Mimics of the self-assembling chlorosomal bacteriochlorophylls: Regio- and stereoselective synthesis and stereoanalysis of acyl(1-hydroxyalkyl)porphyrins

Article abstract of DOI:10.1021/ja905628h

Diacylation of copper 10,20-bis(3,5-di-tert-butylphenylporphyrin) using Friedel-Crafts conditions at short reaction times, high concentrations of catalyst, and 0-4°C affords only the 3,17-diacyl-substituted porphyrins, out of the 12 possible regioisomers. At longer reaction times and higher temperatures, the 3,13-diacyl compounds are also formed, and the two isomers can be conveniently separated by normal chromatographic techniques. Monoreduction of these diketones affords in good yields the corresponding acyl(1-hydroxyalkyl) porphyrins, which after zinc metalation are mimics of the natural chlorosomal bacteriochlorophyll (BChl) d. Racemate resolution by HPLC on a variety of chiral columns was achieved and further optimized, thus permitting easy access to enantiopure porphyrins. Enantioselective reductions proved to be less effective in this respect, giving moderate yields and only 79% ee in the best case. The absolute configuration of the 3¹-stereocenter was assigned by independent chemical and spectroscopic methods. Self-assembly of a variety of these zinc BChl d mimics proves that a collinear arrangement of the hydroxyalkyl substituent with the zinc atom and the carbonyl substituent is not a stringent requirement, since both the 3,13 and the 3,17 regioisomers self-assemble readily as the racemates. Interestingly, the separated enantiomers self-assemble less readily, as judged by absorption, fluorescence, and transmission electron microscopy studies. Circular dichroism spectra of the self-assemblies show intense Cotton effects, which are mirror-images for the two 3 ¹-enantiomers, proving that the supramolecular chirality is dependent on the configuration at the 3¹-stereocenter. Upon disruption of these self-assemblies with methanol, which competes with zinc ligation, only very weak monomeric Cotton effects are present. The favored heterochiral self-assembly process may also be encountered for the natural BChls. This touches upon the long-standing problem of why both 3¹-epimers are encountered in BChls in ratios that vary with the illumination and culturing conditions.

Full text of DOI:10.1021/ja905628h

Published on Web 09/21/2009
Mimics of the Self-Assembling Chlorosomal
Bacteriochlorophylls: Regio- and Stereoselective Synthesis
and Stereoanalysis of Acyl(1-hydroxyalkyl)porphyrins
Teodor Silviu Balaban,*,†,§,⊥ Anil Dnyanoba Bhise,† Gerhard Bringmann,*,#
Jochen Bu¨rck,‡ Cyril Chappaz-Gillot,⊥ Andreas Eichho¨fer,† Dieter Fenske,†,§,|
Daniel C. G Go¨tz,# Michael Knauer,# Tadashi Mizoguchi,∇ Dennis Mo¨ssinger,†
Harald Ro¨sner,† Christian Roussel,*,⊥ Michaela Schraut,# Hitoshi Tamiaki,*,∇ and
Nicolas Vanthuyne⊥
Institute for Nanotechnology and Institute of Biological Interfaces, Forschungszentrum Karlsruhe,
Karlsruhe Institute of Technology, Postfach 3640, D-76021 Karlsruhe, Germany, Center for
Functional Nanostructures and Institut fu¨r Anorganische Chemie, Karlsruhe Institute of Technology,
UniVersita¨t Karlsruhe, D-76131 Karlsruhe, Germany, ISM2-Chirosciences, Faculte´ des Sciences,
UniVersite´ Paul Ce´zanne Aix-Marseille III, UMR 6263, Saint-Je´roˆme Case A62, AVenue Escadrille
Normandie-Niemen, F-13397 Marseille, Cedex 20, France, Institute of Organic Chemistry and
Ro¨ntgen Research Center for Complex Material Systems, UniVersity of Wu¨rzburg, Am Hubland,
D-97074 Wu¨rzburg, Germany, and Department of Biosciences and Biotechnology, Faculty of Science
and Engineering, Ritsumeikan UniVersity, Kusatsu, Shiga 525-8577, Japan
Received July 16, 2009; E-mail: silviu.balaban@int.fzk.de; bringman@chemie.uni-wuerzburg.de;
christian.roussel@univ-cezanne.fr; tamiaki@se.ritsumei.ac.jp
Abstract: Diacylation of copper 10,20-bis(3,5-di-tert-butylphenylporphyrin) using Friedel-Crafts conditions at
short reaction times, high concentrations of catalyst, and 0-4 °C affords only the 3,17-diacyl-substituted
porphyrins, out of the 12 possible regioisomers. At longer reaction times and higher temperatures, the 3,13-
diacyl compounds are also formed, and the two isomers can be conveniently separated by normal
chromatographic techniques. Monoreduction of these diketones affords in good yields the corresponding acyl(1-
hydroxyalkyl)porphyrins, which after zinc metalation are mimics of the natural chlorosomal bacteriochlorophyll
(BChl) d. Racemate resolution by HPLC on a variety of chiral columns was achieved and further optimized,
thus permitting easy access to enantiopure porphyrins. Enantioselective reductions proved to be less effective
in this respect, giving moderate yields and only 79% ee in the best case. The absolute configuration of the
31-stereocenter was assigned by independent chemical and spectroscopic methods. Self-assembly of a variety
of these zinc BChl d mimics proves that a collinear arrangement of the hydroxyalkyl substituent with the zinc
atom and the carbonyl substituent is not a stringent requirement, since both the 3,13 and the 3,17 regioisomers
self-assemble readily as the racemates. Interestingly, the separated enantiomers self-assemble less readily,
as judged by absorption, fluorescence, and transmission electron microscopy studies. Circular dichroism spectra
of the self-assemblies show intense Cotton effects, which are mirror-images for the two 31-enantiomers, proving
that the supramolecular chirality is dependent on the configuration at the 31-stereocenter. Upon disruption of
these self-assemblies with methanol, which competes with zinc ligation, only very weak monomeric Cotton
effects are present. The favored heterochiral self-assembly process may also be encountered for the natural
BChls. This touches upon the long-standing problem of why both 31-epimers are encountered in BChls in ratios
that vary with the illumination and culturing conditions.
Introduction
pigments such as carotenoids and chlorophylls (Chls) in plants
or bacteriochlorophylls (BChls) in bacteria.1 Life on Earth relies
Light-harvesting is the primary event in photosynthesis and
certainly accounts for the vast majority of natural photosynthetic
on photosynthesis either directly through photosynthetic organ-
isms or through their incorporation into intricate food chains.
Of course, all fossil fuels mankind presently consumes are also
photosynthetic byproducts.2,3 Because of different habitats,
photosynthetic organisms have developed tailored light-harvest-
ing (also called antenna) systems adapted to their specific
environments. Thus, plants differ slightly in their light-harvesting
ability from algae living at the water surface, all of which have
Chl-protein complexes, which, in turn, differ from purple
photosynthetic bacteria. In the past decade, some of these protein
complexes have been structurally elucidated to atomic resolution
† Institute for Nanotechnology (INT), Karlsruhe Institute of Technology.
‡ Institute of Biological Interfaces (IBG-2), Karlsruhe Institute of
Technology.
§ Center for Functional Nanostructures (CFN), Universita¨t Karlsruhe
(TH).
| Institut fu¨r Anorganische Chemie, Karlsruhe Institute of Technology,
Universita¨t Karlsruhe (TH).
⊥ ISM2-Chirosciences, Universite´ Paul Ce´zanne Aix-Marseille III.
# Institute of Organic Chemistry and Ro¨ntgen Research Center for
Complex Material Systems, University of Wu¨rzburg.
∇ Department of Biosciences and Biotechnology, Ritsumeikan University.
9
14480 J. AM. CHEM. SOC. 2009, 131, 14480–14492
10.1021/ja905628h CCC: $40.75  2009 American Chemical Society

Mimics of Self-Assembling Chlorosomal Bacteriochlorophylls
A R T I C L E S
Chart 1. (Left) Chlorosomal BChls and Their Atom Numberinga
thanks to fascinating single-crystal X-ray studies.4 Additionally,
nanoscopic techniques applied to photosynthetic membranes
have revealed spectacular nanostructures.5
and (Right) Schematic Representation of Our Synthetic Mimics
Which Self-Assembleb
There exist also green photosynthetic bacteria that can thrive
in much deeper waters.1,6 These organisms evolved some 2.5
billion years ago in an oxygen-free atmosphere. For light-
harvesting, they have developed a curious organelle, the so-
called chlorosome, or “green sac”, which agglomerates BChls
c, d, and e (Chart 1). The chlorosomal architecture is based on
self-assembly and not on proteins binding BChls in specific
sites.7 No single-crystal X-ray diffraction data have yet been
obtained from native chlorosomes, various mutants, or isolated
BChls, leaving room for much discussion on the exact nature
of the supramolecular interactions. Bacteria expressing chlo-
rosomes are real photon scavengers and use near-infrared
(1) (a) Blankenship R. E. Molecular Mechanisms of Photosynthesis;
Blackwell: Oxford, 2002. (b) Balaban, T. S.; Tamiaki, H.; Holzwarth,
A. R. In Supramolecular Dye Chemistry; Wu¨rthner, F., Ed.; Topics
in Current Chemistry 258; Springer Verlag: Berlin, 2005; pp 1-38.
(c) Balaban, T. S. Light-harvesting Nanostructures. In Encyclopedia
of Nanoscience and Nanotechnology; Nalwa, H. S., Ed.; American
Scientific Publishers: Los Angeles, 2004; Vol. 4, pp 505-559.
(2) (a) Basic Research Needs for Solar Energy Utilization DOE Report,
September 2005; www.sc.doe.gov/bes/reports/files/SEU_rpt.pdf. (b)
Crabtree, G. W.; Lewis, N. S. Physics Today 2007, March, 37–42.
(3) Balzani, V.; Credi, A.; Venturi, M. Chem.sEur. J. 2008, 14, 26–39.
(4) (a) The LH2 of the purple bacteria Rhodopseudomonas acidophila at
2.5 Å resolution McDermott, G.; Prince, S. M.; Freer, A. A.;
Hawthornthwaite-Lawless, A. M.; Papiz, M. Z.; Cogdell, R. J.; Isaacs,
N. W. Nature 1995, 374, 517–521. (b) At 2.0 Å resolution: Papiz,
M. Z.; Prince, S. M.; Howard, T.; Cogdell, R. J.; Isaacs, N. W. J.
Mol. Biol. 2003, 326, 1523–1538. (c) The LH2 of the purple bacterium
Rhodospirillum molischianum: Koepke, J.; Hu, X.; Muenke, C.;
Schulten, K.; Michel, H. Structure 1996, 4, 581–597. (d) LH1-RC:
Roszak, A. W.; Howard, T. D.; Southall, J.; Gardiner, A. T.; Law,
C. J.; Isaacs, N. W.; Cogdell, R. J. Science 2003, 302, 1969–1972.
(e) PSI: Jordan, P.; Fromme, P.; Witt, H.-T.; Klukas, O.; Saenger,
W.; Krauss, N. Nature 2001, 411, 909–917. (f) PSI-LH1: Ben-Shem,
A.; Frolow, F.; Nelson, N. Nature 2003, 426, 630–635. (g) Plant LHC-
II at 2.72 Å resolution: Liu, Z.; Yan, H.; Wang, K.; Kuang, J.; Gui,
L.; An, X.; Chang, W. Nature 2004, 428, 287–292. (h) Plant LHC-II
at 2.5 Å resolution: Standfuss, J.; Terwisscha van Scheltinga, A. C.;
Lamborghini, M.; Ku¨hlbrandt, W. EMBO J. 2005, 24, 919–928. (j)
PSII at 3.5 Å resolution: Ferreira, K. N.; Iverson, T. M.; Maghlaoui,
K.; Barber, J.; Iwata, S. Science 2004, 303, 1831–1838. (k) PSII at
3.0 Å resolution: Loll, B.; Kern, J.; Saenger, W.; Zouni, A.; Biesiadka,
J. Nature 2005, 438, 1040–1044. (l) PSII at 2.9 Å resolution: Guskov,
A.; Kern, J.; Gabdulkhakov, A.; Broser, M.; Zouni, A.; Saenger, W.
Nature Stuct. Mol. Biol. 2009, 16, 334–342.
a The fatty alcohol (R-OH) esterifying the 17-propionic acid residue is
farnesol in Chlorobium species which have a mixture of homologues
differing in R8 (Et, n-Pr, i-Bu) and/or R12 (Me, Et); BChl c from Chloroflexus
aurantiacus has R8 ) Et and R12 ) Me, with various R groups derived
from phytol, stearol, cetol, geranyl-geraniol, etc. The formula shows the
(31S)-configuration, but both epimers are present in varying amounts, which
depend not only on the species but also on the culturing and illumination
conditions applied. b M is a central metal atom, S are solubilizing groups
such as long alkyl chains or 3,5-di-tert-butylphenyl residues, and RG1 and
RG2 are recognition groups, which may be identical or different and are
capable of ligating the central metal atom of adjacent molecules.10e
radiation for light-harvesting. Once a stray photon is captured,
its energy gets trapped by rapid energy-transfer steps with near-
unity quantum efficiency in a BChl-protein complex, the so-
called Fenna-Matthews-Olson (FMO) complex, which is
attached to the baseplate of the chlorosome. The FMO complex
actually was the very first (B)Chl-protein complex to be
crystallized,8 and its structure was solved by X-ray crystal-
lography.9 From the FMO complex, light energy is passed to
the special pair of BChls located in the nearby reaction center,
where charge separation occurs. The uniqueness of the chlo-
rosome as an antenna lies in the high chromophore density
achieved by self-assembly. Usually, with nonspecific aggregated
pigments, concentration quenching occurs which impairs the
light-harvesting ability.
We have been interested in mimicking the self-assembly
algorithm programmed into the BChl c, d, and e molecules with
fully synthetic pigments such as porphyrins, chlorins, and, more
recently, phthalocyanines.10 By inducing chromophores to self-
assemble, large functional nanostructures can be accessed
without the need to meticulously assemble a covalent scaffold
(5) Bahatyrova, S.; Frese, R. N.; Siebert, C. A.; Olsen, J. D.; van der
Werf, K.; van Grondelle, R.; Niederman, R. A.; Bullough, P. A.; Otto,
C.; Hunter, C. N. Nature 2004, 430, 1058–1062.
(6) (a) Manske, A. K.; Glaeser, J.; Kuypers, M. M. M.; Overmann, J.
Appl. EnViron. Microbiol. 2005, 71, 8049–8060. (b) Beatty, J. T.;
Overmann, J.; Lince, M. T.; Manske, A. K.; Lang, A. S.; Blankenship,
R. E.; Van Dover, C. L.; Martison, T. A.; Plumley, F. G. Proc. Natl.
Acad. Sci.U.S.A. 2005, 102, 9306–9310.
(7) (a) Bystrova, M. I.; Mal’gosheva, I. N.; Krasnovskii, A. A. Mol. Biol.
1979, 13, 440–451. (b) Smith, K. M.; Kehres, L. A. J. Am. Chem.
Soc. 1983, 105, 1387–1389. (c) Brune, D. C.; King, G. H.; Blanken-
ship, R. E. In Photosynthetic Light-HarVesting Systems; Scheer, H.,
Schneider, S., Eds.; Walter de Gruyter: Berlin, 1988; pp 141-151.
(d) Holzwarth, A. R.; Griebenow, K.; Schaffner, K. Z. Naturforsch.
1990, 45c, 203–206. (e) Balaban, T. S.; Holzwarth, A. R.; Schaffner,
K.; Boender, G.-J.; de Groot, H. J. M. Biochemistry 1995, 34, 15259–
15266. (f) Balaban, T. S.; Leitich, J.; Holzwarth, A. R.; Schaffner, K.
J. Phys. Chem. 2000, 104, 1362–1372. (g) Psˇencˇik, J.; Ikonen, T. P.;
Laurinma¨ki, P.; Merckel, M. C.; Butcher, S. J.; Serimaa, R. E.; Tuma,
R. Biophys. J. 2004, 87, 1165–1172. (h) For an excellent review, see:
Frigaard, N.-U.; Bryant, D. In Microbiology Monographs; Shively,
J. M., Ed.; Springer: Berlin, 2006; Vol. 2, pp 79-114. (i) For a very
recent novel chlorosome model and discussion of previous models,
see: Ganapathy, S.; Oostergetel, G. T.; Wawrzyniak, P. K.; Reus, M.;
Maqueo Chew, A. G.; Buda, F.; Boekema, E. J.; Bryant, D. A.;
Holzwarth, A. R.; de Groot, H. J. M. Proc. Natl. Acad. Sci. U.S.A.
2009, 106, 8525–8530.
(8) Olson, J. M. In The Photosynthetic Bacteria; Clayton, R. K., Sistrom,
W. R., Eds.; Plenum: New York, 1978; pp 161-178.
(9) Fenna, R. E.; Mathews, B. W. Nature 1975, 258, 573–577. For a more
recent structure from a different bacterium, see: Li, Y.-F.; Zhou, W.;
Blankenship, R. E.; Allen, J. P. J. Mol. Biol. 1997, 271, 456–471.
(10) (a) Balaban, T. S.; Bhise, A. D.; Fischer, M.; Linke-Schaetzel, M.;
Roussel, C.; Vanthuyne, N. Angew. Chem., Int. Ed. 2003, 42, 2140–
2144. (b) Balaban, T. S.; Bhise, A. D.; Linke-Schaetzel, M.; Van-
thuyne, N.; Roussel, C. Eur. J. Org. Chem. 2004, 3919–3930. (c)
Balaban, T. S.; Linke-Schaetzel, M.; Bhise, A. D.; Vanthuyne, N.;
Roussel, C.; Anson, C. A.; Buth, G.; Eichho¨fer, A.; Foster, K.; Garab,
G.; Gliemann, H.; Goddard, R.; Javorfi, T.; Powell, A. K.; Ro¨sner,
H.; Schimmel, Th. Chem.sEur. J. 2005, 11, 2267–2275. (d) Balaban,
T. S. Acc. Chem. Res. 2005, 38, 612–623. (e) Balaban, M. C.; Balaban,
T. S. J. Porph. Phthal. 2007, 11, 277–286. (f) Jochum, T.; Malla
Reddy, C.; Eichho¨fer, A.; Buth, G.; Szmytkowski, J.; Kalt, H.; Moss,
D.; Balaban, T. S. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 12736–
12741, and references therein.
9
J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009 14481

A R T I C L E S
Balaban et al.
of proteins11 or dendrimers.12 If practical applications are
envisaged, efficient synthetic methods, including optimized
HPLC separations, and/or stereoselective syntheses have to be
developed for accessing enantiopure compounds in sufficient
quantities. Stereochemical challenges are that BChls c, d, and
e have three chiral centers (asterisks in Chart 1), neglecting the
asymmetric carbon atoms of the lipophilic long-chain alcohol
residue (e.g., in phytol), and that the central magnesium atom
becomes an additional stereogenic element upon axial coordina-
tion.13 In the chlorosomes, both epimers of the 3-(1-hydroxy-
ethyl) group are present in varying amounts, and their exact
role in determining the supramolecular architecture is still poorly
understood. It has been demonstrated that, by exposing bacterial
cultures to different light intensities, the ratio of the 31-epimers
is drastically affected.14
In the present study we have optimized both chiral HPLC
conditions and an enantioselective monoreduction which permits
preparation of enantiopure self-assembling BChl d mimics. In
addition, we have assigned the configuration at the sole
31-stereocenter by independent chemical and spectroscopic
methods, which all lead to the same result. Subsequently, we
have investigated the self-assembly properties of the separated
enantiomers by stationary absorption, fluorescence, and circular
dichroism (CD) spectroscopy and by scanning tunneling electron
microscopy (STEM).
co-workers17 have established novel synthetic methods that
could easily be adapted to obtain self-assembling porphyrins
and chlorins with geminal dimethyl groups which prevent the
sometimes spontaneous aerial oxidation of chlorins to porphy-
rins, thus providing pure and stable compounds with a red-
shifted and increased absorption in the vis-NIR region in
comparison to that of porphyrins.
Results and Discussion
To equip porphyrins with recognition groups capable of
assembling supramolecular architectures, we started with the
preformed porphyrin, which was then functionalized.18 This
strategy allowed for the usual lower-yielding porphyrin-forming
reactions to occur at the beginning of synthetic sequences, which
thus permits improved atom management. Scheme 1 presents
the main synthetic transformations for obtaining the self-
assembling BChl mimics addressed in this work.
Diacetylations of 1-Cu. Porphyrin 1 as a free base (1-H2, M
) H2) can be easily obtained at a gram scale by the condensation
of dipyrromethane with 3,5-di-tert-butylbenzaldehyde with
Lewis (BF3 ·OEt2) or Brønsted (TFA) acid catalysis using the
moderately high dilution method (10-2 M) in dichloromethane,
at room temperature, optimized by Lindsey and co-workers.19
By this method, reproducible yields of 35-40% are attainable.
We usually employ an oven-dried 10-L flask and use technical
dichloromethane previously dried over calcium chloride and
deoxygenated by passing a stream of argon through a sintered
Related work by three other groups deserves mention in this
context: (i) Tamiaki et al.15 and, more recently, (ii) Wu¨rthner
et al.16 have used mainly semisynthetic approaches starting with
algal Chl a to obtain self-assembling chlorins; (iii) Lindsey and
(15) (a) Balaban, T. S.; Tamiaki, H.; Holzwarth, A. R.; Schaffner, K. J.
Phys. Chem. 1997, 101, 3424–3431. (b) Miyatake, T.; Tamiaki, H.;
Holzwarth, A. R.; Schaffner, K. HelV. Chim. Acta 1999, 82, 797–
810. (c) Miyatake, T.; Oba, T.; Tamiaki, H. ChemBioChem 2001, 335–
342. (d) Yagai, S.; Tamiaki, H. J. Chem. Soc., Perkin Trans.1 2001,
3135–3144. (e) Yagai, S.; Miyatake, T.; Tamiaki, H. J. Org. Chem.
2002, 67, 49–58. (f) de Boer, I.; Matysik, J.; Amakawa, M.; Yagai,
S.; Tamiaki, H.; Holzwarth, A. R.; de Groot, H. J. M. J. Am. Chem.
Soc. 2003, 125, 13374–13375. (g) Kunieda, M.; Mizoguchi, T.;
Tamiaki, H. Photochem. Photobiol. 2004, 7, 55–61. (h) Sasaki, S.;
Tamiaki, H. Bull. Chem. Soc. Jpn. 2004, 77, 797–800. (i) Kunieda,
M.; Tamiaki, H. J. Org. Chem. 2005, 70, 820–828. (j) Miyatake, T.;
Tamiaki, H. J. Photochem. Photobiol. C: Photochem. ReV. 2005, 6,
89–107. (k) Sasaki, S.; Tamiaki, H. J. Org. Chem. 2006, 71, 2648–
2654. (l) Mizoguchi, T.; Shoji, A.; Kunieda, M.; Miyashita, H.;
Tsuchiya, T.; Mimuro, M.; Tamiaki, H. Photochem. Photobiol. Sci.
2006, 5, 291–299. (m) Saga, Y.; Kim, T.-Y.; Hisai, T.; Tamiaki, H.
Thin Solid Films 2006, 500, 278–282. (n) Tamiaki, H.; Michitsuji,
T.; Shibata, R. Photochem. Photobiol. Sci. 2008, 7, 1225–1230. (o)
Kunieda, M.; Tamiaki, H. J. Org. Chem. 2008, 73, 7686–7694. (p)
Mizoguchi, T.; Kim, T.-Y.; Sawamura, S.; Tamiaki, H. J. Phys. Chem.
B 2008, 112, 16759–16765.
(11) For polypeptide maquettes or myoglobin-binding tetrapyrrolic chro-
mophores, see: (a) Discher, B. M.; Noy, D.; Strzalka, J.; Moser, C. C.;
Lear, J. D.; Blasie, J. K.; Dutton, P. L. Biochemistry 2005, 44, 12329–
12343. (b) Noy, D.; Discher, B. M.; Rubtsov, I. V.; Hochstrasser,
R. M.; Dutton, P. L. Biochemistry 2005, 44, 12344–12354. (c) Pro¨ll,
S.; Wilhelm, B.; Robert, B.; Scheer, H. Biochim. Biophys. Acta
(Bioenergetics) 2006, 1757, 750–763. (d) Markovic´, D.; Pro¨ll, S.;
Bubenzer, C.; Scheer, H. Biochim. Biophys. Acta (Bioenergetics) 2007,
1767, 897–904.
(12) (a) Gensch, T.; Hofkens, J.; Herrmann, A.; Tsuda, K.; Verheijen, W.;
Vosch, T.; Christ, T.; Basche´, T.; Mu¨llen, K.; De Schryver, F. C.
Angew. Chem., Int. Ed. 1999, 38, 3752–3756. (b) Bhyrappa, P.;
Vaijayanthimala, G.; Suslick, K. S. J. Am. Chem. Soc. 1999, 121, 262–
263. (c) Adronov, A.; Gilat, S. L.; Fre´chet, J. M. J.; Ohta, K.; Neuwahl,
F. V. R.; Fleming, G. R. J. Am. Chem. Soc. 2000, 122, 1175–1185.
(d) Adronov, A.; Gilat, S. L.; Fre´chet, J. M. J.; Ohta, K.; Neuwahl,
F. V. R.; Fleming, G. R. J. Am. Chem. Soc. 2000, 122, 1175–1185.
(e) Weil, T.; Wiesler, U. M.; Herrmann, A.; Baur, R.; Hofkens, J.;
De Schryver, F. C.; Mu¨llen, K. J. Am. Chem. Soc. 2001, 123, 8101–
8108. (f) Weil, T.; Reuther, E.; Mu¨llen, K. Angew. Chem., Int. Ed.
2002, 41, 1900–1904. (g) Lee, L. F.; Adronov, A.; Schaller, R. D.;
Fre´chet, J. M. J.; Saykally, R. J. J. Am. Chem. Soc. 2003, 125, 536–
540. (h) Ballester, P.; Gomila, R. M.; Hunter, C. A.; King, A. S. H.;
Twyman, L. J. Chem. Commun. 2003, 38–39. (i) Schweitzer, G.;
Gronheid, R.; Jordens, S.; Lor, M.; De Belder, G.; Weil, T.; Reuther,
E.; Mu¨llen, K.; De Schryver, F. C. J. Phys. Chem. A 2003, 107, 3199–
3207. (j) Metivier, R.; Kulzer, F.; Weil, T.; Mu¨llen, K.; Basche´, T.
J. Am. Chem. Soc. 2004, 126, 14364–14365. (k) Weil, T.; Reuther,
E.; Beer, C.; Mu¨llen, K. Chem.sEur. J. 2004, 10, 1398–1414. (l)
Herrmann, A.; Mu¨llen, K. Chem. Lett. 2006, 35, 978–986. (m) Flors,
C.; Oesterling, I.; Schnitzler, T.; Fron, E.; Schweitzer, G.; Sliwa, M.;
Herrmann, A.; van der Auweraer, M.; de Schryver, F. C.; Mu¨llen, K.;
Hofkens, J. J. Phys. Chem. C 2007, 111, 4861–4870.
(16) (a) Ro¨ger, C.; Mu¨ller, M.; Lysetska, M.; Miloslavina, Y.; Holzwarth,
A. R.; Wu¨rthner, F. J. Am. Chem. Soc. 2006, 128, 6542–6543. (b)
Ro¨ger, C.; Miloslavina, Y.; Brunner, D.; Holzwarth, A. R.; Wu¨rthner,
F. J. Am. Chem. Soc. 2008, 130, 5929–5939. (c) Huber, V.; Sengupta,
S.; Wu¨rthner, F. Chem.sEur. J. 2008, 14, 7791–7807. (d) Ganapathy,
S.; Sengupta, S.; Wawrzyniak, P. K.; Huber, V.; Buda, F.; Baumeister,
U.; Wu¨rthner, F.; de Groot, H. J. M. Proc. Natl. Acad. Sci. U.S.A.
2009, 106, 11472–11477.
(17) (a) Yao, Z.; Bhaumik, J.; Dhanalekshmi, S.; Ptaszek, M.; Rodriguez,
P. A.; Lindsey, J. S. Tetrahedron 2007, 63, 10657–10670. (b) Ptaszek,
M.; Yao, Z.; Savithri, D.; Boyle, P. D.; Lindsey, J. S. Tetrahedron
2007, 63, 12629–12638. (c) Lindsey, J. S.; Chinnasamy, M.; Fan, D.
U.S. Patent US2008/0029155 A1, February 7, 2008.
(18) (a) Balaban, T. S.; Goddard, R.; Linke-Schaetzel, M.; Lehn, J.-M.
J. Am. Chem. Soc. 2003, 125, 4233–4239. (b) Balaban, T. S.;
Eichho¨fer, A.; Krische, M. J.; Lehn, J.-M. HelV. Chim. Acta 2006,
89, 333–351.
(13) (a) Balaban, T. S. Photosynth. Res. 2005, 86, 251–262. (b) Oba, T.;
Tamiaki, H. Bioorg. Med. Chem. 2005, 13, 5733–5739.
(14) (a) Bobe, F. W.; Pfennig, N.; Swanson, K. L.; Smith, K. M.
Biochemistry 1990, 29, 4340–4348. (b) Borrego, C. M.; Gerola, P. D.;
Miller, M.; Cox, R. P. Photosynth. Res. 1999, 59, 159–166. (c)
Steensgaard, D. B.; Wackerbarth, H.; Hildebrandt, P.; Holzwarth, A. R.
J. Phys. Chem. B 2000, 104, 10379–10386. (d) Ishii, T.; Kimura, M.;
Yamamoto, T.; Kirihata, M.; Uehara, K. Photochem. Photobiol. 2000,
71, 567–573.
(19) (a) Lindsey, J. S.; Schreiman, I. C.; Hsu, H. C.; Kearney, P. C.;
Marguerettaz, A. M. J. Org. Chem. 1987, 52, 827–836. (b) Littler,
B. J.; Miller, M. A.; Hung, C.-H.; Wagner, R. W.; O’Shea, D. F.;
Boyle, P. D.; Lindsey, J. S. J. Org. Chem. 1999, 64, 1391–1396. (c)
Littler, B. J.; Ciringh, Y.; Lindsey, J. S. J. Org. Chem. 1999, 64, 2864–
2872.
9
14482 J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009

Mimics of Self-Assembling Chlorosomal Bacteriochlorophylls
A R T I C L E S
Scheme 1a
lattices, thus permitting good-quality single crystals to be
obtained which are amenable to X-ray analyses; and (v) due to
the intense 1H NMR singlet(s), it allows easy detection at
submicromolar concentrations of products in complex reaction
mixtures and sometimes it gives a hint at the symmetry of the
product due to the aniso- or isochronous character of the tert-
butyl groups (Vide infra).
Diacetylation of 1-Cu with an excess of acetic anhydride and
tin tetrachloride as the Friedel-Crafts catalyst at 0-3 °C for
30 min gave an almost equimolar mixture (44:56) of the 3,17-
and 3,13-diacetyl-substituted porphyrins 2a-Cu and 3a-Cu. The
yields over different runs were consistently over 27%, with
∼35% yields when carbon disulfide was used as the solvent.
Under these conditions, both diacetylated isomers were obtained,
and their separation was achieved by careful column chroma-
tography on silica gel after demetalation. By stirring at room
temperature for 90 min with a degassed mixture of concentrated
sulfuric and trifluoroacetic acids, high demetalation yields (e.g.,
98%) of the corresponding free bases, 2a-H2 and 3a-H2, were
achieved at a 600-mg scale. Yields tended to decrease when
working at larger scales, e.g., 86% for a 1.5-g mixture of 2a-
Cu and 3a-Cu. The chromatographic separation worked better
for the free bases than for their copper or zinc complexes, and
by elution with dichloromethane from silica gel, the reddish
3,17-diacetyl isomer 2a-H2 was eluted first, followed closely
by the greenish 3,13-diacetyl isomer 3a-H2. On a preparative
HPLC SiO2 Dynamax 100-Å column run with 3% 2-propanol
in dichloromethane, the elution order was reversed for the copper
complexes, i.e., 3a-Cu eluted before 2a-Cu. On the same
column, baseline separation was achieved with an isocratic
elution using dichloromethane/n-hexane (1:1 v/v) for the cor-
responding free bases, with 2a-H2 eluting before 3a-H2.
Acetylations of geoporphyrins have been described by Callot
and co-workers.22 By adapting their procedure, with only a slight
excess of acetic anhydride and short reaction times (2-10 min),
monoacetylation of 1-Cu occurred, which permitted isolation
of the 3-acetyl copper porphyrin in fair yields. This compound
could be subjected to a second and different acylation step; its
Vilsmeier formylation failed, however, to give the 3-acetyl-15-
formyl copper porphyrin. The latter compound was conveniently
prepared by reversing the order and performing first the mono-
meso-formylation of 1-Cu, followed by the Friedel-Crafts
monoacetylation.23
a Reagents and conditions for each step are as follow. (i) Diacylation:
H3C-COCl, AlCl3, CS2, 0-5 °C or Ac2O, SnCl4, CS2. (ii) Copper
demetalation: TFA:H2SO4, 1:1 v/v, room temperature. (iii) Statistical
monoreduction: NaBH4, MeOH, CH2Cl2 or CHCl3, room temperature. (iV)
Zinc metalation: Zn(OAc)2, CHCl3:MeOH, 4:1 v/v.
glass frit immersed directly in the reaction vessel, which is
protected from light by an aluminum foil. Oxidation of the
intermediate porphyrinogen is more convenient with 2,3-
dichloro-5,6-dicyano-p-benzoquinonone (DDQ) than with p-
chloranil due to the easier chromatographic separation of the
product 1-H2. For Friedel-Crafts diacetylation or Vilsmeier
diformylation, the free-base porphyrin needs to be metalated
with copper acetate under standard conditions.20 Interestingly,
while 1-Cu is 5-monoformylated and 5,15-diformylatedsi.e.,
the Vilsmeier reagent attacks the more reactive free meso
positionssFriedel-Crafts acylations occur preferentially at the
ꢀ-pyrrolic positions (Vide infra).
The 3,5-di-tert-butylphenyl group, which was introduced into
porphyrin chemistry by Crossley,21 has several advantages: (i)
due to steric hindrance by the tert-butyl groups, the phenyl
substituents are inert in electrophilic substitution reactions, which
can thus attack exclusively the porphyrin macrocycle; (ii) the
2, 8, 12, and 18 positions of the porphyrin are also within the
sterical shielding cone of the frequently rotating (on the NMR
time scale) 10,20-di-tert-butylphenyl groups, so that these
ꢀ-pyrrolic positions are not attacked by acylium ions in
Friedel-Crafts acylations; (iii) it induces high solubility in
medium-polarity and nonpolar solvents and inhibits aggregation
by π-stacking of the macrocyles; (iv) it packs well in crystal
Diacylations of 1-Cu were undertaken by using several acyl
chlorides and AlCl3 in carbon disulfide at 0-5 °C. After short
reaction times and in concentrated CS2 solutions (∼5 mM), in
less than 5 min, the 3,17-diacyl isomers 2a-c-Cu were formed
preferentially and could be isolated virtually free of the 3,13-
isomers. At higher temperatures and longer reaction times, the
3,13-diacyl isomers 3a-c-Cu were formed as the minor products.
In more dilute solutions (<0.1 mM), the reaction is slower and
can be stopped before the 3,13-diacyl isomer starts to be formed.
Other acyl chlorides, e.g., butyroyl to myristoyl chloride,
behaved similarly.23 It was thus quite remarkable that, out of
the 12 possible diacylated isomers of 1-Cu, one can drive the
reaction regiospecifically to obtain the 3,17-diacyl isomer as
(20) (a) Smith, K. M.; Bisset, G. M. F.; Case, J. J.; Tabba, H. D.
Tetrahedron Lett. 1980, 21, 3747–3750. (b) Smith, K. M.; Bisset,
G. M. F.; Tabba, H. D. J. Chem. Soc., Perkin Trans. 1 1982, 1982,
581–585.
(22) (a) Jeandon, C.; Bauder, C.; Callot, H. J. Energy Fuels 1990, 4, 665–
667. (b) Jeandon, C.; Ruppert, R.; Callot, H. J. J. Org. Chem. 2006,
71, 3111–3120.
(21) (a) Crossley, M. J.; Burn, P. L. J. Chem. Soc., Chem. Commun. 1987,
39–40. (b) Crossley, M. J.; Burn, P. L. J. Chem. Soc., Chem. Commun.
1991, 1569–1571. (c) Reimers, J. R.; Lu¨, T. X.; Crossley, M. J.; Hush,
N. S. Nanotechnology 1996, 7, 424–429.
(23) Bhise, A. D. Ph.D. Thesis, Universita¨t Karlsruhe (TH), 2005 (Wis-
senschaftliche Berichte FZKA 7174, ISSN 0947-8620).
9
J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009 14483

A R T I C L E S
Balaban et al.
The 3,17-diacetyl free-base 2a-H2 gave crystals suitable for
X-ray structure analysis, which unequivocally proved the
structure as shown in Figure 3. The two acetyl groups are
oriented in opposite directions, both being coplanar with
the porphyrin macrocycle. This arrangement minimizes the
dipole moment within the monomeric unit. Interestingly,
extensive π-stacking occurs with a 3.55-Å interplanar distance
within strongly coupled dimers having the acetyl groups in non-
overlapping positions. Within such dimers, the total dipole
moment is entirely averaged out. This large degree of π-overlap,
which is uncommon because usually porphyrins tend to stack
with an offset geometry,25 is due to the electron-withdrawing
effect of the two acetyl groups, which makes the porphyrin rings
π-poor, minimizing their repulsion and allowing them to come
close to each other.
Monoreductions of Diacetyl Porphyrins 2a-H2 and 3a-H2.
With the pure diacetyl porphyrins in hand, we performed the
reduction of one acyl group to a hydroxyalkyl substituent. The
two acetyl groups being equivalent (homotopic), the statistical
monoreduction of any of them leads to the same product. These
reactions were easily monitored by TLC, as the retention factors
of the monoalcohol (ketol) and especially of the diol species
were much smaller than those of the starting diacyl compounds.
The reduction with NaBH4 in a dichloromethane or chloroform/
methanol mixture (4:1 v/v) could be stopped after the appearance
of the two spots corresponding to the racemic (RR/SS) and meso-
diol (RS) compounds at about 65-80% conversion of the
diacetylated porphyrins. It was very fortunate that the monoal-
cohol could be isolated. This was probably due to the mutual
activation of the two acetyl groups toward nucleophilic reduc-
tion. Once the first acetyl group was reduced, the second one
reacted more sluggishly.
Among different possible reagents and conditions for the
enantioselective monoreduction of the diacetylated porphyrins
2a-H2 and 3a-H2,26 the Corey-Bakshi-Shibata reagent {CBS,
(3aS)- or (3aR)-1-methyl-3,3-diphenyltetrahydro-3H-pyrrolo[1,2-
c][1,3,2]oxazaborrole}, which had previously been employed
by Montforts in the synthesis of hematoporphyrin stereoiso-
mers,27 gave the best results. Under our optimized conditions,
which consisted of reacting 2a-H2 in toluene at -80 °C for 1 h
with 1 equiv of (R)-CBS and 4 equiv of catecholborane, we
attained a 42% yield and 79% ee of the (S)-3-(1-hydroxyeth-
yl)porphyrin 4a-H2 (for assignments of the configuration at the
31-stereocenter, see the next section). The (S)-CBS reductions
consistently gave lower ee’s; we explain this by the variable
quality of the respective commercial reagents, which in this case
are not derived from the natural chiral pool. To our disappoint-
ment, under the same conditions, reaction of 3a-H2 with 1 equiv
of (R)-CBS, when compared to 4a-H2, proceeded more slowly
and in lower yield (32%) and only to 54% ee. With 1 equiv of
(S)-CBS, the yield was only 22% and the ee was 36%. Higher
reaction temperatures or other commercial amine-borane or
dimethylsulfide-borane complexes, although sometimes giving
higher yields, led to the sacrifice of the ee values.
Figure 1. Mechanistic explanation of preferential formation of the 3,17-
diacylporphyrins. Color coding: carbon, black; oxygen, red; nitrogen, blue;
copper, yellow; aluminum, light gray; and chlorine, green. Hydrogen atoms
have been omitted for clarity.
the only product. A plausible mechanistic explanation for
this astonishing regioselectivity follows and is illustrated in
Figure 1.
The acylium cation attacks the most reactive position on the
porphyrin periphery, which has to be one of the two free meso-
positions. The 5-meso-π-complex can decay to the σ-complex,
which, however, due to steric hindrance with the two adjacent
ꢀ-pyrrolic protons, cannot lead to an acyl group in conjugation
with the porphyrin macrocycle. Indeed, our meso-acetyl com-
pounds, obtained by an indirect method,10a have been proven
by X-ray crystallography to have perpendicular carbonyl groups
and are red colored.10c meso-Formyl porphyrins, on the other
hand, are green colored, and the formyl group is in the porphyrin
plane, fully conjugated,10b which is also supported by FT-IR
data.10f Thus, the 5-meso-π-complex prefers to shift to a 3-acyl-
π-complex, where substitution can now occur with formation
of the conjugated carbonyl group, and this product can also be
isolated. Why does the diacylation favor the formation of the
3,17-diacyl compounds when the 3,13- or even 3,7-diacylated
product should be equally stable according to semiempirical
PM3 calculations?24 We exclude that the square planar copper
ion plays a role in chelating the acylium chloroaluminate
acylating agent. Rather, as usual with stoichiometric AlCl3
acylations, the 3-acyl group will strongly complex AlCl3. This
leads to severe steric hindrance of both a renewed attack in the
meso-5 position and the rotation of the neighboring 20-di-tert-
butylphenyl group, which must reside in a frozen twisted
conformation (see Supporting Information, Figure S17, for
conformational analyses). The second acylium ion will thus
preferentially enter the only other available meso-position (15).
The shift of this latter π-complex can occur more rapidly to
the adjacent 17-ꢀ-pyrrolic position, indicated by the red arrow
in Figure 1, which is opened up by the blocked 20-di-tert-
butylphenyl group. Migration to the 13-position occurs more
slowly due to the more frequently rotating 10-di-tert-butylphenyl
group.
Due to progress in separation techniques, including their up-
scaling, in order to obtain enantiopure BChl mimics, it appears
The structural assignment can be deduced quite straightfor-
wardly from the 1H NMR spectra of the free bases, since
compounds 2a-c-H2 have nonequivalent 10- and 20-(3,5-di-tert-
butylphenyl) groups. The splitting of the phenyl protons is
exemplarily shown in Figure 2 for 2a-H2 in comparison with
3a-H2.
(25) Hunter, C.; Sanders, J. K. M. J. Am. Chem. Soc. 1990, 112, 5525–
5534.
(26) Mo¨ssinger, D. Diploma Thesis, Universita¨t Karlsruhe (TH), 2006.
(27) (a) Kusch, D.; To¨llner, E.; Lincke, A.; Montforts, F.-P. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 784–787. (b) Kusch, D.; Montforts, F.-P.
Tetrahedron: Asymmetry 1995, 6, 867–870. (c) Tamiaki, H.; Kouraba,
M.; Takeda, K.; Kondo, S.; Tanikaga, R. Tetrahedron: Asymmetry
1998, 9, 2101–2111.
(24) HyperChem Program Package, version 7.5; HyperCube Inc.: Gaines-
ville, FL, 2003.
9
14484 J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009

Mimics of Self-Assembling Chlorosomal Bacteriochlorophylls
A R T I C L E S
Figure 2. Aromatic region of the 300-MHz 1H NMR spectra in CDCl3 of 2a-H2 (bottom trace) in comparison with 3a-H2 (top). For the full spectra,
including other examples, see the Supporting Information, Figures S5-S7.
Figure 3. (Left) Monomeric unit of 2a-H2 and (right) the unit cell, which comprises a π-stacked dimer. Details of the data collection and crystallographic
analysis are given in the Supporting Information.
advantageous to use the cheap reducing reagent NaBH4, giving
rise to the racemate, which then can be quantitatively resolved
on a variety of chiral HPLC columns into the pure 31-
enantiomers in quantitative yields, without any loss of material.
Assignment of the Absolute Configuration at the 31-Stereo-
center. Chiral HPLC successfully resolved the enantiomers of
4a-H2 and 5a-H2. Analytical HPLC on 11 chiral stationary
phases and a typical chromatogram for 4a-H2 are described in
the Supporting Information. In particular, the ULMO (S,S)
column, eluted with ethanol, permitted separation of the
enantiomers of 4a-H2 and 5a-H2. Uray et al.,28 the inventors of
the ULMO (S,S) stationary phase, reported the order of elution
for more than 15 benzylic secondary alcohols on this column.
On the basis of the finding that the (R)-enantiomers always
eluted first, we can assume that the “faster” enantiomers of 4a-
H2 and 5a-H2 on ULMO (S,S) also have a (31R) absolute
configuration. However, assignment of the absolute configura-
tion from the elution order obtained by chiral HPLC must be
done cautiously29 and needs to be validated by other methods.
Our above preliminary assignment was actually confirmed
by the fact that the first eluted enantiomers were identical to
those formed stereoselectively by the monoreduction of the
(28) (a) Maier, N. M.; Uray, G. J. Chromatogr. A 1996, 732, 215–230. (b)
Uray, G.; Maier, N. M.; Niederreiter, K. S.; Spitaler, M. M.
J. Chromatogr. A 1998, 799, 67–81. (c) Uray, G.; Stampfer, W.;
Fabian, W. M. F. J. Chromatogr. A 2003, 992, 151–157.
(29) Roussel, C.; Del Rio, A.; Pierrot-Sanders, J.; Piras, P.; Vanthuyne, N.
J. Chromatogr. A 2004, 1037, 311–328.
9
J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009 14485

A R T I C L E S
Balaban et al.
Figure 4. HPLC elution from a chiral ULMO (S,S) column with n-hexane/2-propanol (4:1 v/v), 4 mL/min, and UV-vis absorption spectra of the reduction
products of 4a-H2 with 1 equiv of (R)-CBS (left) and (S)-CBS (right). The sharp Soret band, with higher loadings of the column, sometimes saturates the
diode array detector.
corresponding diacetyl compounds employing the (S)-CBS
reagent (Figure 4). The second eluted enantiomers were those
formed predominantly with the (R)-CBS reagent and, due to
the accepted mechanism of ketone reductions,30 were thus
attributed to the (31S)-configuration.
knowledge, no CD spectra of any monomeric porphyrins bearing
just one chiral center in an aliphatic side chain have so far been
reported in the literature,33 in contrast to numerous significant
CD effects described for various chiral porphyrin derivatives
possessing other elements of chirality or having more than one
porphyrin residue.34 Thus, unlike for the self-assembly experi-
ments (see below), the CD effects could not be accurately
measured and interpreted.
In order to investigate the configuration at the stereogenic
center more closely, two other independent methods were
applied, one based on an oxidative degradation of benzylic
acetates and ensuing stereoanalysis of the resulting lactic
acetates, and the other based on the NMR analysis of the
respective Mosher esters.
For a robust assignment of the absolute stereostructures of
the porphyrins synthesized, measurement of circular dichroism
(CD) spectra and their interpretation by quantum chemical CD
calculations31 seemed to be the method of choice. However,
no significant CD spectra were obtained by standard offline CD
measurements of the enantiomers (R)-4a-Zn (99% ee) and (S)-
4a-Zn (98% ee), which were synthesized by zinc(II) insertion
into the previously resolved free bases (R)-4a-H2 and (S)-4a-
H2 (see above). In none of the solvents usedsdichloromethane,
methanol, or 2-propanolswere any clear, reproducible, and thus
interpretable CD effects observed. To exclude any interference
by possible impurities, we also performed online CD measure-
ments (in the stopped-flow mode), i.e., by HPLC-CD
coupling,31c butsas for the offline measurementssno CD
spectra were obtained.32 Obviously, the influence of the
peripheral, freely rotating stereogenic center on the flat, strongly
absorbing chromophore is too small to result in a measurable
CD curve for the monomeric, nonaggregated species.31 Fully
in line with this result is the fact that, to the best of our
Assignment of the 31-Configuration by Oxidative Degrada-
tion. For this purpose, a ruthenium-mediated oxidative degrada-
tion procedure developed in Wu¨rzburg for chiral benzylic
alcohols35 and related chiral molecules36,37 was applied to each
of the two 31-acetoxy-substituted enantiomers of 4a-H2, resulting
in the respective enantiomers of 2-acetoxypropionic acid (6,
Figure 5), which were both analyzed by GC on a CDX-B chiral
phase. The absolute configurations of these enantiomeric 2-acet-
oxypropionic acids were attributed to the particular signals by
GC comparison with racemic 2-acetoxypropionic acid as a
reference, with the more rapidly eluting enantiomer being (R)-
configured and the “slower” one having the (S)-configuration.
In the case of the 2-acetoxypropionic acid obtained from the
faster eluting enantiomer of 4a-H2, a single peak was obtained
(30) (a) Corey, E. J.; Helal, C. J. Angew. Chem., Int. Ed. 1998, 37, 1986–
2012. (b) Corey, E. J.; Bakshi, R. K. Tetrahedron Lett. 1990, 31, 611–
614.
(31) (a) Bringmann, G.; Busemann, S. In Natural Product Analysis;
Schreier, P., Herderich, M., Humpf, H. U., Schwab, W., Eds.; Vieweg:
Wiesbaden, 1998; pp 195-212. (b) Bringmann, G.; Gulder, T. A. M.;
Reichert, M.; Gulder, T. Chirality 2008, 20, 628–642. (c) Bringmann,
G.; Bruhn, T.; Maksimenka, K.; Hemberger, Y. Eur. J. Org. Chem.
2009, 2717–2727.
(33) For a review on porphyrins as CD reporter groups, see: Huang, X.;
Nakanishi, K.; Berova, N. Chirality 2000, 12, 237–255.
(34) For some selected, recent examples, see: (a) Borovkov, V. V.; Inoue,
Y. In Supramolecular Chirality; Crego-Calama, M., Reinhoudt, D. N.,
Eds.; Topics in Current Chemistry 265; Springer: Berlin, 2006; pp
89-146. (b) Mizumura, M.; Shinokubo, H.; Osuka, A. Angew. Chem.,
Int. Ed. 2008, 47, 5378–5381. (c) Kai, H.; Nara, S.; Kinbara, K.; Aida,
T. J. Am. Chem. Soc. 2008, 130, 6725–6727. (d) Shoji, Y.; Tashiro,
K.; Aida, T. Chirality 2008, 20, 420–424. (e) Bringmann, G.; Go¨tz,
D. C. G.; Gulder, T. A. M.; Gehrke, T. H.; Bruhn, T.; Kupfer, T.;
Radacki, K.; Braunschweig, H.; Heckmann, A.; Lambert, C. J. Am.
Chem. Soc. 2008, 130, 17812–17825.
(32) (a) The results of the quantum chemical CD calculations performed
in parallel thus could not be compared with experimental CD curves
and will be reported elsewhere. (b) In a similar way, strongly colored
compounds often do not give reliable [R]D values, due to their intense
UV absorption. (c) Mutanyatta, J.; Bezabih, M.; Abegaz, B.; Dreyer,
M.; Brun, R.; Kocher, N.; Bringmann, G. Tetrahedron 2005, 61, 8475–
8484. (d) Thomson, R. H. Naturally Occurring Quinones IV Recent
AdVances; Blackie Academic and Professional: London, 1997; p 134.
(e) Sharma, P. K.; Khanna, R. N.; Rohatgi, B. K.; Thomson, R. H.
Phytochemistry 1988, 27, 632–633. (f) Khanna, R. N.; Sharma, P. K.;
Thomson, R. H. J. Chem. Soc., Perkin Trans. 1 1987, 1821–1824. (g)
Capon, R. J.; Groves, D. R.; Urban, S.; Watson, R. G. Aust. J. Chem.
1993, 46, 1245–1253. (h) Bringmann, G.; Ru¨denauer, S.; Go¨tz,
D. C. G.; Gulder, T. A. M.; Reichert, M. Org. Lett. 2006, 8, 4743–
4746.
(35) Bringmann, G.; Mu¨nchbach, M.; Michel, M. Tetrahedron: Asymmetry
2000, 11, 3167–3176.
(36) Bringmann, G.; God, R.; Scha¨ffer, M. Phytochemistry 1996, 43, 1393–
1403.
(37) Bringmann, G.; Mu¨nchbach, M.; Messer, K.; Koppler, D.; Michel,
M.; Schupp, O.; Wenzel, M.; Louis, A. M. Phytochemistry 1999, 51,
693–699.
9
14486 J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009

Mimics of Self-Assembling Chlorosomal Bacteriochlorophylls
A R T I C L E S
Figure 5. Enantiomeric assignment of the degradation product 6 of the faster enantiomer of 4a-H2 (a) and of the more slowly eluting one (c), by GC
analysis in comparison with the standard (b). (i) AcCl, NEt3; (ii) RuCl3, NaIO4; (iii) GC analysis on a chiral CDX-B phase.
(Figure 5a), having the same retention time as that found for
the more rapid (R)-enantiomer of 2-acetoxypropionic acid
(Figure 5b). This was further confirmed by coelution experi-
ments (see Figure S4, Supporting Information), leading to the
assignment of the (R)-configuration for this enantiomer. In the
other case, a peak coeluting with that originating from (S)-2-
acetoxypropionic acid was seen (Figure 5c), resulting in the
determination that the stereogenic center of the more slowly
eluting enantiomer from the ULMO (S,S) and Whelk (S,S)
stationary phases of 4a-H2 was (S)-configured. Therefore, the
degradation experiments unambiguously verified the results
obtained by the CBS reduction.
) δ[(S)-MTPA ester] - δ[(R)-MTPA ester]. The δ values for
the 31-methyl group in (R)- and (S)-MTPA esters were measured
in CDCl3 to be 2.330 and 2.424 ppm, respectively. The value
of ∆ ) 2.424 - 2.330 ) +0.094 ppm was positive. By contrast,
all the ∆ values for the assigned protons on the peripheral
positions of the porphyrin π-system gave consistently negative
values, which tended to decrease in absolute value with
increasing distance from the 31-chiral position (Chart 2).
According to the modified Mosher rule,38b the secondary alcohol
was predicted to have the (R)-configuration at the 31-position.
Moreover, the ∆ value in the methoxy group of the MTPA esters
was positive, +0.174 ppm, indicating that the enantiomerically
pure A2 has the (R)-configuration at the 31-position on the basis
of the other modified Mosher rule.38c
Assignment of the 31-Configuration Using Mosher’s
Esters. Recently, a modified Mosher’s method was reported to
be effective for the configurational assignment of epimerically
pure secondary alcohols, including chlorophyll derivatives.38
Using this method, we investigated three samples resulting from
two independent preparative HPLC separations of 4a-H2. The
two HPLC runs were labeled 1 and 2, and the first eluted
enantiomer was labeled A, while the second eluted enantiomer
was labeled B. Enantiomerically pure A2 was treated with
commercially available (+)-MTPA-Cl in dry pyridine to afford
the corresponding Mosher’s ester, (R)-MTPA ester (see Scheme
2). Similarly, reaction of A2 with (-)-MTPA-Cl gave the
corresponding (S)-MTPA ester. Table S1 in the Supporting
The other enantiomer from the same run, B2, was esterified
with (+)- or (-)-MTPA-Cl to give the corresponding esters
(R)-MTPA and (S)-MTPA, respectively. The modified Mosher
method using ∆δH showed that this enantiomer unambiguously
had the S-configuration at the 31-position (see Chart 2 and Table
S1).
As a double-check, another esterified sample from run 1 (B1)
was compared with B2; it gave identical results (Table S2,
Supporting Information) within 0.004 ppm, allowing us to have
a high degree of confidence in the NMR-derived assignments.
1
Information lists the H chemical shifts (δ values) of the (R)-
MTPA and (S)-MTPA esters together with their differences, ∆
Self-Assembly Experiments. With the separated enantiomers
of 4a-H2 and 5a-H2 after zinc insertion, which afforded 4a-Zn
and 5a-Zn, we performed self-assembly experiments in nonpolar
solvents. The self-assembled species are characterized by red-
shifted maxima, in comparison to the monomers, as in J-
(38) (a) Tamiaki, H.; Kitamoto, H.; Nishikawa, A.; Hibino, T.; Shibata, R.
Bioorg. Med. Chem. 2004, 12, 1657–1666. (b) Ohtani, I.; Kusumi,
T.; Kashman, Y.; Kakisawa, H. J. Am. Chem. Soc. 1991, 113, 4092–
4096. (c) Kelly, D. R. Tetrahedron: Asymmetry. 1999, 10, 2927–2934.
9
J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009 14487

A R T I C L E S
Balaban et al.
Scheme 2. Synthesis of (R)- and (S)-Mosher’s Estersa
a The faster eluting enantiomer of 4a-Zn gave a pair of diastereomeric Mosher’s esters, which proved to have the (31R)-configuration. The 10,20-
substituents R are the 3,5-di-tert-butylphenyl group.
Chart 2. Differences (∆δ × 1000) between Chemical Shifts δ
(ppm) of Diastereoisomeric MTPA Esters of 7a-Zn: (a) ∆δ )
[δ(A2-(S)-MTPA) - δ(A2-(R)-MTPA] and (b) [δ(B2-(S)-MTPA) -
δ(B2-(R)-MTPA]a
favored over a homochiral one. Another possible explanation
could be that the samples, after HPLC separation (which uses
2-propanol) and zinc metalation (which uses methanol), still
contain residual amounts of hydroxylic species which hinder
self-assembly by coordinating the zinc atoms. We could exclude
this hypothesis by preparing the pure enantiomers at a 10 mg
scale, followed by washing their dichloromethane solution with
brine, which completely removes any alcoholic traces, followed
by evaporation of the solvent and prolonged drying under
vacuum (10-2 bar).
Figures 6 and 7 present the absorption spectra of the
racemates and of the separated enantiomers of 4a-Zn and 5a-
Zn, respectively, after inducing their self-assembly in dry
n-heptane by dilution of an initial concentrated dichloromethane
solution, and after addition of a suprastoichiometric amount of
methanol, which leads to a complete disruption of the aggregates
due to the formation of the Zn-methanol adducts having the
Soret band at ∼434 nm. Also presented in Figures 6 and 7 are
the CD spectra of the resolved enantiomers. Figure 6 displays
absorption and CD spectra of quiescent solutions of 4a-Zn
aggregated in dry n-heptane with a path length of 5 mm.
It has been claimed that a collinear arrangement of the
“northern” 31-hydroxy substituent with both the central metal
atom (Mg or Zn) and the carbonyl substituent situated in the
“southern” half of a tetrapyrrolic macrocycle is a sine-qua-non
condition to observe chlorosomal-type self-assemblies.40 The
results presented here, together with our earlier examples,10
clearly demonstrate the self-assembly ability of (rac)-4a-Zn,
which does not satisfy this collinearity condition. These appar-
ently contradicting observations can be reconciled by the fact
that, in absorption spectra, the aggregate maxima of (rac)-4a-
Zn are less prominent and appear only at higher concentrations
than for (rac)-5a-Zn. Additional examples are presented in the
Supporting Information (Figure S9). Consistently, the mono-
a Conclusion: A2, (31R)-configuration; B2, (31S)-configuration. The
10,20-substituents R are the 3,5-di-tert-butylphenyl group.
aggregates,39 but with considerable broadening, which we
ascribe to size heterogeneity. Under the same experimental
conditions, the racemates consistently showed increased ag-
gregate maxima (at ca. 450 and 635 nm) for both 4a-Zn and
5a-Zn as compared to the resolved enantiomers, which showed
more intense Soret bands due to the monomeric forms at ∼422
nm. This supports the fact that a heterochiral self-assembly is
(39) (a) Da¨hne, L. J. Am. Chem. Soc. 1995, 117, 12855–12860. (b) De
Rossi, U.; Da¨hne, S.; Meskers, C. J.; Dekkers, P. J. M. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 760–763. (c) Wang, M.; Silva, G. L.; Armitage,
B. A. J. Am. Chem. Soc. 2000, 122, 9977–9986. (d) Hannah, K. C.;
Armitage, B. A. Acc. Chem. Res. 2004, 37, 845–853.
(40) Tamiaki, H.; Kimura, S.; Kimura, T. Tetrahedron 2003, 59, 7423–
7435.
9
14488 J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009

Mimics of Self-Assembling Chlorosomal Bacteriochlorophylls
A R T I C L E S
Figure 7. (A) 5a-Zn self-assembled in quiescent (motionless) n-heptane.
Black trace: racemate. Green trace: (31R) first-eluted enantiomer on ULMO
(S,S). Red trace: (31S) second-eluted enantiomer. Note that the racemate
has a higher intensity of the aggregate bands at 505 and 639 nm. The
baseline shift is due to light scattering by the nanosized aggregates. The
red and green traces were scaled to the monomer maximum at 422 nm by
multiplying by 1.6 and 3.8, respectively. Upon addition of methanol, the
∼505 and ∼639 nm maxima vanish completely, along with the baseline
shift, due to complete disruption of the aggregates. (B) Corresponding CD
spectra measured in the same cuvettes. Note the more intense Cotton effects
for the (31R)- enantiomer. This is probably due to a higher concentration
of oligomeric and higher-aggregated species than for the self-assemblies
generated by the (31S)-enantiomer. Upon methanol addition, both CD spectra
become almost silent. Due to the intense absorption, very little light actually
reaches the detector, giving spectra comparable to those for the 4a-Zn
regioisomer after MeOH addition (Figure 6C).
meric Soret bands are much more intense for 4a-Zn than in the
5a-Zn isomer. However, in the CD spectra, intense and multiple
Cotton effects are encountered after self-assembly for both 4a-
Zn and 5a-Zn, and these are nearly mirror-images for the
respective resolved enantiomers, proving the presence of oli-
gomeric self-assembled species. Upon addition of methanol the
CD spectra become practically silent. As previously reported,
the chirality of the 31-stereocenter induces the supramolecular
chirality of these self-assemblies.10a,b However, the size het-
erogeneity and its strong concentration dependence lead to
highly variable intensities of the observed Cotton effects. It has
been calculated that, for aggregates of chiral chromophores,
depending on the aggregate size, even a reversal of the Cotton
Figure 6. (A) Absorption spectra of 4a-Zn. Black trace: racemate self-
assembled in n-heptane; the strong baseline shift is due to light scattering
by the nanosized aggregates. Green solid trace: (31R) [first-eluted enantiomer
on ULMO (S,S)] self-assembled in n-heptane with 5 mm path length. Green
dotted trace: (31R) after disassembly with MeOH (1 mm path length). Red
solid trace: 31(S) [second-eluted enantiomer] self-assembled in n-heptane
with 5 mm path length. Magenta dashed trace (superimposed on the green
dotted trace): (31S) after disassembly with methanol (1 mm path length).
(B) Corresponding CD spectra of quiescent solutions in the same 5 mm
cuvette. (C) Enlarged portion of the monomeric zinc-methanol adducts
after disassembly with methanol, which in part B appear almost superim-
posed on the zero line (path length 5 mm). Due to the intense absorption
by the Soret band, very little light reaches the detector, so the signal in this
region, indicated by arrows, is covered by noise.
9
J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009 14489

A R T I C L E S
Balaban et al.
effects can be encountered.41 To summarize, the racemate that
self-assembles most readily under the absence of chirality-
inducing perturbations gave a silent CD spectrum, while resolved
enantiomers gave oligomeric species with approximately mirror-
image Cotton effects.
The same behavior could be reproduced with 4b-Zn and 4c-
Zn (Supporting Information). Suspensions of their self-as-
semblies are stable in n-heptane for much longer periods of time
than those of 4a-Zn. After a few hours, the latter tend to
flocculate out of solutions and settle at the bottom of the cuvettes
used for measuring optical spectra. Gentle shaking of the cuvette
completely disrupts the aggregates, restoring a homogeneous
suspension. This property is quite similar to the one encountered
previously for BChl c aggregates.7f For the measurement of
absorption, fluorescence, or CD spectra, homogeneous suspen-
sions were used, assuring reproducible results.
Ribo´ and co-workers have convincingly described that an
achiral porphyrin, namely meso-tetrakis-4-sulfophenylporphyrin,
forms chiral J-aggregates in aqueous solutions and that their
helicities are induced by the stirring sense.42 More recently,
Tsuda, Aida, and collaborators have demonstarted an astonishing
porphyrin dendrimer that reversibly self-assembles into mac-
roscopic chiral J-aggregates with helicities induced by the
vortexing sense.43 In a preliminary study with enantiopure 4a-
Zn, we observed only negligible effects by reversing the stirring
direction of n-heptane solutions (Supporting Information, Fig-
ures S11 and S12). It may well be that, with longer-chain
compounds (e.g., 4b-Zn, 4c-Zn, or those obtained by diacyla-
tions with other fatty acid chlorides) and with more prolonged
and vigorous stirring, such chirality-inducing effects could be
possible in nonpolar solvents. These are less viscous than water
and thus induce lower shearing forces. In the previous reports
by Ribo´ et al., vigorous vortex stirring was effected for at least
24 h.41 In our case, 10 min of moderately intense stirring failed
to induce the supramolecular chirality of the self-assemblies,
which is thus primarily dictated by the 31-stereocenter. We
explain for 4a-Zn the absence of a large effect of the stirring
sense on the supramolecular chirality by the low content of large,
self-assembled species within the enantiopure samples. A
prerequisite condition for successful induction of supramolecular
chirality by the stirring sense appears to be that a large
proportion of higher macroscopic aggregates has to be present
after the nucleation step and during the growth phase. Shearing
forces can then induce chirality at the macroscopic scale. In
the absence of large aggregates, at the nanoscale, no chirality-
inducing effects could be observed upon reversing the vortex
sense. We compare thus as chirality-inducing effects a sub-
stituent effect versus the stirring sense. Preliminary experiments
using (rac)-4b-Zn and, to a lesser extent, (rac)-4a-Zn show
mirror-image CD spectra of films deposited on the cuvette walls
upon reversing the vortex sense. However, we suspect that a
Figure 8. (A) Black trace: solution fluorescence spectrum of aggregates
from the racemic 5a-Zn in dry n-heptane recorded in a 1 × 1 cm cuvette
with an excitation wavelength of 480 nm. Gray trace: same cuvette after
addition of two drops of methanol. (B) Film of ∼65 nm thickness deposited
on the walls of the same fluorescence cuvette. Excitation wavelength was
480 nm for the black trace and 420 nm for the gray trace.
large contribution to the CD signal comes from linear dichroism
(LD). Further experiments with oriented CD and fluorescence
detected LD are currently in progress.
Typical fluorescence spectra are presented in Figure 8. One
can selectively excite the monomeric (in the Soret band at 422
nm) or the self-assembled species (between 450 and 500 nm).
After addition of methanol, the Zn-methanol adducts present
two intense fluorescence bands, which are more red-shifted
(typically at 630 and 700 nm) than those of the monomers
accompanying the methanol-free self-assembled species. One
can thus easily distinguish between the emission of monomers,
either free or coordinated by methanol, and the emission of the
self-assemblies, which is quenched by a factor of about 10.
Interestingly, if the fluorescence cuvettes have not previously
been dried for a longer period (usually overnight) in an oven at
>110 °C, the adsorbed water film nucleates the formation on
the quartz walls of a thin and ordered aggregate film which is
visible to the eye and has a yellow-greenish color. The film
fluorescence has a red-shifted shoulder at ∼720 nm, as reported
before.44 With randomly oriented chromophores, solid-state
fluorescence is usually almost entirely self-quenched. In our
case, although the films have at least an order of magnitude
less intense fluorescence, this is not entirely quenched due to
their orderly supramolecular arrangement. This phenomenon has
(41) (a) Didraga, C.; Klugkist, J. A.; Knoester, J. J. Phys. Chem. B 2002,
106, 11474–11486. (b) Prokhorenko, V. I.; Steensgaard, D. B.;
Holzwarth, A. R. Biophys. J. 2003, 85, 3173–3186. (c) Pugzlys, A.;
Augulis, R.; van Loosdrecht, P. H. M.; Didraga, C.; Malyshev, V. A.;
Knoester, J. J. Phys. Chem. B 2006, 110, 20268–20276.
(42) (a) Ribo´, J. M.; Crusats, J.; Sagues, F.; Claret, J. M.; Rubires, R.
Science 2001, 292, 2063–2066. (b) Escudero, C.; Crusats, J.; Diez-
Perez, I.; El-Hachemi, Z.; Ribo´, J. M. Angew. Chem., Int. Ed. 2006,
45, 8032–8035. (c) El-Hachemi, Z.; Mancini, G.; Ribo´, J. M.; Sorrenti,
A. J. Am. Chem. Soc. 2008, 130, 15176–15184. (d) El-Hachemi, Z.;
Arteaga, O.; Canillas, A.; Crusats, J.; Escudero, C.; Kuroda, R.; Harada,
T.; Rosa, M.; Ribo´, J. M. Chem.sEur. J. 2008, 14, 6438–6443.
(43) Tsuda, A.; Alam, Md. A.; Harada, T.; Yamaguchi, T.; Ishii, N.; Aida,
T. Angew. Chem., Int. Ed. 2008, 46, 8198–8202.
9
14490 J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009

Mimics of Self-Assembling Chlorosomal Bacteriochlorophylls
A R T I C L E S
been termed “aggregate-induced fluorescence” and has been
studied recently by Swager, Park, and co-workers.45 Figure S12
(Supporting Information) shows the fluorescence spectra of the
separated enantiomers of 5a-Zn, which are quite similar to each
other. Again, no major effect could be seen in the fluorescence
spectra upon short stirring of the self-assemblies in clockwise
or counter-clockwise directions, except for the fact that the
aggregate emission was much smaller than for the racemic
samples. Fluorescence measurements thus also confirmed that
the separated enantiomers do not form as easily and frequently
large aggregates as their racemates, as evidenced by the
monomer peaks and the fact that the 710-720 nm aggregate
peak is much larger for the racemate than for the separated
enantiomers.
Scanning Transmission Electron Microscopy (STEM). STEM
studies were performed on self-assemblies of (rac)-4a-Zn and
on the respective resolved enantiomers in a similar fashion as
described before for 5a-Zn.10c Three conclusions can be derived
from these investigations: First, the micrographs (Figure 9 and
Supporting Information, Figures S14 and S15) show quite large
nanorods, confirming that a collinear arrangement of the 31-
hydroxyalkyl substituent with the central metal atom and the
carbonyl substituent is not mandatory for self-assembly to occur;
second, the propensity of the aggregates is much reduced (almost
2 orders of magnitude) in the separated enantiomers as compared
to the racemate, proving that the heterochiral self-assembly is
thermodynamically favored; and third, all imaged samples, either
racemic or enantiopure, show the same fibrillar fine structure
at high magnifications, with 1.60 ( 0.05 nm spacings, having
the parallel fringes aligned along the long axis of the nanorod.
We cannot exclude that the nanorods are initially hollow (i.e.,
nanotubes or nanotubules) and then, upon drying under ultrahigh
vacuum, become collapsed because the tilting of the sample
does not change the spacings between fringes. It seems that a
rather dense packing of the porphyrins occurs within such rods,
which also suffer very easily from damage by the electron beam.
To obtain high-resolution images, one has to focus the beam
within one range, briefly move the sample, and briskly record
the micrograph of a nearby area, which within a few seconds
then becomes “burnt out”. Fast Fourier transmission of the
fringes allows clear diffraction to be obtained (shown in the
insets in Figure 9), from which the spacing between the fringes
can be calculated.
Figure 9. (Left) TEM bright-field micrographs showing assemblies into
nanorods of 4a-Zn: (A) the racemate; (B) the (31R)-enantiomer, and (C)
the (31S)-enantiomer (scale bars are 200 nm). (Right) The corresponding
high-resolution images (scale bars are 10 nm) showing striations parallel
to the long axis of the nanorod. (Insets) Fast Fourier transform images
displaying diffraction spots corresponding to the periodic structure of the
molecules. The spacings between fringes were found to be 1.59, 1.61, and
1. 55 nm, respectively.
statistical analysis, with samples with various enantiomeric
excesses, could shed light on this dilemma, which touches on
the long-standing problem related to the obligatory existence
of both 31-epimers in the chlorosomal BChls.
The observed substantially different rates and propensities
for the formation of homochiral and heterochiral aggregates are
rather unique and have far-reaching implications. According to
the classification proposed by Soloshonok of major intermo-
lecular interactions leading to the self-disproportionation of
enantiomers under the conditions of achiral-phase chromatog-
raphy, preferential formation of heterochiral over homochiral
oligomers is considered to be the most rare and has not been
observed so far.46
The resolved enantiomers show mirror-image Cotton effects
due to self-aggregation (Figures 6 and 7), but the micrographs
present the same lamellar, fine structure without noticeable
nanoscopic chirality. We cannot entirely exclude that the imaged
nanorods are actually heterochiral assemblies, although the ee’s
after HPLC resolution were >98%. The paucity of nanorods in
the samples of the separated enantiomers could thus be due to
the less probable formation of homochiral self-assemblies. A
(44) (a) Linke-Schaetzel, M.; Bhise, A. D.; Gliemann, H.; Koch, T.;
Schimmel, T.; Balaban, T. S. Thin Solid Films 2004, 451-452, 16–
21. (b) Huijser, A.; Marek, P. L.; Savenije, T. J.; Siebbeles, L. D. A.;
Scherer, T.; Hauschild, R.; Szmytkowski, J.; Kalt, H.; Hahn, H.;
Balaban, T. S. J. Phys. Chem. C 2007, 111, 11726–11733. (c) Balaban,
M. C.; Eichho¨fer, A.; Buth, G.; Hauschild, R.; Szmytkowski, R.; Kalt,
H.; Balaban, T. S. J. Phys. Chem. B. 2008, 112, 5512–5521.
Conclusion
The exact nature of the chlorosomal architecture will probably
remain elusive in the absence of high-resolution single-crystal
(45) (a) Deans, R.; Kim, J.; Machacek, M. R.; Swager, T. M. J. Am. Chem.
Soc. 2000, 122, 8565–8566. (b) Luo, J.; Xie, Z.; Lam, J. W. Y.; Cheng,
L.; Chen, H.; Qiu, C.; Kwok, H. S.; Zhan, X.; Liu, X.; Zhu, D.; Tang,
B. Z. Chem. Commun. 2001, 1740–1741. (c) An, B.-K.; Kwon, S.-
K.; Jung, S.-D.; Park, S. Y. J. Am. Chem. Soc. 2002, 124, 14410–
14415. (d) Bhongale, C. J.; Chang, C.-W.; Diau, E. W.-G.; Hsu, C.-
S.; Dong, Y.; Tang, B.-Z. Chem. Phys. Lett. 2006, 419, 444–449.
(46) (a) Soloshonok, V. A. Angew. Chem., Int. Ed. 2006, 45, 766–769. (b)
Soloshonok, V. A.; Berbasov, D. O. J. Fluorine Chem. 2006, 127,
597–603. (c) Soloshonok, V. A.; Berbasov, D. O. Chim. Oggi/Chem.
Today 2006, 24 (3), 44–47. (d) Soloshonok, V. A.; Ueki, H.;
Yasumoto, M.; Mekala, S.; Hirschi, J. S.; Singleton, D. A. J. Am.
Chem. Soc. 2007, 129, 12112–12113.
9
J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009 14491

A R T I C L E S
Balaban et al.
data. Currently, several controversial models are discussed, such
as the lamellar model of Psˇencˇik et al.7g which is based on small-
angle X-ray scattering data, and the tubular models based on a
combination of solid state NMR-data and electron microscopy
images.7i Various groups have had repeated attempts to crystal-
lize the chlorosomal BChls, but only nondiffracting material
could be isolated after self-assembly.47 Our studies on synthetic
BChl mimics have led to self-assemblies that have spectroscopic
features similar to those of their respective natural counterparts,
and in several of our cases the exact supramolecular architecture
could be gleaned by solving crystal structures on the basis of
synchrotron X-ray data. It appears possible to infer from our
crystal structure(s) the chlorosomal architecture,10d,f although
the solubility-inducing groups, which are involved only in
recessive weak supramolecular interactions, are quite different
from the ones found in BChls. In the present study, after
optimizing regioselective diacylations of porphyrins and HPLC
methods to obtain enantiopure BChl mimics, we have studied
their ability to self-assemble. The spectroscopic and nanoscopic
investigations point toward a more favored heterochiral self-
assembly process in the case of only one stereogenic center
present on the porphyrin periphery as 3-hydroxyalkyl groups.
We infer from this that the natural chlorosomal BChls use the
same mechanism for self-assembly, which leads to larger
structures if both 31-epimers are present. This correlates well
with the observed up-regulation of the biosyntheses under low-
light illumination conditions of the less abundant (31S)-
epimers.14 The bias toward a less favored self-assembly with
epimerically enriched (31R)-homologues could provide a way
to assemble smaller chlorosomes and thus regulate the light-
harvesting ability when sufficient quantities of photons are
available. The chirality of the 31-stereocenter appears thus to
play a key role in regulating a physical property, namely the
antenna function. Ongoing studies in our laboratories should
shed light on this long-standing problem, which has evolutionary
implications, as well as allowing the design of efficient artificial
light-harvesting systems.
Acknowledgment. We gratefully acknowledge Dr. T. Bruhn
and Dr. M. Reichert (University of Wu¨rzburg) for their computa-
tional contributions. We thank the Deutsche Forschungsgemein-
schaft (DFG) for partial support in Karlsruhe through the Center
for Functional Nanostructures (CFN) through projects C3.2 and
C3.5 “Energy Transfer Studies of Artificial Mimics of Natural Light
Harvesting Molecules”. Collaborative work between Karlsruhe and
Marseille was supported by the CNRS through PICS no. 3777. This
work in Japan was partially supported by Grants-in-Aid for
Scientific Research (B) (No. 19350088, to H.T.) from the Japanese
Society for the Promotion of Science (JSPS) and for Young
Scientists (B) (no. 19750148, to T.M.) from the Ministry of
Education, Culture, Sports, Science and Technology. This work in
Wu¨rzburg was supported by the Degussa Stiftung and the Studi-
enstiftung des deutschen Volkes e.V. (fellowships to D.C.G.G.).
We dedicate this work to Prof. Jean-Marie Lehn on the occasion
of his 70th birthday.
Supporting Information Available: Detailed experimental
procedures and additional illustrative spectral data. This material
is available free of charge via the Internet at http://pubs.acs.org.
(47) (a) Fajer, J., and Smith, K. M., personal communications. (b) Over a
period of 15 years, we have been unable to grow single crystals from
a variety of BChl c homologues.
JA905628H
9
14492 J. AM. CHEM. SOC. VOL. 131, NO. 40, 2009