Supramolecular complexation and enantiodifferentiating photocyclodimerization of 2-anthracenecarboxylic acid with 4-aminoprolinol derivatives as chiral hydrogen-bonding templates

Article abstract of DOI:10.1021/jo901792t

(Figure Presented) The photochirogenesis of 2-anthracenecarboxylic acid (AC) complexed to a hydrogen-bonding template (TKS159) was investigated to obtainmechanistic information on how chirogenesis is achieved for the dimerization of AC. Complexation of AC to TKS159 leads to the shielding of one of the two surfaces of the prochiral AC molecule. The two diastereomeric AC-TKS complexes, i.e., re-AC-TKS and si-AC-TKS, were characterized by changes in the UV-vis, fluorescence, and circular dichroism spectra and excited-state lifetimes. The ee is not simply determined by the diastereomeric ratio of the re- and si-AC-TKS complexes but also depends on the relative lifetimes of the diastereomeric complexes. The relative population of the re and si complexes was calculated from the enantiomeric excess (ee) for the products, taking into account the relative lifetimes of the two complexes. These studies established a protocol that can be used to reveal the mechanism for photochirogenesis by investigating the ground state and the excited state behavior of supramolecular systems. 2009 American Chemical Society.

Full text of DOI:10.1021/jo901792t

pubs.acs.org/joc
Supramolecular Complexation and Enantiodifferentiating
Photocyclodimerization of 2-Anthracenecarboxylic Acid with
4-Aminoprolinol Derivatives as Chiral Hydrogen-Bonding Templates
Yuko Kawanami,† Tamara C. S. Pace,‡ Jun-ichi Mizoguchi,§ Toshiharu Yanagi,§
Masaki Nishijima,† Tadashi Mori,† Takehiko Wada,†, Cornelia Bohne,*,‡ and
Yoshihisa Inoue*,†
†Department of Applied Chemistry and Center for Advanced Science and Innovation, Osaka University, 2-1
Yamadaoka, Suita 565-0871, Japan, ‡Department of Chemistry, University of Victoria, P.O. Box 3065,
Victoria, BC, Canada V8W 3V6, and Bio/Fine Chemicals Department, Nagase ChemteX Corporation,
§
2-2-3 Murotani, Nishi-ku, Kobe 651-2241, Japan. Present address: Institute of Multidisciplinary Research
for Advanced Materials, Tohoku University, 2-1-1 Katahira, Aoba, Sendai 980-8577, Japan
bohne@uvic.ca; inoue@chem.eng.osaka-u.ac.jp
Received August 18, 2009
The photochirogenesis of 2-anthracenecarboxylic acid (AC) complexed to a hydrogen-bonding
template (TKS159) was investigatedtoobtain mechanistic information onhow chirogenesisisachieved
for the dimerization of AC. Complexation of AC to TKS159 leads to the shielding of one of the two
surfaces of the prochiral AC molecule. The two diastereomeric AC-TKS complexes, i.e., re-AC-TKS
and si-AC-TKS, were characterized by changes in the UV-vis, fluorescence, and circular dichroism
spectra and excited-state lifetimes. The ee is not simply determined by the diastereomeric ratio of the
re- and si-AC-TKS complexes but also depends on the relative lifetimes of the diastereomeric
complexes. The relative population of the re and si complexes was calculated from the enantiomeric
excess (ee) for the products, taking into account the relative lifetimes of the two complexes. These
studies established a protocol that can be used to reveal the mechanism for photochirogenesis by
investigating the ground state and the excited state behavior of supramolecular systems.
Introduction
understand how chirogenesis can be achieved and manip-
ulated.1-5 A chiral environment is required when prochiral
reactants are used, and this chiral environment can
be achieved in anisotropic media, such as solids6-12 or
Chirality is prevalent in nature and is essential in many
biological functions. In this context, supramolecular
photochirogenesis is one approach being developed to
(6) Chong, K. C. W.; Sivaguru, J.; Schichi, T.; Yoshimi, Y.; Ramamurthy,
V.; Scheffer, J. R. J. Am. Chem. Soc. 2002, 124, 2858–2859.
(7) Joy, A.; Ramamurthy, V.; Scheffer, J. R.; Corbin, D. R. Chem.
Commun. 1998, 1379–1380.
(8) Joy, A.; Scheffer, J. R.; Ramamurthy, V. Org. Lett. 2000, 2, 119–121.
(9) Scheffer, J. R. In Chiral Photochemistry; Inoue, Y., Ramamurthy, V.,
Eds.; Marcel Dekker: New York, 2004.
(10) Braga, D.; Grepioni, F. Angew. Chem. 2004, 43, 4002–4011.
(11) Matsumoto, A. Top. Curr. Chem. 2005, 254, 263–305.
(12) Tanaka, K.; Toda, F. Chem. Rev. 2000, 100, 1025–1074.
(1) Griesbeck, A. G.; Meierhenrich, U. J. Angew. Chem. 2002, 41, 3147–
3154.
(2) Inoue, Y. Chem. Rev. 1992, 92, 741–770.
€
(3) Muller, C.; Bach, T. Aust. J. Chem. 2008, 61, 557–564.
(4) Wada, T.; Inoue, Y. In Chiral Photochemistry; Inoue, Y., Ramamurthy,
V., Eds.; Marcel Dekker: New York, 2004; pp 341-384.
(5) Yang, C.; Inoue, Y. In Cyclodextrin Materials Photochemistry, Pho-
tophysics and Photobiology; Douhal, A., Ed.; Elsevier: New York, 2006;
pp 241-265.
7908 J. Org. Chem. 2009, 74, 7908–7921
Published on Web 09/17/2009
DOI: 10.1021/jo901792t
r
2009 American Chemical Society

Kawanami et al.
JOCArticle
SCHEME 1. Photocyclodimerization of 2-Anthracenecarboxylic Acid (AC) and Chiral Templates TKS159 and TM166 That Are
Epimeric to Each Other
liquid crystals13,14 or in isotropic media with the use
of chiral host systems.4 Examples of supramolecular host
systems in solution for which photochirogenesis has
been investigated include cyclodextrins,15-21 DNA,22 and
hydrogen-bonding templates.23-25
on the amide functionality.23,24,26-34 Host-guest binding
was directed by a small number (two or three) of hydrogen
bonds, and stereochemical control was accomplished for
photochemical reactions by shielding one of the enantiofaces
of the prochiral molecule. Moderate to high enantiomeric
excesses (ee) were observed. Many of the reactions studied
were unimolecular, but examples were also reported for
bimolecular reactions where one reaction partner is com-
plexed to the chiral template.35,36
Photochirogenesis involving bimolecular reactions adds
another layer of complexity because the two reaction
partners can either be in close proximity at the time of
excitation or they have to diffuse and form an encounter
complex before products are formed. Anthracenes photo-
dimerize readily,37,38 and when asymmetrically substituted,
such as in the case of 2-anthracenecarboxylic acid (AC),
form four configurational isomers of which two products
are chiral (Scheme 1). The photodimerization of AC has
been used as a model reaction in several supramolecular
systems.17,18,20,21,39-43
In anisotropic solid media the spatial relationship is well-
defined and dictates the stereochemical outcome of reac-
tions. Mechanistic information is drawn from the knowledge
of the spatial distribution of molecules in the media, but
changes in the stereochemical outcome are frequently diffi-
cult to achieve without changing the whole medium. In
contrast, supramolecular host-guest systems in isotropic
media provide the ability to manipulate the conformations
and orientations of the prochiral reactants in the chiral host
systems by making subtle changes in environmental factors
within the host-guest system without having to change the
bulk environment, i.e., the solvent system. The drawback for
the use of supramolecular systems in isotropic media is that
mechanistic information is not as readily available as that
from the structural determination in anisotropic media.
Therefore, it is essential to develop experimental approaches
to elucidate the mechanisms that operate in photochirogen-
esis in isotropic media to establish the conceptual framework
necessary for the design of new chiral hosts with increased
chiral discrimination ability.
(26) Aechtner, T.; Dressel, M.; Bach, T. Angew. Chem. 2004, 43, 5849–
5851.
€
(27) Bach, T.; Aechtner, T.; Neumulller, B. Chem. Eur. J. 2002, 8, 2464–
2475.
(28) Bach, T.; Bergmann, H. J. Am. Chem. Soc. 2000, 122, 11525–11526.
(29) Bach, T.; Bergmann, H.; Harms, K. Angew. Chem. 2000, 39, 2302–
2304.
(30) Bach, T.; Bergmann, H.; Harms, K. Org. Lett. 2001, 3, 601–603.
Photochirogenesis has been achieved previously with the
use of hydrogen-bonding templates, such as templates based
€
(31) Bauer, A.; Westkamper, F.; Grimme, S.; Bach, T. Nature 2005, 436,
(13) Leigh, W. J.; Workentin, M. S. In Handbook of Liquid Crystals;
Demun, D., Ed.; Wiley-VCH: Weinheim, 1998; pp 839-895.
(14) Ishida, Y.; Kai, Y.; Kato, S.; Misawa, A.; Mano, S.; Matsuoka, Y.;
Saigo, K. Angew. Chem. 2008, 47, 8241–8245.
1139–1140.
(32) Brandes, S.; Seiling, P.; Bach, T. Synlett 2004, 14, 2588–2590.
(33) Selig, P.; Bach, T. J. Org. Chem. 2006, 71, 5662–5673.
€
(34) Muller, C.; Bauer, A.; Bach, T. Angew. Chem. 2009, 48.
(15) Inoue, Y.; Dong, F.; Yamamoto, K.; Tong, L.-H.; Tsuneishi, H.;
Hakushi, T.; Tai, A. J. Am. Chem. Soc. 1995, 117, 11033–11034.
(16) Inoue, Y.; Wada, T.; Sugahara, N.; Yamamoto, K.; Kimura, K.;
Tong, L.; Gao, X.; Hou, Z.; Liu, Y. J. Org. Chem. 2000, 65, 8041–8050.
(17) Nakamura, A.; Inoue, Y. J. Am. Chem. Soc. 2003, 125, 966–972.
(18) Nakamura, A.; Inoue, Y. J. Am. Chem. Soc. 2005, 127, 5338–5339.
(19) Rao, V. P.; Turro, N. J. Tetrahedron Lett. 1989, 30, 4641–4644.
(20) Tamaki, T.; Kokubu, T.; Ichimura, K. Tetrahedron 1987, 43, 1485–
1494.
(21) Yang, C.; Nakamura, A.; Fukuhara, G.; Origane, Y.; Mori, T.;
Wada, T.; Inoue, Y. J. Org. Chem. 2006, 71, 3126–3136.
(22) Wada, T.; Sugahara, N.; Kawano, M.; Inoue, Y. Chem. Lett. 2000,
1174–1175.
(35) Grosch, B.; Orlebar, C. N.; Herdtwec, E.; Kaneda, M.; Wada, T.;
Inoue, Y.; Bach, T. Chem. Eur. J. 2004, 10, 2179–2189.
(36) Grosch, B.; Orlebar, C. N.; Herdtweck, E.; Massa, W.; Bach, T.
Angew. Chem. 2003, 42, 3693–3696.
(37) Bouas-Laurent, H.; Castellan, A.; Desvergne, J.-P.; Lapouyade, R.
Chem. Soc. Rev. 2000, 29, 43–55.
(38) Bouas-Laurent, H.; Castellan, A.; Desvergne, J.-P.; Lapouyade, R.
Chem. Soc. Rev. 2001, 30, 248–263.
(39) Yang, C.; Fukuhara, G.; Nakamura, A.; Origane, Y.; Fujita, K.;
Yuan, D.-Q.; Mori, T.; Wada, T.; Inoue, Y. J. Photochem. Photobiol., A
2005, 173, 375–383.
(40) Yang, C.; Nakamura, A.; Wada, T.; Inoue, Y. Org. Lett. 2006, 8,
3005–3008.
(23) Bach, T.; Bergmann, H.; Grosch, B.; Harms, K. J. Am. Chem. Soc.
2002, 124, 7982–7990.
(24) Cauble, D. F.; Lynch, V.; Krische, M. J. J. Org. Chem. 2003, 68,
15–21.
(25) Mizoguchi, J.; Kawanami, Y.; Wada, T.; Kodama, K.; Anzai, K.;
Yanagi, T.; Inoue, Y. Org. Lett. 2006, 8, 6051–6054.
(41) Wada, T.; Nishijima, M.; Fujisawa, T.; Sugahara, N.; Mori, T.;
Nakamura, A.; Inoue, Y. J. Am. Chem. Soc. 2003, 125, 7492–7493.
(42) Nishijima, M.; Pace, T. C. S.; Nakamura, A.; Mori, T.; Wada, T.;
Bohne, C.; Inoue, Y. J. Org. Chem. 2007, 72, 2707–2715.
(43) Nishijima, M.; Wada, T.; Mori, T.; Pace, T. C. S.; Bohne, C.; Inoue,
Y. J. Am. Chem. Soc. 2007, 129, 3478–3479.
J. Org. Chem. Vol. 74, No. 20, 2009 7909

Kawanami et al.
JOCArticle
The formation of 3 with an ee of 70-80% was observed
when the photodimerization was performed in a smectic
liquid crystal formed using a chiral mesogen. A very low ee
(2%) was observed when 1-anthracenecarboxylate was
incorporated in this liquid crystal.14 A similar strategy
using a chiral cationic gelator has also been applied for
the photodimerization of AC incorporated as counteranion
in the gel to give exclusively the head-to-head cyclodimers
but only -10% ee for 3.44 It is important to note that these
examples constitute a reaction in a “rigid chiral” anisotro-
pic environment. AC photodimerization was studied when
two AC molecules form complexes with γ-cyclodextrin
(γ-CD).17,18,20,21,39,40 This reaction occurs immediately
upon excitation because the two reaction partners are
incorporated in the same γ-CD cavity. The ee for 2 and 3
for the native γ-CD were þ32% and -3%, respectively,
where the “þ” and “-” signs are related to the order of
elution from the HPLC column and are not assigned to an
absolute configuration.17 The ee for 2 was changed to
-58% when a capped γ-CD was employed,40 while the ee
for 3 can be changed to -41% using electrostatic interac-
tions in a γ-CD modified with a pendant arm containing
positive charges.18 These examples show how changes to a
chiral host system in solution lead to significant changes in
the ee for the chiral products of a bimolecular photoreac-
tion. The photodimerization of AC in bovine41,42 and
human serum albumin43,45 is one example where the chir-
ogenesis is the outcome of a dynamic reaction. The proteins
have several binding sites for AC and in the case of bovine
serum albumin the highest ee was observed for AC bound to
a site with modest binding affinity. In the case of human
serum albumin a very high ee was observed (79% for 2 and
88% for 3) at moderate AC/protein ratios (3-5), while a
modest ee (50% for 2 and 14% for 3) persisted even when
most of the AC was bound to weak binding sites of the
protein. The presence of multiple binding sites on the
protein makes it difficult to unravel the mechanistic aspects
that influence the enhancement of photochirogenesis in a
dynamic bimolecular reaction.
FIGURE 1. UV/vis absorption spectra of 0.1 mM AC in CH2Cl2
upon gradual addition of TKS159 at 25 °C. TKS/AC ratios: (a) 0,
(b) 0.2, (c) 0.4, (d) 0.6, (e) 1.6, (f) 3.6, (g) 5.6, (h) 7.6, (i) 9.6, (j) 11.6,
(k) 30, (l) 60, and (m) 120.
The host-guest complex involving the chiral template and
prochiral reactant provide a chiral environment for the
reaction to occur, while the lability of the supramolecular
system is instrumental for the release of the product. These
features are essential48 to achieve the catalytic photochiro-
genesis observed by Bach and co-workers.31,34
The objective of the present work was to obtain mechan-
istic information on how photochirogenesis occurs by com-
bining information from studies on the characterization of
the AC-TKS complex, product distribution, and the photo-
physics of AC. Studies at different TKS/AC ratios and
temperatures were required to find conditions where the
reactivity of free AC was minimized and the formation of
AC hydrogen-bonded dimers, present at low temperatures,
was absent. Complexation of AC to TKS159 led to the
formation of two diastereomeric complexes for which dif-
ferent lifetimes are observed for the singlet excited state of
AC. This observation is the first experimental proof for the
involvement of two diastereomeric precursor complexes
in the photochirogeneses with chiral supramolecular hosts
including chirally modified zeolites, cyclodextrins, hydro-
gen-bonding templates, chiral liquid crystals, chiral gels,
and serum albumins. Furthermore, we have unequivocally
demonstrated that in supramolecular photochirogenesis the
enantioselectivities are not determined simply by the popula-
tion of the diastereomeric precursor complexes in the ground
state but are also affected significantly by their excited-state
lifetimes. This established a standard protocol to fully eluci-
date the photophysics and photochemistry of supramolecu-
lar photochirogenesis and developed a few approaches to
control/enhance the product’s enantioselectivity, which will
be applicable to many other supramolecular photochiro-
genic systems and make significant contribution to this
rapidly developing interdisciplinary area of photochemistry
and supramolecular chemistry.
AC is a prochiral molecule, where reactivity on each face
leads to different isomers and in the case of products 2 and 3
to different enantiomers. A different way to study how
chirogenesis is achieved in the bimolecular dynamic reaction
between two ACs is to impede the access of the second AC
molecule to one of the prochiral faces of the AC excited state.
Preliminary experiments using (2S,4S)-4-amino-5-chloro-2-
methoxy-N-(1-ethyl-2-hydroxymethyl-4-pyrrolidinyl)benz-
amide (TKS159,46,47 Scheme 1) or its enantiomer as tem-
plates showed that the ee values for both 2 and 3 were
enhanced.25 NMR analysis of the AC-TKS complex
showed that TKS159 shielded one face of the aromatic
moiety of AC.25 This templating approach is akin to the
template strategy pioneered by Bach to induce photochir-
ogenesis for unimolecular and bimolecular reactions.23,26-36
Results
(44) Dawn, A.; Fujita, N.; Haraguchi, S.; Sada, K.; Shinkai, S. Chem.
Commun. 2009, 2100–2102.
(45) Pace, T. C. S.; Nishijima, M.; Wada, T.; Inoue, Y.; Bohne, C. J. Phys.
Chem. B 2009, 113, 10445–10453.
(46) Matsuyama, S.; Sakiyama, H.; Nei, K.; Tanaka, C. J. Pharmacol.
Exp. Ther. 1996, 276, 989–995.
Characterization of the AC-TKS and AC-TM Com-
plexes. The binding of AC as a guest to host systems, such
as cyclodextrins17,21,40 and serum albumins,41-43,45 led to
changes in the AC absorption, circular dichroism (CD), and
(47) Yanagi, T.; Kitajima, A.; Anzai, K.; Kodama, K.; Mizoguchi, J.;
Fujiwara, H.; Sakiyama, H.; Kamoda, O.; Kamei, C. Chem. Pharm. Bull.
1999, 47, 1650–1654.
(48) Inoue, Y. Nature 2005, 436, 1099–1100.
7910 J. Org. Chem. Vol. 74, No. 20, 2009

Kawanami et al.
JOCArticle
FIGURE 2. Induced CD spectra of 0.1 mM AC in CH2Cl2 at
various concentrations of TKS159 at 10 °C. TKS/AC ratios: (a) 0,
(b) 0.4, (c) 0.8, (d) 1.2, (e) 2.0, (f) 3.6, (g) 5.2, (h) 7.6, (i) 10.8, (j) 14.0,
and (k) 19.6.
fluorescence spectra. These changes are associated with the
formation of complexes and can be used to determine
complex stability constants (K) and stoichiometries. The
addition of TKS159 (Figure 1) and TM166 (Figure S1 in
the Supporting Information) to AC solutions led to a blue
shift in the AC absorption. TKS does not absorb above
350 nm (Figure S2 in the Supporting Information), and
therefore, the changes observed above 350 nm are due to
the incorporation of AC in an environment different from
the homogeneous solvent, dichloromethane. Inspection of
the absorbance values at the wavelengths close to the cross-
ing points in the spectra showed some variability, which
changed somewhat between independent experiments. This
experimental variability makes it difficult to determine if
isosbestic points are present for all TKS/AC ratios. The same
trends for the changes in the AC absorption spectra with the
addition of TKS were observed at lower temperatures.
AC when added to dichloromethane is in its protonated
form. Addition of triethylamine to AC in dichloromethane
led to a blue shift of the AC absorption spectrum (Figure S3
in the Supporting Information), which is more pronounced
than the blue shift observed when AC is bound to TKS. This
result suggests that binding of AC to TKS led to a partial
deprotonation of AC and shows that hydrogen bonding
plays a role in the complex formation.
AC is an achiral molecule and for this reason does not show
an intrinsic circular dichroism (CD) signal. In the presence of
TKS159 (Figure 2), but not TM166 (Figure S4 in the Support-
ing Information), an induced CD spectrum was observed in
the spectral region where AC absorbs. This result indicates
that AC is located in a chiral environment when bound to
TKS159. In contrast, binding of AC to TM166 does not lead
to a chiral environment around AC. This result is in agreement
with the crystal structure determined for the AC-TM com-
plex, where AC is hydrogen bonded to TM166 forming an
extended complex.25 In this complex, the AC and TM166
chromophores are not in close contact. The CD spectra could
not be measured below 320 nm because the absorption
of TKS159 is too strong at wavelengths shorter than 320 nm
(see Figure S2 in the Supporting Information).
FIGURE 3. Difference spectra at 25 °C of the fluorescence spectra
normalized at 420 nm for AC (0.025 mM) in the presence of varying
amounts of TKS159 and the spectrum in the absence of TKS159
(top, TKS/AC ratios: (a) 0.54, (b) 1.2, (c) 1.6, (d) 2.7, (e) 5.3, (f) 10.7,
(g) 18.7, (h) 26.7, (i) 40.0, (j) 53.3, (k) 80, and (l) 107). TM166
(bottom, TM/AC ratios: (a) 0.54, (b) 1.2, (c) 1.6, (d) 2.7, (e) 5.3,
(f) 10.7, (g) 18.7, (h) 26.7, (i) 40.0, (j) 53.3, and (k) 80).
for free AC in dichloromethane. Normalization of the fluor-
escence intensity at 420 nm showed that the shape of the
fluorescence spectra in the presence of TKS159 and TM166
changed. The fluorescence difference spectra at 25 °C
(Figure 3) did not change shape as the template concentration
was increased, suggesting that in the presence of each template
the environment provided by the template does not change
significantly when the concentration of template is raised.
However, the difference spectra are not the same for
TKS159 and TM166, indicating that AC is bound differently
to each template. The same pattern for the difference of the
normalized fluorescence spectra in the presence of TKS159
was seen at 0 °C, while at -25 °C a small deviation of the
difference spectra was observed at low TKS/AC ratios (see
Figure S6 in the Supporting Information). At -50 °C
(Figure 4), the shape of the difference spectra changes as the
concentration of TKS159 increased. The maximum observed
above 500 nm shifts to shorter wavelengths as the TKS/AC
ratio increases. This result suggests that at low TKS/AC ratios
and low temperature a larger number of emissive species are
available than at the higher temperatures studied.
Formation of AC hydrogen-bonded dimers has been
proposed previously when the photodimerization reaction
of AC in various solvents was investigated.49 For this reason,
The fluorescence of AC decreased with the addition of
TKS159 and TM166 (see Figure S5 in the Supporting
Information). This decrease suggests that the emission effi-
ciency for AC bound to TKS159 and TM166 is lower than that
(49) Mizoguchi, J.; Wada, T.; Inoue, Y. Chem. Lett. 2006, 35, 738–739.
J. Org. Chem. Vol. 74, No. 20, 2009 7911

Kawanami et al.
JOCArticle
FIGURE 5. Plot of the change of ellipticity at 367 nm (solid square)
and 386 nm (open circle) based on the CD titration at 10 °C of AC
with TKS159. The solid line corresponds to the fit of the experi-
mental data to a binding isotherm derived for a 1:1 stoichiometry
when neither the AC nor the TKS159 concentrations are in excess
(see the Experimental Section).
FIGURE 4. Difference spectra at -50 °C of the fluorescence spec-
tra normalized at 420 nm for AC (0.025 mM) in the presence of
varying amounts of TKS159 and the spectrum in the absence of
TKS159 (TKS/AC ratios: (a) 0.16, (b) 0.56, (c) 0.8, (d) 1.0, (e) 1.2,
(f) 1.6, (g) 10.6, and (h) 40).
the changes in the difference spectra at -50 °C could be due
to the formation of the AC hydrogen-bonded dimers. In
order to explore this possibility, the fluorescence spectra for
AC in dichloromethane at various AC concentrations were
measured at different temperatures. The normalized fluor-
escence spectra for AC (2.5-75 μM) were the same at 25 and
0 °C, whereas a change was observed at 75 μM AC at -25 °C
and above 6.25 μM AC at -50 °C (Figure S7 in the
Supporting Information). The change in the AC fluores-
cence spectra at -50 °C as the AC concentration increases
(Figure S7) suggests that a significant amount of AC is
present in solution as an AC hydrogen-bonded dimer.
Therefore, at low concentrations of TKS159 the difference
spectra shown in Figure 4 correspond to changes in the AC
monomer/hydrogen-bonded dimer ratio and the formation
of the AC-TKS complex. At high TKS159 concentrations,
the AC-TKS complex is the major species in solution.
Unless otherwise stated, the following results represent
experimental conditions for which the contribution of the
AC hydrogen-bonded dimer in homogeneous solution is
negligible.
Complex stability constant values can be determined by
studying the changes in the intensities for the absorption, CD,
or fluorescence spectra with an increase in the concentration
of template. A 1:1 guest/template stoichiometry was assumed
for these studies. The changes observed for the fluorescence
intensities were due to various factors (see below), such as
dynamic quenching of free AC by the host, in addition to the
formation of the complex. For this reason, fluorescence data
were not suitable for the determination of the AC-template
complex stability constants. Absorption data are less precise
because of the small changes observed for the absorption of
AC in the presence of TKS, but the values of K recovered are
the same within experimental error to the ones determined
from CD data (see below).
TABLE 1. Stability Constants (K) for AC-TKS and AC-TM Com-
plexes Determined by Direct (TKS159) or Competitive (TM166) Circular
Dichroism Spectral Titration Experiments at Various Temperaturesa
K/M-1
T/°C
TKS159b
TM166c
25
10
0
3400 ( 200 (4)
7400 ( 400 (1)
15000 ( 1000 (1)
1400 ( 100
3000 ( 200
5800 ( 400
aDirect CD titrations were run by adding TKS159 to an AC/CH2Cl2
solution and fitting the data to a binding isotherm assuming a 1:1
stoichiometry. For each independent experiment, the K value was deter-
mined at two wavelengths (367 and 386 nm). Competitive CD titrations
were performed by adding TM166 to an AC/TKS159 solution. The value
for K was calculated from the decrease in the AC-induced CD signal. bFor
experiments performed once, the errors correspond to the error propaga-
tion for the values determined by the data fit at each individual wave-
length. For experimentsperformed more thanonce, the errors correspond
to standard deviations. Number of independent runs are shown in
parentheses. cErrors correspond to error propagation from the averages
for the competitive titration method and the error for K for TKS159.
displaced to bind TM166, and this decrease is used to
calculate the value of K for the AC-TM complex (Table 1).
The values of K increased as the temperature of the
solution was lowered (Table 1). However, attempts to obtain
values at -25 and -50 °C were hampered because of the
unavailability of a suitable cooling system for the CD
spectrometer. For this reason, the values of K at the low
temperatures for a 1:1 complexation stoichiometry were
estimated from the fit of the data at higher temperatures
to the van’t Hoff equation (eq 1) (Figure 6). The K values
for AC-TKS were calculated to be (9 ( 1) ꢀ 104 M-1 and
(8 ( 1) ꢀ 105 M-1 at -25 and -50 °C, respectively. It is
important to note that this extrapolation to lower tempera-
tures assumes that the complex stoichiometry is 1:1 and that
no other types of complexes were formed. From the van’t
Hoff plots (Figure 6), the enthalpy (ΔH°) and entropy
changes (ΔS°) were calculated for the 1:1 complex formation
of AC with TM166 and TKS159 (Table 2).
The complex stability constant (K) values were determined
from the changes in the induced CD intensity of AC (0.1 mM)
with the increase in the concentration of TKS159 (Figure 5,
Table 1). In the case of TM166 no induced CD signal was
observed, and a competitive CD measurement was per-
formed. TM166 was added to a solution containing suffi-
cient TKS159 to bind over 97% of AC. The decrease in the
AC-TKS CD signal is related to the amount of AC
ln K ¼ -ΔH°=RT þΔS°=R
ð1Þ
Job plots, based on the changes in the induced CD signals,
were constructed at 25 and 0 °C ([AC]0 þ [TKS]0=1 mM; see
7912 J. Org. Chem. Vol. 74, No. 20, 2009

Kawanami et al.
JOCArticle
TABLE 3. Fluorescence Lifetimes (τ) and Pre-exponential Factors (A)
for AC-TKS Complexes at Various Temperaturesa
AC
T/°C TKS/AC boundb /% τ2c/ns
A2 τ3e /ns A3
A2/A3
d
d
25 80
0
87
g94
g96
g99
3.7 ( 0.3 0.28f
10 0.69f 0.41 ( 0.03
40-80
3.9 ( 0.5 0.42g 12 0.58g 0.72 ( 0.04
4.5 ( 0.3 0.54g 13 0.46g 1.18 ( 0.07
6.1 ( 0.5 0.54g 18 0.46g 1.18 ( 0.07
-25 10.7-80
-50 10.7-80
aMeasured in aerated CH2Cl2: [AC]=0.025 mM; λex=361 nm; λem
=
440 nm. bCalculated using the complex stability constants assuming a 1:1
binding stoichiometry (see text). cAverage determined from the fit of the
fluorescence decay to the sum of two exponentials. dEstimated errors (
0.02. eEstimated when the concentration of AC free in dichloromethane
was low (see text), estimated error ( 2 ns. fDetermined from the fit of the
decay to the sum of three exponentials. gDetermined from the fit of the
decay to two exponentials where the long-lived component is assigned
exclusively to TKS159-bound AC.
FIGURE 6. van’t Hoff plots for AC-TM and AC-TKS com-
plexation using the complex stability constants shown in Table 1.
diffusion-controlled process in dichloromethane (1.6 ꢀ 1010
M-1 s-1 at 25 °C).50 However, the kq values are sufficiently
high for the decrease observed at lower temperatures to be
mostly due to an increase in the solvent viscosity. This
quenching observed for ACOMe needs to be taken into
account when analyzing the changes in lifetime for the AC/
template systems.
TABLE 2. Thermodynamic Parameters for the Complexation of AC
with Templates TKS159 and TM166 in Dichloromethane at T = 298.15
Ka
template
ΔG°/kJ mol-1
ΔH°/kJ mol-1
TΔS°/kJ mol-1
TKS159
TM166
-(20 ( 4)
-(18 ( 3)
-(40 ( 3)
-(37 ( 2)
-(20 ( 2)
-(19 ( 2)
aErrors correspond to statistical errors for the fit of the data to eq 1.
The decay for AC in dichloromethane followed a mono-
exponential function (Figure S9 in the Supporting
Information), and the lifetime of the singlet excited state of
AC lengthened as the temperature was lowered from 25 °C
(14.9ns) to0°C (16.6 ns) and then leveled off at -25 °C (17.5ns)
and -50 °C (17.0 ns). The decay of the AC fluorescence in
the presence of templates did not follow a monoexponential
decay (Figure S9 in the Supporting Information). A short-
ening of the lifetime was observed, which is consistent
with the decrease in the steady-state fluorescence intensity
observed for the AC/template solutions. Adequate fits were
obtained when fitting the decays to a sum of two exponen-
tials. The pre-exponential factor (Ai) for each species is
related to the concentration of that particular species,
taking into account the species’ excitation efficiency. The
excitation efficiencies for free and bound AC were different
and are unknown because the bandwidth for the single-
photon-counting experiment was 16 nm. However, trends
in the Ai values with changes in the concentration of
TKS159 can be correlated with changes in concentration
of each species.
The lifetime of the long-lived component at 25 °C dec-
reased from 14.9 ns for AC free in dichloromethane to 10.7 ns
when the TKS/AC ratio was 80. The short-lived component
became more prominent as the concentration of TKS159
was increased (Table S1 in Supporting Information). Once
the pre-exponential factor for the short-lived component
was significant (A2 g 0.1), the lifetime for this component
was constant at 3.7 ( 0.3 ns (Table 3). This short-lived
component, which is only present when AC is bound to
TKS159, is assigned to a complex between AC and TKS159.
The long-lived component incorporates the emission of AC
free in dichloromethane; however, its lifetime was shorter
than expected for the bimolecular quenching of AC by
TKS159 based on the kq values determined using ACOMe.
For example, at a TKS/AC ratio of 80 the calculated
Figure S8 in the Supporting Information) to determine the
stoichiometry of the complex between AC and TKS159.
A maximum was observed at 0.5 indicating that the stoichi-
ometry of the complex is 1:1. Measurements at [AC]/([AC] þ
[TKS]) ratios higher than 0.7 and at low temperatures were
not possible because of the presence of an AC precipitate.
Fluorescence Lifetime Measurements. Kinetic measure-
ments for the decay of the singlet excited state of AC in the
presence of template are useful to obtain mechanistic infor-
mation on the photodimerization reaction, provided that the
lifetimes for AC free in dichloromethane and when bound to
the templates are different. Control experiments using the
methyl ester of AC (ACOMe) were performed to determine if
the excited state of ACOMe could be dynamically quenched
by the templates in a bimolecular reaction without the
formation of a ground-state complex. The absorption spec-
trum of ACOMe did not change in the presence of TKS159,
indicating that no complex was formed because ACOMe
lacks the ability to form hydrogen bonds with the template.
A shortening of the singlet excited state lifetime of ACOMe
was observed in the presence of TKS159 and TM166 in-
dicating that quenching occurred. A linear relationship was
observed between the observed decay rate constant (kobs
)
and the quencher concentration ([Q]). The quenching rate
constants (kq) were determined from the fit of the data to eq
2, where ko corresponds to the inverse of the singlet excited
lifetime of ACOMe in the absence of quencher.
kobs ¼ koþkq½Qꢁ
ð2Þ
The values of kq for the quenching of ACOMe by TKS159
decreased as the temperature was lowered: (3.4 ( 0.1) ꢀ 109
M-1 s-1 at 25 °C, (2.8 ( 0.1) ꢀ 109 M-1 s-1 at 0 °C, (2.1 (
0.1) ꢀ 109 M-1 s-1 at -25 °C, and (1.5 ( 0.2) ꢀ 109 M-1 s-1
at -50 °C. The quenching rate constant of ACOMe by
TM166 at 25 °C was determined to be (3.1 ( 0.1) ꢀ 109
M-1 s-1. These quenching rate constants are almost 1 order
of magnitude lower than the rate constant observed for a
(50) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photochemistry;
2nd, revised and expanded ed.; Marcel Dekker, Inc.: New York, 1993.
J. Org. Chem. Vol. 74, No. 20, 2009 7913

Kawanami et al.
JOCArticle
TABLE 4. Product Distribution and Enantiomeric Excess (ee) for the Photodimerization Reaction of AC in Dichloromethane in the Presence and Absence
of TKS159a
product distribution/%
% ee
T/°C
TKS/AC
AC boundb /%
conv/%
1
2
3
4
2
3
HT/HH
anti/syn
(1 þ 4)/(2 þ 3)
25
0
40
120
0
28
0c
10
10
10
0c
3
10
0
97
>99
0
>99
0
>99
>99
>99
0
92
96
97
60
95
12d
93
19
23
3d
33
38
40
36
39
41
37
37
38
33
41
38
22
25
26
20
27
15
24
24
22
13
19
20
24
24
23
28
24
34
28
29
30
39
31
32
21
13
11
16
10
10
11
10
10
15
9
1
0
1.2
1.7
1.9
1.3
1.9
1.3
1.6
1.6
1.5
0.9
1.5
1.4
1.3
1.6
1.7
1.8
1.7
3.0
1.9
1.9
2.1
2.6
2.6
2.3
1.2
1.0
1.0
1.1
1.0
1.0
0.9
0.9
0.9
0.9
1.0
0.9
-28
-25
-2
-33
-3
-37
-39
-41
2
-43
-42
-20
-23
0
-28
0
-36
-38
-44
-2
-49
-48
0
-25
-30e
-40e
-50
>99
>99
88
90
10
a[AC]=0.25 mM, unless noted otherwise. Irradiated at λ g 320 nm for 120 min, errors for product distribution (1% and errors for ee values (2%.
bCalculated using the complex stability constants assuming a 1:1 binding stoichiometry (see text). cPrecipitation of AC observed. dLow conversion due to
the precipitation of AC. e[AC]=0.2 mM, irradiated for 30 min.
lifetime for free AC is 13.7 ns, while the measured lifetime
was 10.7 ns. In addition, if the long-lived component was
related to free AC the A1 value should decrease as the
concentration of TKS is raised, but the A1 value levels off
at 0.7. Both the excessive shortening of τ1 and leveling off
observed for A1 suggest that another AC-TKS complex
was present that had a longer lifetime than the species with
the 3.7 ns lifetime.
from þ25 to -25 °C but remained constant between -25
and -50 °C (Table 3).
The decays for the AC fluorescence at 0 °C for a TKS/AC
ratio of 80 were measured at several wavelengths (395, 440,
and 520 nm). The decays were fit to the sum of two
exponentials because at this TKS159 concentration 97% of
the AC is bound. The temperature of 0 °C was chosen
because conditions could be used when almost all AC was
bound to TKS159 and no AC hydrogen-bonded dimers were
formed at the AC concentration employed (0.025 mM)
(see above and Figure S7 in the Supporting Information).
The two lifetimes recovered at these three wavelengths
ranged from 3.5 to 4.1 ns and 11.0 to 11.4 ns, which are the
same lifetimes as determined from multiple experiments at
various TKS/AC ratios (see Table S2 in the Supporting
Information). The pre-exponential factor for the short-lived
species decreased from 0.64 at 395 nm to 0.43 at 440 nm and
0.12 at 520 nm, while the pre-exponential factors for the
long-lived species increased from 0.36 to 0.57 and 0.88 as the
detection wavelength became longer. This result suggests
that the spectrum corresponding to the long-lived species is
red-shifted compared to the spectrum for the short-lived
species.
The lifetime for the second AC-TKS complex was esti-
mated by fitting the decays to the sum of three exponentials
where the value of τ1 was fixed to the value calculated for the
lifetime of free AC, taking into account the dynamic quench-
ing determined with ACOMe, and by fixing τ2 to 3.7 ns
(Table S1 in the Supporting Information). The lifetime of the
second AC-TKS species (τ3) is close to that for AC in
dichloromethane. For this reason, fitting the decays to the
sum of three exponentials, when a significant portion of the
AC is free in solution, was not satisfactory, leading to A1 and
A3 values that are theoretically not possible (Table S1 in the
Supporting Information). However, at high TKS/AC ratios
the values for A1 are very low because most of the ACs are
bound to the template. The values for τ3, A2, and A3 at 25 °C
were taken from the fit of the data to the sum of three
exponentials at the highest TKS/AC ratio studied (Table 3).
The analysis of the decays at lower temperatures is facili-
tated because conditions could be used where most (g90%)
of the AC was bound to TKS159. At 0 and -25 °C, the value
for τ1 at high TKS/AC ratios was shorter than calculated
from the dynamic quenching of the excited AC by the
template (Tables S2 and S3 in the Supporting Information),
again suggesting that a third emissive species was present.
The fit to the sum of three exponentials is consistent with this
assignment. However, the values for τ3, A2, and A3 reported
are those from the fit to the sum of two exponentials under
conditions where most AC was bound (Table 3). At -50 °C,
the two lifetimes changed little but the pre-exponential
factors changed until constant values were observed when
more than 90% of the AC was bound (Table S4 in the
Supporting Information).
In the case of TM166 only an extended complex is formed
with AC.25 Two lifetimes were observed for the singlet
excited state of AC in the presence of TM166 at 25 °C
(Table S5 in the Supporting Information), and the τ1 value
was shortened from 14.9 ns in the absence of TM166 to 13.5
ns at a TKS/AC ratio of 80. The value calculated for τ1 taking
into account the dynamic quenching of excited AC by
TM166 is 13.7 ns. This result shows that in the case of
TM166 only one complex is formed with AC, which has a
lifetime (5.0 ( 0.2 ns) shorter than the excited AC lifetime in
dichloromethane. The lifetime for the complex increased
from 5 ns at 25 °C to 9 ns at -25 °C and 10 ns at -50 °C.
Product Studies. The distribution of the four stereoiso-
meric products 1-4 and the ee values for chiral products 2
and 3 were determined by chiral HPLC for the photodimeri-
zation in the presence and absence of TKS159 at various
temperatures (Table 4).
The two singlet excited state lifetimes for the AC-TKS
complexes (see the Discussion for a structural assignment)
were lengthened as the temperature was lowered. In addition,
the concentration ratio between these two species, ex-
pressed as A2/A3, increased as the temperature was lowered
In the absence of TKS159, the solubility of AC in dichloro-
methane is limited, particularly at lower temperatures, and a
solution of 0.25 mM AC forms precipitates at temperatures
below -25 °C. For this reason, the photodimerization of AC
7914 J. Org. Chem. Vol. 74, No. 20, 2009

Kawanami et al.
JOCArticle
FIGURE 8. Structure of the AC-TM complex optimized by DFT-
D-B-LYP/TZVP.
FIGURE 7. Eyring plot of the enantiomer ratio obtained upon
irradiation at various temperatures (data from Table 3).
alone was seriously decelerated to give extremely low con-
versions (3-12%) at low temperatures. However, the addi-
tion of an excess amount of TKS159 readily solubilized AC
of 0.25 mM in dichloromethane, even at -50 °C.
Control experiments at þ25, -25, and -50 °C showed that
the product distribution ((1% error), and the ee values
((2% error) were independent of the conversion and the
AC concentration (0.05-0.25 mM, Table S6 in the Support-
ing Information). This result is important because the fluor-
escence lifetime measurements were performed at low AC
concentrations (0.025 mM), while the product studies were
normally performed at higher concentrations (0.25 mM) to
increase the precision of the HPLC analysis.
As shown in Table 4, in the presence of TKS159 the
formation of HT dimers 1 and 2 was appreciably increased
at the expense of HH dimers 3 and 4 at all temperatures
studied. This probably reflects the fact that the TKS159
hydrogen bonding to the carboxyl group (H-bonding end of
the molecule) increases the head-to-head steric repulsion
when these groups are aligned in the formation of the dimer.
Once all AC molecules were bound to TKS159 the ee values
for 2 and 3 did not change within the experimental error at
each temperature examined. The ee values were gradually
enhanced as the temperature was lowered and reached values
of -43% and -49% at -50 °C. It is interesting to note that
the (1 þ 4)/(2 þ 3) ratio is surprisingly constant around unity
(1.0 ( 0.1), irrespective of the reacting species present in
solution (free AC and AC-TKS complex), the temperature,
and the conversion. This result is in sharp contrast to the
behavior of HT/HH and anti/syn ratios and suggests the
presence of a common precursor/process for each pair of
photodimers (see below).
The increase in the magnitude of the ee as the temperature is
lowered can be used to calculate an apparent differential en-
thalpy and entropy for the chirogenesis induced by complexation
of AC to TKS159. It is important to note that these values can
include a difference in the population of precursors, i.e., different
equilibrium constants, and a difference in the reactivity of each
complex. Therefore, this Eyring plot does not provide mechan-
istic information, but it can be used to establish whether the
mechanism changes as the temperature is varied. We assume that
the ee values for 2and 3are the same when more than 99% of AC
is bound, and therefore, we used the average values to construct
the Eyring plot (eq 3, Figure 7). The apparent differential
enthalpy and entropy changes are ΔΔH=-(3.6 ( 0.1) kJ mol-1
FIGURE 9. Structures of the re- and si-AC-TKS complexes opti-
mized by DFT-D/B-LYP/TZVP, where the AC re- or si-face is
accessible.
and ΔΔS=-(7.9(0.4) J mol-1 K-1. The linearity of the Eyring
plot suggests that the same mechanism operates at high and low
temperatures.
ð100þj%eejÞ
ð100 -j%eejÞ
k
k00
ΔΔS ΔΔH 1
ln
¼ ln
¼
-
ð3Þ
R
R T
DFT-D-Optimized Structures of AC-TKS and AC-TM
Complexes. As a consequence of the prochiral nature of
AC, the hydrogen-bonding interaction of AC with chiral
template TKS159 leads to the formation of a pair of
diastereomeric complexes, where either the re- or si-face
of AC (the enantioface is defined at the 2-position of AC,
to which the carboxylic group is attached) is shielded by
J. Org. Chem. Vol. 74, No. 20, 2009 7915

Kawanami et al.
JOCArticle
TABLE 5. Results of the DFT-D Calculations for AC-TM and re- and si-AC-TKS Complexes
˚
distance/A
complex
energya /kJ mol -1
O-H
O
N-H
O
Cl-ACb
face-to-facec
tilt angled (deg)
3 3 3
3 3 3
AC-TM
re-AC-TKS
si-AC-TKS
a
38.4
ꢂ0.0
8.6
1.612
1.658
1.651
1.465
1.593
1.607
11.16
3.60
4.67
13.34
3.72
4.10
59.6
4.3
12.1
Energy relative to the most stable re-AC-TKS complex (for the detailed results including the actual energy, see the Supporting Information).
bDistance between the chlorine atom of the template and the closest carbon atom of AC. cCenter-to-center distance between the anthracene and the
benzene ring of the template. dTilt angle of the anthracene plane and the benzene plane of the template.
the methoxybenzamide moiety of TKS159. These confor-
mations are folded, and there is no possibility for the direct
interconversion between the si and re complexes without
decomplexation of AC from the template. Epimeric
TM166 also forms a pair of diastereomeric complexes with
AC, which possess the extended conformation, as revealed
by X-ray crystallography.25 Hence, there will be no differ-
ence in stability between the two diastereomeric com-
plexes, which can interconvert by free rotation around
the bond between the aromatic ring and the carbon on the
carboxyl group to spontaneously “racemize”. As a con-
sequence, no enantiodifferentiation in the products is
observed, as demonstrated previously.25
Structure optimizations by the dispersion-corrected DFT-
D method at the B-LYP level, which is known to appro-
priately correct for stacking interactions,51 were performed
for two diastereomeric AC-TKS complexes and an ex-
tended AC-TM complex. All structures were treated as
the pyrrolidinium(TKS/TM)-carboxylate(AC) ion pair, as
the X-ray crystallographic study revealed that the AC-TM
complex exists in this salt form.25 The DFT-optimized
structure of the AC-TM complex (Figure 8) nicely repro-
duced the geometry of the structure obtained from X-ray
crystallography justifying the geometry optimization at the
DFT-D/B-LYP level for this kind of hydrogen-bonded
complex. The DFT-optimized structures of diastereomeric
re- and si-AC-TKS complexes have stacked conformations,
in which the π overlap and the tilt angle between the two
aromatic planes are appreciably different (Figure 9). The
calculated energies and some selected geometries of the
AC-TM and two AC-TKS complexes are summarized in
Table 5.
carboxyl group on AC is partially deprotonated as expected
from the formation of a hydrogen bond with the template.
Deprotonation is not complete since the blue shift in the
AC-TKS spectrum compared to the protonated AC is
smaller than that observed for the spectrum of AC in the
presence of triethylamine, where AC is deprotonated. The
environment around AC is different in the complexes with
TKS159 and TM166. AC complexation to TKS159 leads to
strong induced CD signals, while no such signals were
observed for the AC-TM complex. In addition, the fluor-
escence difference spectra are different for AC bound to
TKS159 or TM166.
The driving force for the formation of the AC complex
with TKS159 or TM166 is enthalpic, while an entropy loss,
arising from the loss of degrees of freedom upon molecular
association, is observed in both cases (Table 2). The ΔH° and
ΔS° values are similar for both complexes. For the AC-TM
complex, which has an extended conformation, the hydro-
gen-bonding interaction is the only source of enthalpic gain,
while for the AC-TKS complex hydrogen bonding and π-π
stacking can lead to the stabilization of the complex. The two
diastereomeric re and si complexes of TKS159 with AC are
expected to be different in energy and geometry, while the
TM166 complex will have conformers with similar energy
since bond rotation is not restricted (see above). DFT-D
calculations provide a qualitative estimate of the relative
stability of these complexes (Table 5). The calculations
predict that the AC-TM complex is less stable than the
AC-TKS complexes in line with the equilibrium constant
values determined from CD measurements (Table 1). The
AC-TM complex has shorter hydrogen bonds than calcu-
˚
lated for the two AC-TKS complexes by 0.04-0.05 A for
˚
the OH O distance and by 0.13-0.14 A for the NH
O
3 3 3
3 3 3
distance. The elongation of the hydrogen bonds in the case of
TKS159 is likely to be driven by the structural demand of the
TKS159 template but is energetically compensated by the
π-stacking of two aromatic moieties in the AC-TKS com-
plexes. An additional difference for the re- and si-AC-TKS
complexes is the distance and tilt angle of the two aromatic
rings in each complex (Table 5). Thus, the two aromatic
Discussion
TKS159 and its epimer TM166 were previously shown to
form complexes with AC. The enantiomeric excess for the
photodimerization of AC observed for the AC-TKS com-
plex is due to the shielding by the template of one of the faces
of AC.25 X-ray crystallographic studies, as well as NMR
experiments, showed that the carboxylic group on AC forms
two hydrogen bonds with the pyrrolidine nitrogen and the
hydroxyl group of the templates. In the case of TKS159,
the AC aromatic ring is partially covered by the 2-methoxy-
benzoyl moiety of the template. In contrast, for TM166, no
such stacking occurs due to the 2S,4R configuration of the
4-aminoprolinol moiety.25
˚
˚
rings in the re complex are closer (distance: 3.72 A versus 4.10 A)
and more parallel to each other (tilt angle: 4.3° versus 12.1°)
than those in the si complex, rendering the re complex more
stable than the si complex. It is important to note that we
could not differentiate the K values for the re and si com-
plexes when determining the binding isotherm in the CD
experiments. This result suggests that the difference between
Kre and Ksi is small and the overall K value incorporates the
equilibrium constants for both complexes. In addition, the
shape of the fluorescence difference spectra did not change
for different TKS/AC ratios, suggesting that a similar pro-
portion of re-AC-TKS and si-AC-TKS was present at all
The blue shift observed in the absorption spectra of AC
when bound to TKS159 and TM166 suggests that the
(51) Grimme, S.; Antony, J.; Schwabe, T.; Muck-Lichtenfeld, C. Org.
Biomol. Chem. 2007, 5, 741–758.
7916 J. Org. Chem. Vol. 74, No. 20, 2009

Kawanami et al.
JOCArticle
TKS/AC ratios. The time-resolved experiments established
that the spectra for re-AC-TKS and si-AC-TKS are
slightly different, and a change would have been observed
for the steady-sate fluorescence difference spectra as the
TKS/AC ratio was changed if the ratio for the concentration
of the re- and si- complexes changed significantly.
The observation of two lifetimes could in principle indi-
cate the interconversion between two isomers, in this case the
si and re complexes. However, such a mechanism would
require the dissociation of the complex during the singlet
excited-state lifetime of AC in order for bond rotation to
occur around the bond between the aromatic ring and the
carbon on the carboxyl group. The upper limit at 25 °C
for the dissociation rate constant of the AC-TKS complex is
4.7ꢀ106 s-1, which was calculated from the complex stability
constant of 3400 M-1 and the diffusion rate constant in
dichloromethane of 1.6ꢀ1010 M-1 s-1. Complete dissocia-
tion is too slow compared to the decay of the species with the
shortest lifetime (2.7 ꢀ 108 s-1). In addition, no growth
kinetics of the long-lived and red-shifted fluorescent species
were observed, a result that eliminates the possibility of
partial dissociation of AC and bond rotation during the
excited state lifetime. Therefore, the lifetimes observed for
the complex of AC with TKS159 are the intrinsic lifetimes
of the individual si and re complexes. A quantitative analy-
sis relating the pre-exponential factors with concentration
ratios for the re-AC-TKS and si-AC-TKS is not warranted
because the emission spectra for each complex are not known
and the relative emission quantum yield at different wave-
lengths cannot be established. However, the trend that the
pre-exponential factor for the short-lived species at 440 nm
increases as the temperature is lowered is in agreement with
the assignment of the short-lived species to the most stable
re-AC-TKS complex.
Preliminary product studies at template/AC ratios of 10
using TKS159, TM166, and their enantiomers showed that
AC complexation to TM166 did not lead to any significant
changes in the distribution for products 1-4, and no sig-
nificant ee was observed.25 Therefore, the complexation of
AC to TM166 does not affect the photodimerization selec-
tivity showing that the approach to the two faces of the AC
bound to TM166 is the same as observed for AC in dichloro-
methane. In contrast, an ee was observed for products 2 and
3 for AC in the presence of TKS159. The magnitude of the ee
values were the same for TKS159 and its enantiomer, but the
signs were opposite showing that antipodal products were
formed for each of the enantiomers.25
The ideal experimental conditions to gain mechanistic
information for this system are those where the amount of
AC in solution is minimized and the contribution of AC
hydrogen-bonded dimers in solution is negligible. Binding
of 99% of the AC was achieved at TKS/AC ratios of 120 at
25 °C, while complete binding was achieved at 0 °C for TKS/
AC ratios of 28 or higher. At -25 and -50 °C, formation of
hydrogen-bonded AC dimers occurred at the AC concentra-
tions employed for product studies (0.25 mM) and low
TKS159 concentrations. However, 99% of the AC was
bound to the template when the TKS/AC ratio was 6 or
higher at -25 °C and 3 or higher at -50 °C, and for these
conditions no interference from the hydrogen-bonded AC
dimer occurred for the product studies.
The two faces of AC are prochiral, and the stereochemistry
of the products depends on which faces interact in the
encounter complex. Interaction of an AC re-face with an
AC si-face leads to the formation of achiral products 1 and 4,
while the interaction of ACs through re-re or si-si faces
leads to chiral products 2 and 3 (Figure 10). The products
formed from the re-re interaction are the enantiomers of the
The photophysics of anthracenes is very sensitive to sub-
stitution on the aromatic ring and the nature of the solvent,
and therefore, the structural differences for the re-AC-TKS
and si-AC-TKS complexes affect the lifetime of the singlet
excited state of AC in each complex. The sensitivity of the
photophysics of anthracenes is due to the closeness in energy
of the S1 and T2 states.52-56 Very small changes in the
energies of S1 or T2 lead to significant variations in the
intersystem crossing rate constants which directly affect the
anthracene fluorescence quantum yields and singlet excited-
state lifetimes. In addition, anthracenes can show dual
emission resulting from the interconversion between con-
formational isomers.57-60 This interconversion leads to the
detection of coupled kinetics where growth kinetics are
observed at certain wavelengths indicating the isomerization
to the most stable conformer.
The shortening of the singlet excited state lifetime for AC
bound to TM166 suggests that the lifetime of AC is sensitive
to the degree of hydrogen bonding on the carboxyl group.
It is important to note that in the AC-TM complex no
π-stacking occurs, and therefore, π-stacking is not respon-
sible for the shortening of the AC lifetime. In the case of
TKS159, the two diastereomeric complexes have different
singlet excited-state lifetimes, which are shorter than the
lifetime for AC in dichloromethane. However, only one of
the lifetimes is shorter than the one observed for AC-TM,
while the second lifetime is longer. The different lifetimes
for re-AC-TKS and si-AC-TKS are not likely due to
differences in the degrees of hydrogen bonding because
the calculated distances for the hydrogen bonds are similar
in both complexes (Table 5). The structural differences
between re-AC-TKS and si-AC-TKS are the degree of
π-π stacking and the distance between the AC aromatic
ring and the Cl atom. The chlorine atom is expected to
enhance the intersystem crossing quantum yield through a
heavy atom effect decreasing the lifetime of the singlet
excited state. Indeed, the singlet excited-state lifetimes of
1-chloro- and 2-chloroanthracene are shorter than the life-
time for anthracene.53 Based on these considerations, we
assign the shorter lifetime to the re-AC-TKS complex
where the anthracene moiety is in closer proximity to the
chlorine atom and the overlap of the aromatic systems is
larger.
(52) Bennett, R. G.; McCartin, P. J. J. Chem. Phys. 1966, 44, 1969–1972.
(53) Bohne, C.; Kennedy, S. R.; Boch, R.; Negri, F.; Orlandi, G.;
Siebrand, W.; Scaiano, J. C. J. Phys. Chem. 1991, 95, 10300–10306.
(54) Kellogg, R. E. J. Chem. Phys. 1965, 43, 411–412.
(55) Lim, E. C.; Laposa, J. D.; Yu, J. M. H. J. Mol. Spectrosc. 1966, 19,
412–420.
(56) Ware, W. R.; Baldwin, B. A. J. Chem. Phys. 1965, 43, 1194–1197.
(57) Albrecht, M.; Bohne, C.; Granzhan, A.; Ihmels, H.; Pace, T. C. S.;
Schnurpfeil, A.; Waidelich, M.; Yihwa, C. J. Phys. Chem. A 2007, 111, 1036–
1044.
(58) Brearley, A. M.; Flom, S. R.; Nagarajan, V.; Barbara, P. F. J. Phys.
Chem. 1986, 90, 2092–2099.
(59) Flom, S. R.; Nagarajan, V.; Barbara, P. F. J. Phys. Chem. 1986, 90,
2085–2092.
(60) Furuuchi, H.; Arai, T.; Sakuragi, H.; Tokumaru, K.; Nishimura, Y.;
Yamazaki, I. J. Phys. Chem. 1991, 95, 10322–10325.
J. Org. Chem. Vol. 74, No. 20, 2009 7917

Kawanami et al.
JOCArticle
into account the re- and si-AC-TKS complexes but does
not have to include any reactivity of free AC since the
concentration of the latter species is negligible. The photo-
reaction was carried out under zero-order conditions,
where the concentration of excited states was much lower
than the concentration of reagents and the illumination was
constant. The reaction of the singlet excited state of the
re-AC-TKS complex (re*) and of the singlet excited state of
the si-AC-TKS (si*) can be seen as parallel reactions where
re* and si* are two different reagents that reacted with
either the ground-state complexes re or si. Re-equilibration
between re and si occurred in the ground state because the
lifetimes of re* and si* were short. Therefore, throughout
the reaction the ratio [re]/[si] was constant and was deter-
mined by the equilibrium constants in the ground state of
AC with TKS159. The probability of the formation of the
four possible complexes, i.e., re*-si, si*-re, re*-re, and
si*-si, remains constant during the reaction because
ground-state re-equilibration between re and si occurred.
This analysis is consistent with the fact that the product
distribution does not change with the irradiation time, i.e.,
with the degree of conversion of the reactant. A further
consequence of this analysis is that the ee can be used as a
measure of the ratio for the concentrations of the re and
si complexes because the fraction of complexes react-
ing through the re*-si and si*-re, leading to nonchiral
products 1 and 4, will be constant throughout the photo-
reaction.
Products 2 and 3 are formed from the encounter com-
plexes re*-re and si*-si, where one of the encounter
complexes leads to the formation of products 2- and 3-
while the other encounter complex leads to the formation
of products 2 and 3 . We assigned the short-lived
AC-TKS species to the re complex because calculations
showed that this complex was more stable than the
si-AC-TKS complex, and as the temperature was lowered
the pre-exponential factors for the short-lived species
increased. The ee increased as the temperature was lowered
indicating that it arises from the complex with the short
lifetime, and therefore, products 2- and 3- are formed
from the re*-re encounter complex, leading to the nega-
tive ee observed in the product studies (note that the “-”
and “þ” signs are arbitrary and are related to the elution
time from the HPLC column, “þ” being assigned to the
product that elutes first).
The ee values are determined by the relative amounts of
the re and si complexes, the intrinsic reactivity for photo-
dimerization, and the lifetimes of re* and si*. A very
different intrinsic reactivity for the re*-re and si*-si
encounter complexes, which are diastereomeric in the pre-
sence of chiral template TKS159, would lead to a different
partition between products 2 and 3 for each complex
leading to different values for the ee of 2 and 3. The similar
ee values observed at each temperature suggests that the
partition for the formation of 2 and 3 is similar in the
re*-re and si*-si encounter complexes. Therefore the
major discriminating factors leading to the chirogenesis
observed are the relative amount of the re and si complexes
and the lifetimes for each complex (τre and τsi). The product
ratio from the re*-re and si*-si encounter complexes
are related to the ratio of encounter complexes, which in
turn is related to the concentration of excited states, the
FIGURE 10. Product formation from the encounter complexes of
two AC molecules assuming an “average” perpendicular geometry.
CW and CCW indicate the clockwise and counterclockwise rotation
of the top AC leading to the photodimer product. Si and re indicate
the faces of the AC involved in the formation of the photodimer.
þ
þ
products formed from the si-si interactions. Therefore, the
sum of products 1 and 4 compared to the sum of 2 and 3
indicates if the encounter complex involved opposite or the
same prochiral faces of AC. A (1 þ 4)/(2 þ 3) ratio of 1 is
expected in homogeneous solution if the reactivity of re-si
encounter complexes is the same as the reactivity for the
re-re and si-si encounter complexes. This ratio was 1.2 at
25 °C and 1.0 at 0 °C. Steric or electrostatic repulsion could
affect the reactivity for photodimerization in two different
ways. The repulsive interactions can lead to a higher dis-
sociation of an encounter complex, therefore leading to a
decreased reactivity for certain encounter geometries. The
second effect is that the distribution between 1 and 4 and
between 2 and 3 is dictated after an encounter complex is
formed; i.e., the formation of the photodimer with higher
repulsive interactions is minimized. The fact that the (1 þ
4)/(2 þ 3) ratio is close to unity indicates that repulsive effects
play only a minor role at least in the early stages of the
encounter complex. For this reason, we propose that the
encounter complexes on “average” have a perpendicular
geometry, where electrostatic repulsion is minimized
(Figure 10). The formation of the products is then deter-
mined by the relative rotation of the two AC molecules.
Analysis for the product distribution and ee values at
TKS159 concentrations when all AC is bound has to take
7918 J. Org. Chem. Vol. 74, No. 20, 2009

Kawanami et al.
JOCArticle
excited-state lifetimes, and the concentrations of ground
states (eq 4).
TABLE 6. Calculations of re/si Ratio and the Diastereomeric Excess
(de) upon Complex Formation and the ee0 Obtained When the re and si
Complexes Have the Same Lifetimea
ꢃ
ꢃ
½2-þ3-ꢁ ½re -reꢁ τre½re ꢁ½reꢁ
obsd
calcd
¼
þ
¼
ð4Þ
ꢃ
ꢃ
τsi½si ꢁ½siꢁ
½2 þ3 ꢁ
½si -siꢁ
þ
T/°C
τre/ns
τsi/ns
|ee|/%
[re]/[si]a
de/%
|ee0|/%
The ratio of excited states is equal to the ground-state ratio
for the re and si complexes (see above) leading to a squared
dependence of the product ratio:
25
0
3.7
3.9
4.5
6.1
10
12
13
18
24
2.1 ( 0.5
2.4 ( 0.5
2.5 ( 0.4
2.8 ( 0.4
35
41
43
47
63
70
72
77
30.5
36.5
45
-25
-50
aErrors correspond to the error propagationfor the lifetimes (Table 3)
2
½2-þ3-ꢁ
τre½reꢁ
¼
ð5Þ
and (2% for the ee values.
2
½2 þ3 ꢁ
þ
þ
τsi½siꢁ
combined with kinetic behavior of the excited states provid-
ing detailed information on the mechanism and stereochem-
ical consequences of the photodimerization reaction.
Qualitatively, the ee observed is due to a [re]/[si] ratio
larger than unity because the lifetime for the re complex is
shorter than for the si complex. The [re]/[si] ratio cannot be
calculated from the pre-exponential factors recovered from
the time-resolved experiments because the emission effi-
ciency and the spectra for each complex are unknown.
However, this ratio can be calculated from the ee values
observed (eq 6).
Experimental Section
Materials. 2-Anthracenecarboxylic acid, triethylamine (99%
checked by GC), and dichloromethane (fluorescence grade)
were used as received. Acetonitrile was purchased from Wako
(HPLC grade) or from Kanto (purified by distillation). 4-
Amino-5-chloro-2-methoxy-N-[(2S,4S)-1-ethyl-2-hydroxymet-
hyl-4-pyrrolidinyl]benzamide (TKS159) and 4-amino-5-chloro-
2-methoxy-N-[(2S,4R)-1-ethyl-2-hydroxymethyl-4-pyrrolidinyl]-
benzamide (TM166) were prepared as previously reported.47
2-Anthracenecarboxylic acid methyl ester (ACOMe) was
synthesized in 94% yield by the reaction of 2-anthracenecar-
boxylic acid (0.5 g, 2.25 mM) in dry methanol (30 mL) contain-
ing sulfuric acid (1 mL) at 65 °C and the subsequent workup and
column chromatography on silica gel with a chloroform eluent
½2-þ3-ꢁ 100þj%eej
¼
ð6Þ
½2 þ3 ꢁ
100 -j%eej
þ
þ
Combining eqs 5 and 6 leads to
2
τre½reꢁ
100þj%eej
¼
ð7Þ
2
100 -j%eej
τsi½siꢁ
1
(see Figure S10 in the Supporting Information for H NMR).
This assignment is in agreement with the published character-
The average ee values for 2 and 3 were used to calculate the
[re]/[si] ratio (Table 6), which as expected increased as the
temperature was lowered. The ee values are decreased by the
shorter excited-state lifetime for re-AC-TKS compared to
the si complex. The diastereomeric excess for the re and si
complexes were calculated as well as the ee values assuming
equal lifetimes for the re and the si complexes (|ee0|) (Table 6).
It is important to emphasize the inherent advantage of the
enantiodifferentiating photodimerization for a bimolecular
reaction because the product ee can be much larger than the
diastereomeric excess due to the dependence of the ee on the
square of the concentration ratio for the re and si complexes.
The ee can be further enhanced by a strategy where the
lifetime of the most stable complex is designed to be longer
than the lifetime of the least stable complex, the opposite
situation from that observed for TKS159. Our results clearly
showed for the first time that the product’s enantiomeric
excess (ee) is not simply determined by the diastereomeric
ratio of the ground-state complexes but also depends on the
relative lifetimes of the diastereomeric complexes. Therefore,
the future rational design of supramolecular systems to
achieve photochirogenesis should consider thermodynamic
factors, such as the stability of complexes, and kinetic
factors, such as the lifetime of excited states.
ization of this compound.61
Equipment. Circular dichroism (CD) spectra were measured
on a JASCO J-820S instrument equipped with an ETC505T
temperature controller. Fluorescence spectra were collected
with a JASCO FP6500 spectrometer, and samples were kept at
a constant temperature with a Unisoku cryostat (USP-203HP
model). The bandwidths for the emission and excitation slits
were set to 3 and 1 nm, respectively. An excitation wavelength of
361 nm was used for measurements at 25, 0, and -50 °C, and
362.5 nm was used for measurements at -25 °C. The emission
spectra were collected between 380 and 600 nm.
Fluorescence decays were measured with Edinburgh Instruments
OB 920 or FL 920S single-photon counters using a hydrogen-
filled ns-flash lamp as the excitation light source. The bandwidth
was approximately 16 nm for both the excitation and emission
monochromators. The maximum intensity collected was 2000
counts. The excitation and emission wavelengths were 361 and
440 nm, respectively. The excitation and emission monochromators
were set to the excitation wavelength of the sample in order to
collect the instrument response function (IRF) using the scattered
light from an aqueous solution containing suspended silica gel. For
temperatures at or below 0 °C, the IRF was collected using a
dichloromethane solution instead of the aqueous solution. The
decays were fit to an exponential function (i=1) or to a sum of
exponentials (eq 8) using software from the manufacturer to
deconvolute the IRF from the experimental data.
In conclusion, we have established a protocol, which
should become standard, to fully elucidate the mechanism
for photochirogenesis in supramolecular systems. The full
characterization, i.e., spectral features, lifetimes and relative
abundance, of two distinct excited state diastereomeric
complexes was achieved for the first time by combining
lifetime and product studies. These comprehensive studies
revealed the importance of understanding the ground state
behavior of the system, i.e., differential complex stability,
i
X
The parameter Ai corresponds to the pre-exponential factor
for each emissive species, i, with a different lifetime, where the
Aie-k t
ð8Þ
i
IðtÞ ¼ Io
1
(61) Kriegel, R. M.; Saliba, K. L.; Jones, G.; Schiraldi, D. A.; Collard,
D. M. Macromol. Chem. Phys. 2005, 206, 1479–1487.
J. Org. Chem. Vol. 74, No. 20, 2009 7919

Kawanami et al.
JOCArticle
sum of all pre-exponential factors is equal to 1. The parameter
ki corresponds to the decay rate constant of each fluorescent
species, and its inverse value is equal to the singlet excited state
lifetime of the emissive species. The fit of the experimental data
to eq 8 was considered to be acceptable when the χ2 value was
between 0.9 and 1.2. Visual inspection of the residuals and
autocorrelation was also used to determine the quality of
the fit.62
A 500-W ultrahigh pressure mercury lamp fitted with a UV-35
filter (effective λ > 320 nm) was used for the photoreaction experi-
ments. Temperature was controlled using a Unisoku cryostat
(USP-203HP model). The HPLC system was equipped with a
tandem column of Cosmosil 5C18-AR-II (Nacalai Tesque) and
Chiralcel OJ-RH (Daicel). The eluent was a 64:36 (v/v) mixture of
water and acetonitrile, the latter containing 0.1% trifluoroacetic
acid. The rate of flow was 0.5 mL min-1, and 5 μL aliquots of
sample were injected. The columns were kept at 35 °C. The relative
yield and %ee were determined from the peak areas on the HPLC
Determination of the Complex Stability Constants. The com-
plex stability constants were determined using a nonlinear least-
squares method based on the CD spectral changes upon titrating
TKS159 solutions with AC. The titration curves were fit (eq 9)
using KaleidaGraph. A 1:1 equilibrium was assumed between
AC and TKS159. Neither species is in excess, resulting in the
solution for a quadratic equation
Δθ ¼
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2
ð½ACꢁ0þ½TKSꢁ0þ1=KÞ ( ð½ACꢁ0þ½TKSꢁ0þ1=KÞ -4½ACꢁ0½TKSꢁ0
R
½ACꢁ0
2
ð9Þ
where Δθ is the change in ellipticity, K is the complex stability
constant, [AC]0 and [TKS]o are the initial concentrations of AC
and TKS159, and R is the maximum ellipticity when all AC is
bound.
Competition Experiments. Competition experiments for the
binding of AC with TM166 and TKS159 were performed
because AC binding to TM166 did not lead to a CD signal.
Two 0.1 mM AC solutions, containing TKS159 (12 mM) or
TKS159 (12 mM) þ TM166 (12 mM), were prepared, and their
CD spectra were measured at 25, 10, and 0 °C. Under these
conditions, >97% of added AC are bound to a template.
The solution containing just TKS159 leads to an ellipticity
value when all AC is bound to TKS (θTKS), while the observed
ellipticity (θobs) in the presence of TKS159 and TM166 is
decreased by the amount of AC bound to TM166. The ratio
of equilibrium constants is given by eq 10, while the ratio of
complexed AC to each template is calculated from eq 11, and the
free concentration of TKS159 and TM166 are calculated from
the mass balance eqs 12 and 13, where the subscript “T”
corresponds to total concentrations.
chromatogram observed by the fluorescence detector (λex
=
254 nm, λem = 420 nm). This detection method relies on the
photodecomposition of the AC photodimers by the excitation
light, the efficiency of which was confirmed to be identical for all
photodimers, followed by the excitation of the AC monomers
produced in the detection cuvette.63
Sample Preparation. AC stock solutions (approximately
1 mM) were prepared by dissolving AC in dichloromethane
with sonication. The actual AC concentration was determined
by UV/vis (ε397 =4100 M-1 cm-1). TKS159 or TM166 stock
solutions (4.0, 20, or 60 mM) were prepared by dissolving each
template in dichloromethane. TM166 is less soluble than
TKS159, and the stock solutions of TM166 were sonicated to
achieve solubilization. Solutions for UV/vis, CD, and fluores-
cence experiments were made in dichloromethane by adding the
appropriate amounts of AC and template stock solutions, with
the exception of the UV/vis experiments at template/AC ratios
above 20, where solid TKS159 was added to the solutions to
avoid excessive dilution of AC. The concentrations used for UV/
vis and CD experiments were [AC] = 0.1 mM, [template] =
0.02-12 mM ([template]/[AC]=0.2-120), and the concentra-
tions used for fluorescence experiments were [AC]=0.025 mM,
[template]=0.0125 - 2 mM ([template]/[AC]=0.5 - 80). All
UV/vis, CD, and fluorescence solutions were aerated.
Solutions for photoreactions were prepared by mixing 2 mL
of 0.5 mM AC with 2 mL of a TKS159 solution (0.125-60 mM)
to obtain reaction mixtures with [AC] = 0.25 mM, [TKS] =
0.0625-30 mM (TKS/AC=0.25-120). Solutions were placed in
1ꢀ1ꢀ4.5 cm quartz cells containing a stirrer and sealed with
septa. Solutions were deoxygenated by cooling on ice and
bubbling Ar for at least 5 min. Solutions were stirred during
the photoreactions. After photoirradiation, the dichloro-
methane was evaporated to leave an organic residue, which
was redissolved in 4 mL of a 1:1 mixture of aqueous NaOH
(2 mM) and acetonitrile. This mixture was used for the HPLC
analysis.
KTKS
KTM
½AC TKSꢁ½TMꢁ
3
¼
ð10Þ
½AC TMꢁ½TKSꢁ
3
½AC TKSꢁ
θobs
3
¼
ð11Þ
ð12Þ
ð13Þ
½AC TMꢁ
θTKS -θobs
3
θobs
½TKSꢁ ¼ ½TKSꢁT -½ACꢁT ꢀ
θTKS
ðθTKS -θobs
Þ
½TM166ꢁ ¼ ½TM166ꢁT -½ACꢁT ꢀ
θTKS
Job Plots. Solutions of AC and solutions of TKS159 were
prepared at concentrations of 0.8, 0.6, 0.5, 0.4, and 0.2 mM.
Solutions were mixed to make a solution with a total concentra-
tion of AC þ TKS159 of 1.0 mM. CD spectra of all solutions
were measured at 25 and 0 °C, and the CD spectral changes at
389 nm were plotted against [AC]/([AC] þ [TKS159]).
DFT Calculations. All calculations were performed on Linux-
PCs using the TURBOMOLE 5.9 program suite.64 The resolu-
tion of identity (RI) approximation65-69 was employed in all
geometry optimizations, and the corresponding auxiliary basis
sets were taken from the TURBOMOLE basis set library.
UV/vis, CD, and Fluorescence Spectral Titration. Spectra
were measured upon gradual addition of a TKS159 stock
solution to 3 mL of AC solution. UV/vis and CD spectra were
recorded at the various TKS/AC ratios at 25, 10, and 0 °C.
Fluorescence spectra and decays were obtained at 25, 0, -25,
and -50 °C. The maximum volume of added TKS159 was
300 μL. The measured absorbance and ellipticity values were
corrected for dilution.
€
(64) Ahlrichs, R. et al. TURBOMOLE, version 5.9 Universitat Karlsruhe,
Karlsruhe. See also: http://www.cosmologic.de/QuantumChemistry/main_
turbomole.html, 2005.
€
€
(65) Eichkorn, K.; Treutler, O.; Ohm, H.; Haser, M.; Ahlrichs, R. Chem.
Phys. Lett. 1995, 240, 283–289.
(66) Grimme, S.; Waletzke, M. Phys. Chem. Chem. Phys. 2000, 2, 2075–
2081.
(62) Bohne, C.; Redmond, R. W.; Scaiano, J. C. In Photochemistry in
Organized & Constrained Media; Ramamurthy, V., Ed.; VCH Publishers,
Inc.: New York, 1991; pp 79-132.
(63) Nishijima, M.; Wada, T.; Nagamori, K.; Inoue, Y. Chem. Lett. 2009,
38, 726–727.
€
(67) Hattig, C.; Weigend, F. J. Chem. Phys. 2000, 113, 5154–5161.
€
(68) Vahtras, O.; Almlof, J.; Feyereisen, M. Chem. Phys. Lett. 1993, 213,
514–518.
(69) Weigend, F.; Haser, M. Theor. Chem. Acc. 1997, 97, 331–340.
€
7920 J. Org. Chem. Vol. 74, No. 20, 2009

Kawanami et al.
JOCArticle
Diastereomeric AC-TKS and AC-TM complexes were fully
optimized at the dispersion corrected DFT-D/B-LYP level
without symmetry constraints (C1 symmetry) using an AO basis
set of valence triple-ζ quality with a set of polarization functions
(TZVP) with numerical quadrature multiple grid (m4 in TUR-
BOMOLE terminology). The subsequent single-point energy
calculations of the optimized structures were performed by
using the SCS-MP2 method with the basis-set of TZVPP,70-75
and the result was used for the calculation of the final Boltz-
mann distribution among the conformers.76-78
Promotion of Science for the support of the research carried
out at Osaka University; Y.I. and T.C.S.P. thank the Global
COE Program on Bio-Environmental Chemistry, Osaka
University, for the support of the stay of T.C.S.P. at Osaka
University; T.M. thanks the Mitsubishi Chemical Corpora-
tion Fund and Y.K. thanks the Uomoto International
Scholarship Foundation for the support of their stay at
UVic. T.C.S.P. and C.B. thank Natural Sciences and Engine-
ering Research Council of Canada (NSERC) for support
of the research carried out at the University of Victoria, and
T.C.S.P. thanks NSERC for a CGS-D scholarship.
Acknowledgment. Y.I. and T.M. thank the Japan Science
and Technology Agency and the Japan Society for the
Supporting Information Available: UV-vis, fluorescence
and circular dichroism spectra of AC complexed to TKS159 or
TM166 (Figures S1-S8), time-resolved fluorescence decays for
AC-TKS (Figure S9), lifetimes for AC complexed to TKS159
or TM166 at different AC/template ratios and temperatures
(Tables S1-S5), product studies for AC-TKS (Table S6),
NMR for ACOMe (Figure S10), table (S7) with the atomic
coordinates for the DFT-D-calculated AC-TKS or AC-TM
complexes, and the complete ref 64. This material is available
free of charge via the Internet at http://pubs.acs.org.
(70) Gerenkamp, M.; Grimme, S. Chem. Phys. Lett. 2004, 392, 229–235.
€
(71) Goumans, T. P.; Ehlers, A. W.; Lammertsma, K.; Wulthwein, E.-U.;
Grimme, S. Chem. Eur. J. 2004, 10, 6468–6475.
(72) Grimme, S. J. Chem. Phys. 2003, 118, 9095–9102.
(73) Grimme, S. J. Comput. Chem. 2003, 24, 1529–1537.
(74) Grimme, S. J. Phys. Chem. A 2005, 109, 3067–3077.
(75) Piacenza, M.; Grimme, S. J. Comput. Chem. 2004, 25, 83–99.
(76) Mori, T.; Grimme, S.; Inoue, Y. J. Org. Chem. 2007, 72, 6998–7010.
(77) Mori, T.; Inoue, Y.; Grimme, S. J. Phys. Chem. A 2007, 111, 4222–
4234.
(78) Mori, T.; Inoue, Y.; Grimme, S. J. Phys. Chem. A 2007, 111, 7995–
8006.
J. Org. Chem. Vol. 74, No. 20, 2009 7921