Fluorescence detected exciton coupled circular Dichroism: Development of new fluorescent reporter groups for structural studies

Article abstract of DOI:10.1007/s00706-004-0276-5

In collaboration with Jasco Corporation we have recently developed an FDCD (fluorescence detected circular dichroic) instrument J-465, which eliminates the photoselection artifacts and efficiently collects the emitted light from the sample solution based

Full text of DOI:10.1007/s00706-004-0276-5

Monatshefte f€ur Chemie 136, 367–395 (2005)
DOI 10.1007/s00706-004-0276-5
Fluorescence Detected Exciton Coupled
Circular Dichroism: Development
of New Fluorescent Reporter
Groups for Structural Studies
Katsunori Tanaka1, Gennaro Pescitelli2, Koji Nakanishi1,
and Nina Berova1;ꢀ
1
Department of Chemistry, Columbia University, New York, New York 10027, USA
2
ꢀ
Dipartimento di Chimica e Chimica Industriale, Universita degli Studi di Pisa,
I-56126 Pisa, Italy
Received September 17, 2004; accepted October 30, 2004
Published online March 8, 2005 # Springer-Verlag 2005
Summary. In collaboration with Jasco Corporation we have recently developed an FDCD (fluores-
cence detected circular dichroic) instrument J-465, which eliminates the photoselection artifacts and
efficiently collects the emitted light from the sample solution based on the ideal ellipsoidal mirror
principle. Using the J-465 we have investigated a variety of fluorophores with=without polarization for
exciton-coupled FDCD stereochemical analysis. The following three cases extend the FDCD meth-
odology to new challenging areas beyond the limits of conventional CD: (1) substrates containing C¼C
double bonds, (2) molecules with sterically hindered hydroxyls, and (3) substrates bearing remote
stereogenic centers. The pico- to nano-level FDCD analysis described in this paper leads to promising
opportunities for the stereochemical analysis of a wide range of natural products with minuscule
amounts of sample available.
Keywords. Fluorescence detected exciton coupled CD; Photoselection; Sensitivity enhancement;
Cross metathesis; Styrenoid.
Introduction
Circular dichroic spectroscopy (CD) is a general technique for studying chiral
molecules [1–3]. In particular, electronic CD, which is associated with transitions
between electronic states, is commonly used for structural investigations of organic
and inorganic compounds, polymers, and biomacromolecules. The appeal of CD
is frequently associated with high structural sensitivity and the relative ease in
interpreting or predicting CD spectra, thus making CD the tool of choice for
ꢀ
Corresponding author. E-mail: ndb1@columbia.edu

368
K. Tanaka et al.
complementing NMR and other spectroscopic methods in determining the struc-
tures of chiral compounds [1–3].
The key step in any spectroscopic application consists of translating experi-
mental spectra into structural information. A variety of methods, depending on the
nature of the molecule, the number and type of chromophores, etc., are available
for the interpretation of electronic CD. A most efficient and useful approach, the
‘‘exciton chirality method’’, originates in the spatial interaction between two or
more strong chromophoric electric dipole transitions that are chirally arranged and
close in energy [4, 5]. When the exciton or coupled dipole mechanism of optical
activity holds, the CD of a molecule can be quantitatively predicted on the basis of:
(1) molecular conformation (relative position and orientation between chromo-
phores) and (2) the spectroscopic properties of the constituent chromophores,
i.e., transition energies, dipolar strengths, transition dipole positions, and polariza-
tions. Conversely, the appearance of an exciton coupled CD spectrum offers direct
and straightforward information on the absolute molecular structure [4, 5].
In the simple case of two interacting electric dipole allowed transitions, the
strength of the exciton coupling, i.e., perturbation and mixing term of the excited
states of the two chromophores, is directly proportional to their dipolar strength
(and then to "2), and inversely proportional to the square of their spatial distance R
and to their energy difference DE. The coupling is maximal for the degenerate case
in which the two transitions are equivalent and the excited states are split into two
states separated by the so-called Davydov splitting. The exciton coupling also de-
pends on the mutual arrangement of transition dipoles through a simple dipole-
dipole coupling term, which is zero for collinear or coplanar dipoles, whereas it
reaches a maximum at dihedral angles around 70ꢁ for dipoles lying on quasi-
parallel planes [4, 5].
A simple example of exciton coupling is when two chirally arranged and spa-
tially close chromophores interact without perturbation from other factors. This
results in an exciton-split CD spectrum (a CD couplet) consisting of two Cotton
effects of opposite signs flanking the absorption band; in degenerate cases when the
two chromophores are identical, the couplet crossover point coincides with the
absorption maximum. The sign of the CD couplet is defined by the sign of its long
wavelength component and reflects the absolute sense of twist between two inter-
acting dipoles. The sense of twist can be inferred by viewing the two dipoles along
the direction connecting their centers (the transition dipole positions). If a clock-
wise rotation is required to superimpose the dipole in the front to that in the rear,
the chirality is positive and vice-versa. Thus this exciton chirality rule permits one
to readily assign absolute configurations from exciton coupled CD spectra; Fig. 1
shows a typical couplet exhibited by a chiral diol bis(4-dimethylaminobenzoate).
The couplet amplitude ACD, or the intensity difference between the peak and
trough, depends on the coupling strength, and is proportional to "2=R2 and to a geo-
metrical dipolar factor G; the width of the couplet WCD, Davydov splitting, is
proportional to G ꢂ "=R3. For non-degenerate cases, ACD is also inversely propor-
tional to the energy difference DE between the two isolated transitions.
The exciton chirality approach is widely applicable since the coupled dipole
mechanism is usually very strong, and in the presence of suitably arranged chro-
mophores, it dominates the other possible sources of optical activity [3]. In fact,

Fluorescence Detected Exciton Coupled Circular Dichroism
369
Fig. 1. Exciton-coupled CD spectra of bis(4-dimethylaminobenzoate) of the chiral diol, 5ꢀ-
cholestan-3ꢁ,6ꢀ-diol; D" is the unit of molꢃ1 dm3 cmꢃ1
when the chiral molecule lacks two proper chromophores, rather than resorting to
different tools for interpreting its CD spectrum, it may be convenient to introduce
one or two chromophoric groups by chemical derivatization and proceed with the
exciton method; in such cases, strong and red-shifted chromophores that do not
interact with pre-existing chromophores should be selected [4, 5]. A further recent
approach consists of analyzing the CD of supramolecular assemblies formed
between the chiral compound acting as a guest and achiral bis-chromophoric hosts,
e.g., bis-tetraarylporphyrin tweezers [6–11]. Since the CD signals associated with
the exciton mechanism are usually strong, the technique is also very sensitive, and
microscale applications are possible: a few microgram or nanogram amounts of
sample at micromolar concentration suffices for acquiring a well defined exciton-
split CD [12–16]. A second consequence of this enhanced sensitivity is that exciton
CD spectra of bis-chromophoric compounds can be significant and diagnostic even
˚
when the two chromophores are separated by 40–50 A [4]. Such features have
rendered the exciton chirality method a popular approach for configurational and
conformational analyses of chiral molecules, notably in the field of biologically
relevant compounds [4, 5].
When three or more chromophores are present, the overall CD can be approxi-
mated by summation of all possible pair-wise interacting terms [17]. In more
complicated cases, and also when a more quantitative approach is required, the
entire exciton coupled CD is calculated. One of the most widely employed calcu-
lation is the DeVoe coupled oscillator approach [18–20] for structural elucidations

370
K. Tanaka et al.
of small molecules [21–26], biopolymers [27–29], and theoretical investigation of
exciton interactions [30]. According to this approach each transition dipole is
described as a classical damped oscillator with its energy, strength, position, and
direction. The optical properties of the oscillator aggregate thus depend only on
their dipolar interaction matrix and may be readily calculated.
Despite the advantages of electronic CD, two limitations are relevant.
(i) Sensitivity. CD is defined as the differential absorption of left and right circu-
larly polarized radiation (LCP and RCP), and is commonly measured as the
differential transmission intensity in the direction of the incident light. The fact
that CD is based on difference spectroscopy makes it one or two orders of mag-
nitude less sensitive than absorption spectroscopic methods; in ideal cases the
g-factor (DA=A or D"=" value) is 10ꢃ2. CD is then substantially less sensitive
than other spectroscopies such as fluorescence. Even in the most sensitive ex-
citon split CD arising from interaction between two strong chromophores such
as tetraarylporphyrins, the concentration limits for CD detection are seldom
below 10ꢃ6ꢄ7 M [12–16].
(ii) Selectivity. As an example, the CD analysis of a complex consisting of a chiral
compound (ligand) attached to a chiral polymeric receptor could be quite
challenging since the ligand CD is superimposed on that of the receptor, as
is frequently the case encountered in biomacromolecules, where it is especially
important to determine the local ligand conformation that often play critical
roles in biological processes.
Fluorescence detected circular dichroism (FDCD) combines the structural sen-
sitivity and chiral specificity of CD with the sensitivity and specificity of fluores-
cence [31–36]. It consists in measuring the differential emission of light from a
sample excited with LCP and RCP radiation, and is based on the assumption that
the amount of light emitted depends exclusively on the amount absorbed; in other
words, the excitation spectrum of the fluorophore parallels the absorption spec-
trum, and the molecular quantum yield is the same for both circularly polarized
components (which may not necessarily be the case) [37]. FDCD is then an indirect
way of measuring a differential absorption [31, 32]. It is sensitive to both chiral and
fluorescent molecules, and therefore is far more specific than standard transmission
CD, because in principle the CD associated with a single fluorophoric molecule or
moiety may be selectivity extracted in the presence of many non-fluorescent chro-
mophores. Moreover, similar to the higher sensitivity of fluorescence compared to
absorption spectroscopy, the direct measurement of emitted radiation against a zero
background renders FDCD more sensitive than conventional CD.
The first description of the theoretical basis of FDCD phenomenon and experi-
mental framework date back to 1974 when Turner, Tinoco, and Maestre were able
to estimate the selective contribution arising from the chiral fluorophore tryptophan
in a mixture with the chiral non-fluorescent chromophore cysteine [31]. The ex-
perimental setup used for FDCD detection is as follows: behind a cell placed after
the source of CP light (similar to conventional CD instruments), an FDCD photo-
multiplier (PMT) was placed along the direction orthogonal to the incident ra-
diation (different from CD measurement) to collect the emitted light. A cutoff
long-pass filter was inserted between the cell and PMT to prevent scattered light

Fluorescence Detected Exciton Coupled Circular Dichroism
371
to be detected. In the simple case of a non-viscous solution of a single chiral and
fluorophoric molecule, e.g. 10-camphorsulfonic acid, the measured FDCD may be
converted into CD by Eq. (1) [31, 32] where A is the sample absorption,
DA ¼ cdD" (c is the sample molar concentration and d the cell length in cm), FL
and FR are difference in the emission intensities upon LCP and RCP excitation, and
k is a conversion constant specific for each machine. After the FDCD instrument
outputs FL ꢃ FR and FL þ FR (as AC and DC current components), the absorption
A of the same sample affords the desired CD normarized in D" units; A must be low
to prevent both intermolecular fluorescence quenching and excessive attenuation of
the more absorbed CP component [38]. For more complex samples, i.e., in the
presence of other non-fluorescent CD-active or fluorescent CD-inactive species,
which can give rise to energy transfer with the observed fluorophore, more com-
plicate formulae must be used [32, 38].
DA ¼ 2Sð1 ꢃ 10ꢃAÞ=ln 10 ꢂ 10ꢃA; S ¼ kðFL ꢃ FRÞ=ðFL þ FRÞ
ð1Þ
Even in most straightforward cases, it soon became clear that FDCD would
suffer from a problem not encountered in conventional CD, namely, the anisotropic
excitation, known as photoselection, which depends on temporary orientation of the
molecules during the irradiation process [39, 40]. In fact, the light absorbed by
the system does not equally excite all molecules with various orientations, because
the probability of light absorption by a molecule depends on the direction of its
electric and magnetic dipoles with respect to the incident light wave vector. This
would lead to an emission emerging from a non-isotropic collection of excited
molecules, unless unrestricted Brownian movement causes them to rotate rapidly
and their orientation between excitation and emission phenomena is randomized
leading to an isotropic emission process. The latter is usually the case when the
rotational relaxation time is sufficiently shorter than the fluorescence lifetime.
In the presence of photoselection, the observed fluorescence will be polarized,
that is, the vertical and horizontal components F and F differ and the ratio
?
k
PF ¼ ðFk ꢃ F?Þ=ðFk þ F?Þ, known as fluorescence polarization, is non-vanishing.
Consequently, the measured FDCD will be a complex function of the isotropic CD
and of some particular components of the dipole and rotational strength tensors; in
other words, the observed differential emission will not reflect the differential
absorption in a straightforward manner. The only exception to this behavior is
observed in the uncommon situation when the molecular transition absorption
and emission dipoles either coincide or are mutually perpendicular. In all other
cases, special instrumental setups have been proposed which would minimize the
anisotropic contribution; they consist of placing the PMT along a ‘‘magic’’ direc-
tion or placing a linear polarizer in front of the PMT with a ‘‘magic’’ rotation angle.
In practice, however, the sample photoselection will introduce additional severe
artifacts in FDCD, due to the linear polarization present in the CP light produced
by imperfect optical components in commercial CD instruments. If the LCP=RCP
excitation beams contain some residual linear polarization, a difference in the
fluorescence signal will be detected not arising from the sample optical activity
[41–43]. In practice, the whole measurement will be distorted as with other emis-
sion or scattering spectroscopies such as the circular polarization of luminescence
and Raman spectroscopy [44–46].

372
K. Tanaka et al.
The elimination of these artifacts is possible either upon collection of the
emitted light along more than one direction, or by specific selection of the unpo-
larized component of the emitted light. This has been achieved by using a double
PMT-system [42, 43], introducing a linear polarizer between the sample and PMT
[41], or by conveying the emitted light through an ellipsoidal mirror surrounding
the whole cell [47]. Together with Jasco, we have recently developed an ellipsoidal
mirror prototype, Jasco FDCD J-465, which can be attached to any modern CD
instrument [48]. J-465 not only eliminates artifacts due to fluorescence polariza-
tion, but also increases the signal-to-noise ratio S=N by collecting a large fraction
of emitted radiation, as discussed in the following [48].
The applications of FDCD reported so far may be divided into three main
classes:
(1) cases when the excited states of chiral fluorescent molecules were investigated
and=or their emission properties were exploited;
(2) cases when a selective CD detection of a fluorophore in the presence of other
chromophores was required;
(3) cases when FDCD was used as a more sensitive alternative to CD.
To group (1) belong a few interesting applications such as the measurement of
lifetime-resolved FDCD [49–51], which helps in separating contributions from
multi-component systems, and the quenching detection of chiral molecules also
through multidimensional FDCD experiments [52, 53]. To group (2) belong
numerous examples of mixtures of a fluorophoric probe and a chromophoric spe-
cies, i.e., a fluorescent dye associated with a biopolymer) [54–59], or of a fluor-
ophore inserted into a non-fluorescent chiral matrix such as proteins or nucleic
acids [60]. In this context, FDCD affords important local chiral information: the
FDCD of a single tryptophan within a complex protein may in fact be extracted
and changes in its environment accompanying binding or conformational changes
can be followed [43, 61, 62]; contribution to CD from the backbone and side
chains of poly-tryptophan and poly-tyrosine can be separated [63, 64]; moreover,
fluorescent and non-fluorescent aggregation or conformational states can be
detected [65, 66]. Despite many interesting reports on FDCD selectivity, this
important feature of the method still remains underestimated. Even less exploited
is the opportunity for sensitivity enhancement, the second crucial feature of this
technique. Apart from the promising employment of FDCD as a detection tool in
chromatography [67, 68], electrophoresis [67, 68], and determination
of enantiomeric excesses [69], only a few other reports have provided data regard-
ing FDCD sensitivity enhancement, obviously due to the still prevailing photose-
lection problem [70–75].
FDCD was applied for the first time in 1997 to exciton coupled molecules such
as 2-naphthoate, 6-methoxy-2-naphtoate, and 2-anthroate diesters of 1,2-cyclohex-
anediol and some steroidal di=triesters (Table 1) [74].
It was shown that in these favorable cases, exciton coupled FDCD spectra were
in excellent agreement with conventional CD spectra. More importantly, FDCD
showed satisfactory S=N ratios at concentrations much lower than those in CD,
leading to a 100-fold sensitivity increase; while CD detection limit was of the order
of 10ꢃ6 M, that of FDCD was around 10ꢃ8 M [74]. Further investigations using

Fluorescence Detected Exciton Coupled Circular Dichroism
373
Table 1. UV and fluorescence properties of previously utilized chromophoresa
b
abs
max
em
Fluorophore
Solvent
ꢂ
=nmð"Þ
ꢂ
=nm
max
Ff
Pf
(a) Strong fluorescence with low polarization
MeCN
MeCN
MeCN
234 (58000)
237 (48000)
258 (93000)
359
374
434
0.32
0.65
0.62
0.0032
0.0026
0.0023
(b) Strong fluorescence with high polarization
MeCN 270 (21000)
337
0.66
0.034
(c) Weak or nonfluorescence
MeCN
MeCN
MeCN
306 (23400)
244 (19500)
257 (20400)
380
n.a.
n.a.
0.0005
n.a.
0.4
n.a.
n.a.
n.a.
a
Ff ¼ fluorescence quantum yield, Pf ¼ fluorescence polarization, R ¼ (1R,2R)-trans-1,2-cyclohex-
ane diol monoester; " is the unit of molꢃ1 dm3 cmꢃ1
b
1,2-cyclohexanediol as a chiral scaffold confirmed that similar sensitivity enhance-
ments may be attained for all cases when suitable fluorophores are employed [75].
Such fluorophores should possess, in addition to large " values, a high fluorescence
quantum yield Ff (above 0.3) and negligible or low fluorescence polarization Pf
(below 0.02). In contrast, the FDCD spectra of 4-phenylbenzoate and especially 4-
methoxycinnamate bisesters were not in good agreement with CD, because of ar-
tifacts associated with the relatively large Pf. Finally, FDCD spectra of weakly or
non-fluorescent 4-methoxycinnamate, 4-bromo, and 4-methoxybenzoate bisesters,
as expected, could not be measured. Diesters of 1,2-cyclohexanediol containing
two different chromophores demonstrated that the presence of a single favorable
fluorophore is sufficient to obtain significant and artifact-free FDCD spectra. Thus
the combination of naphthoate-type fluorophores with either 4-methoxybenzoate or
4-phenylbenzoate gave FDCD spectra in which both wings of the exciton couplet
were apparent, in full accord with CD [75]. It appears that when the two chromo-
phores are excitonically coupled, the two mixed excited states both become fluo-
rescent and emit a FDCD signal. In other words, a very fast exciton-type energy
transfer mechanism presumably occurs in such cases [76–78]. It was also verified

374
K. Tanaka et al.
that when a possibility of F€orster-type energy transfer exists, e.g., from 6-methoxy-
2-naphtoate to 2-anthroate fluorophore, a good agreement between FDCD and CD
spectra is still attained [75]. On the other hand, a combination between naphthoate-
type fluorophores and either totally non-fluorescent chromophores, such as
4-bromobenzoate, or fluorophores with high Pf, such as 4-methoxycinnamate,
exhibited FDCD spectra dissimilar to CD due to apparent artifacts. It is important
to notice that a Jasco FDCD prototype (J-357) was used for this study analogous to
the instrument introduced by Tinoco [31]: a single PMT was placed along the
direction orthogonal to the incident light, and no polarizer or light-collecting mir-
rors were introduced. It was thus concluded that [75]:
(1) when suitable fluorophores with high Ff and low Pf are already present or
introduced in the molecule, exciton coupled FDCD spectra represent a power-
ful alternative to conventional CD. In fact, FDCD with a sensitivity increase of
up to 100-fold disclose the same structural information as that resulting from
CD.
(2) Fluorophores with even high Ff values but with significant Pf represent a
challenge for FDCD if the artifacts due to photoselection are not eliminated
by the FDCD device.
It became clear that future applications of FDCD require development not only
of new fluorophores with favorable properties but also of improved FDCD detec-
tors which would minimize photoselection effects.
A recently developed ellipsoidal mirror FDCD device (J-465) has shown prom-
ising features and is expected to solve the problem of polarized and weak fluo-
rescence [48]. First, by averaging de facto radiation emitted all around the sample, it
eliminates the artifacts due to photoselection. Second, by collecting a large fraction
of emitted radiation, not only along a single direction but multiple directions, it
increases the intensity of signal collected by the PMT, thereby leading to sensitive
detection of weakly fluorescent samples. A previous device consisting of a linear
polarizer, with a proper angle rotation, placed between the sample and PMT
(implemented in Jasco FDCD J-405 prototype), was shown to similarly help elim-
inating FDCD artifacts, but suffers from reduced sensitivity and is therefore of
limited applicability [48]. Figure 2 shows the CD and FDCD spectra of (1R,2R)-
trans-cyclohexanediol bis(6-methoxy-2-naphthoate) in the viscous solution of
methanol=glycerol. In this case fluorescence anisotropy, Pf ¼ 0.050, was purposely
increased by employing glycerol in addition to methanol. At 10ꢃ6 M concentration
(Fig. 2a), CD is in good agreement with the artifact-free FDCD obtained by
employing either a linear polarizer rotated by 81ꢁ with respect to the vertical
position (J-405) or the ellipsoidal-mirror light-collecting device (J-465). The
FDCD measured with J-357 device where the photoselection is not eliminated
differs significantly from the CD. At 10ꢃ8 M concentration (Fig. 2b), only FDCD
measured with J-465 shows the CD signals with reasonable S=N ratio, whereas at
this concentration, both conventional CD and FDCD measured with J-405 are
unacceptable.
In the following, three challenging cases so far unexplored by the exciton
coupled CD method have been now solved by application of a recently developed
cross metathesis=CD chiroptical protocol, and further improved by combination

Fluorescence Detected Exciton Coupled Circular Dichroism
375
Fig. 2. CD and FDCD of (1R,2R)-trans-cyclohexanediol bis(6-methoxy-2-naphthoate) in
glycerol=MeOH (Ff ¼ 0.37, Pf ¼ 0.050); CD (line a), FDCD by J-465 (line b, ellipsoidal mirror,
masks 3.0 mm, light-cut filter, 320 nm), FDCD by J-405 (line c, polarizer 81ꢁ), and FDCD by
J-357 (line d, Pf not eliminated); D" is the unit of molꢃ1 dm3 cmꢃ1
with sensitivity enhanced FDCD detection by employment of the ellipsoidal mirror
device J-465. The three cases include: (i) substrates containing C¼C double bonds,
(ii) compounds bearing sterically hindered hydroxyl groups, and (iii) substrates
bearing remote stereogenic centers.

376
K. Tanaka et al.
Results and Discussions
1. Convenient Fluorophores for Exciton Chirality Studies by FDCD
The fluorophores used in this study and their UV-Vis and fluorescence properties
are listed in Table 2. The fluorophores are categorized into three types 1a–1c
according to the methods of introduction into the substrates.
1a. Styrenoid Fluorophores (1, 2, 3a, and 3b) for C¼C
Double Bond Containing Compounds
These styrenoid reagents could be introduced into a variety of substituted C¼C
double bonds by cross metathesis [79, 80] using Grubbs’ ruthenium catalyst
Table 2. Introduction methods, UV-Vis, and fluorescence properties of 1–6a
b
abs
max
em
Fluorophore
Solvent
ꢂ
=nmð"Þ
ꢂ
=nm Ff
max
Pf
(a) Introduced at C¼C Double Bonds by Cross Metathesis
1
MeCN
248 (15000)
304
331
0.16 0.0020
0.65 0.0029
0.12 0.0010
2
MeCN
277 (23700)
3a (X ¼ H2)c
CH2Cl2 418 (420000) 651
3b (X ¼ Mg)d CH2Cl2 425 (560000) 661
0.23 0.0029
(b) Introduced at –OH Group by Acylation
4
MeCN
270 (21000)
337
0.66 0.034
0.11 0.0025
5a (X ¼ H2)
CH2Cl2 418 (440000) 650
5b (X ¼ Mg)d CH2Cl2 425 (540000) 661
0.19 0.0015
(c) Introduced at Sterically Hindered –OH Groups by Acylation and Cross Metathesis
6e
CH2Cl2 419 (387000) 651 0.13 0.002
a
c
Ff ¼ fluorescence quantum yield, Pf ¼fluorescence polarization; b " is the unit of molꢃ1 dm3 cmꢃ1
;
preparation is shown in Scheme 1; d corresponding magnesium porphyrins 3b and 5b were prepared
by the reaction of 3a and 5a with excess amount of MgI2 and Et3N in CH2Cl2 at rt in quantitative
e
yields; preparation is shown in Scheme 1, R ¼ 3-oxo-5ꢀ-androstan-17ꢁ-yl

Fluorescence Detected Exciton Coupled Circular Dichroism
377
[81–83]. Whereas styrene (1) itself is weakly fluorescent, p-vinylbiphenyl (2)
is a more favorable FDCD fluorophore owing to its red shifted absorbance,
higher extinction coefficient, and excellent fluorescent properties (277 nm, " ¼
23700 molꢃ1 dm3 cmꢃ1, Ff ¼ 0.65, Pf ¼ 0.0029).
A fluorophore that possesses a very intense UV-Vis absorption has to be attached
when the stereogenic centers are remote and thus expected to give only weak long
range exciton coupling [84, 85]. For this purpose, we developed the vinyl-TPP (3a,
ꢂ ¼ 418 nm, " ¼ 420000 molꢃ1 dm3 cmꢃ1). Although the fluorescence is weaker
(Ff ¼ 0.12), its high extinction coefficient gives rise to enhanced exciton coupling
and thus sensitive FDCD detection. It is noteworthy that the extinction coefficient of
corresponding magnesium porphyrin 3b is even larger (ꢂ ¼ 425 nm, " ¼
560000 molꢃ1 dm3 cmꢃ1) and the quantum yield (Ff ¼ 0.23) is twice that of 3a
[86]. Since it is known that the radiative lifetime of the porphyrin derivatives is
relatively long (ca. 10 ns) [86], both vinyl-TPP 3a and its magnesium derivative
3b show low polarization. The synthesis of 3a has been reported [87, 88], but this
fluorophore can also be readily prepared in high yields from the commercially avail-
able tetraphenyl carboxylic acid methyl ester 7 as shown in Scheme 1.
1b. Fluorescent Carboxylic Acids (4, 5a, and 5b)
for Derivatization of Hydroxyls
p-Phenylbenzoate 4 shows a high quantum yield (Ff ¼ 0.66), but its relatively
high polarization (Pf ¼ 0.034) [75] has so far restricted its application with
previous FDCD techniques. The commercially available TPP-carboxylic acid (5a,
ꢂ ¼ 418 nm, " ¼ 440000 molꢃ1 dm3 cmꢃ1) employed in observing long range CD
exciton coupling [89], has weak fluorescence (Ff ¼ 0.11) and bears a low polariza-
tion (Pf ¼ 0.0025). In accordance with the trend observed for vinyl-TPP 3a, the
incorporation of magnesium in 5a leads to an increase in both extinction coefficient
Scheme 1

378
K. Tanaka et al.
(5b, " ¼ 540000 molꢃ1 dm3 cmꢃ1) and quantum yield (Ff ¼ 0.19), a two-fold en-
hancement as compared to 5a.
1c. Cinnamate Type Fluorophore (6) for Sterically Hindered Hydroxyls
Cinnamate type fluorophore 6 was particularly developed for derivatizing sterically
hindered hydroxyls that are separated over a large distance and therefore
are difficult to be analyzed with other chromophores. Fluorophore 6 shows
UV-Vis and fluorescent properties compatible to those of 3a and 5a (" ¼
387000 molꢃ1 dm3 cmꢃ1, Ff ¼ 0.13, Pf ¼ 0.002). The synthesis of 6 is shown
in Scheme 1. Steroid 9 containing a sterically congested 17-hydroxyl was first
acylated with acryloyl chloride 10 to give 11 in quantitative yield. The vinyl group
in 11 was then subjected to cross metathesis by vinyl-TPP 3a to provide 6 in 72%
yield. The detailed chemistry and application to sterically hindered hydroxyls are
described in Sections 2c and 2d.
Some of the fluorophores listed in Table 2, such as styrene 1, have already been
used as exciton coupled ‘‘chromophores’’ for structural analysis of allylic alcohols,
amines, and dihydrofuroangelicins (see below). In this study, these fluorophores are
reconsidered as favorable and promising for the sensitivity enhanced FDCD anal-
ysis. Successful examples and applications of 1–6 toward stereochemical analyses
by exciton coupled FDCD are shown in the following Sections. The CD and FDCD
results are summarized in Table 3.
2. Application of Fluorophores 1–6 for Structural Analysis
of Selected Chiral Compounds by FDCD. Scope and Limitations
2a. Stereochemical Analysis of Allylic and Homoallylic Alcohols
The configurations of many biologically active natural products with ene moieties
are unknown, and moreover, some are available in only minuscule amounts
[90–96]. We have recently developed a general microscale CD method where the
double bond is converted into a styrene by cross olefin metathesis. The styrenoid
chromophore then couples with the allylic acylate or a preexisting chromophore to
yield a distinct CD couplet. We have studied ‘‘Corey-lactone’’ 13 as a model for
allylic and homoallylic alcohols (Scheme 2a) [12]. The double bond in 13 (10 ꢃg)
was thus reacted with 1 in the presence of catalyst 12 to give styryl alcohols 14a
and 14b. Acylation with 4-phenylbenzoic acid 4 gave 15a and 15b, the conforma-
1
tion being deduced from H NMR and conformational analysis as shown in
Scheme 2a. The CD of 15a and 15b exhibited the expected negative and positive
couplets, in agreement with the known chirality of 13 (Fig. 3a, Table 3).
An FDCD exciton analysis utilizing the fluorescence of styrene and 4-phenyl-
benzoate moieties was performed. A comparison of detection sensitivity of CD and
FDCD for 15a is shown in Fig. 3a (data for 15b is summarized in Table 3).
Whereas the detection limit for CD measurement is at 10ꢃ6 M, FDCD is still
observable at 10ꢃ8 M, although the fluorescent impurity included in the solvent
(excitation at 245 nm) partially interferes with the clear observation of the positive
couplet in the lower wavelength region (vide infra). It is noteworthy that the

Fluorescence Detected Exciton Coupled Circular Dichroism
Table 3. Summary of CD and FDCD
379
Compounds
Solvent
D"=ðꢂ=nm)a
ACD
CD detection
Pf
FDCD detection
amplitude
limit conc=M
limit conc=M
Allylic and Homoallylic Alcohols
MeCN
þ12 (273)
ꢃ10 (248)
þ21
ꢃ45
10ꢃ6
0.17
10ꢃ8
15b
10ꢃ6
0.11
10ꢃ8
15a
MeCN
ꢃ29 (270)
þ16 (247)
17a (R ¼ NMe2)
MeCN
MeCN
ꢃ41 (312)
þ22 (280)
ꢃ63
ꢃ45
10ꢃ6
0.022
0.18
10ꢃ8
18a (R ¼ Ph)
ꢃ31 (291)
þ17 (263)
10ꢃ6
10ꢃ8
Dihydrofuroangelicins with Preexisting Coumarin Fluorophore
MeCN
MeCN
þ15 (320)
þ32
ꢃ23
10ꢃ6
0.047
0.087
10ꢃ8
sty-19
ꢃ17 (250)
ꢃ11 (320)
þ12 (251)
10ꢃ6
10ꢃ8
sty-20
sty-21
MeCN
ꢃ14 (32)
ꢃ28
ꢃ51
10ꢃ6
0.030
10ꢃ8
þ14 (243)
Ginkgolide Derivatives with Sterically Hindered –OH Groups
MeCN
ꢃ33 (291)
þ18 (263)
10ꢃ6
0.18
10ꢃ8
32
(continued)

380
K. Tanaka et al.
Table 3 (continued)
a
Compounds
Solvent
D"=ðꢂ=nmÞ
ACD
CD detection
Pf
FDCD detection
amplitude
limit conc=M
limit conc=M
Mono- and Bis-steroids with Remote Stereogenic Centers
35a (X ¼ H2)
CH2Cl2
þ83 (424)
ꢃ44 (414)
þ127
þ203
10ꢃ7
ꢄ0
ꢄ0
10ꢃ9
35b (X ¼ Mg) CH2Cl2
þ70 (431)
10ꢃ7
10ꢃ10
ꢃ133 (424)
40
CH2Cl2
ꢃ19 (425)
þ18 (418)
ꢃ37
10ꢃ6
ꢄ0
10ꢃ8
a
D" is the unit of molꢃ1 dm3 cmꢃ1
exciton couplet of 15a can be obtained at 10ꢃ8 M using the J-465 instrument even
in the presence of extremely high polarization of 15a (Pf ¼ 0.11, Table 3). The
results clearly show that the polarization artifacts can be completely eliminated
while retaining the highly sensitive CD analysis in emission, namely a 100-fold
enhancement in sensitivity compared with CD in transmission.
In order to examine the scope and limitation of the FDCD detection for these
allylic type compounds, a more fluorescent 4-phenylstyrene 2 (Ff ¼ 0.65) was
introduced in 13 (Scheme 2b); the double bond in 13 was replaced with a 4-
phenylstyrenoid in the presence of the ‘‘Ru’’ catalyst to provide 16a and 16b,
ca. 60% yields. We then attempted to introduce either a non-fluorophore or fluo-
rophore at 11-hydroxyl of 16a (Scheme 2b); one is the non-fluorescent 4-dimethyl-
aminobenzoate coupled with the fluorophoric 4-phenylstyrenoid [75]. Another is
the highly fluorescent 4-phenylbenzoate (4, Ff ¼ 0.66) [75] aimed at further re-
ducing the FDCD detection limit of 18a down to 10ꢃ9 M. The CD and FDCD data of
17a and 18a are shown in Figs. 3b and c. In accordance with our previously
observed phenomena, the FDCD spectrum of 17a exhibited a couplet over the entire
region with 100-fold enhanced sensitivity over CD; the detection limit for CD was
evaluated at 10ꢃ6 M, whereas for FDCD was at 10ꢃ8 M. This phenomenon would
expand the range of usable chromophores and=or fluorophores pairs for exciton
coupled FDCD leading to a wider application of FDCD in structural analyses.
Although two strong fluorophores, 4-phenylstyrenoid and 4-phenylbenzoate
(Ff >0.6), were employed in 18a, the concentration limit could not go below
10ꢃ8 M; no reliable FDCD couplet signals were observed at 10ꢃ9 M (line c in
Fig. 3c). In this case also, the fluorescent impurities at 245 nm interfered with

Fluorescence Detected Exciton Coupled Circular Dichroism
381
Scheme 2
the FDCD measurements at this very low concentration. Therefore, it is desirable
to use the TPP fluorophore, such as 5a, with much larger extinction coefficients
(" ꢄ 440000 molꢃ1 dm3 cmꢃ1) for performing highly sensitive analyses. In addition,
since porphyrins feature red-shifted UV absorption at ca. 420 nm, the FDCD mea-
surement would not be hampered by excitations caused by the solvent impurities in
the lower wavelength regions even under increased HT voltage conditions. Suc-
cessful examples using these porphyrin fluorophores are discussed below.
2b. Dihydrofuroangelicins – Coumarin as a Preexisting Fluorophore
Although coumarin derivatives are associated with a variety of pharmacological
activities [97–101] and contain chromophores having an experimentally accessible
absorbance band above 300 nm, which are also fluorescent, the recently developed
cross metathesis protocol so far represents the only general method for a
chemical=CD approach for determination of their absolute configurations [24, 26,
102–104]. This has been achieved for a series of dihydrofuroangelicins bearing a

382
K. Tanaka et al.
Fig. 3. (a) CD (line a, c ¼ 1.46 ꢂ 10ꢃ6 M; line c, c ¼ 1.46 ꢂ 10ꢃ7 M) and FDCD (J-465 device, line b,
c ¼ 1.46 ꢂ 10ꢃ8 M, light-cut filter, 340 nm, masks 4.0 mm, HT voltage, 793 V); (b) CD (line a,
c ¼ 2.33 ꢂ 10ꢃ6 M) and FDCD (J-465 device, line b, c ¼ 2.33 ꢂ 10ꢃ8 M, light-cut filter, 360 nm, masks
3.5 mm, HT voltage, 777 V), (c) CD (line a, c ¼ 2.56 ꢂ 10ꢃ6 M) and FDCD (J-465 device, light-cut
filter, 340 nm, masks 3.5 mm; line b, c ¼ 2.56 ꢂ 10ꢃ8 M, HT voltage, 825 V; line c, c ¼ 2.56 ꢂ 10ꢃ9 M,
HT voltage, 900 V); D" is the unit of molꢃ1 dm3 cmꢃ1
variety of C-8 substituted double bonds, namely, 1,2-disubstituted (19), endocyclic
double bond (20), and 1,1-disubstituted (21) double bonds (Scheme 3) [24, 26,
102–104].
Upon derivatization of these compounds to the corresponding styrenoids sty-
19-sty-21 by cross metathesis and Heck reaction [26], moderately intense and
clear-cut CD couplets arising from the coupling between the coumarin and the
styrene chromophores were observed in the 230–350 nm region. A straightforward
Scheme 3

Fluorescence Detected Exciton Coupled Circular Dichroism
383
CD exciton analysis leads to assignment of the absolute configurations of styrenoid
derivatives, and therefore the parent compounds (Table 3).
Since both the preexisting coumarin and the styrenoids are fluorescent
(Ff ¼ 0.16 for styrene and Ff ¼ 0.050 for coumarin), these substrates represent
interesting models for testing the applicability of FDCD for stereochemical anal-
ysis. As summarized in Table 3, sty-19 to sty-21 were analyzed by FDCD with a
100-fold sensitivity enhancement (detection limit 10ꢃ8 M) with respect to the
conventional CD transmission measurements, detection limit at 10ꢃ6 M. Although
the coumarin derivatives sty-19-sty-21 bear large polarization of fluorescence
ranging from 0.030 to 0.087 (Table 3), the J-465 instrument completely elimi-
nates the photoselection artifacts and collects much emitted light from the
sample solution. Figures 4a and b depict the CD and FDCD curves of sty-21
(Pf ¼ 0.030). As shown in Fig. 4a, the use of the J-465 device (ellipsoidal mirror,
line b) and the J-405 device (polarizer at 81ꢁ, line c) successfully eliminate the
photoselection artifacts at the longer wavelength and lead to excellent agreement
between the FDCD and CD spectra at 10ꢃ5 M. At 10ꢃ7 M (Fig. 4b), FDCD mea-
sured by J-465 (line b) remains observable, whereas S=N ratios for both the
conventional CD (line a) and FDCD obtained by J-405 (line c) are significantly
decreased at this concentration. The FDCD detection limit at 10ꢃ8 M evaluated
for sty-19-sty-21 is the same as that obtained for allylic alcohol derivatives
15a–18a; a similar trend for compounds featuring moderate CD amplitudes is
observed (Table 3).
2c. New Ginkgolide Lactol Derivatives. A Protocol
for Sterically Hindered Hydroxyls
Recently, we have found that the lactone groups in ginkgolide derivatives were
smoothly reduced by NaBH4 to give the corresponding lactols creating new stereo-
Fig. 4. CD and FDCD of sty-21; (a) CD (line a) and FDCD by J-465 device (line b, light-cut filter,
380 nm, masks 5.0 mm, HT voltage, 391 V), FDCD by J-405 device (line c, polarizer at 81ꢁ, HT
voltage, 603 V), FDCD by J-357 device (line d, HT voltage, 448 V); (b) CD (line a), FDCD by J-465
device (line b, HT voltage, 652 V), FDCD by J-405 device (line c, polarizer at 81ꢁ, HT voltage,
800 V); D" is the unit of molꢃ1 dm3 cmꢃ1

384
K. Tanaka et al.
Scheme 4
genic centers at the lactol hydroxyls. An example is shown in Scheme 4a. The
NaBH4 treatment of 22, which was degradated from ginkgolide C [105], quantita-
tively provided the inseparable mixture of C-13 and C-11 lactol derivatives 24 and
25, via the possible intermediate 23. The efficient NaBH4 reduction of ginkgolide
lactones will expand the ginkgolide-analog library for their biological evaluation
[105].
Since both the reduced compounds 24 and 25 bear an additional hydroxy group
at C-7 (see Scheme 4a), the stereochemical analysis of newly created stereogenic
centers at the lactol hydroxyls could be studied by the exciton coupled method after
introduction of suitable chromophores and=or fluorophores at the C-7 hydroxyl and
lactol hydroxyls. However, acylation attempts at the ginkgolide 7-hydroxyl with
carboxylic acid chromophores, e.g., benzoic acid, were not successful due to the
steric hindrance caused by the neighboring t-butyl group; the reaction required
excess reagents and proceeded in poor yields. For such sterically hindered hydrox-
yls, a new protocol was developed as shown in Scheme 4b. It utilizes the combined
methods of acylation and metathesis; (i) acylation of the sterically hindered hydrox-
yls with the small and reactive acryloyl chloride [106–110] followed by (ii) cross
metathesis with styrenes. According to the method, the ginkgolide derivative 22

Fluorescence Detected Exciton Coupled Circular Dichroism
385
was first reacted with acryloyl chloride in the presence of diisopropylethylamine in
THF to provide 28 in quantitative yield under mild conditions (Scheme 4c). The
introduced vinyl moiety in 28 was then subjected to cross metathesis with styrene,
and the cinnamate chromophore was efficiently introduced at the sterically hin-
dered ginkgolide C-7 position in 60% yield. Subsequently, the lactone groups at
C-11 and C-13 positions in 29 were smoothly reduced by NaBH4 to give the cor-
responding lactols 30 and 31 as their four diastereoisomers in quantitative yield.
Since these lactol derivatives 30 and 31 could not be separated and existed as its
equiblium mixtures, they were continuously acylated by fluorescent 4-phenylben-
zoic acid in the presence of EDC and DMAP to provide 32 as a main diastereo-
isomer, readily separable from the other isomers by chromatography on silica gel.
The 4-phenylbenzoate group was used not only as a ‘‘well-separating acylate
agent’’ for the hydroxyl-containing compounds [111], but also as one of the exciton
coupled fluorophores for FDCD study.
CD and FDCD of the main isomer 32 are shown in Fig. 5a. Compound 32
shows negative exciton coupling between the cinnamate and 4-phenylbenzoate
chromophores, with CD amplitude of ACD ¼ ꢃ51; this is in agreement with that
expected from conformational analysis (MMFF and Monte Carlo method) obtained
for the 13ꢀ-lactone hydroxyl isomer (Fig. 5b). On the other hand, a similar con-
formational analysis of the diastereomeric 13ꢁ-lactone hydroxyl derivative pre-
dicted positive CD exciton chirality. Therefore, the main isomer 32 obtained in
Scheme 4c was determined to posses the C-13ꢀ configuration. The conclusion
obtained by the CD=FDCD analysis was also supported by NOE between 8-H
and 13-H (see Scheme 4c).
Although 32 showed extremely high polarization of fluorescence (Pf ¼ 0.18
shown in Table 3), FDCD analysis was performed satisfactorily with J-465; as
with the cases of the previously mentioned compounds 15a–18a and sty-19-sty-
21, the modest CD amplitude of 32 (ACD ¼ ꢃ51) enabled the FDCD measurements
Fig. 5. (a) CD and FDCD of 32; CD (line a, c ¼ 2.67 ꢂ 10ꢃ6 M), FDCD (J-465 device, line b,
c ¼ 2.67 ꢂ 10ꢃ8 M, light-cut filter, 340 nm, masks 4.0 mm, HT voltage, 829 V); for the sake of clarity,
noisy CD at 10ꢃ7 M is not shown; (b) minimized conformation and exciton chirality of 32 and its
17ꢁ-isomer; D" is the unit of molꢃ1 dm3 cmꢃ1

386
K. Tanaka et al.
to be performed down to 10ꢃ8 M concentrations, whereas the CD measurement was
possible down to 10ꢃ6 M (Fig. 5a).
By using a variety of fluorescent 4-substituted cinnamates, e.g., 4-phenylcin-
namate or 4-tetraphenylporphyrine (see Section 2d), the application of exciton
coupled FDCD will be further extended to a wide range of the sterically congested
natural products available only in minuscule amount [16, 112–115], and therefore
unsuited for analysis by less sensitive conventional CD analysis.
2d. Steroid and Dimeric Steroid: Exciton Coupled
FDCD Over a Large Distance
In the case of 33 (Scheme 5) [116], in which the C-17 hydroxyl and C-3 olefin are far
apart, we planned to use TPP-CO2H 5a (ꢂ ¼ 418 nm, " ¼ 440000 molꢃ1 dm3 cmꢃ1)
and vinyl-TPP 3a (ꢂ ¼ 418 nm, " ¼ 420000 molꢃ1 dm3 cmꢃ1) for both CD and
FDCD studies. We have already shown that TPP-CO2H 5a, which was introduced
at the hydroxy or amino group via acylates, represents one of the most potent CD
reporter groups, especially for the observation of long range exciton coupling [4, 89].
Scheme 5

Fluorescence Detected Exciton Coupled Circular Dichroism
387
On the other hand, it is challenging to introduce the vinyl-TPP 3a at the C-3 olefin of
33; the cross metathesis of the nitrogen-containing porphyrin substrates is rarely
reported in the literature since the transition metal catalysts such as Grubbs’ ruthe-
nium catalyst are believed to be susceptible to nitrogen poisoning. Nevertheless, we
explored the vinyl-TPP cross metathesis approach keeping in mind that this protocol
might be useful for other natural products containing remote OH=C¼C or
C¼C=C¼C functionalities, as seen in polyether type marine natural products
[90, 91, 117, 118].
We first attached TPP-CO2H 5a at the C-17 hydroxyl of steroid 33 via an ester
linkage (Scheme 5). Substrate 33 was then reacted with TPP-CO2H 5a in the
presence of EDC and DMAP to provide 34 in 75% yield. The cross metathesis
of the double bond in 34 with vinyl-TPP 3a in the presence of Grubbs’ reagent 12
smoothly provided the corresponding bis-TPP derivative 35a in 65% yield. Since
the porphyrin derivatives are easily detectable on the thin layer chromatography
(TLC) by their intense deep red color, the reaction monitoring, isolation, and other
chemical procedures can be readily performed through microgram-scale handling
of the sample. Subsequently, magnesium cations were quantitatively incorporated
inside both porphyrins by treatment of 35a with excess magnesium iodide (10
equivalents) and triethylamine to yield 35b. In order to compare the CD sensitivity
of TPP-chromophores in 35b with that of common aryl chromophores, we also
prepared 37 in which styrene (1) (ꢂ ¼ 248 nm, " ¼ 15000 molꢃ1 dm3 cmꢃ1) and
4-phenylbenzoate 4 (ꢂ ¼ 270 nm, " ¼ 21000 molꢃ1 dm3 cmꢃ1) are introduced at
the C-3 and C-17 positions as shown in Scheme 5.
The CD of 37 (Fig. 6a, dotted line) shows only a positive Cotton effect at
269 nm (D" ¼ þ11.2 molꢃ1 dm3 cmꢃ1), presumably due to the overlap between
the exciton couplet and the individual CD Cotton effects resulting from C-17 4-
phenylbenzoate and C-3 styrenoid chromophores; since the two stereogenic centers
in 37 are separated by a large distance, the exciton coupling between these two
chromophores is not sufficiently strong to override the inherent CD Cotton effects
of each chromophore at the C-3 and C-17, thus failing to give clearcut bisignate
couplets. However, when CD of 36 was subtracted from that of 37 (subtraction of
the inherent CD Cotton effect caused by C-17 4-phenylbenzoate), a small negative
counterpart at 247 nm (D" ¼ ꢃ2.4 molꢃ1 dm3 cmꢃ1) becomes visible in the result-
ing difference CD (Fig. 6a).
On the other hand, clear bisignate positive CD couplets with an amplitude of
þ127 are observed for the bis-porphyrin derivative 35a as shown in Fig. 6b; the
positive exciton chirality of 35a reflects the clockwise arrangement of porphyrin
resulting from the two remote stereogenic centers at C-3 and C-17 separated by
˚
26 A (see minimized conformation in Fig. 6c).
A conformational analysis with MMFF and Monte Carlo method revealed that
35a exists as a mixture of several conformers close in energy; 12 minimum energy
conformations were isolated within 8.4 kJ=mol. They are due to the variation of at
least three degrees of conformational freedom (Fig. 6c): (i) the C3–C¼C rotamer-
ism, with three conformers (C¼C bond eclipsed to C3–O, C3–C2, and C3–C4
bonds); (ii) the aryl-C¼C rotation, with two minima (aryl-C¼C dihedral angle
about ꢅ30ꢁ); (iii) the C17–O rotamerism (C(¼O)–O=C17–H17 dihedral angle
about ꢅ30ꢁ). Although the unrestricted libration of the phenyl-porphyrin bond,

388
K. Tanaka et al.
Fig. 6. (a) CD of 36 and 37 and their difference CD; (b) CD (line a, c ¼ 3.08 ꢂ 10ꢃ7 M) and FDCD of
35a (J-465 device, line b, c ¼ 3.08 ꢂ 10ꢃ9 M, light-cut filter, 520 nm, masks 3.0 mm, HT voltage,
778 V); (c) minimized conformation and exciton chirality of 35a; (d) CD and FDCD of 35b, CD
(line a, c ¼ 2.91 ꢂ 10ꢃ7 M), FDCD (J-465 device, line b, c ¼ 2.91 ꢂ 10ꢃ10 M, light-cut filter, 520 nm,
masks 3.0 mm, HT voltage, 896 V); " and D" are the unit of molꢃ1 dm3 cmꢃ1
with dihedrals in the range 45 to 135ꢁ, should also be considered, this dihedral
angle is found to be 90ꢁ in all minimum energy conformations. Despite this wide
flexibility, it is notable that in all minima the twist defined by 50–150 and 500–1500
porphyrin directions is always positive, with dihedral angles ranging from 32 to
70ꢁ. The porphyrin Soret band may be described as due to two degenerate transi-
tions with mutually orthogonal transition dipoles [30, 89]. However, due to the
libration (largely unrestricted rotation) around the phenyl-porphyrin junction in
tetraarylporphyrins a so-called effective transition moment can be considered to
exist directed along the 5–15 direction, which is in effect responsible for the sign
of exciton coupling in bis-tetraarylporphyrin systems [30, 89]. In conclusion, the
observed positive couplet for 35a is in keeping with a positive chirality defined by
the two 50–150 and 500–1500 effective moments.
Fortunately, the FDCD spectrum was also measurable due to the fluorescent
property of these TPP-derivatives (Fig. 6b). As noted previously, since porphyrin

Fluorescence Detected Exciton Coupled Circular Dichroism
389
fluorophores have large extinction coefficients (" ca. 450000 molꢃ1 dm3 cmꢃ1) and
the CD amplitude of 35a exceeded 100, the FDCD measurement was possible
even at the highly diluted concentration of 10ꢃ9 M despite the long interchromo-
phoric distance. The FDCD detection again achieved a 100-fold enhancement in
sensitivity over conventional CD detection as shown in Fig. 6b. Furthermore,
introduction of magnesium into porphyrin skeletons (Fig. 6d), highly enhances
the FDCD detection, leading to a limit of 10ꢃ10 M, due to the increase in both the
quantum yield and extinction coefficient (Ff ꢄ 0.2, " ¼ 560000 molꢃ1 dm3 cmꢃ1).
On the other hand, the concentration limit for CD remains around 10ꢃ7 M indicat-
ing a 1000-fold enhancement in FDCD sensitivity, the largest increase so far
observed.
Encouraged by the highly sensitive FDCD analysis of 35a and 35b with a long
˚
range coupling distance over 26 A, the dimeric steroid 40 with the interporphyrin
˚
distance increased to 40 A was measured (Scheme 6). However, the simultaneous
bis-acylation of 38 [119] at the two 17-hydroxyls by TPP-CO2H 5a did not work;
the product obtained under several different acylation conditions, e.g., with EDC
and DMAP in dichloromethane, only resulted in a mono-acylated derivative. We
therefore utilized the previously established method for sterically hindered hydrox-
yls employed in Section 2c; bis-steroid 38 was first reacted with the reactive
acryloyl chloride, 77% yield, and the vinyl group introduced in 39 was then func-
tionalized by the ‘‘double cross metathesis’’ with vinyl-TPP 3a (Scheme 6). By this
method, the bis-porphyrin derivative 40 was successfully obtained in ꢄ50% yield
from 38.
The CD spectrum of 40 (Fig. 7a) shows clear negative exciton couplets with an
amplitude of ꢃ37, reflecting the two remote stereogenic centers at 17 and 170
˚
positions over a distance of 40 A (see minimized conformation in Fig. 7b). Ten
minimum energy conformations within 9.6 kJ=mol were found for 40 by a
MMFF=Monte Carlo conformational analysis. The main degrees of conformational
freedom are (Fig. 7b): (i) the C17–O and C170–O0 rotamerisms, CO–O–C17–H17
dihedral angle about ꢅ30ꢁ); (ii) the s-cis=s-trans C¼C–C¼O bond isomerism.
Scheme 6

390
K. Tanaka et al.
Fig. 7. (a) CD and FDCD of 40; CD (line a, c ¼ 1.05 ꢂ 10ꢃ6 M), FDCD (J-465 device, line b,
c ¼ 1.05 ꢂ 10ꢃ8 M, light-cut filter, 520 nm, masks 4.0 mm, HT voltage, 586 V), CD at 10ꢃ7 M with
low S=N ratio is not shown; (b) minimized conformation and exciton chirality of 40; " and D" are the
unit of molꢃ1 dm3 cmꢃ1
Our attention was then restricted to eight minima with Boltzmann population at
room temperature above 4%, by considering also entropic factors resulting from
the C2 symmetry system. All show negative twist between 50–150 and 500–1500
directions, with dihedral angles in the ꢃ32 to ꢃ67ꢁ range, in keeping with the
observed negative CD couplet. Quantitative CD calculations were also performed
on this compound by the DeVoe method to investigate the nature of the exciton
˚
coupling for such large interchromophoric distances of 39.5–42.5 A. An hybrid
approach was used to describe the Soret transition [30, 89], where the 5–15 com-
ponent was given full dipolar strength (77.4 D2, extrapolated from the experimental
absorption spectrum of 5-(4-(2-methoxycarbonylethenyl)phenyl)-10,15,20-triphe-
nylporphyrin in CH2Cl2), and 10–20 component only one half (38.7 D2). We have
previously verified that this approach efficiently describes the exciton interaction in
bis-tetraarylporphyrins, by taking into account the libration around the phenyl-
porphyrin bonds [30]. All minima gave a negative calculated CD couplet with
intensities between ꢃ14 and ꢃ140; the Boltzmann-weighted average was ꢃ85,
in good agreement with experimental results. Despite the large interchromophoric
distance, a moderate exciton coupling is therefore observed for compound 40
owing to the intense extinction coefficient of the Soret bands.
Furthermore, as shown in Fig. 7a, FDCD measurement enabled the exciton
analysis at even 10ꢃ8 M for this extremely long range exciton coupling system.
The 100-fold enhancement in sensitivity observed with 40 is in accord with the
trend obtained for the exciton-systems with the modest CD amplitudes observed in
Sections 2a–2c. The successful microscale analysis performed for both steroid 33
and dimeric steroid 38 provides an intriguing protocol for the stereochemical anal-
ysis of many large natural products carrying a variety of functional groups,
e.g., sterically hindered hydroxyls, amines, or substituted olefins, which frequently
are present in marine toxins [90, 91, 117, 118].

Fluorescence Detected Exciton Coupled Circular Dichroism
391
Conclusions
Fluorescence detection of circular dichroism represents an attractive alternative to
conventional CD detection whenever specificity toward fluorescent molecules or
fluorophoric groups is required, or enhanced sensitivity is desirable because of the
restricted amount of compounds, e.g., natural products. Despite these advantages,
the application of FDCD has remained for many years restricted due to instrumen-
tal inadequacy: namely, residual linear dichroism in circularly polarized light,
arising from imperfect optical components, as well as inherent fluorescence polar-
ization typical of many fluorophores, to produce unacceptable artifacts in the
FDCD spectra. The recently developed FDCD apparatus (J-465), based on an
ellipsoidal mirror device that collects most of the light emitted around the sample,
has opened new perspectives to the FDCD technique.
In this paper we described the application of several new FDCD fluorophoric
reporter groups, such as styrene, 4-vinylbiphenyl, 4-phenylbenzoate, cinnamate, and
porphyrin-based groups which can be conveniently introduced into either olefin
moieties by cross-methatesis or at hydroxyls (including sterically hindered ones)
by acylation. The paper also provides the examples for efficiency of the ellipsoidal
mirror FDCD device, which eliminates photoselection artifacts, augments the inten-
sity of the total fluorescence signal, and overcomes the inconveniencies due to low
quantum yields and strong fluorescence polarization. The experimental results high-
light the advantage of fluorescence detection in providing FDCD spectra not only in
excellent agreement with CD but also at much lower concentrations than conven-
tional CD detection by two-three orders of magnitude reaching 10ꢃ8 M in most
cases, and down to 10ꢃ10 M for favorable examples. We believe the structural diver-
sity of the compounds analyzed in this study demonstrates the versatility of the
exciton chirality approach, and in particular, the fluorescence based approach as a
practical and sensitive technique with a broader applicability.
Experimental
FDCD Measurement
Three different FDCD devices, J-357, J-405, and J-465 attached to Jasco J-810 spectropolarimeter have
been used. The scattered light was prevented by employing a long-pass filter in front of the PMT
(wavelength used for each measurement is shown in Figure captions). The FDCD signals recorded
were converted into the CD spectra based on the equation: D" ¼ "L–"R ¼ 3.032 ꢂ 10ꢃ5 ꢂ S ꢂ (1–
10ꢃA)=c ꢂ d ꢂ 10ꢃA in which A is UV absorbance, c is molar concentration, d is cell length (cm), and
S ¼ kðFL ꢃ FRÞ=ðFL þ FRÞ. ðFL þ FRÞ was measured as DC voltage and k is an instrumental ellip-
ticity constant þ28648 (millideg).
Computational Methods
Molecular modeling was performed with Spartan ’02 (Wavefunction, Inc., Irvine, CA), using default
parameters and convergence criteria. Molecular conformations were calculated with Merck force field
(MMFF) using the Monte Carlo conformational search method with 1000 iteration steps. Electronic
structure calculations were performed with Gaussian 03 (Gaussian, Inc., Pittsburgh, PA) [120] with
ZINDO=S method on input structures minimized with DFT, B3LYP=6-31G(d) level. Coupled oscil-
lator calculations were performed with the DeVoe method using a Fortran program developed by

392
K. Tanaka et al.
W. Hug [27]. MMFF-optimized structures were used as input geometries; average CD spectra were
calculated as Boltzmann-weighted at 300 K. Spectral parameters were taken from experimental UV-Vis
spectra of the constituent chromophores and ZINDO calculations.
General Procedure of Acryloylation of Sterically Hindered Hydroxyls
Acryloylated Monosteroid (11, C22H32O3)
To a solution of 50 mg 17ꢁ-hydroxy-5ꢀ-androstan-3-one (9) (0.172 mmol) and 60.0mm3 diisopropyl-
ethylamine (0.344mmol) in 1.0cm3 CH2Cl2 was slowly added 28.0mm3 acryloyl chloride 10
(0.344 mmol) at 0ꢁC. After the mixture was warmed to room temperature and stirred for an additional
h, H2O and saturated aqueous NH4Cl solution were added, and the resulting mixture was extracted
with ethyl acetate. The organic layers were combined, washed with brine, dried (Na2SO4), filtered, and
concentrated in vacuo to give the crude products. Column chromatography on silica gel (gradually
from 9 to 25% ethyl acetate in n-hexane) gave 60mg 11 (100%) as a white solid. 1H NMR (400 MHz,
CDCl3): ꢄ ¼ 0.73–1.80 (m, 16H), 0.84 (s, 3H), 1.02 (s, 3H), 1.99–2.04 (m, 1H), 2.07–2.12 (m, 1H),
2.16–2.43 (m, 4H), 4.68 (dd, J ¼ 8.8, 8.8 Hz, 1H), 5.81 (dd, J ¼ 10.4, 1.3 Hz, 1H), 6.12 (dd, J ¼ 17.3,
10.4Hz, 1H), 6.38 (dd, J ¼ 17.4, 1.3 Hz, 1H) ppm; HRFABMS: m=z ¼ calcd for C22H33O3 [Mþ H]þ
345.2429, found 345.2435.
General Procedure of Cross Metathesis
Porphyrincinnamate (6, C66H60O3N4)
To a solution of 3.0 mg 11 (8.71 ꢃmol) in 2.0 cm3 CH2Cl2 was added 1.7 mg Grubbs’ second genera-
tion ruthenium catalyst 12 (1.99 ꢃmol) and 33.5mg 3a (52.3 ꢃmol) at room temperature, and the
mixture was stirred at 80ꢁC for 3 h. The reaction mixture was concentrated in vacuo to give the crude
products which were purified by column chromatography on silica gel (ethyl acetate in n-hexane,
gradually from 9 to 20%) to afford 6.0 mg 6 (72%) as a deep red colored solid. The reaction could be
also performed using 100 ꢃg of 11. 1H NMR (300 MHz, CDCl3): ꢄ ¼ 0.97 (s, 3H), 1.05 (s, 3H), 0.82–
2.45 (m, 22H), 4.85 (dd, J ¼ 7.9, 7.9 Hz, 1H), 6.75 (d, J ¼ 16.0Hz, 1H), 7.72–7.79 (m, 9H), 7.93 (d,
J ¼ 8.1 Hz, 2H), 8.00 (d, J ¼ 16.1 Hz, 1H), 8.20–8.26 (m, 8H), 8.83–8.87 (m, 8H) ppm; HRFABMS:
m=z ¼ calcd for C66H61O3N4 [Mþ H]þ 957.4743, found 957.4766.
Acknowledgements
We are grateful to the Jasco Corporation for generously providing us with a new FDCD J-465
instrument and to Dr. T. Takakuwa, Mrs. C. Ohshima, Dr. H. Masago, Dr. A. Wada, all of Jasco
Corporation, and to Dr. T. Nehira of Hiroshima University for helpful discussions. We appreciate also
the help of Mrs. K. D. Kester in the synthesis described in Scheme 2b and 4c. This research is
supported by the National Institutes of Health, NIH Grant GM 34509 and MH 068817 (K. N. and
N. B.). K. T. is grateful to the JSPS Postdoctoral Fellowships for Research Abroad.
References
[1] Berova N, Nakanishi K, Woody RW (2000) Circular Dichroism: Principles and Applications.
Wiley-VCH, New York
ꢁ
[2] Rodger A, Norden B (1997) Circular Dichroism & Linear Dichroism. Oxford University Press,
Oxford
[3] Mason SF (1982) Molecular Optical Activity and the Chiral Discrimination. Cambridge
University Press, Cambridge
[4] Berova N, Nakanishi K (2000) Principles and Applications of Exciton Chirality Method. In:
Berova N, Nakanishi K, Woody RW (eds) Circular Dichroism: Principles and Applications.
Wiley-VCH, New York, p 337

Fluorescence Detected Exciton Coupled Circular Dichroism
393
[5] Harada N, Nakanishi K (1983) Circular Dichroic Spectroscopy – Exciton Coupling in Organic
Stereochemistry. University Science Books, Mill Valley, CA
[6] Huang X, Rickman BH, Borhan B, Berova N, Nakanishi K (1998) J Am Chem Soc 120: 6185
[7] Huang X, Borhan B, Rickman BH, Nakanishi K, Berova N (2000) Chem Eur J 6: 216
[8] Kurtan T, Nesnas N, Koehn FE, Li Y-Q, Nakanishi K, Berova N (2001) JAm Chem Soc 123: 5974
[9] Kurtan T, Nesnas N, Li Y-Q, Huang X, Nakanishi K, Berova N (2001) J Am Chem Soc 123: 5962
[10] Proni G, Pescitelli G, Huang X, Quraishi NQ, Nakanishi K, Berova N (2002) Chem Commun
1590
[11] Proni G, Pescitelli G, Huang X, Nakanishi K, Berova N (2003) J Am Chem Soc 125: 12914
[12] Tanaka K, Nakanishi K, Berova N (2003) J Am Chem Soc 125: 10802
[13] Jiang H, Huang X, Nakanishi K, Berova N (1999) Tetrahedron Lett 40: 7645
[14] Kawamura A, Berova N, Nakanishi K, Voigt B, Adam G (1997) Tetrahedron 53: 11961
[15] Zhou P, Berova N, Nakanishi K, Knani Mh, Rohmer M (1991) J Am Chem Soc 113: 4040
[16] Wiesler WT, Berova N, Ojika M, Meyers HV, Chang M, Zhou P, Lo LC, Niwa M, Takeda R,
Nakanishi K (1990) Helv Chim Acta 73: 509
[17] Dong J-G, Akritopoulou-Zanze I, Guo J, Berova N, Nakanishi K, Harada N (1997) Enantiomer
2: 397
[18] DeVoe H (1964) J Chem Phys 41: 393
[19] DeVoe H (1965) J Chem Phys 43: 3199
[20] Rosini C, Zandomeneghi M, Salvadori P (1993) Tetrahedron Asymm 4: 545
[21] Di Bari L, Pescitelli G, Salvadori P (1999) J Am Chem Soc 121: 7998
[22] Di Bari L, Pescitelli G, Reginato G, Salvadori P (2001) Chirality 13: 548
[23] Di Bari L, Lelli M, Pintacuda G, Pescitelli G, Marchetti F, Salvadori P (2003) J Am Chem Soc
125: 5549
[24] Pescitelli G, Berova N, Xiao TL, Rozhkov RV, Larock RC, Armstrong DW (2003) Org Biomol
Chem 1: 186
[25] Superchi S, Giorgio E, Rosini C (2004) Chirality 16: 422
[26] Tanaka K, Pescitelli G, Di Bari L, Xiao TL, Nakanishi K, Armstrong DW, Berova N (2004) Org
Biomol Chem 2: 48
[27] Cech CL, Hug W, Tinoco I Jr (1976) Biopolymers 15: 131
[28] Bode KA, Applequist J (1996) J Phys Chem 100: 17825
[29] Bode KA, Applequist J (1997) Macromolecules 30: 2144
[30] Pescitelli G, Gabriel S, Wang Y, Fleischhauer J, Woody RW, Berova N (2003) J Am Chem Soc
125: 7613
[31] Turner DH, Tinoco I Jr, Maestre M (1974) J Am Chem Soc 96: 4340
[32] Tinoco I Jr, Turner DH (1976) J Am Chem Soc 98: 6453
[33] Turner DH (1978) Methods Enzymol 49: 199
[34] Tinoco I Jr (1979) NATO ASI Series, Series C: Mathematical and Physical Sciences 48: 57
[35] Thomas MP, Patonay G, Warner IM (1991) Pract Spectr 12: 421
[36] Brittain HG (2000) App Spectr Rev 35: 175
[37] Tran CD, Fendler JH (1979) J Am Chem Soc 101: 1285
[38] White TG, Pao Y-H, Tang MM (1975) J Am Chem Soc 97: 4751
[39] Ehrenberg B, Steinberg IZ (1976) J Am Chem Soc 98: 1293
[40] Tinoco I Jr, Ehrenberg B, Steinberg IZ (1977) J Chem Phys 66: 916
[41] Lobenstine EW, Turner DH (1979) J Am Chem Soc 101: 2205
[42] Lobenstine EW, Turner DH (1980) J Am Chem Soc 102: 7786
[43] Lobenstine EW, Schaefer WC, Turner DH (1981) J Am Chem Soc 103: 4936
[44] Steinberg IZ, Ehrenberg B (1974) J Chem Phys 61: 3382
[45] Hug W (1981) App Spectrosc 35: 115
[46] Shindo Y, Nakagawa M (1985) App Spectrosc 39: 32
[47] Bicknese SE, Maestre MF (1987) Rev Sci Instr 58: 2060

394
K. Tanaka et al.
[48] Nehira T, Tanaka K, Takakuwa T, Ohshima C, Masago H, Pescitelli G, Wada A, Berova N
(2005) App Spectrosc (in press)
[49] Wu K, McGown LB (1991) App Spectrosc 45: 1
[50] Geng L, McGown LB (1992) Anal Chem 64: 68
[51] Wu K, Geng L, Joseph MJ, McGown LB (1993) Anal Chem 65: 2339
[52] Thomas MP, Patonay G, Warner IM (1987) Anal Lett 20: 717
[53] Thomas MP, Nelson G, Patonay G, Warner IM (1988) Spectrochim Acta B: Atom Spectr 43B:
651
[54] Dahl KS, Pardi A, Tinoco I Jr (1982) Biochemistry 21: 2730
[55] Lamos ML, Turner DH (1985) Biochemistry 24: 2819
[56] Lamos ML, Lobenstine EW, Turner DH (1986) J Am Chem Soc 108: 4278
[57] Lamos ML, Walker GT, Krugh TR, Turner DH (1986) Biochemistry 25: 687
[58] Thomas MP, Patonay G, Warner IM (1987) Anal Biochem 164: 466
[59] Meadows F, Narayanan N, Patonay G (2000) Talanta 50: 1149
[60] Turner DH, Tinoco I Jr, Maestre MF (1975) Biochem 14: 2794
[61] Wasylewski M, Malecki J, Wasylewski Z (1995) J Protein Chem 14: 299
[62] Park S, Ajtai K, Burghardt TP (1996) Biophys Chem 63: 67
[63] Muto K, Yoshida R, Yashida R, Ishii T, Handa T (1986) J Am Chem Soc 108: 6416
[64] Watanabe K, Muto K, Ishii T (1997) Biospectroscopy 3: 103
[65] Reich C, Tinoco I Jr (1980) Biopolymers 19: 833
[66] Egusa S, Takagi J, Sisido M, Imanishi Y (1986) Bull Chem Soc Jap 59: 2195
[67] Synovec RE, Yeung ES (1986) J Chromatogr 368: 85
[68] Christensen PL, Yeung ES (1989) Anal Chem 61: 1344
[69] Geng L, McGown LB (1994) Anal Chem 66: 3243
[70] Sugimoto T, Ikemoto K, Murata S, Tazawa M, Nomura T, Hagino Y, Ichinose H, Nagatsu T
(2001) Helvetica Chimica Acta 84: 918
[71] Sugimoto T, Ikemoto K, Murata S, Tazawa M, Nomura T, Hagino Y, Ichinose H, Nagatsu T,
Wada A (2001) Heterocycles 54: 283
[72] Chen N, Ikemoto K, Sugimoto T, Murata S, Ichinose H, Nagatsu T (2002) Heterocycles 56: 387
[73] Dai Z, Proni G, Mancheno D, Karimi S, Berova N, Canary JW (2004) J Am Chem Soc 126:
11760
[74] Dong J-G, Wada A, Takakuwa T, Nakanishi K, Berova N (1997) J Am Chem Soc 119: 12024
[75] Nehira T, Parish CA, Jockusch S, Turro NJ, Nakanishi K, Berova N (1999) JAm Chem Soc 121:
8681
€
[76] Forster T (1965) Mod Quantum Chem 93
[77] Scholes GD, Ghiggino KP (1994) J Chem Phys 101: 1251
[78] Harcourt RD, Scholes GD, Ghiggino KP (1994) J Chem Phys 101: 10521
[79] Chatterjee AK, Choi T-L, Sanders DP, Grubbs RH (2003) J Am Chem Soc 125: 11360
[80] Connon SJ, Blechert S (2003) Angew Chem Int Ed 42: 1900
[81] Chatterjee AK, Toste FD, Choi T-L, Grubbs RH (2002) Adv Synth Catal 344: 634
[82] Scholl M, Ding S, Lee CW, Grubbs RH (1999) Org Lett 1: 953
[83] Scholl M, Trnka TM, Morgan JP, Grubbs RH (1999) Tetrahedron Lett 40: 2247
[84] Chen S-ML, Harada N, Nakanishi K (1974) J Am Chem Soc 96: 7352
[85] Canceill J, Collet A, Jacques J (1982) J Chem Soc Perkin 2, 83
[86] Gradyushko AT, Tsvirko MP (1971) Opt Spectrosc (USSR) 31: 291
[87] Sumi K, Kimura M, Nakamura I (1994) J Polym Sci Part A: Polym Chem 32: 1243
[88] Aramata K, Kajiwara A, Kamachi M (1995) Macromolecules 28: 4774
[89] Matile S, Berova N, Nakanishi K, Fleischhauer J, Woody RW (1996) JAm Chem Soc 118: 5198
[90] Faulkner DJ (1999) Nat Prod Rep 16: 155
[91] Faulkner DJ (2002) Nat Prod Rep 19: 1

Fluorescence Detected Exciton Coupled Circular Dichroism
395
[92] Zhang W-H, Williams ID, Che C-T (2001) Tetrahedron Lett 42: 4681
[93] Torres YR, Berlinck RGS, Magalhaes A, Schefer AB, Ferreira AG, Hajdu E, Muricy G (2000)
J Nat Prod 63: 1098
[94] Iijima T, Yaoita Y, Kikuchi M (2003) Chem Pharm Bull 51: 894
[95] Iwagawa T, Nakamura K, Hirose T, Okamuea H, Nakatani M (2000) J Nat Prod 63: 468
[96] Chai M-C, Wang S-K, Dai C-F, Duh C-Y (2000) J Nat Prod 63: 846
[97] Guilet D, Helesbeux J-J, Seraphin D, Sevenet T, Richomme P, Bruneton J (2001) J Nat Prod 64:
563
[98] Thastrup O, Fjalland B, Lemmich J (1983) Acta Pharmacol Tox 52: 246
[99] Tovar-Miranda R, Cortes-Garcia R, Santos-Sanchez NF, Joseph-Nathan P (1998) J Nat Prod 61:
1216
[100] Bell RG, Sadowski JA, Matschiner JT (1972) Biochemistry 11: 1959
[101] Rodigliero G (1992) J Photochem Photobiol B Biol 14: 1
[102] Rozhkov RV, Larock RC (2002) Org Lett 5: 797
[103] Rozhkov RV, Larock RC (2003) J Org Chem 68: 6314
[104] Xiao TL, Rozhkov RV, Larock RC, Armstrong DW (2003) Anal Bioanal Chem 377: 639
[105] Stromgaard K, Nakanishi K (2004) Angew Chem Int Ed 43: 1640
[106] Olsson T, Stern K, Westman G, Sundell S (1990) Tetrahedron 46: 2473
[107] Mizutani H, Watanabe M, Honda T (2002) Tetrahedron 58: 8929
[108] Langer P, Albrecht U (2002) Synlett 1841
[109] Honda T, Namiki H, Kaneda K, Mizutani H (2004) Org Lett 6: 87
[110] Wang C, Russell GA (1999) J Org Chem 64: 2066
[111] Corey EJ, Albonico SM, Koelliker U, Schaaf TK, Varma RK (1971) J Am Chem Soc 93: 1491
[112] Takeda R, Park MH, Nakanishi K (1989) Isr J Chem 29: 239
[113] Takeda R, Zask A, Nakanishi K, Park MH (1987) J Am Chem Soc 109: 914
[114] Nakanishi K, Kuroyanagi M, Nambu H, Oltz EM, Takeda R, Verdine GL, Zask A (1984) Pure
Appl Chem 56: 1031
[115] Chang M, Meyers HV, Nakanishi K, Ojika M, Park JH, Park MH, Takeda R, Vazquez J, Wiesler
WT (1989) Pure Appl Chem 61: 1193
[116] Ngatcha BT, Luu-The V, Poirier D (2000) Bioorg Med Chem Lett 10: 2533
[117] Yasumoto T, Murata M (1993) Chem Rev 93: 1897
[118] Satake M, Murata M, Yasumoto T (1993) J Am Chem Soc 115: 361
[119] Cerny I, Pouzar V, Budesinsky M, Drasar P (2000) Collect Czech Chem Commun 65: 1597
[120] Frisch MJ et al. (2003) Gaussian 03; Gaussian, Inc., Pittsburgh, PA, USA, http:==www.
gaussian.com