Green self-assembling porphyrins and chlorins as mimics of the natural bacteriochlorophylls c, d, and e

Article abstract of DOI:10.1002/ejoc.200400151

Novel porphyrins and chlorins that self-assemble in nonpolar solvents in a manner similar to that of the bacteriochlorophylls c, d, and e have been synthesized by a common protective group approach. The supramolecular assemblies have broad and red-shifted absorption spectra in comparison to those of the monomeric building blocks. The presence of a carbonyl group in conjugation with the tetrapyrrolic macrocycle produces green colors both in the free bases and in their zinc complexes, which, after self-assembly, are thus perfect artificial mimics of the chlorosomal antennas encountered in green photosynthetic bacteria. Enantiopure building blocks produce large helical aggregates with M or P helicity determined by the chirality of the 1-hydroxyethyl substituent. It is demonstrated that the groups essential for self-organization to occur - namely the hydroxy group, the zinc metal atom and the carbonyl group - do not have to be collinear, as has been presumed until now. Surprisingly, the chlorin derivative does not show hyperchromicity relative to similarly substituted porphyrins. This fact allows us to conclude that the more readily available porphyrins may be used for efficient artificial antennas of potential use in solar devices; otherwise it is necessary to increase the number of synthetic steps in order to incorporate into chlorins the annulated cyclopentanone ring, a structural feature carefully optimized by evolution in all (bacterio) chlorophylls. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2004.

Full text of DOI:10.1002/ejoc.200400151

FULL PAPER
Green Self-Assembling Porphyrins and Chlorins as Mimics of the Natural
Bacteriochlorophylls c, d, and e
Teodor Silviu Balaban,*[a] Myriam Linke-Schaetzel,[a] Anil Dnyanoba Bhise,[a]
Nicolas Vanthuyne,[b] and Christian Roussel[b]
Keywords: Self-assembly / Bacteriochlorophyll / Porphyrinoids / Light-harvesting / J-aggregates / Circular dichroism /
Chlorosome
Novel porphyrins and chlorins that self-assemble in nonpolar
ganization to occur − namely the hydroxy group, the zinc
solvents in a manner similar to that of the bacteriochloro- metal atom and the carbonyl group − do not have to be collin-
phylls c, d, and e have been synthesized by a common pro- ear, as has been presumed until now. Surprisingly, the chlorin
tective group approach. The supramolecular assemblies have
broad and red-shifted absorption spectra in comparison to
those of the monomeric building blocks. The presence of a
derivative does not show hyperchromicity relative to sim-
ilarly substituted porphyrins. This fact allows us to conclude
that the more readily available porphyrins may be used for
carbonyl group in conjugation with the tetrapyrrolic macro- efficient artificial antennas of potential use in solar devices;
cycle produces green colors both in the free bases and in
their zinc complexes, which, after self-assembly, are thus
perfect artificial mimics of the chlorosomal antennas encoun-
tered in green photosynthetic bacteria. Enantiopure building
blocks produce large helical aggregates with M or P helicity
determined by the chirality of the 1-hydroxyethyl substitu-
otherwise it is necessary to increase the number of synthetic
steps in order to incorporate into chlorins the annulated
cyclopentanone ring, a structural feature carefully optimized
by evolution in all (bacterio)chlorophylls.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
ent. It is demonstrated that the groups essential for self-or- Germany, 2004)
Introduction
with robust and easily available pigments, for the goal of
artificial photosynthesis.
The ubiquitous green color of light-harvesting apparata
is encountered in (bacterio)chlorophyll-based photosyn-
thetic organisms. In conjunction with other pigments, such
as carotenoids, the light absorption characteristics, which
include broad wavelength ranges and high extinction coef-
ficients, have been optimized during evolution in order to
ensure conversion of light into biochemical energy by adap-
tation to different habitats. Thus, bacteria that live under
the water surface at depths of over 50 m have evolved
antenna systems different to those of bacteria or algae liv-
ing at the surface, and these differ in turn from terrestrial
plants in their light-harvesting systems. While the last, more
highly evolved, of these species have developed protein
complexes to bind chromophores, in the early green photo-
synthetic bacteria, due to the pressures of synthetic and
genetic economy, self-assembly of bacteriochlorophylls c, d,
and e (Figure 1) is used.[1,2] This much simpler architectural
construct, which is fully functional, is worth mimicking
Through the use of different spectroscopic techniques,
we[3] and others[4] were involved fairly early on in the inves-
tigation of this intriguing self-assembly, although in the ab-
sence of definite crystallographic proof, structural details
are still a matter of debate. Essential for the self-assembly
algorithm to operate correctly are: (i) the 3-(1-hydroxyethyl)
group (which may have either R or S stereochemistry), (ii)
the 131-carbonyl group in the annulated five-membered
ring, (iii) the central magnesium atom (although this may
be replaced by other divalent diamagnetic metal ions such
as zinc or cadmium), and (iv) the πϪπ interactions between
the conjugated macrocycles. The close but ordered interac-
tion between the chromophoric systems gives rise to broad-
ened and red-shifted absorption maxima, which are ben-
eficial for light-harvesting and, forms strongly fluorescent
aggregates which can pass on the radiant energy to a trap
from where the reaction center can be triggered for charge
separation. The fluorescence quantum yields usually enco-
untered with highly aggregated pigments are only very low,
due to concentration quenching, a fact that hampers their
use as photosenzitisers in photovoltaic cells.
[a]
Forschungszentrum Karlsruhe, Institute for Nanotechnology,
Postfach 3640, 76021 Karlsruhe, Germany
Fax: (internat.) ϩ 49-(0)724-782-8298
E-mail: silviu.balaban@int.fzk.de
[b]
´
Universite Aix-Marseille III, UMR: Chirotechnologie: catalyse
Here we present novel, fully synthetic pigments Ϫ namely
a zinc porphyrin 10-Zn and a zinc-chlorin 11-Zn Ϫ capable
et biocatalyse,
Centre de Saint Jerome, 13397, Marseille, Cedex 20, France
Eur. J. Org. Chem. 2004, 3919Ϫ3930
DOI: 10.1002/ejoc.200400151
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3919

T. S. Balaban, M. Linke-Schaetzel, A. D. Bhise, N. Vanthuyne, C. Roussel
FULL PAPER
(originating either from an achiral hydroxymethyl or from
a 1-hydroxyethyl group), the zinc atom, and the carbonyl
group arranged collinearly. A very intriguing result is that
the isomeric zinc porphyrin 18-Zn failed to give red-shifted
and broad absorption spectra upon self-assembly, which led
the authors to speculate that collinearity of these three
groups is an essential condition for the self-assembly to oc-
cur.[8] As this observation was at odds with the fact that
both 15-Zn and 16-Zn could self-assemble, we have now
prepared 10-Zn, with an angular disposition of these
groups. We observed that it is the amount of π-overlap of
the macrocycles in the self-assembled species that is respon-
sible for the broadening and the red-shift of the absorption
maxima, so the collinearity mentioned above is not a sine
qua non for induction of self-organization.
Furthermore, we have now also synthesized the chlorin
11-Zn, which because of its reduced symmetry was expected
to display more intense Q-absorption bands than porphy-
rins, in a quest for an optimum self-assembling chromoph-
oric system for hybrid solar cells.[6] These are mixed or-
ganic-inorganic solar cells in which an antenna system ad-
ditionally assures a large photon capture cross section for
improved light-to-current conversion efficiencies under low
light illumination conditions.
Figure 1. Self-assembling bacteriocholorphylls encountered in the
light-harvesting organelles known as chlorosomes (green sacs) of
green photosynthetic bacteria
Results and Discussion
Syntheses: Scheme 1 presents the synthetic transform-
ations, starting from the known copper porphyrin 1[10] with
of self-assembly in similar manner to the bacteriochloro- two meso 3,5-di-tert-butylphenyl groups[11] and making use
phylls c, d, and e possessing the same functional character- of 2,2-dimethyl-1,3-propanediol to protect meso-formyl
istics (i-iv). In a preliminary communication we described groups, to enable the reduction both of a 3-acetyl group
the structurally very similar porphyrins 12-ZnϪ16-Zn (Fig- and of the porphyrin to the chlorin. While the isolated in-
ure 2), which were recently synthesized in our group and termediates in the synthetic sequence of 10-Zn are rep-
showed intense fluorescence after self-assembly.[5] Further- resented by even numbers, odd numbers are used for inter-
more, these could also be induced to self-assemble onto mediates in the reaction steps leading to chlorin 11-Zn.
nanocrystalline titania.[5,6] These were the first fully syn- Monoformylation of 1 gave 2, which was subsequently ace-
thetic pigments capable of self-assembly similarly to that tylated in a novel FriedelϪCrafts type of acylation. After
displayed by the natural (chlorosomal) bacteriochloro- demetallation, the more reactive aldehyde group in 4 was
phylls. Tamiaki and co-workers had previously synthesized protected in order to allow selective monoreduction of the
several self-assembling mimics, but these, being derived acetyl group. Removal of the protective group and final zinc
either from chlorophyll a or from bacteriochlorophyll a, are remetallation afforded 10-Zn in a remarkable global yield
products of semisyntheses and thus involved rather tedious of 43% over eight steps. For the synthesis of chlorin 11-
isolation and purification steps.[7] Since our initial report,[5] Zn, diformylation of 1 was performed, followed by double
Tamiaki’s group, using an elegant synthesis, have im- protection of the two meso-formyl groups in order to permit
plemented the same design principles to obtain the self-as- selective reduction of the porphyrin to the chlorin. Again,
sembling octaethylporphyrin-derived porphyrin 17-Zn.[8] removal of the protective groups (vide infra) and final zinc
The authors rightly argue that 17-Zn, possessing only alkyl metallation provided the desired chlorin 11-Zn, albeit in a
substituents, is more closely related than the meso-aryl-sub- much lower global yield due to the simultaneous transform-
stituted porphyrins 12-ZnϪ16-Zn to the natural BChls. If ations in both the northern and southern part of the macro-
practical applications are envisaged,[9] however, our com- cycle.
pounds require fewer synthetic steps and so are more easily
Like our previous porphyrins 12Ϫ16-Zn[5] and Tamiaki’s
available, although octaethylporphyrin is a commercially octaethylporphyrin-based compound 17-Zn,[8] 10-Zn and
available compound. As judged from the broad, red-shifted 11-Zn are fully synthetic mimics of the natural bacterio-
absorption spectra, structurally very similar self-assembly chlorophylls and are capable of self-assembly. Our synthetic
patterns were obtained from 12-Zn, 13-Zn, 14-Zn, and 17- concept is to use preformed porphyrins, which are available
Zn.[5,8] These four compounds each have the hydroxy group in multigram quantities, and, by means of selective trans-
3920
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3919Ϫ3930

Green Self-Assembling Porphyrins and Chlorins as Mimics of the Natural Bacteriochlorophylls c, d, and e
FULL PAPER
Figure 2. Synthetic mimics of the self-assembling BChls c, d, and e
formations, to equip these with groups programming a self-
Natural chlorins have been shown to photosenzitize nan-
assembly algorithm.[12] The final structure, and its proper- ocrystalline titania effectively,[14] and Lindsey’s group has
ties are encoded in the architecture of the monomeric build- recently developed rational synthetic routes to chlorins.[15]
ing block, or tecton.[13]
We have applied the protective group strategy shown in
Figure 3 shows the NMR spectra for the monomeric Scheme 1 to allow reduction of the porphyrin followed by
methanol adducts of the novel 10-Zn porphyrin in compari- unmasking of the now nonequivalent formyl groups. Sev-
son with the isomeric compound 15-Zn, which nicely reflect eral protecting strategies were tried [bis(m)ethyl acetals or
the structural differences between the two isomers. It is 1,3-dioxolane], but only the gem-dimethyl-substituted diox-
interesting to note that the aromatic signals of the two non- ane group, also introduced to porphyrin chemistry by Lind-
equivalent 3,5-di-tert-butylphenyl groups appear well sepa- sey,[16] gave satisfactory results. The porphyrin reduction
rated at 300 MHz in the case of 15-Zn but are almost super- works smoothly with tosylhydrazide[17] for all protective
imposed for Zn-10. This is probably due to the influence of groups tried, but the deprotection step requires gentle con-
the carbonyl group, which makes the meso-aryl groups dif- ditions under which only the Lindsey 1,3-dioxane is cleaved
fer more when it is present in the 13-position (as an acetyl properly. In this case we were able to suppress the compet-
group) in Zn-15 than when it is present in the 15-meso- ing ‘‘decarbonylation’’ reaction in which the formyl pro-
position (as a formyl group) in Zn-10.
tected group is replaced by a meso-hydrogen atom. This ser-
Eur. J. Org. Chem. 2004, 3919Ϫ3930
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3921

T. S. Balaban, M. Linke-Schaetzel, A. D. Bhise, N. Vanthuyne, C. Roussel
FULL PAPER
Scheme 1. Reagents and conditions: a: POCl3/DMF, ClCH2CH2Cl, 40 °C, 45 min, 83%;[5] b: POCl3/DMF, ClCH2CH2Cl, 60 °C, 14 h,
43%;[5] c: TFA/H2SO4 (1:1 v/v), Ͼ85%;[5] d: Ac2O, SnCl4/CS2, 0 °C, 3 min, 80% (after demetallation as in c);[5] e: 2,2-dimethyl-1,3-
propanediol, PTSA, toluene, ∆, 65 min Ϫ6 h, 75Ϫ96%; f: NaBH4, MeOH, CH2Cl2, room temp., 0.5Ϫ19 h, 50Ϫ72%; g: HCl, 1,4-dioxane,
Ar, room temp., 17 h, 94%; h: Zn(OAc)2, CHCl3/MeOH (5:3, v/v), room temp., 3 h, 93Ϫ95%; i: TsϪNHϪNH2, K2CO3, β-picoline, ∆,
15 h, 55%; j: TiCl4, CHCl3, room temp., 15 min, 66%.
3922
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3919Ϫ3930

Green Self-Assembling Porphyrins and Chlorins as Mimics of the Natural Bacteriochlorophylls c, d, and e
FULL PAPER
Figure 3. Low-field region of the 300 MHz NMR spectra of 10-Zn (lower trace) and 15-Zn (upper trace) in CDCl3:CD3OD 10:1 (v/v).
Assignments are based on COSY and NOESY spectra; insets show the 3-CHOHCH3 signals (lower trace) and the 13-COCH3 singlet
(upper trace); the peaks indicated by asterisks do not belong to the sample, being either 13C-satellites and spinning side bands of the
CHCl3 signal or a trace of acetone.
endipitously found reaction sequence might prove to be of LH2 and LH3 light-harvesting complexes of purple bac-
some synthetic interest for selective removal of meso-substi- teria, which have been characterized by X-ray crystallogra-
tuted formyl groups, but this was not in our interest in the phy,[20,21] conjugation of an acetyl group in a β-pyrrolic po-
current study.
sition in a chlorin ring (in bacteriochlorophyll a) can be
Remarkable in our case is the potential to reduce the 5- finely tuned by the protein matrix. This shifts the absorp-
formyl group in 9 selectively, leaving the 15-formyl group in tion spectrum from 850 nm for the B850 ring in LH2 to
the 17,18-dihydroporphyrin intact.[18] With excess reagent 820 nm in LH3, just through titling of the acetyl group out
and prolonged reaction times, the 15-formyl group starts to of conjugation (dihedral angles of 15° in LH2 and 43° in
become reduced. We were not able to separate the other LH3). This has profound implications for the light-har-
isomeric formyl-hydroxymethylene-chlorin isomer cleanly vesting ability of these bacteria, which produce LH3 com-
from these reaction mixtures as the dihydroxymethylene plexes only under low light illumination conditions.
compound is also formed concomitantly. As in the case of
UV/Vis Absorption Properties of the Assemblies: That
the related dicarbonyl porphyrins,[5] it is interesting to note self-assembly is fully operational in 10-Zn can be seen in
that the possibility of isolating the monoreduced hydroxy- nonpolar solvents by the broad and red-shifted absorption
oxo compounds in relatively high yields is due to the fact maxima of both the Soret and Q bands (Figure 4, part A)
that one carbonyl group activates the other. In the case of and by the strong Rayleigh light scattering. Also presented
symmetrical compounds (i.e., equivalent carbonyl groups), here for the first time, for comparison, are the correspond-
after the reduction of the first CO group, the second is re- ing absorption spectra of the isomeric compound 15-Zn
duced much more sluggishly, accounting for this unexpec- (Figure 4, part B). This allows direct comparison between
ted product selectivity, With nonequivalent CO groups, as the achiral hydroxymethylene group and the hydroxyethyl
is the case in 9, the less sterically hindered one (i.e., the 5- group as ligands for the zinc atoms in the self-assembled
CHO group) is reduced more rapidly.
structures. Complete disassembly is ensured by addition of
An unhindered meso-carbonyl group is not only more re- minute (but suprastoichiometric) amounts of a competing
active but also produces a green color not only in chlorins solvent, such as methanol, for Zn-ligation, which is also
but even in the free base porphyrins.[19] It is thus interesting expected to disrupt any intermolecular hydrogen bonding
to note that while a meso-formyl group situated in the por- (Figure 4, dotted traces). It is evident from the spectra that
phyrin plane gives green colored free base porphyrins, a almost no monomeric 15-Zn (shoulder at 405 nm) is pre-
meso-acetyl group is forced out of conjugation with the sent, whereas for 10-Zn there are considerable amounts of
macrocycle, due to buttressing of the methyl group by the β- monomers present (sharp Soret band at 420 nm), in ad-
pyrrolic protons. This translates into a hypsochromic shift, dition to self-assembled species characterized by the broad
resulting in the ‘‘normal’’, purple colored free base and zinc and red-shifted absorptions. Three bands are diagnostic for
metalated porphyrins such as 13-Zn.[5] An acetyl group in the self-assembly, one corresponding to the Soret band and
an unhindered β-pyrrolic position, however, can adopt two smaller ones corresponding to the Qx and Qy bands
planar conformations and thus again give rise to green col- and indicated by arrows in Figure 4. Titration of MeOH
ored compounds such as 14-Zn, 15-Zn, or 16-Zn. In the indicates that a monomeric ZnϪMeOH adduct is formed,
Eur. J. Org. Chem. 2004, 3919Ϫ3930
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3923

T. S. Balaban, M. Linke-Schaetzel, A. D. Bhise, N. Vanthuyne, C. Roussel
FULL PAPER
Figure 5. Stacking interactions in 10-Zn (left) and 15-Zn (right);
R ϭ 3,5-di-tert-butylphenyl; only two porphyrin rings are shown
in each case, together with the directions of the Qy transition dipole
moments (arrows); the geometries were obtained by successive opti-
mizations with the semiempirical force-field PM3 and molecular
[23]
mechanics (MMϩ) by use of the HyperChem
program pack-
age and should be regarded only as a crude approximation
molecular models shown in Figure 5. Subtle variations in
the π-overlap and/or optical properties of the assemblies
can be engineered by varying the positions of attachment
to the porphyrin macrocycle of the anchoring and carbonyl
groups. We can rule out formation of closed dimers by
double coordination of the zinc atoms by the hydroxy-
(m)ethyl groups, which would also lead to parallel but op-
posing Qy transition dipole moments and should thus be
blue-shifted. More probably, the large assemblies are
formed of extended stacks of different lengths, as implied
by the dotted lines in Figure 5. The different lengths ac-
count for the heterogeneity and thus the broad absorption
maxima.
The temperature dependence of the absorption spectra
for 10-Zn shows that gradual disassembly (vide supra) oc-
curs upon heating (Figure 6). In heptane, even at 95 °C, a
considerable amount of aggregates still persist but their size
has been trimmed as observed from the step-wise decrease
in light scattering. Only relative large aggregates and small
Figure 4. Self-assembled 10-Zn (50 µM, part A) and 15-Zn (11 µM,
part B) in dry n-heptane at room temperature (full traces); the
dotted traces are for the same samples after addition of supra-
stoichiometric amounts of methanol; path-length was 0.5 cm for
all traces; for the 10-ZnϪMeOH adduct (part A, dotted trace) the
solution was diluted with an equal volume of n-heptane
its Soret band being slightly red-shifted by ഠ8 nm in re- units, which could be, for instance, either monomers or di-
lation to the non-ligated monomers. It is, however, instruc- mers, seem to be present in equilibrium.
tive to compare the wavelength difference between the ag-
No quantitative description of the aggregation is yet pos-
gregate bands and those of these methanol adducts: sible, due to the uncertain size of the aggregates. We there-
∆λmax. ϭ 52 nm and 22 nm for the corresponding Soret fore limit ourselves here to a qualitative description, show-
bands of 10-Zn and 15-Zn, respectively. The two Q bands ing not only that the temperature decreases the aggregation
have different intensities because of the different transition tendency (cf. Figure 6), but also that the concentration
probabilities in these two desymmetrized porphyrins. Inter- plays an important role (inset in Figure 6); a critical nucleus
estingly, in the self-assembled species, the most strongly red- must be formed so that aggregation proceeds; that is, more
shifted Q band is also the more intense one, as in the 10- molecules can become part of the aggregate more rapidly
ZnϪMeOH adduct. In 15-Zn the red-shifted aggregate than others are being disassembled.
band at 630 nm is also more intense, but a reversal occurs
The above examples clearly show that a linear arrange-
for the 15-ZnϪMeOH adduct. This implies a ∆λmax.
ϭ
ment of the zinc-chelating hydroxy group and the carbonyl
75 nm for this transition, which allows us to conclude that group is not necessary for induction of the self-assembly
very similar self-assembled structures are actually obtained algorithm. It is thus surprising that Tamiaki’s compound
in both cases, giving rise to red-shifted spectra as for J- 18-Zn failed to show self-assembly. In our opinion sterical
aggregates.[22] The anchoring group is hydroxy(m)ethyl, hindrance by interlocked ethyl groups cannot be invoked,
which binds a zinc atom of a neighboring porphyrin, while as the isomeric compound 17-Zn self-assembles readily.
the direction of the Qy transition is determined by the car- Rather, we suspect that the concentration range studied was
bonyl group, which must thus have slightly different relative too dilute for red-shifted absorption maxima to be detect-
orientations in the two supramolecules, as indicated by the able and that stack formation must also occur, although less
3924
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3919Ϫ3930

Green Self-Assembling Porphyrins and Chlorins as Mimics of the Natural Bacteriochlorophylls c, d, and e
FULL PAPER
stituents. This fact again supports the notion that the main
interaction is the zinc coordination by the hydroxy substitu-
ent, and that this determines the amount of π-stacking in-
teractions after interdigitation of the peripheral substitu-
ents. We are currently pursuing single-crystal growth and X-
ray diffraction studies on several of our compounds, which
hopefully will shed light on the details of the self-as-
sembled architectures.[24]
Figure 8 presents the absorption spectra of the chlorin
11-Zn, both after self-assembly and as the monomeric
methanol adduct. Similarly to the case of the porphyrin 10-
Zn, several broad, almost featureless bands (indicated by
the arrows), are encountered for the self-assembled species.
However, more transitions are clearly discernible. This is
due to the more strongly desymmetrized chlorin and to con-
siderable variations in the size of the aggregates, which re-
sult in inhomogeneous broadening. The broad Soret bands
in the aggregates almost overlap with the Qx bands in the
monomers or dimers. A broad wavelength absorption range
is advantageous, as it allows a larger portion of the solar
spectrum to be used for light-harvesting.
Figure 6. Variable temperature absorption spectra of self-assembled
(rac)Ϫ10-Zn in n-heptane (70 µ); the 1-cm path-length cuvette
was heated in 10° increments and the trace at 65 °C is not shown;
note the gradual decrease in the scattering contribution, which usu-
ally produces a shift of the baseline; Inset Ϫ concentration depen-
dence of the aggregation at 25 °C; traces have been scaled vertically
at the Soret maximum of the monomer by multiplication with the
following scaling factors (from the top trace to the bottom one): 1,
1.6, 2.3, 3.3, 13.2; the samples were prepared by dilution of a
3.7 m CH2Cl2 solution of 10-Zn with 2 mL dry n-heptane and
had the following concentrations calculated for monomeric species:
70, 40, 30, 20, and 10 µM
readily, with 18-Zn. In support of this hypothesis, here we
present the absorption spectra for 16-Zn, self-assembled at
two different concentrations, and after disassembly with
methanol (Figure 7). While the 478, 510, and 640 aggregate
maxima are barely discernible in the more dilute samples,
they are clearly evident at higher concentration. The con-
tent of monomers is much higher in 16-Zn, however, and so,
as with Tamiaki’s compound 18-Zn, self-assembly, although
possible, is somewhat inhibited. The bulky 3,5-di-tert-bu-
tylphenyl groups present in our case are surely much more
sterically demanding than the readily rotating ethyl sub-
Figure 8. UV/Vis spectra of the self-assembled chlorin Zn-11 in dry
n-heptane. Dotted trace is after disassembly with methanol (c ϭ 17
µ); both spectra were recorded at room temperature with a path-
length of 1 cm
Another important observation can be made from the
absorption spectra in Figure 4 and 8 on the basis of the
extinction coefficients of the chlorin 11-Zn in comparison
to the porphyrins 10-Zn either as the racemate or as sepa-
rated enantiomers (vide infra). While for the Q band the
logε value is practically the same (4.16 versus 4.19), the So-
ret band is markedly decreased, by a factor of almost 3, in
the chlorin (logε ϭ 5.01) in relation to the porphyrin
(logε ϭ 5.47). Contrary to expectations based on nonquan-
titative absorption data, simple self-assembling chlorins, al-
though they do present broad absorptions, thus allowing a
large wavelength palette of the solar spectrum to be har-
vested (Figure 8), should not be superior to porphyrins as
artificial antenna systems. In (bacterio)chlorophylls, while
the Soret band is as intense as in our chlorin (logε ϭ 5.01
in diethyl ether), the Q-band is markedly increased by a
factor of 5.9 (logε ϭ 4.93 versus 4.16).[25] It thus appears
that the cyclopentanone moiety in the natural chlorophylls
Figure 7. Absorption spectra of Zn-16 under conditions similar to
those of Figure 4 and 6, and with a path-length of 0.5 cm; dotted
lines after disassembly with methanol; inset: a five times more con-
centrated sample (50 µ) exhibiting more self-assembled species,
which give the red-shifted absorption maxima indicated by the ar-
rows
Eur. J. Org. Chem. 2004, 3919Ϫ3930
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3925

T. S. Balaban, M. Linke-Schaetzel, A. D. Bhise, N. Vanthuyne, C. Roussel
FULL PAPER
is an important structural element carefully optimized by
evolution. Although excellent synthetic methodologies for
annulation of such oxo-five-membered rings to porphyrins
and chlorins have been developed, these involve several ad-
ditional synthetic steps.[26] This structural element also for-
ces the 131-carbonyl group into co-planarity with the tetra-
pyrrolic macrocycle.
CD-Spectroscopy and the Supramolecular Chirality of
Self-Assembly: The separated enantiomers of the free base
10, after zinc metallation and self-assembly, give helical
superstructures which display giant (or PSI-type[27]) circular
dichroism (CD) with mirror imaged exciton couplets in the
region of the red-shifted Soret aggregate band. From the
positive or negative Cotton effect appearing at the longer
wavelength, a clockwise or counterclockwise disposition of
the transition dipole moments[28] in the π-stacked porphy-
rins can be assigned to the assemblies formed from the en-
antiomers eluted first and second, respectively (Figure 9).
Addition of polar solvents such as methanol again leads to
disassembly, which is accompanied by the disappearance of
the intense CD signals and a markedly increased, blue-
shifted Soret absorption band (Figure 10).
Figure 10. UV/Vis spectra of self-assembled 10-Zn; red trace Ϫ
first eluted enantiomer of 10 after metallation with zinc acetate and
self-assembly in n-heptane (28 µ solution); blue trace Ϫ second
eluted enantiomer of 10 after Zn insertion and self-assembly in n-
heptane (22 µ solution); the red trace represents more self-as-
sembled species with maxima at 472 and 635 nm than the blue
trace; the spectra were recorded at room temperature with a path-
length of 1 cm; dotted traces are after disassembly with methanol;
the red dotted trace was recorded with a 0.1-cm path length, while
the blue dotted trace was recorded with a 0.5-cm path length; note
the much smaller ratio between the aggregate maxima and the
bands of the monomers in comparison with the rac-10-Zn, shown
in Figure 4, part A and in Figure 6
strongly dependent on the initial concentration of the
monomers. It has also recently been shown,[29] and theoreti-
cally explained,[30] that the CD signals are also strongly de-
pendent on the size of the aggregates, and even that sign
reversals of the longest-wavelength Cotton effect may occur,
simply with increasing length of tubular aggregates of chro-
mophores.
From ongoing work in Tamiaki’s group, which has as-
signed the stereochemistry of 3-hydroxyethyl groups at-
tached to porphyrins, bacteriochlorophylls, and bacterioch-
lorins through correlation of NMR chemical shifts in
Mosher-type esters,[31] we tentatively assign the 31-R con-
figuration to the enantiomer eluted first and the 31-S con-
figuration to the enantiomer eluted second. Assuming that
no strange size effects occur in the CD spectra, the first
and the second eluted enantiomers produce P- and M-type
helical suprastructures, respectively. In future work we will
use complementary methods to verify this assignment.
Figure 9. CD spectra of self-assembled 10-Zn; red trace Ϫ first
eluted enantiomer of 10 after metallation with zinc acetate and self-
assembly in n-heptane (28 µ solution); blue trace Ϫ second eluted
enantiomer of 10 after Zn insertion and self-assembly in n-heptane
(22 µ solution); the strongest Cotton effects are slightly truncated
due to saturation of the detector; in more dilute solutions the self-
assembly was less pronounced, while with thinner cuvettes the Cot-
ton effects of the Q-bands were less visible; dotted traces, superim-
posing the zero line, are after disassembly with methanol; in these
cases practically no light passed through the sample at ഠ 430 nm
anymore, due to the intense Soret absorption of the monomers; the
spectra were recorded at room temperature with a path length of
1 cm; the inset shows a typical preparative chiral HPLC trace (for
details see the Exp. Sect.)
Conclusion
We have extended the self-assembly algorithm which op-
erates in the chlorosomes of photosynthetic bacteria to syn-
At present, we cannot tell whether the self-assembly pro- thetic porphyrins and chlorins, thus establishing its gener-
cess with rac-10-Zn occurs equally well between the sepa- ality. The separated enantiomers of 10-Zn form chiral sup-
rated R and S enantiomers or if a hetero-enantiomeric pro- rastructures displaying opposite chiralities after self-as-
cess is energetically preferred. However, it appears that the sembly. The ‘‘molecular’’ chirality of the hydroxyethyl
separated enantiomers are less prone to self-assembly than group is thus transferred at the ‘‘supramolecular’’ level.
similarly treated racemate samples. We can rule out impurit- Interestingly, the separated enantiomers appear to self-as-
ies leaking from the chiral HPLC column or traces of polar semble less readily than the racemate at the same concen-
solvents inhibiting their aggregation, as these were sub- trations. This observation might explain why in the chloro-
sequently removed in a further purification step. An exper- somes both enantiomers of the 3-hydroxyethyl group are
imental observation is that the size of the aggregates is present within various homologues.
3926
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3919Ϫ3930

Green Self-Assembling Porphyrins and Chlorins as Mimics of the Natural Bacteriochlorophylls c, d, and e
FULL PAPER
were obtained on a Voyager Instrument from Applied Biosystems,
with anhydrous glycerol, HABA [2Ј-(4-hydroxyphenylazo)benzoic
acid], or 1,8,9-anthracenetriol matrices. HR-FAB-MS were re-
corded with 3-nitrobenzyl alcohol (NBA) as the matrix on a Finni-
gan MAT 90 machine. CD spectra were measured with an JASCO
J-810 CD spectrometer with 2 s integration time and with a 2 nm
bandwidth with baseline subtraction. The baseline was recorded
with n-heptane alone in a cuvette identical to that used for re-
cording the samples. Thin layer chromatography was performed on
silica gel plates from MachereyϪNagel. Column chromatography
Our novel self-assembling zinc porphyrins present at-
tractive optical properties, which might make them useful in
artificial light-harvesting devices. The presence of carbonyl
groups in conjugation with the macrocycle produce a green
color both in the free bases and in the zinc complexes, thus
mimicking the natural antenna systems. Surprisingly, simi-
lar simple chlorins do not show higher extinction coef-
ficients for the Q bands and have Soret bands of reduced
intensity. If only the relative proportions are considered,
then the Q bands in chlorins indeed appear to be more in- was performed with Merck silica gel 40Ϫ63 µm.
tense than those of the more symmetric porphyrins. In
3-Acetyl-10,20-bis(3,5-di-tert-butylphenyl)-15-(5,5-dimethyl-1,3-
dioxan-2-yl)porphyrin (6): (Transformation e in the Scheme). 3-Ace-
tyl-10,20-bis(3,5-di-tert-butylphenyl)-15-formylporphyrin
(4)[5]
(0.200 g, 0.26 mmol), neopentyl glycol (0.255 g, 2.44 mmol), p-tolu-
enesulfonic acid (0.018 g, 0.094 mmol), and toluene (140 mL) were
chlorophylls the annulated cyclopentanone ring has a spe-
cial functional role in further desymmetrizing the macrocy-
cle and at the same time it ensures a coplanar carbonyl
group, fully conjugated with the macrocycle. Whether it is
worth implementing this extra ring for artificial devices is a placed in a 250 mL flask fitted with a DeanϪStark trap and a
reflux condenser, and the mixture was heated to reflux for 65 min.
The reaction mixture was then cooled and was washed with a satu-
rated NaHCO3 solution and then with water. The organic layer was
then dried with Na2SO4. After solvent evaporation, 6 was obtained
as a purple solid (0.215 mg, 96%). 1H NMR (300 MHz, CDCl3):
δ ϭ 11.30 (s, 1 H, 5-H), 10.10 (d, 3J ϭ 4.8 Hz, 1 H, 13-H), 9.80 (d,
matter of device performance and ultimately cost. We
would like to encourage studies in this area, stressing that
for organic solar cells the cost-limiting factor is not the
chemical synthesis of the components involved but rather
the modest performance (ഠ 3%).[32] A doubling of this
value Ϫ through the use of artificial antennas as in natural
light-harvesting, for instance Ϫ might prove an economi-
cally viable alternative to the currently still costly silicon-
based solar cells, which can, however, attain ഠ14% energy
conversion efficiencies in industrially produced devices.
3
3J ϭ 4.8 Hz, 1 H, 17-H), 9.45 (d, J ϭ 4.8 Hz, 1 H, 7-H), 9.40 (s,
3
3
1 H, 2-H), 9.11 (d, J ϭ 5.1 Hz, 1 H, 12-H), 9.01 (d, J ϭ 4.5 Hz,
3
4
1 H, 18-H), 8.94 (d, J ϭ 4.8 Hz, 1 H, 8-H), 8.10 (d, J ϭ 1.8 Hz,
2 H, 2Ј, 6Ј), 8.06 (d, 4J ϭ 1.8 Hz, 2 H, 2ЈЈ, 6ЈЈ), 7.97 (s, 1 H, Hacetal),
7.87 (t, 4J ϭ 1.5 Hz, 1 H, 4Ј), 7.83 (t, 4J ϭ 1.8 Hz, 1 H, 4ЈЈ),
4.33 (s, 4 H, CH2 acetal), 3.13 (s, 3 H, COCH3), 1.93 (s, 3 H,
CH3 eq-acetal), 1.12 (s, 3 H, CH3ax-acetal), Ϫ2.83 (s, 2 H, NH) ppm.
UV/Vis (CH2Cl2) : λmax. (nm) ϭ 428, 534, 576, 603, 661. HR-FAB-
MS: found m/z ϭ 843.5195 [M ϩ H]ϩ, calcd. for C56H67N4O3
843.5213.
Experimental Section
General Remarks: Solvents were dried and freshly distilled before
use as follows: dichloromethane from calcium hydride; toluene and
n-heptane from sodium metal; [D]chloroform from phosphorus
pentoxide. NMR spectra were recorded at 300 MHz (1H) with a
Bruker DPX 300 Avance spectrometer. Chemical shifts are given in
ppm relative to the signal of CHCl3 which was taken as δ ϭ 7.26
(for 1H) and 77.00 ppm (for 13C). Analytical chiral HPLC analyses
of 10, 13, 14, and 16 were performed on a screening unit composed
of a Merck D-7000 system manager, MerckϪLachrom L-7100
pump, MerckϪLachrom L-7360 oven capable of accommodating
rac-10,20-Bis(3,5-di-tert-butylphenyl)-15-(5,5-dimethyl-1,3-dioxan-
2-yl)-3-(hydroxyethyl)porphyrin (8): (Transformation
f in the
Scheme). Compound 6 (160 mg, 0.1897 mmol), NaBH4 (143.5 mg,
0.379 mmol), methanol (12 mL), and dichloromethane (25 mL)
were stirred in a 100 mL single-necked flask at room temp. for 19
h. The reaction mixture was then washed twice with brine and dried
with Na2SO4, after which the solvents were evaporated to leave a
crude product, which was then purified by column chromatography
(SiO2, eluted with CH2Cl2) to give pure purple compound 8
(115 mg, 72%). 1H NMR (300 MHz, CDCl3): δ ϭ 10.42 (s, 1 H, 5-
12 columns fed by
a
Valco 12 positions valve, and
MerckϪLachrom L-7400 UV detector. Generally, the free bases
were better resolved than the corresponding Zn complexes. Separ-
ation of 10 was obtained on a chiral (S,S)-WhelkϪO1 (250 ϫ
46 mm, Regis) column with n-hexane/ethanol (9:1, v/v) at 25 °C at
2 mL/min and with UV detection at 420 nm. Retention times were
R(t1) ϭ 7.36 min and R(t2) ϭ 8.36 min; retention factors k1 ϭ 3.91
and k2 ϭ 4.58; selectivity factor α ϭ k2/k1 ϭ 1.17 and resolution
Rs ϭ 1.70. Preparative chiral HPLC was performed on the same
column under the same conditions with use of a MerckϪHitachi
LiChrograph L-6000 pump, MerckϪHitachi L-4000 UV detector,
and Merck D-7000 system manager. n-Hexane and ethanol were
HPLC grade from SDS (Peypin, France). The solvents for the chro-
matography experiments were degassed and filtered through a
Millipore membrane (0.45 µm) before use. Further details of the
3
H), 10.02 and 9.97 (two d, 2 H, J ϭ 4.5 Hz, 13- and 17-H), 9.34
3
(d, J ϭ 4.5 Hz, 1 H, 7-H), 9.03 (two partially overlapping d, 3 H,
12, 18 and 8-H), 8.91 (s, 1 H, 2-H), 8.09 and 8.08 (two d, 4J ϭ
1.8 Hz, 4 H, 2Ј, 6Ј and 2ЈЈ, 6ЈЈ-H), 8.03 (s, 1 H, 2-Hacetal), 7.84 (t,
4J ϭ 1.8 Hz, 2 H, 4Ј,4ЈЈ-H), 6.65 (quint, 1 H, HOϪCHϪCH3), 4.34
3
3
(s, 4 H, CH2-acetal), 2.59 (d, J ϭ 4.5 Hz, 1 H, OH), 2.26 (d, J ϭ
6.3 Hz, 3 H, 6.6 Hz, HOϪCHϪCH3), 1.95 (s, 3 H, CH3-eq-acetal),
1.57 (two s, 2 ϫ 18 H, C(CH3)3], 1.13 (s, 3 H, CH3-ax-acetal), Ϫ3.05
(s, 2 H, NϪH) ppm. UV/Vis (CH2Cl2): λmax. (nm) ϭ 414, 510, 573,
and other two less intense Q bands. HR-FAB-MS: m/z ϭ 845.5357
[M ϩ H]ϩ, calcd. for C56H69N4O3 845.5370.
rac-10,20-Bis(3,5-di-tert-butylphenyl)-15-formyl-3-(1-hydroxyethyl)-
porphyrin (10 ): (Transformation g in the Scheme). Aqueous HCl
chiral separations are included in the CHIRBASE database.[33] (20 mL, 0.5 ⁾ was added to a solution of compound 8 (100 mg,
UV/Vis spectra were measured with a Varian Cary 500 instrument.
The variable temperature spectra were performed with the aid of a
0.118 mmol) in 1,4-dioxane (30 mL) under inert atmosphere, and
the mixture was stirred at room temp. for 17 h. After completion
Peltier heating unit allowing equilibration of temperature before of the reaction (TLC monitoring), the mixture was washed with
each curve was measured while the solution in the quartz cuvette
was stirred with a magnetic stirrer. MALDI-TOF mass spectra
brine and neutralized with a saturated aqueous NaHCO3 solution.
The organic layer was extracted with dichloromethane and dried
Eur. J. Org. Chem. 2004, 3919Ϫ3930
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3927

T. S. Balaban, M. Linke-Schaetzel, A. D. Bhise, N. Vanthuyne, C. Roussel
FULL PAPER
1
with Na2SO4. The solvent was evaporated to provide a crude prod-
uct, which after column chromatography on SiO2 eluted with
CH2Cl2 afforded green compound 10 (83.4 mg, 94% yield). 1H
hexane, 1:1, v/v) to leave pure 7 (11 mg; 55%) as a green solid. H
NMR (300 MHz, CDCl3): δ ϭ 9.75 (d, 3J ϭ 5.1 Hz, 1 H, 3-H),
3
3
9.53 (d, J ϭ 4.5 Hz, 1 H, 13-H), 9.28 (d, J ϭ 4.8 Hz, 1 H, 7-H),
NMR (300 MHz, CDCl3): δ ϭ 12.58 (s, 1 H, 15-CHO), 10.49 (s, 1 8.68 (d, J ϭ 3.6 Hz, 1 H, 4-H), 8.51 (d, J ϭ 4.5 Hz, 1 H, 8-H),
3
3
H, 5-H), 10.08 (d, 3J ϭ 4.5 Hz, 1 H, 2 H, 13-H), 10.04 (d, 3J ϭ
4.5 Hz, 2 H, 17-H), 9.31 (d, 3J ϭ 4.8 Hz, 1 H, 7-H), 9.09 (two d, 2
8.27 (d, J ϭ 4.5 Hz, 1 H, 2-H), 7.92 (d, J ϭ 1.8 Hz, 2 H, 2Ј,6Ј-
H), 7.74 (t, J ϭ 1.8 Hz, 1 H, 4Ј-H), 7.69 (t, J ϭ 1.8 Hz, 1 H, 4ЈЈ-
H), 7.65 (d, J ϭ 1.8 Hz, 2 H, 2ЈЈ,6ЈЈ-H), 7.62 (s, 1 H, Hacetal), 7.02
3
4
4
4
H, 3J ϭ 4.5, 12ϪH and 18-H), 8.93 (d, 3J ϭ 4.5 Hz, 1 H, 8-H),
4
4
8.82 (s, 1 H, 2-H), 8.08 (m, 4 H, 2Ј,6Ј and 2ЈЈ,6ЈЈ-H), 7.88 (t, J ϭ (s, 1 H, Hacetal), 4.74 (m, 2 H, 17-CH2), 4.20 (m, 2 H, 18-CH2),
1.8 Hz, 2 H, 4Ј,4ЈЈ-H), 6.59 (q, 3J ϭ 6.3 Hz, 1 H, HOϪCHϪCH3),
1.82 (s, 3 H, CH3-eq-acetal), 1.77 (s, 3 H, CH3-eq-acetal), 1.49 [s, 18 H,
2.58 (s, 1 H, OH), 2.24 (d, 3J ϭ 6.6 Hz, 3 H, HOϪCHϪCH3), 1.56 C(CH)3]3), 1.48 [s, 18 H, C(CH)3]3), 1.05 (s, 3 H, CH3-ax-acetal), 1.01
[m, 36 H, C(CH3)3], Ϫ2.36 (s, 2 H, NH) ppm UV/Vis (CH2Cl2) : (s, 3 H, CH3-ax-acetal), Ϫ1.30 (s, 1 H, NH), Ϫ1.55 (s, 1 H, NH) ppm.
λmax. (nm) ϭ 424, 525, 565, 596, 653. HR-FAB-MS: found m/z ϭ
UV/Vis (CHCl3): λmax. (nm) ϭ 417, 514, 541, 602, 655. HR-FAB-
MS: found m/z ϭ 945.4504 [M ϩ H]ϩ, calcd. for C50H57N4O2
945.4481.
759.4653 [M ϩ H]ϩ, calcd. for C51H59N4O2 759.4638.
rac-10,20-Bis(3,5-di-tert-butylphenyl)-15-formyl-3-(1-hydroxyethyl)-
porphinatozinc (10-Zn): (Transformation h in the Scheme). Zinc
acetate (43 mg, 0.023 mmol) was added to a solution of compound
10,20-Bis(3,5-di-tert-butylphenyl)-5,15-bis(formyl)chlorin (9):
(Transformation j in the Scheme). A solution of TiCl4 in CHCl3
10 (25 mg, 0.0329 mmol) in CHCl3 (20 mL) and MeOH (12 mL), (0.1 , 0.5 ml, 0.05 mmol) was added dropwise by syringe, under
and the reaction mixture was stirred at room temp. for 3 h under
inert atmosphere. The reaction mixture was then washed with brine
argon, to a solution of 10,20-bis(3,5-di-tert-butylphenyl)-5,10-
bis(5,5-dimethyl-1,3-dioxan-2-yl)chlorin (6) (30 mg; 0.032 mmol) in
and extracted with dichloromethane. The organic layer was dried dry CHCl3 (5 mL). The reaction mixture was stirred at room tem-
with Na2SO4 and evaporated to leave the green compound Zn-10
perature for 15 min and then hydrolyzed by addition of a saturated
in high purity and yield (25.2 mg, 93.3%). 1H NMR (300 MHz, aqueous NaHCO3 solution (5 mL). The crude reaction mixture was
CDCl3ϩCD3OD, 10:1, v/v): δ ϭ 12.54 (s, 1 H, 15ϪCHO), 10.37 extracted with CH2Cl2, and the solvents were evaporated to dry-
(s, 1 H, 5-H), 10.02 and 10.01 (two d, J ϭ 4.9 Hz, 2 H, 13- and 17-
H), 9.25 (d, J ϭ 4.5 Hz, 1 H, 7-H), 9.05 and 9.03 (two d, J ϭ 4.7,
ness. Purification by column chromatography on SiO2 (eluent:
CH2Cl2/n-hexane, 1:1, v/v) gave pure 9 (19 mg; 66%) as a green
4.9 Hz, 2 H, 12- and 18-H), 8.89 (d, J ϭ 4.3 Hz, 1 H, 8-H), 8.79 solid. 1H NMR (300 MHz, CDCl3): δ ϭ 12.23 (s, 1 H, 5-CHO),
(s, 1 H, 2-H), 8.01 (m, 4 H, 2Ј, 6Ј, 2ЈЈ, 6ЈЈ-H), 7.57 (t, J ϭ 1.8 Hz,
11.90 (s, 1 H, 15-CHO), 9.79 (d, 3J ϭ 6 Hz, 1 H, 3-H), 9.57 (d,
3
3
2 H, 4Ј, 4ЈЈ-H), 6.55 (q, J ϭ 6.5 Hz, 1 H, HOϪCHϪCH3), 3.40 3J ϭ 5.1 Hz, 1 H, 13-H), 9.39 (d, J ϭ 4.5 Hz, 1 H, 7-H), 8.72 (d,
3
(s, 1 H, OH), 2.20 (d, J ϭ 6.4 Hz, 3 H, HOϪCHϪCH3), 1.53 [two
3J ϭ 5.1 Hz, 1 H, 4-H), 8.56 (d, J ϭ 4.8 Hz, 1 H, 8-H), 8.35 (d,
s, 36 H, C(CH3)3] ppm. UV/Vis (CH2Cl2 ϩ n-heptane) : λmax. 3J ϭ 5.1 Hz, 1 H, 2-H), 7.95 (d, J ϭ 1.8 Hz, 2 H, PhϪo-H), 7.82
4
4
4
(nm) ϭ 638, 594, 480, 420. After methanol addition: 607, 563, 428.
(t, J ϭ 1.8 Hz, 1 H, PhϪp-H), 7.76 (t, J ϭ 1.8 Hz, 1 H, PhϪp-
UV/Vis (CH2Cl2, 0.23 m): λmax (lg εmax.) ϭ 600 (4.19), 557 (4.05), H), 7.65 (d, 4J ϭ 1.8 Hz, 2 H, PhϪo-H), 4.91 (m, 2 H, 17-CH2),
424 (5.47). HR-FAB-MS: found m/z ϭ 821.3782, [M ϩ H]ϩ, calcd.
4.23 (m, 2 H, 18-CH2), 1.54 [s, 18 H, C(CH3)3], 1.51 [s, 18 H,
C(CH3)3], Ϫ0.35 (s, 1 H, NH), Ϫ0.47 (s, 1 H, NH) ppm. UV/Vis
(CHCl3): λmax. (nm) ϭ 425, 505, 531, 578, 660, 718. HR-FAB-MS:
found m/z ϭ 915.5803 [M ϩ H]ϩ, calcd. for C60H75N4O4ϭ
915.5788.
for C51H56N4O2Zn 821.3773.
10,20-Bis-(3,5-di-tert-butylphenyl)-5,15-bis(5,5-dimethyl-1,3-dioxan-
2-yl)porphyrin (5): (Transformation e in the Scheme). A solution of
10,20-bis(3,5-di-tert-butylphenyl)-5,15-bis(formyl)
(47 mg, 0.0632 mmol), 2,2-dimethylpropanediol
porphyrin[5]
(100 mg; 10,20-Bis(3,5-di-tert-butylphenyl)-15-formyl-5-(hydroxymethyl)-
0.99 mmol), and p-toluenesulfonic acid as catalyst in toluene (100
mL) was heated to reflux in a DeanϪStark apparatus for 6 h. After
evaporation of the solvent, the residue was taken up in CH2Cl2
and washed with a saturated aqueous NaHCO3 solution, and the
solvents were evaporated off to dryness. The crude product was
purified by column chromatography on SiO2 (eluent: CH2Cl2/n-
chlorin (11): (Transformation f in the Scheme). NaBH4 (0.9 mg;
0.024 mmol) was added to a solution of 10,20-bis(3,5-di-tert-butyl-
phenyl)-5,15-bis(formyl)chlorin (9) (22 mg; 0.024 mmol) in CHCl3
(10 mL) and MeOH (2 mL). The mixture was stirred for 30 min at
room temperature, and then water (5 mL) was added. The crude
reaction mixture was extracted with CH2Cl2 and the solvents were
evaporated to dryness. Purification by column chromatography on
hexane, 1:1, v/v) to leave pure 5 (43 mg; 75%) as a red solid. 1H
3
NMR (300 MHz, CDCl3): 9.90 (d, J ϭ 4.8 Hz, 4 H, 3,7,13,17-H), SiO2 (eluent: CH2Cl2/hexane, 50:50) gave pure 11 (11 mg; 50%). 1H
3
4
8.94 (d, J ϭ 5.1 Hz, 4 H, 2,8,12,18-H), 8.03 (d, J ϭ 1.8 Hz, 4 H,
NMR (300 MHz, CDCl3): δ ϭ 11.76 (s, 1 H, 15-CHO), 9.55 (d,
4
3
Ph-ortho-H), 7.93 (s, 2 H, Hacetal), 7.81 (t, J ϭ 1.8 Hz, 2 H, Ph-
3J ϭ 5.4 Hz, 1 H, 3-H), 9.17 (d, J ϭ 5.1 Hz, 1 H, 13-H), 9.02 (d,
para-H), 4.31 (m, 8 H, CH2 acetal), 1.92 (s, 3 H, CH3-eq-acetal), 1.54 3J ϭ 4.8 Hz, 1 H, 7-H), 8.66 (d, J ϭ 4.8 Hz, 1 H, 4-H), 8.38 (d,
3
[s, 36 H, C(CH3)3], 1.11 (s, 3 H, CH3-ax-acetal), Ϫ2.95 (s, 2 H, NH) 3J ϭ 4.5 Hz, 1 H, 8-H), 8.12 (d, J ϭ 4.8 Hz, 1 H, 2-H), 7.90 (d,
3
4
ppm. UV/Vis (CHCl3) : λmax. (nm) ϭ 416, 512, 544, 586, 641. HR-
FAB-MS: found m/z ϭ 915.5803, calcd. for C60H77N4O4 915.5788.
4J ϭ 1.8 Hz, 2 H, 10-PhϪo-H), 7.82 (t, J ϭ 1.8 Hz, 1 H, PhϪp-
4
H), 7.70 (t, J ϭ 1.8 Hz, 1 H, PhϪp-H), 7.64 (d, 4J ϭ 1.8 Hz, 2 H,
20-PhϪo-H), 6.53 (d, 3J ϭ 5.7 Hz, 2 H, CH2OH), 4.81 (m, 2 H,
17-CH2), 4.13 (m, 2 H, 18-CH2), 2.40 (t, 1 H, CH2OH), 1.50 [s, 18
H, C(CH3)3], 1.48 [s, 18 H, C(CH3)3], 0.05 (s, 1 H, NH), Ϫ0.05 (s,
1 H, NH) ppm. UV/Vis (CHCl3) : λmax. (nm) ϭ 418, 516, 555, 578,
627, 681. HR-FAB-MS: found m/z ϭ 747.4628 [M ϩ H]ϩ, calcd.
for C59H59N4O2 747.4638.
10,20-Bis(3,5-di-tert-butylphenyl)-5,15-bis(5,5-dimethyl-1,3-dioxan-
2-yl)chlorin (7): (Transformation i in the Scheme). A solution of
10,20-bis(3,5-di-tert-butylphenyl)-5,15-bis(5,5-dimethyl-1,3-dioxan-
2-yl)porphyrin (5) (20 mg; 0.022 mmol), K2CO3 (90 mg;
0.65 mmol), and tosylhydrazide (40 mg, 0.215 mmol) in β-picoline
(5 mL) was heated at reflux for 15 h. Water was then added, and
the crude reaction mixture was extracted with CH2Cl2. The organic
phase was washed once with brine, twice with HCl (5%), and twice
with water, and was then evaporated to dryness. The residue was
10,20-Bis(3,5-di-tert-butylphenyl)-15-formyl-5-(hydroxymethyl)-
chlorinatozinc (11-Zn): (Transformation
h in the Scheme).
Zn(OAc)2 (2 mg, 0.0108 mmol) was added to a solution of 10,20-
purified by column chromatography on SiO2 (eluent: CH2Cl2/n- bis(3,5-di-tert-butylphenyl)-15-formyl)-5-(hydroxymethyl)chlorin
3928 www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2004, 3919Ϫ3930

Green Self-Assembling Porphyrins and Chlorins as Mimics of the Natural Bacteriochlorophylls c, d, and e
FULL PAPER
[7]
[7a] S. Yagai, H. Miyatake, H. Tamiaki, J. Org. Chem. 2002, 67,
(11) (5 mg, 0.0054 mmol) in CHCl3 (3 mL) and MeOH (1 mL).
The mixture was stirred for 2 h at room temperature, and water (5
mL) was then added. The crude reaction mixture was extracted
with CH2Cl2 and the solvents were evaporated to dryness. Purifi-
cation by column chromatography on SiO2 (eluent: CH2Cl2) gave
pure 11-Zn (5 mg; 95%). 1H NMR (300 MHz, CDCl3/CD3OD,
[7b]
49.
For a review of earlier work on semisynthetic BChl c
[7c]
mimics see: H. Tamiaki, Coord. Chem. Rev. 1996, 148, 183.
H. Tamiaki, T. Miyatake, R. Tanikaga, A. R. Holzwarth, K.
Schaffner, Angew. Chem. Int. Ed. Engl. 1996, 35, 772; Angew.
[7d]
Chem. 1996, 108, 810.
T. Miyatake, T. Oba, H. Tamiaki,
Chem. Bio. Chem. 2001, 335.
3
[8]
[9]
10:1 v/v): δ ϭ 11.48 (s, 1 H, 15-CHO), 9.09 (d, J ϭ 4.8 Hz, 1 H,
H. Tamiaki, S. Kimura, T. Kimura, Tetrahedron 2003, 59, 7423.
T. S. Balaban, J.-M. Lehn, S. Lebedkin, M-J. Brienne, Ger.
Offen. 2003, Patent DE 101 46 970 A1.
K. Susumu, T. Shimidzu, K. Tanaka, H. Segawa, Tetrahedron
Lett. 1996, 37, 8399.
3
3
3-H), 8.89 (d, J ϭ 4.8 Hz, 1 H, 13-H), 8.74 (d, J ϭ 4.5 Hz, 1 H,
3
3
7-H), 8.34 (d, J ϭ 4.8 Hz, 1 H, 4-H), 8.07 (d, J ϭ 4.5 Hz, 1 H,
3
4
[10]
8-H), 7.74 (d, J ϭ 4.8 Hz, 1 H, 2-H), 7.73 (d, J ϭ 1.8 Hz, 2 H,
10-PhϪo-H), 7.68 (t, 4J ϭ 1.8 Hz, 1 H, PhϪp-H), 7.59 (t, 4J ϭ
1.8 Hz, 1 H, PhϪp-H), 7.49 (d, 4J ϭ 1.8 Hz, 2 H, 20-PhϪo-H),
[11] [11a]
M. J. Crossley, P. L. Burn, J. Chem. Soc., Chem. Commun.
[11b]
3
1987, 39.
M. J. Crossley, P. L. Burn, J. Chem. Soc., Chem.
6.26 (d, J ϭ 5.7 Hz, 2 H, CH2OH), 4.66 (m, 2 H, 17-CH2), 4.04
[11c]
Commun. 1991, 1569.
We propose to nickname the 3,5-
(m, 2 H, 18-CH2), 1.48 [s, 18 H, C(CH3)3], 1.42 [s, 18 H, C(CH3)3]
ppm. UV/Vis (CHCl3 ϩ MeOH) : λmax. (nm) ϭ 419, 639. UV/Vis
(CH2Cl2 ϩ MeOH, 0.13 m) : λmax. (lg εmax.) ϭ 642 (4.16), 599
(3.77), 522 (3.51), 421 (5.01). HR-FAB-MS: found m/z ϭ 809.3760
[M ϩ H]ϩ, calcd. for C50H57N4O2Zn 809.3772.
di-tert-butylphenyl substituent ‘‘Crossley’s crutch’’, as it helps
porphyrin chemists in four ways: (i) it solubilizes complex
architectures and minimizes π-stacking of meso-aryl porphy-
rins, thanks to its steric bulge, (ii) it prevents electrophilic sub-
stitution in the adjacent β-pyrrolic positions and on the ben-
zene ring, thus enabling very selective functionalization of the
porphyrins, (iii) it packs well into crystals, enabling X-ray dif-
fraction studies to be carried out[12,24] but at the same time
enables visualization by scanning probe techniques[34] for
noncrystalline samples, and (iv) thanks to the sharp NMR tert-
butyl singlet, visible even in nanomolar concentrations, it al-
lows easy identification of compounds and of their symmetry
in order to allow monitoring of complex reaction mixtures.
T. S. Balaban, R. Goddard, M. Linke-Schaetzel, J.-M. Lehn, J.
Am. Chem. Soc. 2003, 125, 4233.
Acknowledgments
Professor Nils Metzler-Nolte is thanked for enabling us to record
the CD spectra. Professor Hitoshi Tamiaki kindly shared unpub-
lished results[31] and enriching discussions with us. Matthias
Fischer is thanked for technical assistance, and Ingrid Roßnagel for
HR-FAB-MS measurements. This work was partially financed by
the DFG-Center for Functional Nanostructures at the University
of Karlsruhe, through project C3.2.
[12]
[13]
J.-M. Lehn, Supramolecular Chemistry: Concepts and Perspec-
tives, VCH Verlagsgesellschaft, Weinheim, 1995.
A. Kay, M. Grätzel, J. Phys. Chem. 1993, 97, 6272.
J.-P. Strachan, D. F. O’Shea, T. Balasubramanian, J. S. Lindsey,
[14]
[15]
[1] [1a]
T. S. Balaban, A. R. Holzwarth, K. Schaffner, G.-J.
J. Org. Chem. 2000, 65, 3160.
[1b]
Boender, H. J. M. de Groot, Biochemistry 1995, 34, 15259.
[16] [16a] D. T. Gryko, C. Clausen, J. S. Lindsey, J. Org. Chem. 1999,
B. J. van Rossum, D. B. Steensgaard, F. M. Mulder, G. J.
Boender, K. Schaffner, A. R. Holzwarth, H. J. M. de Groot,
Biochemistry 2001, 40, 1587.
[16b]
64, 8635.
A. Osuka, G. Noya, S. Taniguchi, T. Okada, Y.
Nishimura, I. Yamazaki, N. Mataga, Chem. Eur. J. 2000, 6, 33.
[17] [17a] H. W. Whitlock Jr., R. Hanauer, M. Y. Oester, B. K. Bower,
[2] [2a]
R. E. Blankenship, J. M. Olson, M. Miller, ‘‘Antenna com-
[17b]
J. Am. Chem. Soc. 1969, 91, 7485.
M. O. Senge, W. W.
plexes from green photosynthetic bacteria’’ in Anoxygenic
photosynthetic bacteria (Eds.: R. E. Blankenship, M. T. Madi-
[17c]
Kalish, S. Runge, Tetrahedron 1998, 54, 3781.
M. Pineiro,
A. L. Carvalho, M. M. Pereira, A. M. d’A. Rocha Gonsalves,
L. G. Arnaut, S. J. Formosinho, Chem. Eur. J. 1998, 4, 2299.
[17d] Y. Ferrand, L. Bourre, G. Simonneaux, S. Thibaut, F. Odo-
[2b]
gan, C. E. Bauer), Kluwer Academic, 1995; pp. 399Ϫ435.
R. E. Blankenship, Molecular Mechanisms of Photosynthesis,
Blackwell Science, Oxford, 2002. [2c] T. S. Balaban, ‘‘Light-har-
vesting nanostructures’’ in Encyclopedia of Nanoscience and
Nanotechnology (Ed.: H. S. Nalwa), American Scientific Pub-
lishers, Los Angeles, 2004, vol. 4, pp. 505Ϫ559.
bel, Y. Lajat, T. Patrice, Bioorg. Med. Chem. Lett. 2003, 13,
[17e]
833.
J. Laville, T. Figueiredo, B. Loock, S. Pigaglio, R.
Maillard, D. S. Grierson, D. Carrez, A. Croisy, J. Blais, Bioorg.
Med. Chem. 2003, 11, 1643.
An analogy to the (bacterio)chlorophyll numbering is pre-
served here. Confusion may arise due to systematic nomencla-
ture, which places a hydroxy(m)ethyl group after formyl, thus
reversing the numbering sense of the macrocycle between the
different compounds in the Scheme.
[3] [3a]
T. S. Balaban, A. R. Holzwarth, K. Schaffner, J. Mol.
[18]
[19]
[3b]
Struct. 1995, 349, 183.
N. Griebenow, T. S. Balaban, A. R. Holzwarth, K. Schaffner,
J. Chiefari, K. Griebenow, F. Fages,
[3c]
J. Phys. Chem. 1995, 99, 1357.
T. S. Balaban, H. Tamiaki,
A. R. Holzwarth, K. Schaffner, J. Phys. Chem. 1997, 101, 3424.
[3d]
T. S. Balaban, J. Leitich, A. R. Holzwarth, K. Schaffner, J.
Not only the color, but also the shift of the n˜CϭO frequencies
give strong indications of whether an acyl substituent has the
carbonyl group in conjugation with the macrocycle (A. D.
Bhise, M. Linke-Schaetzel, D. Moss, T. S. Balaban, manuscript
in preparation).
Phys. Chem. B 2000, 104, 1362.
[4] [4a]
P. Cheng, P. A. Liddell, S. X. C. Ma, R. E. Blankenship,
Photochem. Photobiol. 1993, 58, 290. [4b] T. Mizoguchi, L. Lim-
antara, K. Matsuura, K. Shimada, Y. Koyama, J. Mol. Struc.
[4c]
1996, 379, 249.
Z.-Y. Wang, M. Umetsu, K. Yoza, M. Ko-
[20] [20a]
G. McDermott, S. M. Prince, A. A. Freer, A. Haw-
thornthwaite-Lawless, M. Z. Papiz, R. J. Cogdell, N. W. Isaacs,
bayashi, M. Imai, Y. Matsushita, N. Niimura, T. Nozawa, Bio-
[4d]
chim. Biophys. Acta 1997, 1320, 73.
Z.-Y. Wang, M.
[20b]
Nature 1995, 374, 517.
J. Koepke, X. Hu, C. Muenke, K.
Umetsu, M. Kobayashi, T. Nozawa, J. Am. Chem. Soc. 1999,
121, 9363. [4e] Z.-Y. Wang, M. Umetsu, M. Kobayashi, T. Noz-
Schulten, H. Michel, Structure 1996, 4, 581.
K. McLuskey, S. M. Prince, R. J. Cogdell, N. Isaacs, Biochemis-
try 2001, 40, 8783.
[21]
[4f]
awa, J. Phys. Chem. B. 1999, 103, 3742.
T. Mizoguchi, K.
Ogura, F. Inagaki, Y. Koyama, Biospectroscopy 1999, 5, 63.
T. S. Balaban, A. D. Bhise, M. Fischer, M. Linke-Schaetzel, C.
Roussel, N. Vanthuyne, Angew. Chem. Int. Ed. 2003, 42, 2140;
Angew. Chem. 2003, 115, 2189.
M. Linke-Schaetzel, A. D. Bhise, H. Gliemann, T. Koch, T.
Schimmel, T. S. Balaban, Thin Solid Films 2004, 451/452,
16Ϫ21.
[5]
[6]
[22] [22a]
[22b]
L. Dähne, J. Am. Chem. Soc. 1995, 117, 12855.
M.
Wang, G. L. Silva, B. A. Armitage, J. Am. Chem. Soc. 2000,
[22c]
122, 9977.
C. Thalacker, F. Würthner, Adv. Funct. Mater.
2002, 12, 209. [22d] A. P. H. J. Schenning, J. v. Herrikhuyzen, P.
Jonkheijm, Z. Chen, F. Würthner, E. W. Meijer, J. Am. Chem.
[22e]
Soc. 2002, 124, 10252.
H. von Berlepsh, S. Kirstein, R.
Eur. J. Org. Chem. 2004, 3919Ϫ3930
www.eurjoc.org
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 3929

T. S. Balaban, M. Linke-Schaetzel, A. D. Bhise, N. Vanthuyne, C. Roussel
FULL PAPER
Hania, C. Didraga, A. Pugzˇlys, C. Böttcher, J. Phys. Chem. B
[28b]G. Pescitelli, S. Gabriel, Y. Wang, J. Fleischhauer, R. W.
Woody, N. Berova, J. Am. Chem. Soc. 2003, 125, 7613.
V. I. Prokhorenko, D. B. Steensgard, A. R. Holzwarth, Biophys.
J. 2003, 85, 3173.
[22f]
2003, 107, 14176; for reviews see:
D. Möbius, Adv. Mater.
A. Mishra, R. K. Behera, P. K. Behera, B.
[22g]
[29]
[30]
1995, 7, 437.
K. Mishra, G. M. Behera, Chem. Rev. 2000, 100, 1973.
HyperChem, Hypercube Inc. Release 7, Gainesville, Florida,
[23]
C. Didraga, J. A. Klugkist, J. Knoester, J. Phys. Chem. B 2002,
106, 11474.
2002.
[31] [31a] H. Kitamoto, Master Thesis 2002, Ritsumeikan University,
[24] [24a]
T. S. Balaban et al. unpublished results. [24b] T. S. Balaban,
[31b]
Kusatsu, Shiga, Japan.
H. Tamiaki, H. Kitamoto, T. Wat-
M. Linke-Schaetzel, A. D. Bhise, N. Vanthuyne, C. Roussel,
C. E. Anson, G. Buth, A. Eichhöfer, K. Foster, G. Garab, H.
Gliemann, R. Goddard, T. Javorfi, A. K. Powell, H. Rösner,
Th. Schimmel, submitted for publication.
M. Umetsu, Z.-Y. Wang, M. Kobayashi, T. Nozawa, Biochim.
Biophys. Acta 1999, 1410, 19.
[31c]
anabe, R. Shibata, submitted for publication.
H. Tamiaki,
H. Kitamoto, A. Nishikawa, T. Hibino, R. Shibata, Bioorg.
Med. Chem. 2004, 12, 1657.
[32]
For a review on organic solar cells see: C. J. Brabec, N. S. Saric-
iftci, J. C. Hummelen, Adv. Funct. Mater. 2001, 11, 15.
[25]
[33] [33a]
C. Roussel, J. Pierrot-Sanders, I. Heitmann, P. Piras, in
Chiral Separation Techniques Ϫ A Practical Approach, G. Sub-
[26] [26a]
K. M. Smith, G. M. F. Bisset, M. Bushell, J. Org. Chem.
[26b]
1980, 45, 2218.
B. Gerlach, S. E. Brantley, K. M. Smith,
[33b]
ramanian (Ed.), Wiley-VCH, Weinheim, 2001.
The data
J. Org. Chem. 1998, 63, 2314.
can be obtained free of charge upon request mentioning the
article’s reference at: CHIRBASE Project, Case A-62, Univer-
sity Aix-Marseille III, 13397 Marseille, Cedex 20, France.
[27] [27a]
D. Keller, C. Bustamante, J. Chem. Phys. 1986, 84, 2972.
[27b] M.-H. Kim, L. Ulibarri, D. Keller, M. F. Maestre, C. Bus-
[34] [34a] T. A. Jung, R. R. Schlittler, J. K. Gimzewski, Nature 1997,
tamante, J. Chem. Phys. 1986, 84, 2981.
[34b]
[28] [28a] N. Berova, K. Nakanishi, in Circular Dichroism. Principles
and Applications, 2nd ed (Eds.: N. Berova, K. Nakanishi, R. W.
Woody), John Wiley & Sons, New York, 2000, pp. 337Ϫ382.
386, 696.
283, 1683.
J. K. Gimzewski, C. Joachim, Science 1999,
Received March 8, 2004
3930
 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
Eur. J. Org. Chem. 2004, 3919Ϫ3930