Solid-state and solution photolyses of tetracyanobenzene with benzyl cyanides or benzyl alcohols

Article abstract of DOI:10.1016/S0040-4020(00)00631-1

Intermolecular photoreactions of tetracyanobenzene (TCNB) with benzyl cyanide (BzCN), benzyl alcohol (BzOH) and various others were investigated in the solid state (cocrystal) and in solution. The new solid-state photocoupling reaction found for the cocrystal TCNB·BzCN, giving a stilbene derivative followed by the solution isomerization into an isoindole derivative, is a very limited reaction. On the other hand, its solution photocondensation to give products of the diphenylmethane type occurred quite generally, probably under acidic conditions. (C) 2000 Elsevier Science Ltd.

Full text of DOI:10.1016/S0040-4020(00)00631-1

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 7139–7152
Solid-State and Solution Photolyses of Tetracyanobenzene
with Benzyl Cyanides or Benzyl Alcohols
a
b
b
Yoshikatsu Ito,a, Hironari Nakabayashi, Shigeru Ohba and Hiroyuki Hosomi
*
aDepartment of Synthetic Chemistry & Biological Chemistry, Graduate School of Engineering, Kyoto University, Kyoto 606-8501, Japan
bDepartment of Chemistry, Faculty of Science and Technology, Keio University, Yokohama 223-8522, Japan
Received 1 May 2000; accepted 14 July 2000
Abstract—Intermolecular photoreactions of tetracyanobenzene (TCNB) with benzyl cyanide (BzCN), benzyl alcohol (BzOH) and various
others were investigated in the solid state (cocrystal) and in solution. The new solid-state photocoupling reaction found for the cocrystal
TCNB·BzCN, giving a stilbene derivative followed by the solution isomerization into an isoindole derivative, is a very limited reaction. On
the other hand, its solution photocondensation to give products of the diphenylmethane type occurred quite generally, probably under acidic
conditions. ᭧ 2000 Elsevier Science Ltd. All rights reserved.
Introduction
Incidentally, we found that the colored charge-transfer
crystals between TCNB and anthracene (1:1 complex),6
trans-stilbene (2:1 complex),7 durene (1:1 complex),8
biphenyl (1:1 complex),9 dimethylaniline (1:1 complex),6
and others2 were photostable in the solid state.10
Studies on photochemical reactions occurring in two-
component crystals or in solid mixtures are one of the
current topics.1,2 In our recent paper,3–5 we have described
briefly that a crystalline 1:2 complex between tetracyano-
benzene (TCNB) and benzyl cyanide (BzCN), which is
represented here by TCNB·BzCN, undergoes selectively a
photoreductive coupling reaction at the cyano group of
TCNB, resulting in production of a trans-stilbene derivative
1 in the solid state. By contrast, the photoreaction of TCNB
in BzCN solution was very different from that of
TCNB·BzCN, giving condensation products 3a, 4a and 5.3
In order to investigate the generality of these solid-state and
solution-state photoreactions, we subsequently studied
bimolecular photoreactions of TCNB with various partners
that are analogous or homologous to BzCN. In this paper,
we summarize the details of the results obtained so far.
Scheme 1 lists main compounds that we employed and
Scheme 2 shows the structures for various solid-state and
solution-state photoproducts isolated in the present work.
Results and Discussion
Preparation of cocrystals
The main compounds that were tested to prepare the two-
component crystals of tetracyanobenzene (TCNB) are
shown in Scheme 1. Thus, TCNB was cocrystallized with
the second component such as benzyl cyanides (BzCNs),
benzyl alcohols (BzOHs) and benzyl methyl ether
(BzOMe). When these second components are liquid at
ordinary temperatures, the cocrystallization solvent was
the second component itself. Since both o-methylbenzyl
alcohol (o-MeBzOH) and p-methylbenzyl alcohol
(p-MeBzOH) are solid, acetonitrile was used in these
Scheme 1. Main compounds studied.
Keywords: cyano compounds; electron transfer; crystalline complexes; photochemistry.
* Corresponding author. Tel.: ϩ81-075-753-5670; fax: ϩ81-075-753-5670; e-mail: yito@sbchem.kyoto-u.ac.jp
0040–4020/00/$ - see front matter ᭧ 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00631-1

7140
Y. Ito et al. / Tetrahedron 56 (2000) 7139–7152
Scheme 2. Photoproducts.
cases as the cocrystallization solvent. Crystalline adducts
were formed as colorless crystals in four combinations:
TCNB·BzCN, TCNB·o-MeBzCN, TCNB·BzOH, TCNB·
m-MeBzOH. The former two were 1:2 complexes, while
the latter two were 1:1 complexes. These were characterized
by IR, 1H NMR, and elemental analyses. Their crystal struc-
tures were determined by single crystal X-ray diffraction
(Figs. 1–4).
glass. A novel coupling product 1 was selectively formed
(Table 1: 65% yield, ϳ100% based on reacted TCNB).
1
Other products were not detectable by H NMR. While 1
was nearly stable either as a solid or in a DMSO solution, it
was unstable in solvents like MeOH, EtOH, acetone and
MeCN or in the presence of BzCN, rearranging into an
isoindole derivative 2 readily or in the course of several
days at room temperature. It was sparingly soluble in
ether or insoluble in CHCl3. The structures of 1 and 2
were determined by single crystal X-ray diffraction.3,4
The crystal structure of TCNB·BzCN was already
published.3,4 Two molecules of BzCN and one molecule
of TCNB form a unit cell and are arranged like a
BzCN···TCNB···BzCN sandwich (Fig. 1). There is a center
A probable reaction mechanism is presented in Scheme 3.
Since irradiation through either a Pyrex filter (Ͼ280 nm) or
a solution filter (7 g/L of BiCl3 in 10% HCl, Ͼ355 nm)
produced 1 with comparable efficiencies, the initial step
must be CT excitation of TCNB·BzCN. The diffuse reflect-
ance spectrum of TCNB·BzCN4 is redrawn in Fig. 5. The
CT excitation accompanied by a proton transfer will result
in a hydrogen abstraction by the cyano group of TCNB.
Coupling of the photogenerated radical pair and a subse-
quent 1,3-proton shift will give rise to 1. Although similar
CT excitation and concomitant proton transfer processes
were proposed to occur upon irradiation (at 254 nm) of
TCNB in an acetonitrile or isobutyronitrile solution,11 the
present reaction is the first example of the photoreductive
coupling reaction of the cyano group. The observed high
stability of 1 in the solid state is understandable, because
the amino group and the relevant ortho cyano group are
fixed in an anti relationship in the rigid crystalline environ-
ment. In MeOH, EtOH, acetone, MeCN or BzCN solution,
however, the conformation of 1 will be able to change from
anti to syn, where cyclization to 2 may well occur.13 The
proton transfer seems to be a rate-determining step, since
the irradiation of cocrystal TCNB·BzCN-d2 (BzCN-
d2ˆPhCD2CN) under similar conditions afforded only a
small yield of 1 (Table 1: 3% yield). From the X-ray
analysis, the crystal structures of TCNB·BzCN and
TCNB·BzCN-d2 were found to be isomorphous.
˚
of symmetry at TCNB. The interplane distance is ca. 3.6 A.
The crystal structure of TCNB·o-MeBzCN5 was very simi-
lar to that of TCNB·BzCN. As displayed in Fig. 2, there is a
center of symmetry at TCNB and o-MeBzCN···TCNB···o-
MeBzCN are sandwiched with the interplane distance of ca.
˚
3.7 A.
Fig. 3 shows the crystal structure of TCNB·BzOH. Any
molecules of TCNB and BzOH lie at the crystallographic
inversion center. In other words, the position and orientation
of BzOH is disordered. TCNB and BzOH are stacked alter-
˚
nately along the a axis. The interplane distance is ca. 3.8 A.
The crystal structure of TCNB·m-MeBzOH is shown in
Fig. 4. There is a positional disorder at the alcohol oxygen
atom and the occupancy factors of the O1 and O2 atoms
were assumed to be 50% each. Like TCNB·BzOH, TCNB
and m-MeBzOH are stacked alternately along the a axis.
˚
The interplane distance is ca. 3.8 A.
Solid-state photolyses of cocrystals
Crystals of TCNB·BzCN were crushed and were irradiated
for 20 h with a high-pressure mercury lamp through Pyrex

Y. Ito et al. / Tetrahedron 56 (2000) 7139–7152
7141
Figure 1. Crystal structure of TCNB·BzCN.3
Interatomic distance in crystal is the first measure to
evaluate the solid-state reactivity.2,14,15 The hydrogen
abstraction reaction by the carbonyl or nitro group in the
solid state has been well studied.14,15 It was shown from
these studies that the optimum value of distance for intra-
molecular hydrogen atom abstraction by ketone oxygen was
around the sum of the van der Waals radii for oxygen and
resultant radical pair to occur, because the van der Waals
˚
sum of the N and H atoms is 2.75 A and that of the C and C
˚
˚
atoms 3.4 A. The crystallographic N···H distance (2.77 A)
˚
is close to the van der Waals sum (2.75 A).
Unlike TCNB·BzCN, TCNB·o-MeBzCN was photoinert
under the same reaction conditions (Table 1). Its crystal
structure was found to be similar to that of TCNB·BzCN
(Figs. 1 and 2), but the distance between the cyano nitrogen
of TCNB and the nearest benzylic hydrogen N···H is as long
hydrogen (2.72 A) and the upper limit was 3.1 A.14 For the
cocrystal TCNB·BzCN, the distance between one of the
cyano nitrogens of TCNB and the nearest benzylic hydrogen
˚
˚
˚
˚
˚
of BzCN N···H is 2.77 A and the pertinent C···C distance is
as 3.45 A (the pertinent C···C, 3.95 A) (Fig. 6b). Therefore,
the absence of the short N···H contact may be responsible
for the photoinertness of TCNB·o-MeBzCN.
˚
3.65 A (Fig. 6a). Although the solid-state hydrogen abstrac-
tion by the cyano group is unprecedented, these distances
are probably short enough for the hydrogen abstraction by
the cyano group as well as the radical coupling of the
Upon photolysis of pulverized crystals of TCNB·BzOH or

7142
Y. Ito et al. / Tetrahedron 56 (2000) 7139–7152
Figure 2. Crystal structure of TCNB·o-MeBzCN.5
TCNB·m-MeBzOH, the surfaces of the sample readily
turned into dark brown or fairly brown, respectively. Isola-
tion of the colored substances, however, were unsuccessful
due to their so tiny amounts. Only the oxidation of the
BzOH component to benzaldehyde or m-tolualdehyde,
respectively, could be observed (Table 1). Further irradia-
tion did not appreciably increase the product yields, prob-
ably due to the internal filter effect by the colored substances
Figure 3. Crystal structure of TCNB·BzOH.

Y. Ito et al. / Tetrahedron 56 (2000) 7139–7152
7143
Figure 4. Crystal structure of TCNB·m-MeBzOH.
Table 1. Photolyses of two-component crystals involving TCNB in the solid state: irradiation time, 20 h
Reactants
Products (%)a
PhCHO
Recovered (%)
BzCNs or BzOHS
1
3–7
m-Me-PhCHO
TCNB
TCNB·BzCN (1:2 complex)
TCNB·BzCN-d2 (1:2 complex)
TCNB·o-MeBzCN (1:2
complex)
65
3
0
0
0
0
0
0
35
97
100
66
99
100
No reaction
TCNB·BzOH (1:1 complex)
TCNB·m-MeBzOH (1:1
complex)
0
0
0
0
1
0
0
15
ϳ100
95
99
85
a Yields were estimated by NMR or HPLC and are based on TCNB initially employed.
formed over the crystal surface. Irradiation under an argon
(Table 1) or oxygen stream did not affect results. Inspection
of the crystal structures for TCNB·BzOH and TCNB·m-
MeBzOH showed that the distance between the cyano nitro-
gen of TCNB and the nearest benzylic hydrogen N···H was
In the hope that other unusual photoreactions such as
TCNB·BzCN!1 may be found, considerable efforts were
made to prepare cocrystals of TCNB. Besides those
compounds listed in Scheme 1, p-hydroxybenzyl cyanide,
p-methoxybenzyl cyanide, p-xylylene dicyanide, p-chloro-
benzyl cyanide, p-nitrobenzyl cyanide, benzylamine, benzyl
chloride, ethylbenzene, propylbenzene, cumene, allyl-
benzene, phenylacetylene, 2-phenylethylamine, 2-phenyl-
˚
3.01 and 2.93 A, respectively (Fig. 6c and d). Although
these distances are rough estimation because of their dis-
ordered crystal structures, they appear to be short enough for
the hydrogen transfer to occur (vide supra). The resultant
propionitrile,
3-phenylpropionitrile,
1-phenylethanol,
⅐
ArCH(OH) radical (ArˆC6H5 or m-MeC6H4) will possibly
2-phenylethanol, (2-chloroethyl)benzene, phenylacetalde-
hyde, o-, m-, p-pyridylacetonitrile, and o-, m-, p-pyridyl-
carbinol were tested. As mentioned above, when these
compounds are liquid at ordinary temperatures, they were
decay to generate the observed product ArCHˆO. Because
we were unable to identify the colored products, we can not
tell more about the reaction mechanism.
Scheme 3. Solid-state photolysis of TCNB·BzCN.

7144
Y. Ito et al. / Tetrahedron 56 (2000) 7139–7152
Figure 5. Diffuse reflectance spectra of TCNB and TCNB·BzCN.4
used as the recrystallization solvent. When these are solid,
acetonitrile was used as the recrystallization solvent. In all
cases, however, we failed to obtain crystalline adducts.
trace of TCNB was recovered (the first entry in Table 2). On
the other hand, irradiation of a solution of TCNB (0.039 M)
in non-acid-treated BzCN furnished 6 (31%) and 4a (9%)
along with a quantitative recovery of TCNB (60%) (the
second entry in Table 2).§ Scheme 4 describes a tentative
mechanism for the solution-state photolysis of TCNB in
acid-treated BzCN and the discussion follows.
Solution photolyses
As summarized in Table 2, photolyses of TCNB in solution
by using BzCNs, BzOHs, or BzOMe as the solvent afforded
various condensation products 3–7. The results are very
different from those of the solid-state photolyses (Table
1). The structures of 3a4 and 7 (Fig. 7) were established
by the X-ray analysis. Despite that the products 3–7 appear
to have been formed via complicated steps, material
balances were good in all cases. By contrast, irradiation of
an acetonitrile solution containing TCNB and BzCN (molar
ratio 1:2), TCNB and o-MeBzCN (1:2), TCNB and
m-MeBzCN (1:2), TCNB and p-MeBzCN (1:2), TCNB
and BzOH (1:1), TCNB and m-MeBzOH (1:1), or TCNB
and BzOMe (1:1) (0.1–1 M in each reactants) afforded no or
little reaction products. In the course of these studies, we
found that the photochemistry of TCNB in BzCN was
strongly dependent on acid or amine impurities in BzCN.
Long multistep pathways are considered in Scheme 4. The
first part of the mechanism (TCNBϩBzCN!!A) is essen-
tially the same as the one previously presented.11 The initial
event, i.e. the photochemical hydrogen transfer to TCNB
from BzCN, is probably a common process in view of the
fact that the TCNB excited state is a strong oxidant.11,12,16
Although it seems that intermediates A and B might be
1
isolable, no evidence was obtained by H NMR analysis
for the presence of these primary products. Final products
3a, 4a, and 5 were formed from early stages of irradiation.
Before the product analysis, however, the resultant photoly-
sate was directly evaporated under reduced pressure at high
temperatures (ϳ100ЊC) to remove the solvent BzCN. There-
fore, transformations of some intermediates such as A!B
and B!C may have proceeded thermally in the presence of
acid impurities. In fact, albeit a low yield (12%), a
compound analogous to A was previously isolated as a
final product from the photolysis of TCNB in acetonitrile
or isobutyronitrile.11 The final part of the mechanism
(C!!3aϩ4aϩ5) is highly speculative. However, irradia-
tion of TCNB (0.03 M) in acid-treated BzCN containing
CD3OD or D2O (BzCN/CD3OD or D2Oˆ10:1 v/v) led to
production of 3a–d and 4a–d (quantitative incorporation of
Upon dissolving TCNB in BzCN which was purchased from
several commercial sources (Wako, Nakarai, Tokyo Kasei,
and Aldrich), the solution became dark orange-red in a few
hours. This coloring is probably due to amine impurities in
BzCN.†,‡ In fact, after a sulfuric acid-treatment of com-
mercial BzCN, the solution of TCNB remained colorless
for days and was pale yellow even after weeks.
1
one deuterium from H NMR analyses) as expected from
Irradiation of a solution of TCNB (0.042 M) in acid-treated
BzCN afforded 3a (45%), 5 (18%), and 4a (4%). Only a
this final part of the mechanism. Since the deuterium
incorporation into 5 was not observed, the formation
mechanism for 5 and even the source of its methoxy
group are still unclear.
†
An acetonitrile solution containing TCNB and amine such as benzyla-
mine, 2-phenylethylamine, aniline, and N,N-dimethylaniline, readily
became dark colored.
‡
§
In the previous report,3 we used long-shelved BzCN (Nakarai), which
Two-component crystal TCNB·BzCN was obtained from either acid-
had been left in an old bottle. This BzCN gave results similar to those of
acid-treated BzCN.
treated BzCN or non-acid-treated BzCN. Solid-state photolyses of these
samples, however, gave indistinguishable results.

Y. Ito et al. / Tetrahedron 56 (2000) 7139–7152
7145
Figure 6. The distances between the TCNB nitrogen and the benzylic hydrogens nearby (N···H) as well as the relevant distances (C···C, N···C):
(a) TCNB·BzCN; (b) TCNB·o-MeBzCN; (c) TCNB·BzOH; (d) TCNB·m-MeBzOH.
Table 2. Solution photolyses of TCNB in BzCNs, BzOHs and BzOMe: irradiation time, 15–20 h
Solvent ([TCNB])
Products (%)a
Recovered TCNB (%)
1
3
4
5
Others
BzCN (0.042 M)b
0
0
0
–
–
–
–
–
–
3a, 45
0
0
0
4a, 4
4a, 9
Trace
18
0
0
0
0
0
0
0
64
–
Trace
60
ϳ100
68
76
50
3
15
BzCN (0.041 M)c
6, 31
–
–
x, 3
–
PhCHO, 21
BzCN-d2 (0.034 M)
o-MeBzCN (0.035 M)
m-MeBzCN (0.058 M)
p-MeBzCN (0.034 M)
BzOH (0.015 M)
4b, 9; 4b0, 23
0
4c0, 19
3d, 10
3a, 70
3c, 33
0
4d0, 40
0
0
0
m-MeBzOH (0.05 M)
BzOMe (0.045 M)
7, 22; m-MePhCHO, 30
Ph(OMe)CH–]2, 21
0
a Yields were estimated by NMR or HPLC and are based on TCNB initially employed.
b Acid-treated BzCN was used.
c Non-acid-treated BzCN was used.

7146
Y. Ito et al. / Tetrahedron 56 (2000) 7139–7152
Figure 7. Crystal and molecular structures of compound 7.
Production of 3a in a good yield (70%) from the photolysis
of TCNB (0.015 M) in BzOH (Table 2) can be explained
with a similar route involving the intermediate B, i.e.
TCNBϩBzOH!!B!C!!3a. Formation of 7 (22%) in
addition to 3c (33%) from irradiation of TCNB in
m-MeBzOH is also understandable in terms of a similar
mechanism. The product 7 may be formed through oxida-
tion/hydrolysis of an intermediate like B or C, because D
seems to be a precursor of 7. Irradiation of TCNB (0.045 M)
in BzOMe produced 5 in 64% yield (Table 2). This reaction
can be accommodated by a mechanism similar to the first
part of Scheme 4, where A should correspond to the final
product 5.
Formation of a biphenyl derivative 6 upon photolysis of a
solution of TCNB in non-acid-treated BzCN is somewhat
surprising. At present, we are unable to propose a plausible
mechanism for this unexpected reaction.
Scheme 4. A possible mechanism for the solution-state photolysis of TCNB in acid-treated BzCN.

Y. Ito et al. / Tetrahedron 56 (2000) 7139–7152
7147
Figure 8. Calculated conformations for the cation radicals of o-, m-, and p-MeBzCN.
Photolyses of TCNB in o-, m-, and p-MeBzCN yielded 4b0
(23%), 4c0 (19%), and 4d0 (40%), respectively, as a pre-
dominant product (Table 2). Formation of these products
can be understood by a mechanism similar to Scheme 4. It
is to be noted, however, that the hydrogen transfer to TCNB
occurred from the methyl group rather than the cyanomethyl
group of MeBzCN, although this is unexpected from con-
sideration of the radical stability (vide infra). o-, m-, and
p-MeBzCN employed here were not treated with acid but
the reaction products (4bϩ4b0, xϩ4c0, 3dϩ4d0, respec-
tively) were analogous to those (3aϩ4aϩ5) obtained from
acid-treated BzCN. Biphenyl derivatives corresponding to 6
were not produced, judging from 1H NMR analyses.
Furthermore, coloring of a TCNB solution in o-, m-, or
p-MeBzCN was far less appreciable than in non-acid-
treated BzCN. It seems on the basis of these observations
that unlike BzCN, amine impurities in commercial o-, m-,
and p-MeBzCN are negligible.
coplanar with the benzene ring in each case (C4–C2–C5–
C8ˆϪ10.9, 0.4 and 0.5Њ, respectively), although more
rigorous calculations are required since energy changes
from rotation about the C2–C5 bond are small (Ͻ1 kcal/
mol). With [p-MeC6H4CH2CN]ϩ⅐ (Fig. 8c), a C–H bond to
one of the Me hydrogens (H18) is nearly perpendicular to
the benzene ring, while neither of the bonds to the cyano-
methyl hydrogens (H1 and H3) is. As a result, H18 may be
eliminated more easily than H1 or H3, because an incipient
lone-pair orbital being generated from this proton release
can better overlap with the benzene p-orbital. Similar
consideration based on the orbital overlap may be applic-
able for [o-MeC6H4CH2CN]ϩ⅐ and [m-MeC6H4CH2CN]ϩ⅐.
As seen from Fig. 8a and b, the dihedral angle against the
benzene plane is nearer to the right angle for one of the Me
hydrogens (H18 and H17, respectively) than for the cyano-
methyl hydrogens (H1 and H3 for both of the cation
radicals). Therefore, as with [p-MeC6H4CH2CN]ϩ⅐, the
methyl proton may be transferred more easily than H1 or
H3 in agreement with the experimental data.
As confirmed by MO calculations,k the cyano(tolyl)methyl
⅐
radical NC(o-, m-, or p-MeC6H4)CH is more stable than the
⅐
cyanomethylbenzyl radical o-, m-, or p-NCCH2C6H4CH2 by
In summary, TCNB formed a 1:2 cocrystal with BzCN or
o-MeBzCN and a 1:1 cocrystal with BzOH or m-MeBzOH.
Upon photolysis of the cocrystal TCNB·BzCN in the solid
state, TCNB and BzCN reacted through reductive coupling
of the cyano group of TCNB, leading to a stilbene derivative
1 (65%) as a sole product, which in solution (except in
DMSO) isomerized spontaneously to an isoindole deriva-
tive 2. The cocrystals TCNB·o-MeBzCN, TCNB·BzOH,
and TCNB·m-MeBzOH, on the other hand, did not give
significant products. Crystal structures of these four co-
crystals were determined by single crystal X-ray diffraction
5 kcal/mol. Accordingly, the predominant formation of
4b0 –4d0 instead of 3b–d or 4b–d upon photolysis of
TCNB in MeBzCN (Table 2) cannot be explained from
consideration of the relative stability of radical inter-
mediates. These products may be rationalized by inspection
of the optimized geometry for the cation radicals of
MeBzCN (Fig. 8a–c).k The cyano group tends to be
k
Calculations were carried out by using the AM1 semi-empirical model in
the MacSpartan Plus molecular modeling program.

7148
Y. Ito et al. / Tetrahedron 56 (2000) 7139–7152
and their solid-state photoreactivities were interpreted in
terms of the interatomic distances. In contrast with the
solid-state photolysis, irradiation of TCNB in BzCN
which had been treated with sulfuric acid, produced conden-
sation products 3a (45%), 4a (4%), and 5 (18%). Irradiation
of TCNB in non-acid-treated BzCN afforded a different
product 6 (31%) in addition to 4a (9%). Irradiations of
TCNB in o-MeBzCN, m-MeBzCN, p-MeBzCN, BzOH,
m-MeBzOH, and BzOMe also produced reasonable yields
of condensation products 3, 4, 5, and 7, i.e. the same type of
reactions as in acid-treated BzCN occurred. Probably,
commercial BzCN was contaminated with amine impuri-
ties, and the above solution reactions seem to be very sensi-
tive to acid or base impurities in the solvent.
(13.5 mmol) of potassium tert-butoxide in 60 mL of D2O
was stirred for 8 h at room temperature. Extraction with
ether (60 mL×2), drying with Na2SO4 and then evaporation
of ether gave a,a-dideuterated benzyl cyanide BzCN-d2
(92% D content from 1H NMR). This was further deuterated
by repeating the same procedure and was finally distilled
under reduced pressure to give 5.4 g of BzCN-d2 (97% D
content).
Preparation of two-component crystals
(a) TCNB·BzCN. TCNB (320 mg) was dissolved in hot
BzCN (10 mL).§ After the solution cooled down, colorless
plates soon appeared and were collected by filtration
(740 mg, 100% yield). Since the crystals slowly effloresce
by loss of BzCN and gradually become pale yellow in the
light, they were stored in the dark in a well-capped vial. The
crystalline adduct thus obtained showed no clear mp; solid
TCNB and liquid BzCN separated above 66ЊC and the
TCNB part melted at 273–275ЊC (lit.6 mp of TCNB 270–
272ЊC). The C, H, N analyses (C, 75.82; H, 3.96; N, 19.68)
were consistent with those calculated for (TCNB)(BzCN)2.2
(i.e. C, 76.05; H, 4.02; N, 19.92). The 1H NMR spectrum in
DMSO-d6 was a simple sum of the spectra for TCNB and
BzCN (1:2.3). The IR spectrum in a KBr disc exhibited
strong absorptions at 3113, 3033, 2919, 2254, 2243, 1494,
1451, 1406, 1282, 943, 935, 736, 694 cmϪ1. The peaks at
943 and 935 cmϪ1 were found neither for TCNB nor BzCN.
A CT transition was observed around 367 nm by diffuse
reflectance spectroscopy (Fig. 5)4 and the crystal structure
was determined by X-ray diffraction (Fig. 1).3
Conclusion
Cocrystals will be intriguing treasury for specific unusual
bimolecular reactions which are unlikely to occur in solu-
tion. In addition to the present reaction TCNB·BzCN!1
(!2), we have recently reported: (a) cross photodimeriza-
tion occurring in the hydrogen-bonded 1:2 cocrystal
between fumaric acid and trans-cinnamamide17; and (b)
cross photocycloaddition observed for the crystalline salt
of tryptamine with trans-3-nitrocinnamic acid.18 At this
stage, however, it is difficult to increase the generality of
these reactions, because we do not have enough knowledge
to design cocrystals of desired structure.
Experimental
Likewise, TCNB·BzCN-d2 was obtained as colorless plates;
mp 65–267ЊC (not clear); IR (KBr) 3035, 2255, 2242, 1494,
1447, 1283, 1034, 936, 818, 715 cmϪ1. Its 1H NMR
(DMSO-d6, 200 MHz) was in accord with a mixture of
TCNB and BzCN-d2 (approximately 1:2). The X-ray analy-
sis demonstrated that it was isomorphous with TCNB·
BzCN.
1H NMR spectra were measured on a Varian Gemini-200 or
JEOL JNM-A400 spectrometer. IR spectra were recorded in
a JASCO FTIR-5M or SHIMADZU FTIR-8100A spectro-
meter. Mass spectra were obtained in a JEOL JMS-DX 300
or JEOL JMS-SX 102A spectrometer. Diffuse reflectance
spectra were gathered with a SHIMADZU UV-2400PC
spectrometer equipped with a diffuse-reflectance attach-
ment. HPLC analyses were performed with a SHIMADZU
LC-5A chromatograph by using a Cosmosil 5C18-AR
column (4.6 mm i.d. ×150 mm). Analytical or preparative
TLC was carried out on Merck Silica gel 60 F254 plastic
sheets or 2 mm Merck Silica gel 60 F254 PLC plates, respec-
tively. All the irradiation experiments were carried out by
using a 400 W high-pressure mercury lamp through Pyrex.
Other two-component crystals were similarly prepared and
characterized. Their crystal structures were determined by
single crystal X-ray diffraction.
(b) TCNB·o-MeBzCN. Colorless plates; mp 57–270ЊC (not
clear); IR (KBr) 3114, 3051, 2242, 1485, 1459, 1278, 912,
760 cmϪ1. 1H NMR (DMSO-d6, 200 MHz) and the C, H, N
analyses were in accord with a mixture of TCNB and
o-MeBzCN (approximately 1:2). The crystal structure was
determined by X-ray diffraction (Fig. 2).5
Materials
All reagents were commercially available. BzCNs and
BzOH were distilled under reduced pressure prior to use.
Since commercial BzCN was found to be contaminated with
amines (see text), it was treated with acid. Thus, BzCN
(30 mL) was mixed with 10% aqueous sulfuric acid,
extracted with 10 mL of ethyl ether twice, dried with
Na2SO4, and was then distilled under reduced pressure.
TCNB was recrystallized from EtOH.
(c) TCNB·BzOH. Colorless plates; mp 66–277ЊC (not
clear); IR (KBr) 3460, 3118, 3050, 2247, 1495, 1452,
1
1282, 1196, 1033, 1021, 923, 750, 702 cmϪ1. H NMR
(DMSO-d6, 200 MHz) and the C, H, N analyses were in
accord with a mixture of TCNB and BzOH (approximately
1:1). The crystal structure was determined by X-ray diffrac-
tion (Fig. 3).
Deuteration of benzyl cyanide
(d) TCNB·m-MeBzOH. Colorless prisms; mp 62–267ЊC
(not clear); IR (KBr) 3380, 3115, 3049, 2245, 1487, 1279,
1034, 918, 789, 698 cmϪ1. 1H NMR (DMSO-d6, 200 MHz)
and the C, H, N analyses were in accord with a mixture of
In order to deuterate the benzyl position of BzCN,19
mixture of 10.14 g (86.7 mmol) of BzCN and 1.51 g
a

Y. Ito et al. / Tetrahedron 56 (2000) 7139–7152
7149
TCNB and m-MeBzOH (approximately 1:1). The crystal
structure was determined by X-ray diffraction (Fig. 4).
Irradiation of TCNB·BzCN-d2 under similar conditions for
10 h or 20 h produced 1 in a yield, respectively, only 1/25 or
1
1/15 that of TCNB·BzCN (from H NMR analyses).
Solid-state photolysis of TCNB·BzCN
Solid-state photolyses of TCNB·o-MeBzCN,
TCNB·BzOH, and TCNB·m-MeBzOH
The crystals of TCNB·BzCN (40 mg) were crushed and
spread between two Pyrex plates. This sample was placed
in a photolysis vessel,1 and irradiated for 20 h under argon.
During the irradiation, the photolysis vessel was cooled
from the outside by circulation of cold water (4ЊC). After
the irradiation, the orange photolysate was immediately
evaporated in vacuo at 60ЊC to remove unreacted BzCN.
The residue, where 1 was present as a sole product judging
Solid-state photolyses of TCNB·o-MeBzCN, TCNB·BzOH,
and TCNB·m-MeBzOH were similarly carried out and the
products were analyzed by 1H NMR, TLC (1:1 ether–
CHCl3), and HPLC (2:3 methanol–water). As summarized
in Table 1, no coupling or condensation reactions occurred
in these cases.
1
from H NMR measurement in DMSO-d6 and the conver-
Photolysis of TCNB in BzCN:
sion of TCNB was 65%, was fractionally recrystallized with
MeCN (0.6 mL) to furnish 3 mg of pure 1 as orange-yellow
plates: mpϾ300ЊC (dec). The separation of 1 from
unreacted TCNB was not easy, because 1 often cyclized
into 2 during the work-up. In a separate experiment, the
orange photolysate was directly subjected to fractional
recrystallization with 1 mL of MeCN, followed by further
recrystallization from 10 mL of acetone to afford 6 mg of
2 as orange-yellow needles: mpϾ300ЊC. The structures of
1 and 2 were unequivocally determined by the X-ray
analysis.3,4
(a) A solution of TCNB (112 mg, 0.63 mmol) in 15 mL of
acid-treated BzCN was irradiated for 15 h at 20ЊC under
argon bubbling. The mixture was evaporated under reduced
pressure at ca. 100ЊC to remove BzCN. The resulted dark-
brown viscous residue was analyzed by 1H NMR (DMSO-d6
and CDCl3). Then it was subjected to repeated preparative
TLC (CH2Cl2 and 30–50:1 CH2Cl2–acetone) to afford
72 mg (44%) of 3a, 5 mg (3%) of 4a, 29 mg (17%) of 5,
and a trace of TCNB. Product 3a was recrystallized from
12 mL of 5:1 CH2Cl2–hexane, followed by aqueous MeCN
as yellow needles. Product 4a was recrystallized from ether
as colorless plates, while 5 was recrystallized from 2 mL of
1:1 ether–hexane as white crystals. The structure of 3a was
established by single crystal X-ray analysis.4
a-Amino-b,2,4,5-tetracyanostilbene (1). 1H NMR
(DMSO-d6, 400 MHz) d 7.29–7.34 (1H, m), 7.38 (2H, s,
disappeared on deuteration), 7.45–7.47 (4H, m), 8.72 (1H,
s), 8.98 (1H, s);13C NMR (DMSO-d6, 100 MHz) d 114.31,
114.55, 114.57, 116.64, 116.66, 118.75, 121.19, 127.12
(CH), 127.77 (CH), 129.16 (CH), 132.67, 135.31 (CH),
138.24 (CH), 144.90, 151.80; IR (KBr) 3460, 3362, 2241,
2198, 1622, 1583, 1529, 1493, 1446, 1411, 1275, 925, 915,
770, 704 cmϪ1; MS (EI) m/z (rel intensity) 295 (92, Mϩ),
294 (100); UV/VIS (DMSO) l 435 (e 2500), 292
(14200) nm.
2-Benzyl-4,5-dicyanobenzamide (3a). Mp 182–183ЊC; 1H
NMR (CDCl3, 400 MHz) d 4.28 (2H, s), 5.72 and 5.77 (2H,
two br s, NH2, disappeared on deuteration), 7.14 (2H, finely
split d, Jˆ7.2 Hz), 7.27–7.36 (3H, m), 7.61 (1H, finely split
(0.4 Hz) s, 7.86 (1H, s);13C NMR (CDCl3, 100 MHz) d
38.68 (CH2), 113.90, 114.62, 114.71 and 117.22, 127.41
(CH), 129.11 (CH), 129.19 (CH), 131.90 (CH), 135.77
(CH), 137.41, 139.93, 146.08, 167.41; IR (KBr) 3434,
2235, 1661, 1612, 1406, 742, 701 cmϪ1; MS (EI) m/z (rel
intensity) 261 (25, Mϩ), 244 (100), 215 (35); HRMS calcd
for C16H11ON3 261.0902, found 261.0877. All the NMR
signals were assigned and were confirmed by NOEDF and
2D NMR (NOESY, HMQC and HMBC) measurements.4
1-Amino-3-(a-cyanobenzylidene)-5,6-dicyano-3H-iso-
1
indole (2). H NMR (DMSO-d6, 400 MHz) d 7.40 (1H, t,
Jˆ7.5 Hz), 7.49 (2H, t, Jˆ7.8 Hz), 8.20 (2H, d, Jˆ7.8 Hz),
8.65 (1H, s), 8.81 (1H, s), 8.88 (1H, s, disappeared on
deuteration), 9.19 (1H, s, disappeared on deuteration);13C
NMR (DMSO-d6, 100 MHz) d 115.06, 115.82, 116.08,
116.15, 119.55, 126.70 (CH), 127.46 (CH), 128.55 (CH),
128.81 (CH), 129.63 (CH), 133.00, 136.54, 142.85,
155.95, 165.47, 165.56; IR (KBr) 3464, 2235, 2205, 1671,
1528, 1444, 1392, 1324, 1272, 1194, 772, 697 cmϪ1; MS
(EI) m/z (rel intensity) 295 (87, Mϩ), 294 (100); HRMS
calcd for C18H9N5 295.0858, found 295.0845; UV/VIS
(MeCN) l 429 (e 13300), 278 (9800), 239 (25500), 206
(19400) nm; UV/VIS (DMSO) l 446 (e 24100), 285
(19700) nm.
1
1-Benzyl-2,4,5-tricyanobenzene (4a). Mp 151–153ЊC; H
NMR (CDCl3, 200 MHz) d 4.31 (2H, s), 7.16–7.4 (5H, m),
7.65 (1H, s), 8.06 (1H, s); IR (KBr) 2241, 1600, 1496, 1486,
1458, 1446, 907, 749, 733, 705 cmϪ1; MS (EI) m/z (rel
intensity) 243 (100, Mϩ), 116 (45); HRMS calcd for
C16H9N3 243.0796, found 243.0809.
1,2,4-Tricyano-5-(a-methoxybenzyl)benzene (5). Mp 135–
137ЊC; 1H NMR (CDCl3, 400 MHz) d 3.42 (3H, s, Hd), 5.61
(1H, s, Hc), 7.35–7.44 (5H, m, He–Hg), 8.00 (1H, d,
Jˆ0.6 Hz, Ha), 8.19 (1H, t, Jˆ0.6 Hz, Hb);13C NMR
(CDCl3, 100 MHz) d 57.31 (CH3), 81.80 (CH), 113.54,
114.18, 114.34, 115.48 and 115.86, 119.73, 127.46 (CH),
129.32 (CH), 129.45 (CH), 132.13 (CH), 136.97 (CH),
137.04, 152.19; IR (film) 2240, 1600, 1486, 1453, 1194,
1098, 913, 736, 701 cmϪ1; MS (EI) m/z (rel intensity) 273
(40, Mϩ), 258 (100), 242 (86), 215 (51), 180 (30), 121 (71);
HRMS calcd for C17H11ON3 273.0902, found 273.0914.
Incidentally, a solid mixture, which was obtained by grind-
ing 36 mg (0.20 mmol) of TCNB and 47 mg (0.40 mmol) of
BzCN in an agate mortar and pestle for 10 min, showed the
characteristic IR peaks of TCNB·BzCN at 943 and
935 cmϪ1. This solid mixture (15 mg) was irradiated for
15 h. The NMR analysis of the resulting orange photolysate
indicated that it was a mixture of 2, TCNB and BzCN and
that 33% of TCNB was converted to 2.

7150
Y. Ito et al. / Tetrahedron 56 (2000) 7139–7152
1
The structure was confirmed by NOEDF measurements as
well as by gated decoupling and selective decoupling
experiments.4
160ЊC; H NMR (CDCl3, 200 MHz) d 2.17 (3H, s), 4.28
(2H, s), 7.03 (1H, d, Jˆ8 Hz), 7.22–7.33 (3H, m), 7.36 (1H,
s), 8.06 (1H, s); IR (film) 2241, 1599, 1493, 1460, 1379,
935, 743 cmϪ1; MS (EI) m/z (rel intensity) 257 (100, Mϩ),
129 (95), 91 (51); HRMS calcd for C17H11N3 257.0953,
found 257.0937.
(b) A solution of TCNB (234 mg, 1.31 mmol) in 32 mL of
non-acid-treated BzCN was irradiated for 18 h at 20ЊC
under argon bubbling. The mixture was evaporated under
reduced pressure at ca. 100ЊC. The residue was analyzed by
1H NMR (DMSO-d6 and CDCl3) and then separated by
preparative TLC (4:1 CHCl3–ether) to afford 19 mg (6%)
of 4a, 63 mg (18%) of 6, and 122 mg (52%) of TCNB.
Product 6 was recrystallized from methanol to afford pale
brown needles.
1,2,4-Tricyano-5-[o-(cyanomethyl)benzyl]benzene (4b0).
1H NMR (CDCl3, 200 MHz) d 3.67 (2H, s), 4.34 (2H, s),
7.04 (1H, d, Jˆ7 Hz), 7.38–7.48 (3H, m), 7.46 (1H, s), 8.09
(1H, s); MS (EI) m/z (rel intensity) 282 (100, Mϩ), 254 (79),
242 (44), 228 (44), 116 (39); HRMS calcd for C18H10N4
282.0905, found 282.0901.
2,4,5-Tricyano-30-(cyanomethyl)biphenyl (6). Mp 164–
Photolysis of TCNB in m-MeBzCN
1
167ЊC; H NMR (CDCl3, 200 MHz) d 3.87 (2H, s), 7.51–
1
7.63 (4H, m), 7.97 (1H, s), 8.18 (1H, s); H NMR (C6D6,
A solution of TCNB (366 mg, 2.06 mmol) in 35 mL of
m-MeBzCN was irradiated for 20 h at 20ЊC under argon
bubbling. The mixture was evaporated under reduced pres-
400 MHz) d 2.65 (2H, s), 6.38 (1H, finely split s), 6.44 (1H,
finely split s), 6.73–6.78 (3H, m or finely split s at d 6.73
(ϳ0.5H), 6.75 (ϳ1H) and 6.77 (ϳ1.5H)), 6.88 (1H, t,
Jˆ7.6 Hz); IR (neat film) 2241, 1597, 1537, 1488, 1431,
1266, 918, 803, 710 cmϪ1; MS (EI) m/z (rel intensity) 268
(100, Mϩ), 178 (62); HRMS calcd for C17H8N4 268.0749,
found 268.0758. 2D NMR (HSQC) was measured to
confirm the meta substitution.
1
sure at ca. 100ЊC. The residue was analyzed by H NMR
(DMSO-d6 and CDCl3) and then separated by preparative
TLC (4:1 CHCl3–ether) to afford 78 mg (14%) of 4c0 and
190 mg (52%) of TCNB along with one uncharacterized
product x (15 mg) whose tentative structures are shown in
Scheme 2. 4c0 was recrystallized from ether–CH2Cl2 (1:1)
and x was recrystallized from ether–acetone (9:1).
(c) A solution of TCNB (6 mg, 0.034 mmol) in 1 mL of
BzCN-d2 was irradiated for 16 h at 20ЊC under argon
bubbling. The mixture was evaporated under reduced
1,2,4-Tricyano-5-[m-(cyanomethyl)benzyl]benzene (4c0).
Mp 138–139ЊC; 1H NMR (CDCl3, 200 MHz) d 3.76 (2H, s),
4.30 (2H, s), 7.17 (1H, d, Jˆ8 Hz), 7.18 (1H, s), 7.31 (1H, d,
Jˆ8 Hz), 7.40 (1H, t, Jˆ8 Hz), 7.62 (1H, s), 8.06 (1H, s); IR
(film) 2239, 1610, 1489, 1441, 1400, 914, 764 cmϪ1; MS
(EI) m/z (rel intensity) 282 (88, Mϩ), 281 (100), 254 (38),
116 (21); HRMS calcd for C18H10N4 282.0905, found
282.0898.
1
pressure at ca. 100ЊC. The H NMR analysis of the residue
indicated little occurrence of the reaction.
(d) Irradiation of TCNB in BzCN containing CD3OD or
D2O. A solution of TCNB (12 mg, 0.069 mmol) in 2 mL
of acid-treated BzCN containing 0.2 mL of CD3OD
(Aldrich, 99.8% D) was irradiated for 6 h at 20ЊC under
argon bubbling. The mixture was evaporated under reduced
x. Mp 148–150ЊC; 1H NMR (CDCl3, 200 MHz) d 2.35 (3H,
s), 4.21 (2H, s), 6.93–7.35 (4H, m), 7.62 (1H, s), 8.04 (1H,
s); IR (film) 3437, 2239, 1654, 1605, 1487, 1387, 1052, 909,
836, 734 cmϪ1; MS (EI) m/z (rel intensity) 293 (18, Mϩ),
292 (35), 291 (54), 290 (73), 257 (50), 256 (100); HRMS
(for MϩϪ2H2O) calcd for C17H11N3 257.0953, found
257.0942.
1
pressure at ca. 100ЊC. The H NMR analysis of the residue
(DMSO-d6 and CDCl3) indicated the presence of 3a–d,
4a–d, 5, and TCNB in molar ratios 0.8:0.3:1:0.2. Approxi-
mately one atom of deuterium incorporation was observed
for 3a and 4a at their benzylic positions, but no deuterium
incorporation was observed for 5. A solution of TCNB
(11 mg, 0.062 mmol) in 2 mL of acid-treated BzCN
containing 0.2 mL of D2O (Aldrich, 99.9% D) was
irradiated for 5 h at 20ЊC under argon bubbling. The mixture
was evaporated under reduced pressure at ca. 100ЊC. The 1H
NMR analysis of the residue gave similar results.
Photolysis of TCNB in p-MeBzCN
A solution of TCNB (60 mg, 0.34 mmol) in 10 mL of
p-MeBzCN was irradiated for 20 h at 20ЊC under argon
bubbling. The mixture was evaporated under reduced pres-
Photolysis of TCNB in o-MeBzCN
1
sure at ca. 100ЊC. The residue was analyzed by H NMR
A solution of TCNB (190 mg, 1.07 mmol) in 30 mL of
o-MeBzCN was irradiated for 18 h at 20ЊC under argon
bubbling. The mixture was evaporated under reduced pres-
(DMSO-d6 and CDCl3) and then separated by preparative
TLC (4:1 CHCl3–ether) to afford 7 mg (8%) of 3d, 32 mg
(34%) of 4d0, and 27 mg (45%) of TCNB. 3d and 4d0 were
purified by recrystallization from MeCN–water (3:1) and
ether–MeCN (10:1), respectively.
1
sure at ca. 100ЊC. The residue was analyzed by H NMR
(DMSO-d6 and CDCl3) and then separated by preparative
TLC (CHCl3 and 100:1 CHCl3–acetone) to afford 19 mg
(7%) of 4b, 63 mg (21%) of crude 4b0, and 101 mg (53%)
of TCNB. Because of the instability on silica gel, 4b
could not be purified enough. 4b was recrystallized from
methanol.
4,5-Dicyano-2-(p-tolylmethyl)benzamide (3d). Mp 150–
1
155ЊC; H NMR (CDCl3, 200 MHz) d 2.31 (3H, s), 4.21
(2H, s), 5.71 (2H, br s, NH2), 7.00 and 7.12 (4H, AA0BB0 q,
Jˆ7.9 Hz), 7.58 (1H, s), 7.84 (1H, s); IR (KBr) 3676
(strong), 2234, 1660 (strong), 1599, 1389, 921, 904,
803 cmϪ1; MS (EI) m/z (rel intensity) 275 (14, Mϩ), 258
1,2,4-Tricyano-5-(o-methylbenzyl)benzene (4b). Mp 155–

Y. Ito et al. / Tetrahedron 56 (2000) 7139–7152
7151
(100), 243 (18), 215 (25); HRMS calcd for C17H13ON3
275.1059, found 275.1056.
NMR (DMSO-d6 and CDCl3) and was then separated by
preparative TLC (CH2Cl2) to give 56 mg (46%) of 5 and
23 mg (21%) of meso-Ph(MeO)CH–]2. The latter product
was recrystallized from methanol–water (9:1) as yellow
crystals.
1,2,4-Tricyano-5-[p-(cyanomethyl)benzyl]benzene (4d0).
1
Mp 146–147ЊC; H NMR (CDCl3, 200 MHz) d 3.75 (2H,
s), 4.29 (2H, s), 7.21 and 7.35 (4H, AA0BB0 q, Jˆ8.4 Hz),
7.63 (1H, s), 8.05 (1H, s); IR (KBr) 2240, 1516, 1489,
1425,1384,922 cmϪ1; MS (EI) m/z (rel intensity) 282
(100, Mϩ), 255 (44), 240 (19), 116 (43); HRMS calcd for
C18H10N4 282.0905, found 282.0919.
(1R,2S)-1,2-Dimethoxy-1,2-diphenylethane Ph(MeO)CH–]2
1
(meso). Mp 141.5–145ЊC (lit.20 mp 144–145ЊC); H NMR
(CDCl3, 200 MHz) d 3.14 (6H, s), 4.29 (2H, s), 7.12–7.18
(4H, m), 7.23–7.29 (6H, m); IR (neat film) 1177, 1106, 941,
755, 697 cmϪ1; MS (EI) m/z (rel intensity) 121 (100, Mϩ/2), 77
(13); HRMS calcd for C8H9O1 (ˆMϩ/2) 121.0653, found
121.0661.
Photolysis of TCNB in BzOH
A solution of TCNB (104 mg, 0.58 mmol) in 40 mL of
BzOH was irradiated for 20 h at 20ЊC under argon bubbling.
The mixture was evaporated under reduced pressure at ca.
X-Ray crystal structure determination
1
100ЊC. The residue was analyzed by H NMR (DMSO-d6
X-Ray intensities were measured on Rigaku AFC-5 or AFC-
7R diffractometer with graphite monochromatized MoKa
and CDCl3). Then, 107 mg (70%) of 3a was isolated with
preparative TLC (CHCl3–acetone). The yield of PhCHO
was determined by HPLC (2:3 methanol–water) before
evaporating the photolysate.
˚
radiation (lˆ0.71073 A). The structures were solved by
direct methods and were refined using texsan software.21
ꢀ
Crystal data. TCNB·BzCN: triclinic P1; aˆ7.767(1),
˚
Photolysis of TCNB in m-MeBzOH
bˆ10.957(1), cˆ7.682(2) A, aˆ93.80(1)Њ, bˆ117.61(1)Њ,
3
gˆ103.05(1)Њ, Vˆ553.2(2) A , Zˆ1, Dxˆ1.238 g/cm3,
˚
Rˆ0.053.3 TCNB·o-MeBzCN: triclinic P1, aˆ8.387(1),
ꢀ
A solution of TCNB (45 mg, 0.25 mmol) in 5 mL of
m-MeBzOH was irradiated for 20 h at 20ЊC under argon
bubbling. The mixture was evaporated under reduced pres-
˚
˚
bˆ10.780(1), cˆ7.206(1) A, aˆ103.40(1)Њ, bˆ109.24(1)Њ,
3
3
gˆ79.12(1)Њ, Vˆ593.9(1) A , Zˆ1, Dxˆ1.232 g/cm , Rˆ
1
0.057.5 TCNB·BzOH: monoclinic P21/n, aˆ7.552(1), bˆ
sure at ca. 100ЊC. The residue was analyzed by H NMR
3
˚
˚
(DMSO-d6 and CDCl3). Then, by using preparative TLC
(10:1 CH2Cl2–acetone), 16 mg (23%) of 3c and 13 mg
(18%) of 7 were isolated. Compounds 3c and 7 were
recrystallized from MeCN–water or benzene–acetone–
ether, respectively. m-MePhCHO was analyzed with
HPLC (2:3 methanol–water) before evaporating the
photolysate. The structure of 7 was established by the single
crystal X-ray analysis. It crystallized with one molecule of
benzene included (Fig. 7).
7.182(2), cˆ14.356(1) A, bˆ101.76(1)Њ, Vˆ762.3(2) A ,
Zˆ2, Dxˆ1.247 g/cm3, Rˆ0.066. TCNB·m-MeBzOH: tri-
ꢀ
˚
clinic P1; aˆ7.475(3), bˆ14.930(4), cˆ7.400(2) A,
3
˚
aˆ99.24(2)Њ, bˆ90.76(2)Њ, gˆ102.72(3)Њ, Vˆ794.1(5) A ,
3
ꢀ
Zˆ2, Dxˆ1.256 g/cm , Rˆ0.082. 7: triclinic P1, aˆ9.605(2),
˚
bˆ13.072(3), cˆ9.406(3) A, aˆ100.37(2)Њ, bˆ114.94(2)Њ,
3
3
˚
gˆ100.50(2)Њ, Vˆ1008.3(5) A , Zˆ2, Dxˆ1.210 g/cm ,
Rˆ0.066.
4,5-Dicyano-2-(m-tolylmethyl)benzamide (3c). Mp 198–
198.5ЊC; H NMR (CDCl3, 400 MHz) d 2.31 (3H, s), 4.21
Acknowledgements
1
(2H, s), 5.68 and 5.76 (2H, two br s, NH2), 6.90 (1H, d,
Jˆ8 Hz), 6.92 (1H, s), 7.07 (1H, d, Jˆ8 Hz), 7.20 (1H, t,
Jˆ8 Hz), 7.59 (1H, s), 7.85 (1H, s); IR (film) 3397, 2235,
1667 (strong), 1607, 1590, 1403, 917 cmϪ1; MS (EI) m/z
(rel intensity) 275 (19, Mϩ), 258 (100), 243 (16), 215 (26);
HRMS calcd for C17H13ON3 275.1059, found 275.1059.
This work was supported by the Grant-in-Aid for Scientific
Research from the Japanese Government (Nos. 08221213,
10440215, 10132229).
References
5,6-Dicyano-3-hydroxy-3-(m-tolyl)isoindolin-1-one (7).
Mp 194–195.5ЊC; 1H NMR (CDCl3, 200 MHz) d 2.35
(3H, s), 3.55 (1H, broad s), 6.96 (1H, broad s), 7.20–7.23
(1H, m), 7.27–7.35 (3H, m), 7.77 (1H, s), 8.15 (1H, s); IR
(film) 2239, 1717, 1692, 1424, 1303, 1159, 1082, 1067,
759 cmϪ1; MS (EI) m/z (rel intensity) 289 (33, Mϩ), 272
(55), 271 (53), 256 (100), 198 (25), 180 (31), 119 (33), 91
(27); HRMS calcd for C17H11O2N3 289.0851, found
289.0849.
1. Ito, Y. Synthesis 1998, 1–32.
2. Ito, Y. Molecular and Supramolecular Photochemistry 1999, 3,
1–70.
3. Ito, Y.; Endo, S.; Ohba, S. J. Am. Chem. Soc. 1997, 119, 5974–
5975.
4. Supplementary materials in Ref. 3.
5. Hosomi, H.; Ohba, S.; Ito, Y.; Nakabayashi, H. Acta Crystallogr.
C 1997, 53, IUC 9700031.
6. Biley, A. S.; Henn, B. R.; Langdon, J. M. Tetrahedron 1963, 19,
161–167.
Photolysis of TCNB in BzOMe
7. Agostini, G.; Corbaja, C.; Giacometti, G.; Pasimeni, L.;
Clemente, D. A. J. Phys. Chem. 1988, 92, 997–1003.
8. Lefebvre, J.; Miniewicz, A.; Kowal, R. Acta Crystallogr. C
1989, 45, 1372–1376.
9. Pasimeni, L.; Guella, G.; Corvaja, C.; Clemente, D. A.;
Vicentini, M. Mol. Cryst. Liq. Cryst. 1983, 91, 25–38.
A solution of TCNB (80 mg, 0.45 mmol) in 10 mL of
BzOMe was irradiated for 16 h at 20ЊC under argon
bubbling. The mixture was evaporated under reduced pres-
1
sure at ca. 100ЊC. The brown residue was analyzed by H

7152
Y. Ito et al. / Tetrahedron 56 (2000) 7139–7152
10. Unpublished results.
11. Tsujimoto, K.; Abe, K.; Ohashi, M. J. Chem. Soc., Chem.
Commun. 1983, 984–985.
12. Tsujimoto, K.; Nakao, N.; Ohashi, M. J. Chem. Soc., Chem.
Commun. 1992, 366–367.
17. Ito, Y.; Hosomi, H.; Ohba, S. Tetrahedron 2000, 36, 6833–
6844.
18. Ito, Y.; Fujita, H. Chem. Lett. 2000, 36, 6833–6844.
19. Nishiyama, K.; Oda, J.; Inouye, Y. Bull. Inst. Chem. Res.
Kyoto. Univ. 1974, 52, 499–505.
13. Meyers, A. I.; Sircar, J. C. The Chemistry of the Cyano Group;
Rappoport, Z. Ed.; Wiley-Interscience: New York, 1970; pp 400–
404.
14. Ihmels, H.; Scheffer, J. R. Tetrahedron 1999, 55, 885–907.
15. Ramamurthy, V.; Venkatesan, K. Chem. Rev. 1987, 87, 433–481.
16. Mella, M.; Fagnoni, M.; Freccero, M.; Fasani, E.; Albini, A.
Chem. Soc. Rev. 1998, 81–89.
20. Merz, A.; Eichner, M.; Tomahogh, R. Liebigs Ann Chem.
1981, 1774–1784.
21. Molecular Structure Corporation. texsan, Single Crystal
Structure Analysis Software. Version 1.10. MSC, 3200 Research
Forest Drive, The Woodlands, TX 77381, USA, 1999.