Synchronous enantiomeric enrichment of both reactant and product by absolute asymmetric synthesis using circularly polarized light. Part 3. Numerical simulation and experimental verification of the reversible asymmetric photoisomerization between methyl norbornadiene-2-carboxylate and methyl quadricyclane-1-carboxylate

Article abstract of DOI:10.1039/b100753j

In the first paper of this series, we proposed a new absolute asymmetric synthesis (NAAS) with circularly polarized light (CPL), which can be classified into two subcategories: (a) reversible NAAS (CPL excites both the reactant and product), (b) irreversible NAAS (the reactant only is excited by CPL). In the present paper we deal with reversible NAAS. The numerical simulations, which are based upon the results of the preceding paper of this series, are performed. Using the reversible photoisomerization between methyl norbornadiene-2-carboxylate (I) and methyl quadricyclane-1-carboxylate (II), it was experimentally verified that the enantiomeric excesses (ee's) of both the reactant and product were critically dependant not only on the anisotropy factors (g = Δε/ε) of both components, but also on the relative quantum yields of the forward and reverse reactions. In the case of reversible NAAS, the ee of the reactant and the product will concurrently increase according to the conversion, and attain appreciable values at the photostationary state if we appropriately select the system and the CPL-irradiation wavelength in such a way that if the g factors of the product and the reactant are large enough, the sign of the g factor of the product is opposite to that of the reactant, and the photochemical equilibrium constant K is less than unity.

Full text of DOI:10.1039/b100753j

View Article Online / Journal Homepage / Table of Contents for this issue
Synchronous enantiomeric enrichment of both reactant and
product by absolute asymmetric synthesis using circularly polarized
1
light. Part 3. Numerical simulation and experimental verification
of the reversible asymmetric photoisomerization between methyl
norbornadiene-2-carboxylate and methyl quadricyclane-1-
carboxylate
2
a
a
b
b
a
Hideo Nishino, Asao Nakamura, Hiroshi Shitomi, Hideo Onuki and Yoshihisa Inoue*
a
Inoue Photochirogenesis Project, ERATO, JST, 4-6-3 Kamishinden, Toyonaka, 560-0085,
Japan
Quantum Radiation Division, Electrotechnical Laboratory, 1-1-4 Umezono, Tsukuba,
b
3
05-8568, Japan
Received (in Cambridge, UK) 22nd January 2001, Accepted 21st June 2001
First published as an Advance Article on the web 14th August 2001
In the first paper of this series, we proposed a new absolute asymmetric synthesis (NAAS) with circularly polarized
light (CPL), which can be classified into two subcategories: (a) reversible NAAS (CPL excites both the reactant and
product), (b) irreversible NAAS (the reactant only is excited by CPL). In the present paper we deal with reversible
NAAS. The numerical simulations, which are based upon the results of the preceding paper of this series, are
performed. Using the reversible photoisomerization between methyl norbornadiene-2-carboxylate (I) and methyl
quadricyclane-1-carboxylate (II), it was experimentally verified that the enantiomeric excesses (ee’s) of both the
reactant and product were critically dependant not only on the anisotropy factors (g = ∆ε/ε) of both components, but
also on the relative quantum yields of the forward and reverse reactions. In the case of reversible NAAS, the ee of the
reactant and the product will concurrently increase according to the conversion, and attain appreciable values at the
photostationary state if we appropriately select the system and the CPL-irradiation wavelength in such a way that if
the g factors of the product and the reactant are large enough, the sign of the g factor of the product is opposite to
that of the reactant, and the photochemical equilibrium constant K is less than unity.
Absolute asymmetric synthesis (AAS) using circularly polar-
ized light (CPL) as a physical chiral source, enabling enantio-
meric enrichment without using any chemical chiral source, is
an intriguing and unique methodology among asymmetric syn-
thetic procedures, and has recently been discussed in relation to
as a function of the conversion upon CPL irradiation. These
formulae were experimentally verified by the one-way photo-
isomerization of methyl norbornadiene-2-carboxylate (I) to
methyl quadricyclane-1-carboxylate (II) (Scheme 1) induced by
CPL irradiation at 290 nm which is absorbed only by I.
2–5
the origin of the homochirality in the biosphere. In AAS,
asymmetric induction is brought about through the preferential
excitation of one of the enantiomers of a racemic mixture
upon CPL irradiation. The variety of AAS reactions hitherto
reported have been classified into three categories according
to the manner of enantiodifferentiation, i.e. (a) asymmetric
photodestruction, (b) photochemical deracemization, and (c)
6–8
photochemical asymmetric fixation.
In the first paper of this series, we proposed a new category
of absolute asymmetric synthesis (NAAS) shown in Scheme
1
4
of the preceding paper, where both reactant and product
Scheme 1 Photoisomerization between I and II.
are chiral and enantiomeric enrichment occurs in both com-
ponents. NAAS is classified into two subcategories according
to the nature of the photochemical reversibility between the
reactant and the product: in class (a), there is a photochemical
equilibrium between the reactant and the product because the
reaction mixture is irradiated with CPL which excites both the
reactant and the product; in class (b), the reaction is irreversible
because only the reactant is excited by CPL but the product is
not excited because the product does not absorb incident light
at the excitation wavelength. For the irreversible NAAS system
In the present study, we deal with the more complicated
reversible NAAS system, i.e. class (a), where both the reactant
and the product absorb CPL and the photoequilibrium is estab-
lished upon prolonged irradiation. We have simulated numeric-
ally the development of the ee’s of both reactant and product
upon CPL irradiation in the reversible NAAS system. We have
examined the validity of the results of these simulations by
performing the reversible photoisomerization between I and II
upon excitation by monochromatic CPL generated by an
Onuki-type polarizing undulator installed in a synchrotron
radiation source. In the first paper of this series we used a xenon
arc lamp as the light source for CPL generation. However, this
(
class (b)), in the first paper of this series, we derived the
theoretical equations for formulating the developing patterns of
the enantiomeric excesses (ee’s) of both reactants and product
1
706
J. Chem. Soc., Perkin Trans. 2, 2001, 1706–1713
This journal is © The Royal Society of Chemistry 2001
DOI: 10.1039/b100753j

View Article Online
source could not be used at wavelengths less than 280 nm
because of low intensity. With the aid of the synchrotron
undulator we were able to study the CPL irradiation in the far
UV region. The results of the CPL irradiation of the I–II
system verifies the theoretical conclusions that the ee’s of both
the reactant and the product and the conversion critically
depend upon both the anisotropy (g) factors of the reactant
and the product but also on the relative rates of the forward and
reverse reactions. We will further demonstrate that the method
of simulation that is derived in this study is not restricted to
NAAS but can be universally applied to AAS.
and the product at the PSS by using eqns. (1) and (2). At the
PSS eqn. (6) holds from which eqns. (7) and (8) can be derived.
(6)
(
(
7)
8)
Results and discussion
By using these equations and the mass balance for the
photoisomerization (see Appendix for details) we obtain eqn.
Numerical simulations
(
9), where eqns. (10)–(14) hold.
General scheme for the reaction. We consider a photo-
chemical system in which the enantiomers of the reactant (A)
are isomerized by irradiation with CPL to the enantiomers of
the product (B) without racemization (Scheme 2).
(
9)
(
(
10)
11)
(
(
(
12)
13)
14)
Scheme 2 A general scheme for the reversible photoisomerization
between reactant (A) and product (B).
For the reversible photoisomerization the rate equations are
expressed as shown in eqns. (1)–(5), where cAR, cAS, cBR, and cBS
(
1)
Method of simulation. For obtaining the numerical solution
of the differential equations a fourth-order Runge–Kutta
9
method was employed. If we use the reduced reaction time,
t* = tI {(ε ϩ εAS)/2}(Sl/V)Φ (ln 10), as a measure of the
reaction time, eqns. (1) through (4) are transformed into
(
2)
ex
AR
A
eqns. (13)–(16) where ε = (ε ϩ εAS)/2, ε = (ε ϩ ε )/2, and
A
AR
B
BR
BS
eqn. (5) holds.
(
3)
(13)
(
(
4)
5)
(
(
(
14)
15)
16)
are the concentrations of the enantiomers of the reactant (AR
and A ) and the product (B and B ), I is the incident photonic
flux expressed as einstein cm s , S is the area of the cross
section of the incident light, V is the volume of the solution, A
is the absorbance of the reaction mixture in the reaction
cuvette, εAR, εAS, εBR, and εBS are molar extinction coefficients, l
S
R
S
ex
Ϫ2 Ϫ1
(
5)
is the path length of the reaction cuvette, ΦA and Φ are the
B
quantum yields of the forward (A to B) and back (B to A)
reactions.
For the numerical integration we need the values of six
In the reversible reaction there should be a photostationary
state (PSS). We can estimate the concentrations of the reactant
parameters (εAR, εAS, εBR, ε , K, and l ) and the initial values of
the concentrations (cAR0, cAS0, cBR0, and cBS0). If we know the
BS
J. Chem. Soc., Perkin Trans. 2, 2001, 1706–1713
1707

View Article Online
values of ε , g , ε , g , c , and A , we can calculate the values we
A
A
B
B
0
0
need by using eqns. (17)–(23).
(
(
(
(
(
(
(
17)
18)
19)
20)
21)
22)
23)
The direct estimation of the value of K is relatively difficult,
because the evaluation of Φ’s is difficult. However, we can
estimate the value of K indirectly by using eqn. (9). Eqn. (9)
shows that the parameter K is a function of P, M, and the
conversion at the PSS. Therefore, by using the values of g , g
A
B
and the conversion at the PSS, we can estimate the value of
K. Once the values of the concentrations are obtained, we
can calculate the values of the ee of the reactant (A) and the
product (B) by using eqns. (24) and (25).
(
(
24)
25)
Fig. 1 Effect of the sign of g factors of reactant (A) and of product
(
B) on the enantiomeric enrichment induced by CPL irradiation of the
reversible photoisomerization system. Numerical simulations were
carried out for four different combinations of the signs of g factors.
K = 0.1. (a) The evolution of the enantiomeric excess of reactant A
with the progress of the reaction. (b) The enantiomeric excess of the
product B.
Effect of the sign and the absolute value of g factors on ee.
Fig. 1 is the result of the simulations for examining how the sign
of the g factor of the reactant and the product affects the ee of
the reactant and the products. The simulation was done for
hypothetical cases: case 1 (g × g > 0: g = 1 and g = 1 or
1
(II) by CPL irradiation. In one-way photoisomerization, the
relationship between the ee of the reactant, the product and the
conversion is naturally dependent upon only the g factor of
the starting material. The g factor of the product does not
affect the ee of the reactant and the product because the
product is not excited by the CPL irradiation. The ee of the
product decreases with the progress of the reaction, with the ee
of the product being g/2 at the initial stage of the reaction and
approaching zero at the final stage.
A
B
A
B
g = Ϫ1 and g = Ϫ1; K = 0.072), and case 2 (g × g < 0: g = 1
A
B
A
B
A
and g = Ϫ1 or g = Ϫ1 and g = 1; K = 0.072).
B
A
B
As shown in Fig. 1a, in case 1, the ee of the reactant (A)
increases at the beginning of the reaction, and decreases with
the progress of the reaction. It reaches zero at the PSS. In case
2
, the ee of the reactant (A) increases monotonically and
reaches its maximum at the PSS.
As shown in Fig. 1b, in case 1, the ee of the product (B)
gradually decreases from the beginning of the reaction
and reaches zero at the PSS. In case 2 the ee of the product
At the initial stage of the reversible photoisomerization the
behavior of the ee of the reactant and the product is similar to
the one-way photoisomerization by CPL irradiation. The ee of
the reactant should be zero and the ee of the product should be
g/2 at the beginning of the reaction, however, with the progress
of the photoisomerization, the product begins to be excited by
the CPL irradiation. Thus, the effect of g factor of the product
begins to contribute to the ee of the reactant.
If the sign of the g factor of the reactant is the same as that
of the product, the enantiomeric enrichment of the reactant is
depressed with the progress of the reaction, however, if the sign
of the g factor of the product is opposite to that of the reactant,
the enantiomeric enrichment of the reactant is amplified with
the progress of the reaction. Thus, in AAS, the combination
pattern of the signs of the g factors of the reactant and product
at the CPL-irradiation wavelength will give a remarkable effect
on the relationship between the ee’s of the reactant and product
and the conversion.
(
B) increases with the progress of the reaction and reaches a
maximum at the PSS.
In case 2, both the ee of the reactant (A) and the ee of
the product (B) increase with the progress of the reaction. In
general, both the ee of the reactant and the ee of the product
increase if the signs of the g factors of the reactant and product
are opposite to each other. The synchronous enantiomeric
enrichment of the reactant and product on prolonged irradi-
ation is the result of the preferential excitation of the minor
product isomer of the back reaction. The preferential excitation
of the minor product isomer on the back reaction enhances the
enrichment of the major product isomer.
If the signs of the g factors of the reactant and product are
the same, the back reaction reduces the accumulation of the
major product isomer by the preferential excitation of the same
isomer. This leads to unsuccessful enantiomeric enrichment of
both the reactant and the product at the PSS.
Effect of the value of K on ee. Fig. 2 was obtained from
simulations for the case g × g < 0: g = 1.0 and g = Ϫ1.0 or
These behaviors of the ee of product (B) are quite different
from the results reported in the preceding paper dealing with
the one-way photoisomerization of the reactant (I) to product
A
B
A
B
g = Ϫ1.0 and g = 1.0. The simulations show how the value of
A
B
K affects on the relationship between the ee of the reactant,
1
708
J. Chem. Soc., Perkin Trans. 2, 2001, 1706–1713

View Article Online
Table 1 Sign of g factors of reactant and product at the wavelength
of CPL irradiation
Reactant
Product
Irradiation
Case
wavelength
l-CPL
r-CPL
l-CPL
r-CPL
1
2
3
245 nm
255 nm
245 nm
gII > 0
gII > 0
gI = 0
gII < 0
gII < 0
gI = 0
gI = 0
gI < 0
gII > 0
gI = 0
gI > 0
gII < 0
Simulation using the parameters for the I–II system
In the above simulations, the values of the parameters were
arbitrarily given. Next, the relationship between the ee of
the reactant and product and the conversion was simulated for
the real system based on the actual chiroptical properties of
I and II. The wavelength of CPL irradiation (λCPL) was selected
by taking account of the chiroptical properties of I and II
1
shown in Fig. 2 of the preceding paper. The definition of
the anisotropy factor g is also the same as described in the
1
preceding paper.
By considering the results derived from the above dis-
cussions, we selected three wavelengths as shown in Table 1.
Case (1), λCPL = 245 nm (A (reactant) = II, B (product) = I;
g = Ϫ0.0074 and g = 0 for r-CPL, or g = 0.0074 and g = 0 for
II
I
II
I
l-CPL, conversion at PSS = 0.4, K = 0.667); Case (2), λCPL = 255
nm (A = II, B = I; g = Ϫ0.0074 and g = 0.005 for r-CPL, or
II
I
Fig. 2 Effect of the value of K on the enantiomeric enrichment
induced by CPL irradiation of the reversible photoisomerization
system. Numerical simulations were carried out for three different
values of K (0.072, 1.0 and ∞), assuming two different combinations of
the values of g factors, gA = Ϫ1 and g = 1 or g = 1 and g = Ϫ1. (a)
The evolution of the enantiomeric excess of reactant A with the
progress of the reaction. (b) The enantiomeric excess of the product B.
gII = 0.0074 and gI = Ϫ0.005 for l-CPL, conversion at
PSS = 0.23, K = 0.299); Case (3), λCPL = 245 nm (A = I, B = II;
g = 0 and g = Ϫ0.0074 for r-CPL, or g = 0 and g = 0.0074
I
II
I
II
for l-CPL, conversion at PSS = 0.6, K = 1.50).
In case (1), II is irradiated with CPL at 245 nm. Although II
is excited enantioselectively by CPL, no selection works on the
B
A
B
excitation of I, because g is zero at 245 nm. The function of the
I
CPL for I is the same as the linearly polarized or non-polarized
light, that is CPL isomerizes I to II non-enantioselectively.
When II is irradiated with CPL at 255 nm (case (2)), the CPL
not only acts as a chiral source for II, but it also acts as a chiral
source for I. The sign of the g factor of II is the opposite to that
of I at this wavelength.
In case (3), we start the reaction by taking I as the reac-
tant. In this case I is excited non-enantioselectively by CPL,
but II is excited enantioselectively by CPL. The enantiomeric
enrichment is promoted only by the back reaction through the
preferential excitation of one of the product isomers.
product and the conversion for the case, where the sign of the
g factor of the product is opposite to that of the reactant. Fig.
2
a shows that if K is less than 1, the ee of the reactant (A)
increases with the progress of the reaction, and that if K is
greater than 1, the ee of the reactant decreases with the progress
of the reaction. In an extreme case, if K is ∞, which means that
the reaction proceeds in one way (class (b) in the NAAS), the ee
of the reactant becomes 100% at the end of the reaction. How-
ever, Fig. 2b shows that the larger the K becomes, the more
gently the ee of the product (B) decreases, thus if K is ∞, the ee
of the product becomes 0 at the end of the reaction.
Thus, K also gives a remarkable effect on the relationship
between the ee of the reactant, product and the conversion,
especially on the relationship for the reactant. The ratio of the
rates of the forward and back reactions determines the extent
of the enhancement of the enantiomeric enrichment of the
product by the back reaction.
Evaluation of the simulations by comparing with experimental
results
Chiroptical properties and stereochemical outcome of the photo-
isomerization of the I–II system. The chiroptical properties of
1
II and I in acetonitrile have been reported.
Effect of the absorbance of the solution on the relationship
between the ee and the conversion. In the preceding paper we
have reported that, in the one-way photoisomerization, the
absorbance of the solution does not affect the relationship
between the ee’s of the reactant and product and the conver-
II irradiation at 245 nm (case (1)). In the first paper of this
series we reported that (Ϫ)-I isomerizes to (ϩ)-II when I is
irradiated at 290 nm, and that I quantitatively isomerizes to II
1
without any side reactions. Also for the irradiation at 245 nm
the absence of any side reactions was verified. The reaction
mixture was analyzed using GC with biphenyl as an internal
standard. Other signals except for I and II were not observed on
the GC charts and the total amount of I and II was retained. In
the UV absorption spectra of the reaction mixture during the
reaction two isosbestic points were observed at 208 nm and 212
nm. These results showed that II photochemically isomerized to
I without any side reactions.
Therefore, the conversions of II to I were determined with
the intensities of the UV absorption at 280 nm reflecting the
concentration of product I. Fig. 3 shows the relationship
between the conversion and the dose when the solutions of II
are irradiated with CPL and LPL at 245 nm.
1
sion. Here, for the reversible photoisomerization, the effect of
the absorbance of the solution on the relationship was exam-
ined by numerical simulation. In this simulation, in order to get
the same conversion at the PSS, we kept the value of K con-
stant. When we change the value of the initial absorbance of
the solution (i.e. the value of ε ), the value of εB or ΦB was
A
tuned in order to keep the value of K constant. However, for the
same values of g factors and K, the obtained curves indicating
the relationship between the ee’s of the reactant, product and
the conversion were exactly the same. We could not find any
effect of the absorbance on the relationship between the ee’s of
the reactant and product and the conversion.
J. Chem. Soc., Perkin Trans. 2, 2001, 1706–1713
1709

View Article Online
Fig. 3 Relationship between the absorbance at 280 nm and the dose
Fig. 6 Relationship between the ellipticity of the solution of II
irradiated with CPL and the dose: (᭺), irradiation with l-CPL,
ellipticity at 280 nm; (ᮀ), l-CPL, ellipticity at 245 nm; (᭹), r-CPL,
ellipticity at 280 nm and (᭿), r-CPL, ellipticity at 245 nm; (ꢀ), LPL
ellipticity at 280 nm. The curve drawn by interpolation: —, 280 nm;
for the solution of II ([II] = 0.87 mM) irradiated with CPL and LPL at
0
2
45 nm: (᭺); l-CPL, r-(᭹); CPL, (ꢀ); LPL.
-
--, 245 nm.
in Fig. 5. In the same manner, when the solution of II is irradi-
ated with l-CPL at 245 nm, the changes in the CD spectra
should be those of the evolution of the CD spectra composed
of (Ϫ)-II and (Ϫ)-I in 1 : 1 ratio. Indeed, the changes in the CD
spectra observed on CPL irradiation (Fig. 4) were nearly identi-
cal to the simulated spectra.
By using the changes in CD spectra, we can evaluate the
degree of the enantiomeric enrichment of I and II. The enan-
tiomeric enrichment of I was determined from the ellipticity at
2
80 nm, because II does not show any CD signal at 280 nm. The
enantiomeric enrichment of II was determined from the ellip-
ticity at 245 nm because the magnitude of the ∆ε of (ϩ)- and
(Ϫ)-I was incidentally zero at 245 nm.
Fig. 6 shows the relationship between the ellipticity and the
dose. Upon CPL irradiation the ellipticity at 245 and 280 nm
first increased and then became constant with time. As shown in
Fig. 5, the intensities of the UV absorption at 280 nm also
became constant after a large dose of CPL irradiation at 245
nm. This means that the system reached a photostationary state
after a large dose of irradiation.
Fig. 4 Typical changes in CD spectrum when the solution of II in
acetonitrile was irradiated at 245 nm with l- and r-CPL. The CD spectra
1
were measured after dilution of the reaction mixture to
mM).
3
([II] = 0.87
0
The relationships between the ee and the conversion (the
progress of the photoisomerization of II to I) are shown in Fig.
7
. The values of the ee and the conversion were obtained from
1
the results derived from the CD and UV measurements.
Fig. 7 shows the experimental and simulation results. The ee
of II increases with the increasing conversion, and correspond-
ingly the ee of I decreases with increasing conversion. The
results of the simulations showed good agreement with the
experimental results. In this case the dependence of the ee’s on
the conversion is very much like those observed for the one-way
1
Fig. 5 Spectral simulation of the CD spectrum of the solution of II
after irradiation at 245 nm with r-CPL.
photoisomerization.
II irradiation at 255 nm (case (2)). The UV spectral changes
of the solution of II irradiated with CPL at 255 nm were very
similar to those shown in Fig. 3. Upon irradiation with CPL at
255 nm, the absorption at 280 nm became more intense. Two
isosbestic points in the UV spectra were also observed at 208
nm and 212 nm. GC analysis showed no byproducts. The
results showed that the photoreaction proceeded without any
side reactions. These observations were quite similar to those in
case (1). The conversions of II to I were determined with the
UV absorbance at 280 nm in a similar manner to case (1).
The changes in the CD spectra were also very similar to those
in case (1), shown in Fig. 5. Therefore, the enantiomeric
enrichment of I was determined from the ellipticity at 280 nm
and the enantiomeric enrichment of II was determined from the
ellipticity at 245 nm.
The changes in a typical CD spectrum for the solutions of II
irradiated with l- and r-CPL at 245 nm are shown in Fig. 4.
Upon CPL irradiation, two peaks emerged at 215 nm and 270
nm and the intensities became greater with the progress of the
reaction.
Fig. 5 is the result of the spectral simulation of the change in
the CD spectrum of II upon irradiation with r-CPL at 245 nm.
When the solution of II was irradiated with r-CPL at 245 nm,
(
Ϫ)-II was preferentially excited and consumed to leave (ϩ)-II.
1
The excited (Ϫ)-II isomerizes to (ϩ)-I. Because the photo-
isomerization is a unimolecular reaction, the consumption of
one mole of (Ϫ)-II should be accompanied by the production
of one mole of (ϩ)-I. Therefore, the changes in the CD spectra
when the solution of II is irradiated with r-CPL at 245 nm
should be the evolution of the CD spectra composed of (ϩ)-II
and (ϩ)-I in a 1 : 1 ratio, which is depicted as the bold-solid line
Fig. 8 shows the experimental results and the results of
the simulation. The ee of II increases with the progress of the
1
710
J. Chem. Soc., Perkin Trans. 2, 2001, 1706–1713

View Article Online
Fig. 7 Numerical simulation and experimental data for the evolution
of the enantiomeric excess of the reactant II and the product I with the
progress of the reaction upon the irradiation of the solution of II with
r-CPL and l-CPL at 245 nm. Numerical simulations were carried out by
using the following values of the parameters: g = Ϫ 0.0074 and g = 0
Fig. 8 Numerical simulation and experimental data for the evolution
of the enantiomeric excess of the reactant II and the product I with the
progress of the reaction upon the irradiation of the solution of II with
r-CPL and l-CPL at 255 nm. Numerical simulations were carried out
II
I
by using the following values of the parameters: g = Ϫ0.0074 and
II
for r-CPL irradiation; and g = 0.0074 and g = 0 for l-CPL irradiation:
II
I
g = 0.005 for r-CPL irradiation; and g = 0.0074 and g = Ϫ 0.005 for
I
II
I
K = 0.667. (a) % ee of the reactant (II), (b) % ee of product (I).
Experimental data: (᭺), l-CPL; (᭹), r-CPL. Simulation curve: —,
l-CPL; ---, r-CPL.
l-CPL irradiation: K = 0.299. (a) % ee of the reactant (II), (b) % ee of
product (I). Experimental data: (᭺), l-CPL; (᭹), r-CPL. Simulation
curve: —, l-CPL; ---, r-CPL.
reaction. The ee of I also increases with the progress of
the reaction in contrast to the behavior of the ee of I in the case
of the reaction. However, there is still disagreement between the
results of the simulations and the experimental results. The
reason for the discrepancy is not clear, but it might be caused
by the wide (full width at half maximum is about 25 nm) and
asymmetrically shaped spectrum of the CPL generated by the
Onuki-type polarizing undulator installed in the electron stor-
(
1) experiment. The synchronous enantiomeric enrichment of
the reactant and product on prolonged irradiation is the result
of the preferential excitation of the minor product isomer in
the back reaction. For the irradiation with r-CPL, for example,
the forward reaction produces (ϩ)-I preferentially, and the
back reaction consumes (Ϫ)-I preferentially. Therefore the
accumulation of (ϩ)-I is enhanced by the back reaction at
the prolonged irradiation. The experimental results and the
results of the simulation showed a good agreement with each
other.
10,11
age ring NIJI-II.
In the spectral profile of the CPL the con-
tribution of the light at the longer wavelength is larger than that
at the shorter wavelength. The chiroptical properties of I and II
may enhance this effect. The g factor is zero at 245 nm, but, at
both sides of this wavelength, the g factor is not zero. The
magnitude of the g factor is larger at the longer wavelength half
of the band than at the shorter wavelength half. If the band-
width of the CPL source is large, the g value at the longer
wavelength half of the spectrum contributes much more to the
reaction. In order to check this possibility, we tried simulations
using the estimated average of the values of the g factor over
I irradiation at 245 nm (case (3)). As case (3) we have selected
the photoisomerization of I to II by CPL at 245 nm (g = 0 and
I
gII = ±0.0074, K = 1.50). Under these conditions the enantio-
meric enrichment is promoted only by the back reaction
through the preferential excitation of the product.
the spectrum of CPL (g = ±0.002, gII = ±0.0074). The results of
I
The UV spectral changes of the solution of I irradiated with
the simulations showed a good agreement with the experi-
mental results.
CPL at 245 nm were very similar to those shown in the first
1
paper of this series. Upon irradiation with CPL at 245 nm, the
absorption at 280 nm became smaller, and two isosbestic points
in the UV spectra were observed at 208 nm and 212 nm, which
indicate that the photoreaction proceeded without any side
reactions.
The changes in the CD spectra were analogous to cases (1)
and (2). Therefore we used the same method for determining
the enantiomeric excess of the reactant and product.
Fig. 9 is the result of the simulation and the experimental
examination of the case. The ee of I and II concurrently
increases with the conversion. The results of the simulations
reproduce the trends in the changes in the ee’s with the progress
Conclusions
We have proposed in this work a reversible new absolute asym-
metric synthesis (NAAS) which has been examined using a
photochemical system producing chiral product. The systems
are photochemically reversible between the reactant and the
product, with enantiomers that do not interconvert with each
other photochemically or thermally at ambient temperature.
NAAS may be classified into two subcategories according to
the photochemical reversibility between the reactant and the
product: in class (a), there is a photochemical equilibrium
J. Chem. Soc., Perkin Trans. 2, 2001, 1706–1713
1711

View Article Online
and attain some appreciable values at the photostationary state.
This shows that Kagan’s speculation, which suggests that the
enantiomeric purity of the products decreases throughout
the course of the photolysis, is incorrect in class (b) NAAS
12
reactions.
Our conclusion reveals that the concurrent
enantiomeric enrichment of both the reactant and the product
will occur with the progress of the photoreaction using CPL
irradiation. To further extend the understanding of this
approach, an examination of the molecular structure and
how this may influence the ee of the reactant and product is
currently being undertaken.
Experimental
General
UV spectra were measured with a JASCO V-560 spectrometer.
CD spectra were measured with a JASCO J-725 CD spectro-
meter. Gas chromatographic analysis was performed on a
Shimadzu GC14-A instrument.
Materials
Methyl norbornadiene-2-carboxylate (methyl bicyclo[2.2.1]-
hepta-2,5-diene-2-carboxylate, I) was synthesized according to
1
methods reported in Part 1. Methyl quadricyclane-1-carboxyl-
2,7 4,6
ate (methyl tetracyclo[3.2.0.0 .0 ]heptane-1-carboxylate, II)1
was also synthesized according to methods reported in Part 1.
Acetonitrile (Merck, UV spectroscopic grade) was used as
received.
Fig. 9 Numerical simulation and experimental data for the evolution
of the enantiomeric excess of the reactant I and the product II with the
progress of the reaction upon irradiation of the solution of I with r-
CPL and l-CPL at 245 nm. Numerical simulations were carried out by
using the following values of the parameters: g = 0 and g = Ϫ0.0074
Photolysis
Acetonitrile solutions of II (0.72–0.87 mM) and I (0.73 mM)
were sealed in a rectangular quartz cell (1 × 1 cm) under an
argon atmosphere, and were irradiated by l- and r-CPL gener-
ated by an Onuki-type polarizing undulator installed in the
electron storage ring NIJI-II operated at the Optical Radiation
Laboratory of the Electrotechnical Laboratory at Tsukuba,
producing highly pure, intense, monochromatic CPL at any
wavelengths down to the vacuum UV region, and therefore is
I
II
for r-CPL irradiation; and g = 0 and g = 0.0074 for l-CPL irradiation:
I
II
g = Ϫ0.002 and g = Ϫ0.0074 for r-CPL irradiation; and g = 0.002 and
I
II
I
gII = 0.0074 for l-CPL irradiation K = 1.50. (a) % ee of the reactant (II),
b) % ee of product (I). Experimental data: (᭺), l-CPL; (᭹), r-CPL.
Simulation curve: —, g = 0 and g = Ϫ0.0074, r-CPL; ---, g = 0 and
(
I
II
I
g = 0.0074, l-CPL: – - –, g = Ϫ0.002 and g = Ϫ0.0074, r-CPL; – – – –,
II
I
II
g = 0.002 and g = 0.0074, l-CPL.
I
II
10,11,13,14
the most suitable CPL source for the NAAS studies.
between the reactant and the product because the reaction
mixture is irradiated with CPL which excites both the reac-
tant and the product; in class (b), the reaction is irreversible
because the reactant is excited by CPL but the product is not
excited because the product does not absorb incident light at
the excitation wavelength.
Photolyses were carried out for several irradiation periods for
each series of samples of I and II. The photolyzed solutions
were subjected to UV and CD spectral analysis as well as gas
chromatographic analysis over a CBP-1 column (Shimadzu).
The CD spectra were measured using a cylindrical Suprasil cell
(1.3 cm i.d., 1 cm path).
In this report, for class (a), the numerical simulations based
on the rate expressions for the reaction were used to obtain the
relationship between the ee of the product and the conversion
by CPL irradiation. The validity of the results of the simu-
lations were checked by performing reversible photoisomeriz-
ation between I and II upon excitation by the monochromatic
CPL generated by an Onuki-type polarizing undulator installed
in a synchrotron radiation source. The simulations showed
good agreement with the experimental results. These results
experimentally verify that the relationship between the ee of the
reactant and product and the conversion remarkably depends
upon the g factors of the reactant and the product and the ratio
of the rates of forward and back reactions.
The above consideration will be very helpful in predicting the
enantiomeric enrichment upon CPL irradiation only from the
chiroptical properties of the reactant and the product, even if
a system other than the I–II system is used.
Furthermore, the results obtained from the theoretical and
experimental verification suggest that, if the g factors of the
product and reactant are large enough, the sign of the g factor
of the product is opposite to that of the reactant at the irradi-
ation, and K is less than unity, the ee of the reactant and the
product will concurrently increase according to the conversion
Appendix
The derivation of eqn. (9) to form eqns. (7) and (8) is as
follows.
By using the eqns. (7), (8), and (17) to (20), we obtain eqns.
(
26) and (27).
(
(
26)
27)
The mass balance for the isomerization gives us the relation-
ship shown in eqns. (28) and (29). By using these equations we
derive eqns. (30) and (31).
(
(
28)
29)
1
712
J. Chem. Soc., Perkin Trans. 2, 2001, 1706–1713

View Article Online
Photochirogenesis Project) for discussions and insights, as well
as corrections and improvements arising from preliminary
reading of this manuscript.
(
(
30)
31)
References
1
(a) Part 1: H. Nishino, A. Nakamura and Y. Inoue, J. Chem. Soc.,
Perkin Trans. 2, 2001 (DOI: 10.1039/b100751n); (b) Part 2:
A. Nakamura, H. Nishino and Y. Inoue, J. Chem. Soc., Perkin
Trans. 2, 2001 (DOI: 10.1039/b100749l).
By definition the conversion at PSS is expressed as eqn. (32),
where eqns. (10)–(12) hold.
2
3
4
5
W. A. Bonner, Stud. Phys. Theor. Chem., 1979, 7, 5.
W. A. Bonner and E. Rubenstein, BioSystems, 1987, 20, 99.
W. A. Bonner, AIP Conf. Proc., 1996, 379, 17.
Y. Inoue, H. Tsuneishi, T. Hakushi, K. Yagi, K. Awazu and
H. Onuki, Chem. Commun., 1996, 2627.
(
32)
6
7
8
Y. Inoue, Chem. Rev., 1992, 92, 741.
H. Rau, Chem. Rev., 1983, 83, 535.
S. R. L. Everitt and Y. Inoue, in Asymmetric Photochemical
Reactions in Solution, ed. V. Ramamurthy and K. S. Schanze, Marcel
Dekker, New York, 1999.
(
(
(
10)
11)
12)
9
M. J. Pilling and P. W. Seakins, Oxford University Press, Oxford,
1
995.
1
0 M. Yuri, K. Yagi, T. Yamada and H. Onuki, J. Electron Spectrosc.
Relat. Phenom., 1996, 80, 425.
1 H. Onuki, K. Yagi, K. Awazu, T. Yamada, T. Saito, M. Yuri,
Y. Inoue and S. Ishizaka, Proc. SPIE Int. Soc. Opt. Eng., 1996,
1
2
873, 90.
2 G. Balavoine, A. Moradpour and H. B. Kagan, J. Am. Chem. Soc.,
974, 96, 5152.
13 H. Onuki, Nucl. Instrum. Methods Phys. Res., Sect. A, 1986, 246,
4.
1
1
Acknowledgements
9
We thank Mr Hiroshi Hayakawa for adjusting the JASCO
J-725 CD spectrometer, and Dr Simon Everitt (Inoue
1
4 H. Onuki, N. Saito and T. Saito, Appl. Phys. Lett., 1988, 52,
173.
J. Chem. Soc., Perkin Trans. 2, 2001, 1706–1713
1713