Photoreaction of 9-methylbenz[c]acridine with diphenylacetic acid in the chiral cocrystal

Article abstract of DOI:10.3987/COM-03-9848

Despite 9-methylbenz[c]acridine and diphenylacetic acid are achiral molecules, chiral cocrystal was obtained by crystallization from the solution. Irradiation of the enantiomorphous cocrystals caused photodecarboxylation to afford almost racemic photodecarboxylative condensation product as one of the minor products. Reason for the unsuccessful absolute asymmetric photoreaction was discussed based on the molecular arrangement in the lattice.

Full text of DOI:10.3987/COM-03-9848

HETEROCYCLES, Vol. 60, No. 10, 2003, pp. 2251 - 2258
Received, 7th July, 2003, Accepted, 22nd August, 2003, Published online, 25th August, 2003
PHOTOREACTION
OF
9-METHYLBENZ[c]ACRIDINE
WITH
DIPHENYLACETIC ACID IN THE CHIRAL COCRYSTAL
Hideko Koshima,* Ayaka Nakata, Masaki Nagano, and Haitao Yu
Department of Applied Chemistry, Faculty of Engineering, Ehime University,
Matsuyama 790-8577, Japan
E-mail: koshima@eng.ehime-u.ac.jp
Abstract – Despite 9-methylbenz[c]acridine and diphenylacetic acid are achiral
molecules, chiral cocrystal was obtained by crystallization from the solution.
Irradiation of the enantiomorphous cocrystals caused photodecarboxylation to
afford almost racemic photodecarboxylative condensation product as one of the
minor products. Reason for the unsuccessful absolute asymmetric photoreaction
was discussed based on the molecular arrangement in the lattice.
INTRODUCTION
Absolute asymmetric synthesis by solid-state reaction of chiral crystal spontaneously formed from achiral
molecule is a promising methodology for asymmetric synthesis.1-3 Since our success of absolute
asymmetric photodecarboxylative condensation of acridine with diphenylacetic acid in the chiral
cocrystal,4 we would like to design absolute asymmetric synthesis with high reliability. Because
spontaneous chiral crystallization cannot be predicted at the present time,5,6 we have prepared several
series of chiral cocrystals by combining two different achiral molecules, based on the working hypothesis
that chiral crystallization can occur by freezing two molecules into chiral conformations through
hydrogen bonding. These include propeller-type chiral crystal from phenanthridine with diphenylacetic
acid7 and helical-type chiral crystals from tryptamine with various achiral carboxylic acids,8
phenanthridine with 3-indolepropionic acid,9 and 2-amino-5-nitropyridine with achiral benzenesulfonic
acids.10 Herein we report a new chiral cocrystal (1•2) of 9-methylbenz[c]acridine (1) with
diphenylacetic acid (2) and its solid-state photoreaction.

N
COOH
chiral cocrystal 1•2
•
H
CH
3
2
1
RESULTS AND DISCUSSION
The 1:1 chiral cocrystal (1•2) was prepared by recrystallization from the solution of 1 and 2 in acetonitrile.
Seed crystals of both enantiomorphous crystals of 1•2 were initially obtained by spontaneous
crystallization from the slightly supersaturated solutions in several vessels, and were easily discriminated
by the measurement of solid-state circular dichroism (CD) spectra as Nujol mulls, giving A or B curve for
the crystals in each vessel (Figure 1). Then, the left and right handed crystals could be separately
prepared by the seeding on a large scale. The chiral nature was confirmed by X-Ray crystallographic
analysis due to the acentric space group P1. However, the absolute structure could not be determined
because the oxygen atom is not enough heavy to obtain large X-Ray anomalous dispersion using Cu Ka
radiation. Figure 2 shows the molecular arrangement in 1•2. Hydrogen bonding between 1 and 2 is
formed with an H•••N distance of 1.46 Å. Diphenylacetic acid molecule has a propeller-like
conformation, and 9-methylbenz[c]acridine molecule is not a flat plane but some twisted. The important
feature is that only a kind of molecular pair of the absolute configuration presents
noncentrosymmetrically in a unit cell (Z = 1) to induce the crystal chirality. Namely, the molecules in
themselves are achiral but the molecular shapes are frozen in chiral forms within the lattice.
+
B
DCD
0
A
–
300
200
400
Wavelength (nm)
Figure 1. Solid-state CD spectra of the enantiomorphous cocrystals of 1•2

c
b
0
Figure 2. Molecular arrangement in the cocrystal (1•2)
The cocrystals of 1•2 prepared by seeding (curve A in Figure 1) were pulverized and irradiated to cause
mainly photodecarboxylation, giving four products (Scheme 1). Decarboxylative condensation product
(3) having a chiral center carbon was obtained in 13% yield.
Decarboxylated compounds
diphenylmethane (4) as a main product and the dimer tetraphenylethane (5) were produced in 34% and
3% yields, respectively. Dehydration product benzophenone (6) was also obtained in 13% yield.
On the other hand, solution photolysis of 1 and 2 in acetonitrile afforded racemic 3 as a main product in
53% yield and 4 and 5 in 2% and 12% yields, showing different photoreactivity and product selectivity
from the solid-state photoreaction. The molecular structure of 3 was confirmed by X-Ray
crystallographic analysis using a single crystal of 1:1 inclusion complex of racemic 3 with MeOH (Figure
3).
N
H H
COOH
H
hn
N
*
H
Ph CH
Ph C=O
+
+
Ph CHCHPh
+
2
2
2
•
2
2
CH
cocrystal
3
5
4
6
CH
3
1•2
3
hn
MeCN
+
+
3
4
5
Scheme 1. Photoreaction of 1 and 2 in the cocrystal and in the solution

O
N
C*
Figure 3. Molecular structure of the product (3•MeOH)
In order to know whether absolute asymmetric synthesis was achieved or not, the condensation product
(3) obtained by the solid-state photoreaction was submitted to HPLC analysis using a chiral column.
However, the enantiomeric excess could not be accurately determined due to the smallness. Hence,
enantiomorphous single crystals (2–3 mg) (A and B in Figure 1) were irradiated piece by piece followed
by HPLC determination. The results are summarized in Table 1. Irradiation of crystal A at room
temperature with a high-pressure mercury lamp and a superhigh-pressure mercury lamp gave both plus
and minus signals of low ees (Entries 1-3), showing not significant. In the case of crystal B, similar
results were obtained (Entries 6-8). Even upon irradiation under cooling at –20 ~ –30 °C, the
enantiomeric excesses did not increase (Entries 4, 5 and 9). So we concluded that absolute asymmetric
synthesis was not achieved by the solid-state photoreaction of this chiral cocrystal.
Table 1. Solid-state photoreaction of the chiral single cocrystal (1•2)
Irradiation
Conversion (%)
Yield based on consumed 2 (%)
Eed (%)
Entry Crystala
(°C)
18b
18b
(min)
2.5
1
2
3
4
5
6
3
1
2
3
4
5
6
7
8
9
A
A
A
A
A
B
B
B
B
18
91
5
71
0.1
9
(+) 1
5
10
30
30
2.5
5
34
19
13
2
92
62
38
34
91
92
67
37
4
2
2
2
4
4
1
2
53
39
61
47
61
59
61
38
0.1
0.5
0.3
0
11
6
(+) 2
(–) 4
(–) 4
(–) 4
(+) 3
(+) 6
(–) 4
(+) 3
15c
–21c
–33c
18b
18b
15c
4
2
18
34
19
8
1
8
0.4
0.3
1
10
7
10
30
–32c
0
a
b
Single crystals corresponding to the CD curves A and B in Figure 1.
Irradiation with a 400 W high-
Signals of ees were termed
c
d
pressure Hg lamp.
Irradiation with a 500 W superhigh-pressure Hg lamp.
(+) for the first peak of HPLC, (–) for the second peak.

The reason for unsuccessful absolute asymmetric photodecarboxylative condensation can be discussed
based on the molecular arrangement in the crystal lattice as follows (Scheme 2). Irradiation of 1•2 can
cause electron transfer from 2 to 1 followed by proton transfer to afford a carboxylate radical (7) and
hydrobenz[c]acridine radicals (8) and (9).4,11 Then decarboxylation of the carboxylate radical produces a
•CHPh2 diphenylmethyl radical (10). Next radical coupling between the •CHPh2 radical (10) and the
hydrobenz[c]acridine radical (9) forms 3. Four directions for the radical coupling are possible in the
crystal (Figure 4). Absolute configuration of 3 obtained by •C1---•C2 coupling with the shortest
distance (4.42 Å) is opposite to those by couplings between •C1 and the other three •C3, •C4 and •C5
radicals with the longer distances (6.49, 6.44 and 6.50 Å), to lead to almost racemic mixture of (+)- and
(–)-3.
CHPh
2
2
H
N
H H
O
O
H
N
CHPh
2
•
H
hn
+
+
O
O
•
N
•
cocrystal
CH
H C
3
3
7
8
9
CH
3
1
–CO
+
2
+
+
+
• CHPh
8
9
3
5
4
1
2
H-abstraction
radical
coupling
10
Scheme 2. Possible reaction mechanism
C4
C1
C5
C2
H1
N1
C3
Figure 4. Possible reaction path in the cocrystal (1•2)

In addition, the formation of diphenylmethane (4) is caused by hydrogen abstraction by the •CHPh2
radical (10) from the hydrobenz[c]acridine radical (8) because the •C1---H1 distance (3.50 Å) is short
enough to abstract the hydrogen H1.11 The regeneration of 1 is the reason for the lower conversion of 1
than that of 2 (Table 1). In conclusion, solid state photoreactivity is controlled by the crystal structure.
EXPERIMENTAL
General Procedure
1H-NMR spectra were measured on a JEOL JNM-GSX270 spectrometer with tetramethylsilane as an
internal standard. IR spectra were recorded on a JASCO FT/IR-300E spectrophotometer. Melting
points (mp) were not corrected. Elemental analysis was carried out with a Yanaco CHN Corder MT-5.
The solvents for gelation experiments were of analytical grade. All the solvents and reagents were
commercially available and used as received.
Preparation of cocrystal (1•2)
9-Methylbenz[c]acridine (1) was synthesized by the reaction of 1-naphthylphenylamine and acetic acid in
the presence of zinc chloride under microwave irradiation for several minutes.12 Chiral cocrystal (1•2)
was prepared by slow evaporation of 1:1 solution of 1 and diphenylacetic acid 2 in MeCN at rt. Both
enantiomorphous crystals could be prepared by the seeding.
1•2. yellow prisms; mp 137-139 °C; IR (KBr) 1705, 1962, 2461, 3432 cm–1.
Preparative photoreaction of 1 with 2 in MeCN
A solution of 1 (622 mg, 2.56 mmol) and 2 (542 mg, 2.56 mmol) in MeCN (100 mL) was internally
irradiated with a l00 W high-pressure mercury lamp through Pyrex for 2 h under argon bubbling at rt.
After the irradiation, the reaction mixture was submitted to silica gel column chromatography (toluene-
ethyl acetate, 5:1) to give 3 (866 mg), 4 (16 mg) and 5 (164 mg) in 53, 2 and 12% yields, respectively, at
78% conversion of 1. The IR and 1H NMR spectral signals of 4 and 5 were in accordance to those of the
authentic samples.
1
3. yellow prisms, mp 140-146 °C (from MeCN); IR (KBr) no NH band; H NMR (CDCl3) d 8.13 (d, J =
8.4 Hz, 1H), 7.97 (d, J = 8.4 Hz, 1H), 7.67 (t, J = 7.7 Hz, 1H), 7.58 (t, J = 6.8 Hz, 2H), 6.96-7.26 (m,
12H), 6.81 (d, 7.0 Hz, 1H), 3.81-3.88 (m, 1H), 3.69 (d, J = 11.6 Hz, 1H), 3.28 (dd, J = 16.0, 4.1 Hz, 1H),
3.18 (dd, J = 13.0, 3.0 Hz, 1H), 3.02 (s, 3H). Anal. Calcd for C31H25N: C 92.42, H 4.83, N 2.55. found:
C 92.42, H 4.83, 2.55.

Preparative solid-state photoreaction of chiral crystal (1•2)
Enantiomorphous crystals of 1•2 (1.5 g) prepared by seeding was pulverized, and 100-mg powders were
separately placed between two Pyrex plates and irradiated with a 400 W high-pressure mercury lamp for
4 h under argon at rt. The irradiated mixture was treated with CH2N2 and submitted to silica gel column
chromatography (toluene-ethyl acetate, 5:1) to give 3 (144 mg), 4 (191 mg), 5 (28 mg) and 6 (79 mg) in
13, 34, 3 and 13% yields, respectively, at 53% conversion of 1. The IR and 1H NMR spectral signals of
6 were in accordance to those of the authentic samples.
HPLC study of solid-state photoreaction
One single crystal (2–3 mg) was pulverized, placed between two Pyrex plates and irradiated with a 400 W
high-pressure or 500 W superhigh-pressure mercury lamp under argon at appropriate temperatures and
times. The irradiated sample was methylated with CH2N2 and submitted to HPLC analysis using chiral
column (Daicel Chiralpak AD). The results are shown in Table 1.
X-Ray structure analysis
Two crystals of 1•2 and 3•MeOH were submitted to X-Ray crystallographic analysis. Data were
collected on a Rigaku RAXIS RAPID imaging plate diffractometer with Cu Ka radiation to a maximum
2q of 136.5°. A total of 30 x 5.00° oscillation images was collected, each being exposed for 15.0 min.
The reflection data were corrected for Lorentz-polarization effects and secondary extinction. The
structures were solved by direct methods and expanded using Fourier techniques. The non-hydrogen
atoms were refined anisotropically and hydrogen atoms were not refined. The final cycle of full-matrix
least-squares refinement was based on all reflections (2q < 136.5°) and variable parameters and
converged with unweighted and weighted agreement factors of: R = SÍÍFoÍ–ÍFcÍÍ/SÍFoÍ for I > 2.0s(I)
2
2 2
and Rw = {Sw(Fo –Fc2)2/Sw(Fo ) }1/2. All calculations were performed by using a teXsan crystallographic
software, Molecular Structure Corporation.
1•2. C32H25NO2, Mw = 455.56, yellow prism crystal (0.23 x 0.19 x 0.14 mm3), triclinic, space group P1
with a = 5.683(1), b = 9.110(2), c = 12.111(2) Å, a = 101.443(6)°, b = 98.399(6)°, g = 103.08(1)°, V =
586.5(2) Å3, Z = 1, rcalc = 1.290 g cm–3, F(000) = 240.00, m = 6.27 cm–1. A total of 6528 reflections
were measured, 1990 independent. Final R = 0.061, Rw = 0.144, GOF = 1.89 for 318 parameters.
3•MeOH: C32H29NO, Mw = 443.59, yellow prism crystal (0.73 x 0.55 x 0.25 mm3), orthorhombic, space
group Pbca with a = 18.995(1), b = 12.2607(6), c =21.294(1) Å, V =4959.3(4) Å3, Z = 8, rcalc = 1.188 g
cm–3, F(000) = 1888.00, m = 5.45 cm–1. A total of 42752 reflections were measured, 15738 independent.
Final R = 0.0730, Rw = 0.228, GOF = 1.65 for 308 parameters.
ACKNOWLEDGEMENTS

This work was supported by a Grant-in Aid for scientific research from the Ministry of Education, Sports,
Culture, Science and Technology, Japan (No. 14340220) and Kurata Memorial Hitachi Science and
Technology Foundation in Japan.
REFERENCES
1. H. Koshima, ‘Organic Solid State Reactions,’ ed. by F. Toda, Kluwer Academic publishers,
Dordrecht, 2002, pp. 189-268.
2. B. L. Feringa and R. A. van Delden, Angew. Chem., Int. Ed., 1999, 39, 3418.
3. M. Sakamoto, Chem. Euro. J., 1997, 3, 684.
4. H. Koshima, K. Ding, Y. Chisaka, and T. Matsuura, J. Am. Chem. Soc., 1996, 118, 12059.
5. J. Jacques, A. Collet, and S. H. Wilen, ‘Enantiomers, Racemates, and Resolutions,’ Wiley-
Interscience, New York, 1981.
6. a) H. Koshima and T. Matsuura, J. Synth. Org. Chem., 1998, 56, 268 (Japanese). b) H. Koshima and
T. Matsuura, J. Synth. Org. Chem., 1998, 56, 466 (Japanese).
7. H. Koshima, T. Nakagawa, T. Matsuura, H. Miyamoto, and F. Toda, J. Org. Chem., 1997, 62, 6322.
8. a) H. Koshima, S. I. Khan, and M. A. Garcia-Garibay, Tetrahedron: Asymmetry, 1998, 9, 1851. b) H.
Koshima and S. Honke, J. Org. Chem., 1999, 64, 790. c) H. Koshima, S. Honke, and J. Fujita, J.
Org. Chem., 1999, 64, 3916. c) H. Koshima, S. Honke, and M. Miyauchi, Enantiomer, 2000, 5, 125.
d) H. Koshima, M. Miyauchi, and M. Shiro, Supramol. Chem., 2001, 13, 137.
9. a) H. Koshima, E. Hayashi, T. Matsuura, K. Tanaka, F. Toda, M. Kato, and M. Kiguchi,
Tetrahedron Lett., 1997, 38, 5009. b) H. Koshima, E. Hayashi, and T. Matsuura, Supramol. Chem.,
1999, 11, 57.
10. H. Koshima, M. Hamada, I. Yagi, and K. Uosaki, Cryst. Growth Des., 2001, 1, 467.
11. a) H. Koshima, K. Ding, and T. Matsuura, J. Chem. Soc., Chem. Commun., 1994, 2053. b) H.
Koshima, K. Ding, Y. Chisaka, T. Matsuura, I. Miyahara and K. Hirotsu, J. Am. Chem. Soc., 1997,
119, 10317.
12. H. Koshima and K. Kutsunai, Heterocycles, 2002, 57, 1299.