins, vol 2. Academic Press, New York, p 131
[2] Smith KM (1975) Porphyrins and Methalloporphyrins. Elsevier, Amsterdam, Holland
[3] (a) Matsumoto T, Sato M, Hirayama S (1972) Chem Phys Lett 13: 13; (b) Bensasson R, Land EJ
(1977) Nouveau Journal de Chimie 2: 503
[4] Fischer H, Orth H (1937) Die Chemie des Pyrrols, vol I. Akademische Verlagsgesellschaft,
Leipzig, p 297
[5] Caughey WS, Alben JO, Fujimoto WY, York L (1966) J Org Chem 31: 2631
[6] Inhoffen HH, Bliesener C, Brockmann Jr H (1967) Tetrahedron Lett 727
[7] (a) Inhoffen HH, Fuhrhop J-H, Voigt H, Brockmann Jr H (1966) Liebigs Ann Chem 695: 133; (b)
Old®ed D, Johnson AW (1966) J Chem Soc C 794

Benzoylation of a Zinc(II) Deuteroporphyrin
79
[8] (a) Fischer H, Orth H (1937) Die Chemie des Pyrrols, vol I. Akademische Verlagsgesellschaft,
Leipzig, p 230; (b) Gradyushko AT, Kozhich DT, Sokolev KN, Tsvirko MP (1974) Opt Spectrosc
37: 510; (c) Treibs A (1952) Naturwissenschaften 39: 281; (d) Minnetian OM, Morris IK, Snow
KM, Smith KM (1989) J Org Chem 54: 5567; (e) Bonnett R, Campion-Smith IH, Kozyrev AN,
Mironov AF (1990) J Chem Res (M) 1015
[9] Dinello RK, Chang CK (1979) In: Dolphin D (ed) The Porphyrins, vol 1. Academic Press, New
York, p 294
[10] (a) Reeves PC (1988) Organic Syntheses Coll VI: 625; (b) Gore PH, Hoskins JA (1964) J Chem
Soc 5666; (c) Kulka M (1954) J Am Chem Soc 76: 5469
[11] Fukui K, Yonezawa T, Nataga C (1957) J Chem Phys 26: 831
[12] Fuhrhop J-H, Smith KM (1975) In: Smith KM, Porphyrins and Metalloporphyrins. Elsevier,
Amsterdam, Holland, p 798
[13] Perrin DD, Armarego WLF, Perrin DR (1988) Puri®cation of Laboratory Chemicals, 2nd edn.
Pergamon Press, New York
[14] Spartan version 4.1.1, Wavefunction, Inc. 18401, Von Karman Ave. 370. Irvine, CA 92715 USA,
1995
[15] (a) Stewart JJP (1989) J Comp Chem 10: 209; (b) Stewart JJP (1989) J Comp Chem 10: 221
[16] Darwent JR, Douglas P, Harriman A, Porter G, Richoux M-C (1982) Coord Chem Rev 44: 83
Received June 23, 1997. Accepted (revised) July 23, 1997