Solvent-dependent effect by carbon dioxide on the photoreactions of (9-anthryl)alkylamines

Article abstract of DOI:10.1016/j.tet.2007.09.069

The effect of CO₂ on a photoreaction was first studied systematically by using (9-anthryl)alkylamines (APA, AEA, and AMA) as the starting compound. From close scrutiny of the results, the CO₂ effect was clearly observed and was well rationalized by the previously reported novel solvent dependence of the amine-CO₂ reversible reactions. For instance, the yield of the dimer (h-t from APA or AEA, h-t+h-h from AMA) obtained in MeOH or DMSO was higher under CO₂ than under argon and this was ascribed to formation of either ammonium bicarbonate/carbonate in MeOH or carbamic acid in DMSO, which will prevent the nitrogen lone pair from being involved in electron-transfer reactions. In fact, the electron-transfer side reactions producing 1-3 in DMSO were strongly inhibited under CO₂. Also, due to formation of noncovalent linkage between the ammonium cation and the carbamate anion in 2-PrOH, the proportion of h-h relative to h-t produced from AMA in 2-PrOH was increased by carrying out the reaction under CO₂.

Full text of DOI:10.1016/j.tet.2007.09.069

Tetrahedron 63 (2007) 12286–12293
Solvent-dependent effect by carbon dioxide
on the photoreactions of
(9-anthryl)alkylamines
*
Masahiro Horiguchi and Yoshikatsu Ito
Department of Synthetic Chemistry and Biological Chemistry, Graduate School of Engineering,
Kyoto University, Katsura, Kyoto 615-8510, Japan
Received 5 September 2007; revised 27 September 2007; accepted 27 September 2007
Available online 29 September 2007
2
Abstract—The effect of CO on a photoreaction was first studied systematically by using (9-anthryl)alkylamines (APA, AEA, and AMA) as
the starting compound. From close scrutiny of the results, the CO effect was clearly observed and was well rationalized by the previously
2
reported novel solvent dependence of the amine–CO reversible reactions. For instance, the yield of the dimer (h–t from APA or AEA,
2
2
h–t+h–h from AMA) obtained in MeOH or DMSO was higher under CO than under argon and this was ascribed to formation of either
ammonium bicarbonate/carbonate in MeOH or carbamic acid in DMSO, which will prevent the nitrogen lone pair from being involved in
electron-transfer reactions. In fact, the electron-transfer side reactions producing 1–3 in DMSO were strongly inhibited under CO . Also,
due to formation of noncovalent linkage between the ammonium cation and the carbamate anion in 2-PrOH, the proportion of h–h relative
2
to h–t produced from AMA in 2-PrOH was increased by carrying out the reaction under CO
Ó 2007 Elsevier Ltd. All rights reserved.
2
.
ꢀ
þ
1
. Introduction
RNHCO
2
H þ RNH
2
%RNHCO RNH
ð2Þ
ð3Þ
2
3
Many amines react with carbon dioxide to form carbamic
acids (Eq. 1). The acids are unstable and easily dissociate
back to amines and CO , or react with another amine mole-
2
cule to form ammonium carbamates (Eqs. 1 and 2). In aque-
ous solution, ammonium bicarbonates/carbonates are also
formed (Eq. 3). These reactions are reversible and have
recently been utilized for preparation of a variety of revers-
ible supramolecular materials such as switchable solvent
systems, switchable surfactants, switchable polymeric
hosts, reversible organogels, and reversible organosilicas.
They have also been used for controlling the underlying
thermal reactions, e.g., enhancement of the reaction rate
and the product selectivity.
1
þ
ꢀ
RNH
2
þ CO
2
þ H
2
3
O%RNH HCO
3
For the past few years we have studied the solvent depen-
dence of the formation of carbamic acid species (i.e., car-
bamic acid and ammonium carbamate) from particular
amines and CO : for our previous papers on the amine–
2
CO reaction system, see Ref. 8. In protophilic, dipolar,
2
2
3
4
5
6
7
aprotic solvent such as DMSO, DMF, pyridine, or dioxane
(unlike in water), the equilibrium of Eq. 2 lay so far to the
left for the amines studied.
8
a,b,e
Concurrently, we attemp-
ted to apply this solvent dependence to manage photo-
chemical reactions. Now, we wish to report the effect of
CO2 on the photoreactions of u-(9-anthryl)alkylamines.
The original aim was to control their photodimerization
regioselectivity by using the ammonium carbamate ionic
linkage as a noncovalent linker (see the structure A, which
is proposed in Scheme 5). Although we could achieve
only a limited success for this purpose, the overall CO2
effects observed here were consistent with the previously
RNH
2
þ CO
2
%RNHCO
2
H
ð1Þ
Keywords: Carbon dioxide; Anthryl amine; Carbamic acid; Ammonium
carbamate; Solvent dependence; Photoreaction; Reaction control.
8
a,b
reported novel solvent dependence in the amine–
CO reversible reactions (Eqs. 1–3). To our knowledge,
*
Corresponding author. Tel./fax: +81 75 383 2718; e-mail: yito@sbchem.
kyoto-u.ac.jp
2
0
040–4020/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2007.09.069

M. Horiguchi, Y. Ito / Tetrahedron 63 (2007) 12286–12293
12287
NH2
NHCO H
2
α
β
γ
CO2
in DMSO
APA
carbamic acid (100%)
in MeOH containing
in 2-PrOH
H O as impurity
CO2
CO2
2
NHCO2– +H N
3
NH3+ HCO3– and/or CO32–
α
β
α'
β'
α
β
γ
γ'
γ
ammonium carbamate (100%)
ammonium bicarbonate or carbonate
2
reaction.
Scheme 1. Solvent dependence of the 3-(9-anthryl)propylamine (APA)–CO
the photochemistry of the carbamic acid species has not
been systematically studied so far.
was observed, i.e., formation of carbamic acid in DMSO, am-
monium carbamate in 2-PrOH, and ammonium bicarbonate/
carbonate in MeOH (Scheme 1). In diethyl ether, the ammo-
nium carbamate precipitated as a pale yellow solid. The main
features of the NMR and IR spectra that were obtained after
y
2
. Results and discussion
the CO bubbling, are given below ((a)–(c)). These are
2
8
a
2
.1. Reaction of APA with CO2
exactly analogous with those found for NPA.
We have already performed detailed investigations about the
solvent dependence of the formation of carbamic acid spe-
cies from u-(1-naphthyl)alkylamines. The results were
obtained mainly on the basis of NMR and IR analyses before
(a) In DMSO. The presence of only three signals corre-
sponding to the three methylene groups could be seen
8
a,b
1
13
from the H and C NMR spectra (i.e., the carbamic
acid yield is 100%). The a-methylene proton was con-
siderably shifted to a lower field upon bubbling of
and after bubbling of CO through the amine solution. In
2
a DMSO solution, 3-(1-naphthyl)propylamine (NPA) was
exclusively converted to the corresponding carbamic acid,
whereas in 2-PrOH or in MeOH it was converted to the
ammonium carbamate or to the ammonium bicarbonate/
CO (Dd 0.43 ppm) and resonanced as a quasi-quartet
2
at d 3.18. The NH proton appeared as a broad triplet at
d 6.97. The carboxy carbon of the carbamic acid
appeared at d 157.0 ppm. An HMBC cross peak was ob-
served between the a-methylene proton and the carboxy
8
a
carbonate, respectively.
carbon. The stretching frequency n
of NCOOH was
C]O
ꢀ
1
Here the same experiment was repeated for 3-(9-anthryl)pro-
pylamine (APA) and exactly the parallel solvent dependence
1700 cm
(b) In 2-PrOH. There were six methylene-proton signals
.
1
13
and six methylene-carbon signals in the H and
C
NMR spectra. This corresponds to the formation of
the carbamate anion and the ammonium cation. The
y
The decarboxylation of several N-aralkyl- and N-arylcarbamic acids,
0
a-methylene proton appeared at d 3.33 and the a -meth-
ylene proton appeared at d 3.00. H NMR can neither
which are formed in situ from the corresponding amines in CO
2
-saturated
was investigated by us. It was promoted by irradiation
through Pyrex at room temperature) and an especially large acceleration
8
a,b
8e
1
DMSO-d
6
,
(
distinguish between the carbamate anion and the coexis-
tent carbamic acid, nor between the ammonium cation
was observed for certain N-arylcarbamic acids. For example, the indoli-
ne-derived carbamic acid 4 was decarboxylated in 85% yield into indoline
8
a
and the coexistent free amine. However, the observed
six methylene-proton signals were almost equal in inten-
sity. Hence, the yield of the ammonium carbamate may
be nearly 100%. A broad signal of the carboxy carbon of
the carbamate anion resonanced at d 160.5 ppm. There
was an HMBC cross peak between the a-methylene pro-
ton and the carboxy carbon.
(
5) after 2 h of irradiation (Eq. 4). In the dark, the decarboxylation was
negligible (Eq. 5). The slow decarboxylation in the dark is plausible, be-
cause it is an endergonic reaction (DG¼9 kcal/mol) as mentioned below.
Unlike N-aralkylamines such as APA (Scheme 1) and 3-(1-naphthyl)pro-
pylamine (NPA), N-arylamines can be converted only partly to the car-
bamic acid in CO The
content of 4, however, increased at lower temperatures and reached nearly
8
a,b
2
-saturated DMSO-d
6
, e.g., 4 vs 5¼22:78.
ꢁ
1
13
1
1
00% at ꢀ52 C in CO
2 7
-saturated DMF-d (see H and C NMR in Figs.
and 2). The temperature dependence of [5]/[4] (Fig. 1) was thermody-
(
c) In MeOH. Like in DMSO, three methylene signals were
namically analyzed (Fig. 3) and the free energy change DG for the reac-
tion 4%5+CO was estimated as +9 kcal/mol. The theoretical evaluation
of DG for NH COOH%NH +CO afforded +10.5 kcal/mol.
2 3 2
1
13
observed by the H and C NMR measurements. The
downfield shift of the a-methylene proton was relatively
small (Dd 0.29 ppm) and it resonanced at d 3.12. The
formation of bicarbonate or carbonate was indicated
2
9,1
h , 2 h
in DMSO
5%
+
CO2
(4)
(5)
1
3
N
H
by the C NMR peak at d 161.3 ppm and by the strong
IR bands at 1636 and 1310 cm . However, the amount
8
ꢀ1
5
N
CO H
of the residual free amine cannot be estimated, because
it is indistinguishable from the ammonium species by
H NMR.
2
4
dark, 2 h
in DMSO
negligible decarboxylation
1
8a

1
2288
M. Horiguchi, Y. Ito / Tetrahedron 63 (2007) 12286–12293
R
APA
h
R = (CH ) NH
2
3
2
Ar or CO2
R
h-t
R
NH
AEA
h
h-t
+
+
R = (CH ) NH
2
2
2
Ar or CO2
1
AMA
h
h-t
R
+
R = CH NH
2
2
Ar or CO2
R
h-h
+
2
3
2
Scheme 2. Photoreactions of (9-anthryl)alkylamines under argon or CO .
2
.2. Photoreactions of (9-anthryl)alkylamines
mechanism of this interesting cyclization reaction is not elu-
cidated, a similar reaction occurred for some u-arylalkyl-
amines by photosensitization with some electron acceptors
The anthryl chromophore generally undergoes rapid photo-
With this in mind, the effect of CO on
1
0,11
12
dimerization.
such as 1,4-dicyanonaphthalene or 1,3-dicyanobenzene.
2
the photodimerization reactivity of APA, 2-(9-anthryl)ethyl-
amine (AEA), and (9-anthryl)methylamine (AMA) was
examined. The solvent employed is MeOH, DMSO, or 2-
PrOH. The irradiation was carried out with a 400-W high
pressure mercury lamp (Pyrex) under argon or CO2.
In the case of AMA, small amounts of lepidopterene (2)
and biplanene (3) were obtained. These byproducts 2 and
3 may be formed through the competitive C–N bond homoly-
sis followed by the coupling of two 9-anthrylmethyl
1
1a,13,14
radicals.
The results are summarized in Scheme 2 and Table 1. It is
known that irradiation of a 9-substituted anthracene gives
predominantly a head-to-tail dimer (h–t) with minor forma-
tion of a sterically more hindered head-to-head dimer
2
.3. Close scrutiny of the results (Table 1)
Probably, a simple glance into Table 1 will not disclose the
important effect of CO , which may thus be overlooked.
From the closer inspection, however, it can be seen that irre-
spective of the starting anthrylamines, the total yield of the
dimers (h–t and h–h) obtained in MeOH or DMSO is higher
1
0,11
(
h–h).
and (CH ) NH , respectively) without any detectable forma-
APA and AEA yielded only h–t (R¼(CH ) NH
2
2
3
2
2
2
2
tion of h–h (Table 1). In contrast, AMA produced both h–h
z
(
tional product, AEA gave an intramolecular photoamination
R¼CH NH ) and h–t (R¼CH NH ). As an isolated addi-
2
2
2
2
under CO than under argon (see runs 1–4, 7–10, and 13–
2
1
6), whereas this is not the case for the total dimer yield
product 1 in DMSO under argon atmosphere. Although the
obtained in 2-PrOH (see runs 5, 6, 11, 12, 17, and 18). The
former outcome may be ascribed to the formation of either
ammonium bicarbonate/carbonate (–NH HCO /CO ) or
carbamic acid (–NHCOOH) in MeOH or DMSO, respec-
tively, in the presence of CO (Scheme 1). When the amino
group is protonated or carboxylated, a single electron trans-
fer from the lone pair of the nitrogen atom will become hard
to happen. Hence, potential side reactions that are caused
by the resultant ion radicals will decrease, leading to higher
yields for the photodimerization (Scheme 3). On the other
hand, the carbamate anion (–NHCOO ) formed in 2-PrOH
+
3
ꢀ
2ꢀ
z
3
3
Under argon, h–h was produced presumably via an amine dimer cation
radical, which may be generated through sequential formation of an intra-
molecular exciplex (or a radical ion pair) and a three-electron s bond
(Eq. 6): see Lewis, F. D.; Li, L.-S.; Kurth, T. L.; Kalgutkar, R. S. J. Am.
2
Chem. Soc. 2000, 122, 8573–8574, and the references therein. Much
more experiments will be required to prove the participation of this tri-
plex-like intermediate. The intramolecular electron transfer seems less ef-
ficient in the case of APA and AEA because of their longer
methylene-chain length.
8a
ꢀ
NH2
H
N
H
H
will not be able to restrain the electron transfer from the
amine moiety so effectively, because the electron withdraw-
ing property of the dissociated carboxyl group (COO )
AMA
AMA
h
N
N
h-h
ð6Þ
ꢀ
–
H
H
H
An
An
An
should be much smaller as compared with the undissociated
AMA

M. Horiguchi, Y. Ito / Tetrahedron 63 (2007) 12286–12293
12289
Table 1. Photoreactions of (9-anthryl)alkylamines under argon or CO
2
in three different solvents
Conversion (%)
a
Product yield (%)
Run
Amine
Solvent
Atmosphere
Irradiation
h–t/h–h
ꢁ
h
C
h–t
71
86
78
93
65
78
h–h
Others
1
2
3
4
5
6
APA
MeOH
DMSO
2-PrOH
Ar
CO
Ar
CO
Ar
CO
2
2
2.3
2.3
2
3
3
59
62
w100
87
w100
66
0
0
0
0
0
0
—
—
—
—
—
—
100:0
100:0
100:0
100:0
100:0
100:0
2
2
2
21
21
15
15
2
7
8
9
AEA
MeOH
DMSO
2-PrOH
Ar
CO
Ar
CO
Ar
CO
2
2
2
2
2
2
3
3
21
21
15
15
66
59
61
86
94
80
w100
w100
78
w100
59
50
0
0
0
0
0
0
—
—
1 (18%)
—
—
100:0
100:0
100:0
100:0
100:0
100:0
2
2
2
10
11
12
—
13
14
15
16
17
18
AMA
MeOH
DMSO
2-PrOH
Ar
CO
Ar
CO
Ar
CO
2
2
2
2
2
2
3
3
21
21
15
15
93
96
65
93
32
37
34
57
46
29
24
23
28
37
38
40
2 (trace), 3 (trace)
2 (trace), 3 (trace)
2 (10%), 3 (11%)
2 (trace), 3 (trace)
2 (7%), 3 (6%)
57:43
62:38
55:45
61:39
55:45
42:58
2
2
2
w100
w100
2 (10%), 3 (7%)
a
Yields are based on the consumed reactant.
(
CH ) -NH
2 n 2
h
electron transfer-induced
side reactions
h-t
+
h-h
+
APA, AEA, or AMA
CO2 in DMSO or MeOH
CO2 in 2-PrOH
(CH2)n-NHCO2H
or
(CH2)n-NH3+ HCO3–
h
electron transfer-induced
side reactions
h-t
+
h-h
+
+
(
decrease)
higher yields
–
+
H3N (CH2)n
.
(
CH ) -NHCO
2
2
n
h
electron transfer-induced
side reactions
h-t
+
h-h
(not much affected)
variable yields
Scheme 3. Effect of electron transfer-initiated side reactions on the photodimerization.
8
a
carboxyl group (COOH). Therefore, the observed CO2
effect is variable in this solvent.
MeOH) may be utilized to restrain its amine function from
causing competing electron-transfer side reactions.
Furthermore, byproducts such as 1–3 are thought to be
formed through electron transfer from the amino group or
With regard to the ratio of h–t vs h–h in the photoreaction of
AMA in MeOH (runs 13 and 14) or DMSO (runs 15 and 16),
the proportion of h–h decreased when the irradiation was
1
2,15a,16
through exciplex involving the amino group.
ally, the production of 1 (from AEA) or 2+3 (from AMA)
Actu-
carried out under CO instead of argon. This is probably be-
2
in DMSO was inhibited under CO (runs 10 and 16). On
2
cause the formation of the bicarbonate/carbonate or the car-
bamic acid not only makes the 9-substituent more bulky, but
also suppresses the formation of the amine dimer cation rad-
ical depicted in Eq. 6. On the other hand, in the 2-PrOH so-
lution (runs 17 and 18) the proportion of h–h increased in the
presence of CO . This opposite effect by CO may originate
the other hand, the production of 2+3 in 2-PrOH was not
quenched by CO (run 18). Consideration to the CO effect
2
2
on the byproducts 1–3 is illustrated in Scheme 4. The same
line of discussion was performed to explain the changes in
fluorescence intensities of u-(1-naphthyl)alkylamines,
8
a
2
2
where the fluorescence was quenched via intramolecular
electron transfer from the amine moiety to the naphthalene
moiety at various degrees depending on the solvent (see
Scheme 3 in Ref. 8a). Therefore, CO saturation of an amine
2
solution in suitable solvent (DMSO, DMF, pyridine or
in the formation of ammonium carbamate in 2-PrOH. Ionic
noncovalent linkage between the ammonium cation and the
carbamate anion may enforce two anthracene rings to ar-
range in a head-to-head fashion (Scheme 5, structure A),
thereby favoring the route to the head-to-head dimer.

12290
M. Horiguchi, Y. Ito / Tetrahedron 63 (2007) 12286–12293
NH2
NH2
h
1
–
e
18% yield
in DMSO
AEA
CO2 in DMSO
NHCO H
2
h
1
slow – e
not produced
NH2
NH2
CH2
h
–
e
in DMSO or
AMA
2
-PrOH
CO2 in DMSO
2
+ 3
NHCO H
21% (DMSO)
2
CO2 in 2-PrOH
13% (2-PrOH)
h
2
+ 3
slow – e
trace
NHCO – H N
2
3
h
2
+ 3
7%
relatively
slow – e
1
2
Scheme 4. Consideration to the CO effect on the byproducts 1–3, which are formed through electron transfer-initiated side reactions.
NHCO H
NH3+ HCO – and/or CO32–
2
3
or
in DMSO
or MeOH
h
CO2
lower h-h/h-t ratio
due to the steric effect or by disturbing
the amine dimer cation radical formation
NH2
h
h-h
+
h-t
higher h-h/h-t ratio
AMA
h
CO2
in 2-PrOH
H N
3
H
N
O C
2
A
2
Scheme 5. Control of the (9-anthryl)methylamine (AMA) photodimerization by using CO .
3. Conclusion
the yield for the dimers (h–t from APA or AEA, h–t+h–h
from AMA) obtained in MeOH or DMSO was higher under
A systematic attempt to control photochemistry by using
CO was studied for the first time. Although not necessarily
big, the overall effect of CO on the photoreactions of
2
(9-anthryl)alkylamines (APA, AEA, AMA) was clearly
observed on the basis of close analysis of the results. Thus,
CO than under argon and this was ascribed to formation of
2
either ammonium bicarbonate/carbonate in MeOH or car-
bamic acid in DMSO. It is considered that the formation
of these species suppresses side reactions, which are initi-
ated by the electron transfer from the amino group. In fact,
2

M. Horiguchi, Y. Ito / Tetrahedron 63 (2007) 12286–12293
12291
2
0
production of the electron-transfer-induced byproducts 1–3
in DMSO was strongly inhibited under CO . Also, due to
the formation of ammonium carbamate in 2-PrOH, the pro-
portion of h–h relative to h–t produced from AMA in
conditions of the literature method. Thus, a solution of
9-(chloromethyl)anthracene (20.75 g, 92 mmol) in 300 mL
of DMF was added dropwise to a solution of NaCN (9.01 g,
2
ꢁ
182 mmol) in 400 mL of DMF at 140 C and was kept at this
2
CO . In this case, the ionic linkage between the ammonium
-PrOH was increased by carrying out the reaction under
temperature for 4 h with stirring. After the reaction mixture
was cooled to room temperature, ice-water was added and
the resulting yellow precipitate was collected by filtration.
This was recrystallized from ethanol to afford 12.26 g
(56 mmol, 62% yield) of 9-anthrylacetonitrile (6) as yellow
2
cation and the carbamate anion is probably operating as
a noncovalent linker, leading to the head-to-head photodi-
merization. It is noticeable that this paper not only substan-
8
a,b
ꢁ
20
tiates the previous spectroscopic finding about the solvent
dependence of the amine–CO reactions, but also presents
crystals. Compound 6: mp 158–159 C (lit. mp 158–
159 C); H NMR (400.4 MHz, CDCl ) d 4.59 (s, 1H),
ꢁ
1
2
3
the first example showing that such solvent dependence
can be successfully applied to the reaction control.
7.52 (ddd, J¼8.8, 6.4, 1.2 Hz, 2H), 7.63 (ddd, J¼8.8, 6.4,
1.2 Hz, 2H), 8.05 (dd, J¼8.4, 0.4 Hz, 2H), 8.16 (dd,
1
3
J¼8.8, 0.8 Hz, 2H), 8.50 (s, 1H) ppm;
C NMR
100.7 MHz, CDCl ) d 16.34, 117.56, 120.61, 122.81,
(
125.18, 127.12, 128.70, 129.36, 129.64, 131.28 ppm; IR
3
4
. Experimental
(KBr) 735.8, 790.8, 885.3, 901.7, 1154.3, 1346.3, 1448.4,
1623.8, 2243.1, 3026.1, 3053.5 cm
ꢀ
1
4
.1. General
.
1
13
H and C NMR spectra were measured on JEOL or Varian
FT NMR spectrometers. Mass and IR spectra were recorded
on JEOL JMS-HX110A and SHIMADZU FTIR-8400 spec-
The nitrile 6 obtained above was dissolved in 200 mL of dry
THF and this was added dropwise to a suspension of LiAlH4
(5.80 g, 153 mmol) in 400 mL of dry THF with stirring. The
mixture was refluxed for additional 2 h. After the reaction
mixture was cooled to room temperature, 50 mL of MeOH/
water was added and the resultant solid was removed by fil-
tration. The filtrate was then evaporated and the residue was
dissolved in a small amount of benzene. Then, concentrated
hydrochloric acid was added and the precipitated AEA$HCl
was collected by filtration. This was recrystallization from
MeOH/AcOEt to afford pale yellow crystals: 3.45 g (26%
1
trometers, respectively. The in situ measurements of H and
C NMR, HMBC, and IR spectra in MeOH, DMSO, and
1
3
2
of APA were carried out as described previously.
-PrOH before and after bubbling CO through the solution
2
8
a
4
.2. Materials
4
.2.1. 3-(9-Anthryl)propylamine (APA). It was prepared
ꢁ
1
according to the literature methods from 3-(9-anthryl)pro-
pionitrile, which had been prepared from 9-anthraldehyde
via 3-(9-anthryl)acrylonitrile. APA$HCl: 4.34 g (53%
yield) was obtained by LiAlH reduction of 3-(9-anthryl)-
propionitrile (6.95 g) as pale yellow prisms after recrystalli-
yield), mp 262–264 C; H NMR (300.1 MHz, DMSO-d )
6
d 3.06 (br s, 2H), 3.97 (t, J¼9.0 Hz, 2H), 7.53 (ddd, J¼8.1,
6.3, 0.9 Hz, 2H), 7.61 (ddd, J¼8.4, 6.3, 1.5 Hz, 2H), 8.10
(d, J¼7.8 Hz, 2H), 8.36 (br s, NH), 8.44 (d, J¼8.7 Hz, 2H),
1
7–19
4
1
3
8.56 (s, 1H) ppm; C NMR (75.5 MHz, DMSO-d )
6
ꢁ
19
zation with MeOH/AcOEt; mp >236 C dec (lit. mp
2
d 25.39, 39.29, 123.95, 125.21, 126.33, 126.62, 129.10,
129.23, 129.49, 131.05 ppm; IR (KBr) 736.2, 785.9, 839.0,
878.5, 934.4, 954.4, 1007.7, 1015.9, 1143.7, 1590.6,
ꢁ
42–244 C dec); H NMR (400.4 MHz, DMSO-d ) d 1.99
6
1
(
2
tt, J¼8.0, 8.0 Hz, 2H), 3.04 (br s, 2H), 3.68 (t, J¼8.0 Hz,
ꢀ
1
+
H), 7.55 (ddd, J¼8.8, 6.4, 1.6 Hz, 2H), 7.50 (ddd, J¼8.0,
.4, 1.2 Hz, 2H), 8.07 (dd, J¼8.0, 1.2 Hz, 2H), 8.14 (br s,
2918.9, 3005.9 cm ; MS (EI) m/z (%) 221 ([MꢀHCl] ,
6
15), 192 (100), 191 (38); HRMS (EI) calcd for C H N
1
6 15
1
3
NH), 8.38 (dd, J¼8.0, 0.8 Hz, 2H), 8.48 (s, 1H) ppm; C
NMR (100.7 MHz, DMSO-d ) d 24.29, 28.86, 38.83,
221.1204, found 221.1208.
6
1
1
9
1
24.19, 124.91, 125.55, 125.68, 128.84, 128.88, 130.90,
33.30 ppm; IR (KBr) 631.6, 731.9, 788.8, 839.9, 878.5,
53.7, 1147.6, 1323.1, 1339.7, 1387.7, 1444.6, 1491.8,
The corresponding free amine was obtained by neutraliza-
tion of AEA$HCl with aqueous NaOH, as described above.
1
AEA: H NMR (400.4 MHz, DMSO-d ) d 2.86 (t, J¼8.0 Hz,
6
ꢀ1
598.9, 1620.5, 2800–3200 (br), 3423.8 (br) cm ; MS
2H), 3.68 (t, J¼8.0 Hz, 2H), 7.48 (ddd, J¼8.4, 6.4, 0.8 Hz,
2H), 7.53 (ddd, J¼8.8, 6.4, 1.6 Hz, 2H), 8.05 (dd, J¼8.0,
1.2 Hz, 2H), 8.38 (d, J¼8.8 Hz, 2H), 8.44 (s, 1H) ppm.
+
(FAB) m/z (%) 236 (M , 100), 191 (15); HRMS (FAB) calcd
for C H N 236.1439, found 236.1437.
17 18
An aqueous solution of APA$HCl was made alkaline by ad-
dition of aqueous NaOH and then extracted with ether. The
4.2.3. (9-Anthryl)methylamine (AMA). It was prepared
either (a) from the reaction of 9-(chloromethyl)anthracene
with hexamethylenetetramine by using the previous proce-
dure or (b) through reduction of 9-anthraldehyde oxime.
Since the method (a) gave the compound in a poor yield,
the procedures for the method (b) are described below.
ether solution was dried with MgSO and the solvent was re-
4
1
21
moved by evaporation to afford the free amine APA: H
NMR (400.4 MHz, DMSO-d ) d 1.76 (tt, J¼8.4, 6.4 Hz,
6
2
H), 2.75 (t, J¼6.4 Hz, 2H), 3.61 (t, J¼8.0 Hz, 2H), 7.48
(
1
ddd, J¼8.0, 6.8, 1.6 Hz, 2H), 7.53 (ddd, J¼8.8, 6.8,
.6 Hz, 2H), 8.05 (dd, J¼8.0, 1.6 Hz, 2H), 8.34 (d,
A solution of 9-anthraldehyde (20.69 g, 100 mmol) in
500 mL of pyridine was added to a solution of hydroxyl-
amine hydrochloride (30.58 g, 443 mmol) in 200 mL of pyr-
idine and the mixture was refluxed for 3 h. Then, the mixture
was evaporated and the residue was dissolved in CHCl3.
After removal of an insoluble material by filtration, the fil-
trate was washed successively with 0.5 N HCl and a saturated
1
3
J¼8.8 Hz, 2H), 8.43 (s, 1H) ppm; C NMR (100.7 MHz,
DMSO-d ) d 24.82, 35.30, 41.83, 124.27, 124.80, 124.96,
6
1
25.39, 128.74, 128.83, 130.92, 134.96 ppm.
4.2.2. 2-(9-Anthryl)ethylamine (AEA). It was prepared
from 9-(chloromethyl)anthracene by modifying the reaction

12292
M. Horiguchi, Y. Ito / Tetrahedron 63 (2007) 12286–12293
NaHCO solution, and dried over Mg SO . Evaporation of
3
and returned quantitatively to AMA: the conversion in
DMSO-d at room temperature was 74% in 21 h. The photo-
2
4
the solvent afforded 9-anthraldehyde oxime as yellow nee-
dles: 14.17 g (64% yield); H NMR (269.6 MHz, CDCl )
6
1
lysis was repeated under similar conditions and the effects
by CO were confirmed.
3
d 7.41 (ddd, J¼7.7, 6.6, 1.1 Hz, 2H), 7.46 (ddd, J¼8.5,
2
6
.6, 1.6 Hz, 2H), 7.91 (d, J¼7.4 Hz, 2H), 8.28 (d,
J¼8.7 Hz, 2H), 8.35 (s, 1H), 9.09 (s, 1H) ppm; IR (KBr)
For the purpose of characterizing the photodimers, the re-
spective hydrochlorides of APA, AEA, and AMA were irra-
diated in MeOH under argon. The dihydrochlorides of the
dimers precipitated. They were separated by filtration and
their spectral data were measured.
7
1
1
87.9, 803.5, 844.8, 877.6, 885.3, 918.1, 952.8, 970.1,
023.2, 1069.4, 1096.4, 1157.2, 1181.3, 1261.4, 1294.1,
441.7, 1623.9, 2962.4, 3031.6, 3052.1, 3261.1 cm ; MS
ꢀ
1
+
(FAB) m/z (%) 222 (78), 221 (M , 100), 204 (47). This
was used for the next step without further purification.
The characterization data for lepidopterene (2) and bipla-
1
1a
The oxime (10.18 g, 46 mmol), which was obtained above
was dissolved in 150 mL of dry THF and was added dropwise
to a suspension of LiAlH (5.00 g, 132 mmol) in 200 mL
of dry THF. The mixture was heated under reflux for 1 h,
and then 50 mL of MeOH/water was carefully added. After
removal of an insoluble material by filtration and then
evaporation of the solvent, the residue was dissolved in
nene (3) were reported in our previous work. In order to
isolate compound 1, a solution of AEA (199.8 mg) in
DMSO (40 mL) was irradiated for 8 h under argon. Then,
the solvent was removed by rotary-evaporation and the res-
idue was separated by preparative TLC (silica gel, CHCl3),
followed by HPLC (JAIGEL GS-320, MeOH), to afford
30 mg (14%) of 1.
4
2
50 mL of CHCl and subsequently 200 mL of 1 N HCl
3
was added. The resulted precipitate was collected by filtra-
tion and recrystallized from MeOH/AcOEt to afford
AMA$HCl as yellow crystals (3.12 g, 28% yield): mp
4.4. Spectral data for photoproducts
4.4.1. 5,12:6,11-Di-o-benzenodibenzo[a,e]cyclooctene-
5,11(6H,12H)-bis(propanamine)dihydrochloride, h-t$
2HCl (R=(CH ) NH ). H NMR (399.7 MHz, C D N)
ꢁ
19
ꢁ
1
>
221 C dec (lit.
mp 218–220 C dec); H NMR
1
(
6
400.4 MHz, DMSO-d ) d 5.04 (s, 2H), 7.58 (ddd, J¼8.4,
6
2 3
2
5 5
.4, 0.8 Hz, 2H), 7.67 (ddd, J¼10.0, 6.4, 1.6 Hz, 2H), 8.16
d 1.58 (quin, J¼7.3 Hz, 4H), 2.87 (t, J¼7.3 Hz, 4H), 2.99 (t,
13
(
NH), 8.74 (s, 1H) ppm; C NMR (100.7 MHz, DMSO-d )
d, J¼8.4 Hz, 2H), 8.44 (dd, J¼8.8, 0.8 Hz, 2H), 8.62 (br s,
J¼6.1 Hz, 4H), 4.03 (s, 2H), 4.77 (br, NH), 6.88–6.96 (multi,
1
3
8H), 7.01 (d, J¼5.3 Hz, 4H), 7.23(d, J¼5.3 Hz, 4H) ppm; C
6
d 34.17, 123.94, 124.95, 125.28, 126.73, 128.86, 128.95,
NMR (100.4 MHz, C D N) d 29.46, 35.16, 43.28, 56.79,
5
5
1
9
1
3
3
29.92, 130.68 ppm; IR (KBr) 668.4, 731.9, 788.8, 893.0,
30.6, 1110.9, 1162.0, 1187.6, 1264.2, 1340.2, 1448.4,
489.9, 1554.8, 1597.9, 1624.5, 2921.5, 3002.0, 3046.4,
65.16, 125.51, 125.51, 126.95, 127.91, 143.67, 144.49 ppm;
MS (FAB) m/z (%) 471 ([Mꢀ2HCl+H] , 57), 236 (98), 219
+
(27), 191 (100), 178 (23); HRMS (FAB) calcd for C H N
4 35 2
3
ꢀ
1
+
417.4 (br) cm ; MS (FAB) m/z (%) 207 ([MꢀHCl] ,
471.2800, found 471.2801.
2), 191 (100).
4
.4.2. 5,12:6,11-Di-o-benzenodibenzo[a,e]cyclooctene-
5,11(6H,12H)-bis(ethanamine)dihydrochloride, h-t$
2
2
The corresponding free amine was obtained by neutraliza-
tion of AMA$HCl with aqueous NaOH, as described above.
1
2HCl (R=(CH ) NH ). H NMR (399.7 MHz, DMSO-d )
2
2
2
6
1
AMA: H NMR (400.4 MHz, DMSO-d ) d 4.66 (s, 2H), 7.49
6
d 2.64 (br s, 4H), 3.03 (br t, J¼6.7 Hz, 4H), 3.97 (s, 2H),
(
ddd, J¼8.4, 6.4, 1.2 Hz, 2H), 7.54 (ddd, J¼8.8, 6.4, 1.6 Hz,
6.83 (t, J¼6.8 Hz, 4H), 6.86 (td, J¼7.4, 1.3 Hz, 4H), 6.93
13
2
1
1
1
H), 8.06 (d, J¼8.0 Hz, 2H), 8.40 (d, J¼8.8 Hz, 2H), 8.49 (s,
(d, J¼6.8 Hz, 4H), 7.12 (d, J¼6.6 Hz, 4H), 8.08 (br s,
1
3
H) ppm; C NMR (100.7 MHz, DMSO-d ) d 37.63,
6
NH) ppm; C NMR (100.4 MHz, DMSO-d ) d 33.14,
35.55, 54.31, 62.85, 125.41, 125.41, 125.70, 127.41,
141.52, 142.85 ppm; IR (KBr) 647.4, 690.8, 763.9, 784.4,
6
24.38, 124.84, 125.55, 125.86, 128.62, 128.87, 130.97,
35.34 ppm.
8
18.2, 917.8, 960.3, 1045.6, 1139.4, 1314.7, 1452.5,
1473.6, 1602.1, 2923.6, 3006.9, 3435.0 cm ; MS (FAB)
ꢀ
1
4
.3. Photolysis
+
m/z (%) 443 ([Mꢀ2HCl+H] , 100), 222 (71), 205 (89);
The photolysis was carried out as follows and the results are
listed in Table 1.
HRMS (FAB) calcd for C H N 443.2487, found
2 31 2
3
443.2486.
A solution containing (9-anthryl)alkylamine (APA, AEA, or
AMA: ca. 0.05 M) in MeOH, DMSO-d , or 2-PrOH was di-
4.4.3. 5,12:6,11-Di-o-benzenodibenzo[a,e]cyclooctene-
5,11(6H,12H)-bis(methanamine)dihydrochloride, h-t$
2HCl (R=CH NH ). H NMR (399.7 MHz, CD OD)
6
1
vided into two parts. These were placed either in two Pyrex
tubes (in the case of the MeOH or 2-PrOH solution) or in two
NMR tubes (in the case of the DMSO-d solution) and were
2
2
3
d 4.07 (s, 2H), 5.22 (s, 4H), 6.94 (td, J¼7.4, 1.5 Hz, 4H),
6.99 (td, J¼7.5, 1.6 Hz, 4H), 7.06 (dd, J¼7.7, 1.2 Hz, 4H),
6
1
3
irradiated with a 400-W high pressure mercury lamp (Pyrex
filter, >290 nm) for 2–2.3 h under argon and CO , respec-
7.10 (dd, J¼7.1, 1.6 Hz, 4H) ppm; C NMR (100.4 MHz,
CD OD) d 42.09, 57.08, 60.91, 125.40, 127.85, 127.93,
3
2
tively. During the irradiation, the solutions were maintained
at specified temperatures with thermostated circulating
water. After the irradiation, the samples in DMSO-d were
129.67, 141.25, 144.05 ppm; IR (KBr) 646.1, 732.9, 756.7,
772.4, 783.2, 897.8, 969.9, 982.7, 1108.1, 1136.9, 1459.5,
1476.2, 1596.6, 2920.8, 2984.6, 3422.0 (br) cm ; MS
ꢀ
1
6
1
+
directly analyzed by H NMR. The samples in MeOH or
2
(FAB) m/z (%) 415 ([Mꢀ2HCl+H] , 32), 207 (23), 191
ꢁ
dues were immediately analyzed by H NMR in DMSO-
-PrOH were rotary-evaporated below 30 C, and the resi-
(100); HRMS (FAB) calcd for C H N 415.2174, found
3
0 27 2
1
415.2173. The structure was confirmed by NOESY and
HMBC.
d . The head-to-head dimer h–h (R¼CH NH ) was unstable
6
2
2

M. Horiguchi, Y. Ito / Tetrahedron 63 (2007) 12286–12293
12293
4
5
2
.4.4. 5,12:6,11-Di-o-benzenodibenzo[a,e]cyclooctene-
,6(11H,12H)-bis(methanamine)dihydrochloride, h-h$
HCl (R=CH NH ). H NMR (399.7 MHz, DMF-d7)
Kranemann, C. L.; Rische, T.; Eilbracht, P.; Kluwer, S.;
Ernsting, J. M.; Elsevier, C. J. Chem.—Eur. J. 2001, 7, 4584–
4589; F €u rstner, A.; Ackermann, L.; Beck, K.; Hori, H.; Koch,
D.; Langemann, K.; Liebl, M.; Six, C.; Leitner, W. J. Am.
Chem. Soc. 2001, 123, 9000–9006; Minakata, S.; Yoneda, Y.;
Oderaotoshi, Y.; Komatsu, M. Org. Lett. 2006, 8, 967–969.
8. (a) Masuda, K.; Ito, Y.; Horiguchi, M.; Fujita, H. Tetrahedron
2005, 61, 213–229; (b) Ito, Y.; Ushitora, H. Tetrahedron
2006, 62, 226–235; (c) Ito, Y. Tetrahedron 2007, 63, 3108–
1
2
2
d 4.78 (s, 2H), 5.03 (s, 4H), 6.96–7.02 (multi, 8H), 7.09
dd, J¼5.9, 2.7 Hz, 4H), 7.67 (d, J¼9.0 Hz, 4H), 8.81 (br
(
s, NH) ppm; C NMR (100.4 MHz, DMF-d ) d 36.60,
1
3
7
5
1
3.70, 59.93, 126.73, 127.10, 127.29, 128.84, 141.38,
44.94 ppm. As expected,
1
1b,c
the bridgehead proton of h–
h (d 4.78) appeared at a lower field than that of h–t (d 4.07).
3
114; (d) Ito, Y.; Uozu, Y.; Matsuura, T. J. Chem. Soc.,
1
4
.4.5. 2,3-Dihydro-1H-naphtho[1,2,3-de]quinoline (1). H
Chem. Commun. 1988, 562–564; (e) Ito, Y. Photochemistry
2002, 33, 205–212.
NMR (500.0 MHz, DMSO-d ) d 3.51 (d, J¼5.1 Hz, 2H),
6
3
.53 (d, J¼4.9 Hz, 2H), 6.44 (br s, NH), 6.52 (dd, J¼6.0,
.2 Hz, 1H), 7.21 (dd, J¼13.0, 8.4 Hz, 1H), 7.21 (d,
9. Kaur, D.; Kaur, R. P.; Kaur, P. Bull. Chem. Soc. Jpn. 2006,
79, 1869–1875.
2
J¼6.2 Hz, 1H), 7.44 (td, J¼6.8, 3.2 Hz, 1H), 7.45 (td, J¼6.8,
10. (a) Bouas-Laurent, H.; Castellan, A.; Desvergne, J.-P.;
Lapouyade, R. Chem. Soc. Rev. 2001, 30, 248–263; (b) Becker,
H. D. Chem. Rev. 1993, 93, 145–172; (c) Bouas-Laurent, H.;
Desvergne, J.-P. Photochromism: Molecules and Systems;
D €u rr, H., Bouas-Laurent, H., Eds.; Elsevier: Amsterdam, 1990;
Chapter 4, pp 561–630; (d) Bouas-Laurent, H.; Castellan, A.;
Desvergne, J.-P. Pure Appl. Chem. 1980, 52, 2633–2648.
11. (a) Horiguchi, M.; Ito, Y. J. Org. Chem. 2006, 71, 3608–3611;
(b) Ito, Y.; Fujita, H. J. Org. Chem. 1996, 61, 5677–5680; (c)
Ito, Y.; Olovsson, G. J. Chem. Soc., Perkin Trans. 1 1997,
127–133.
3
.2 Hz, 1H), 7.97 (dt, J¼9.7, 3.5 Hz, 1H), 8.15 (dd, J¼6.4,
1
3
3
DMSO-d ) d 25.72, 39.97, 103.20, 115.10, 120.41, 123.30,
.7 Hz, 1H), 8.22 (s, 1H) ppm; C NMR (100.4 MHz,
6
1
1
7
1
2
23.64, 124.46, 125.05, 126.29, 126.40, 128.38, 129.07,
31.06, 131.83, 144.92 ppm; IR (KBr) 668.0, 731.0, 750.7,
61.8, 836.3, 880.4, 1141.1, 1151.4, 1306.7, 1336.3,
360.7, 1528.4, 1559.9, 1569.9, 1617.8, 2851.1, 2924.4,
ꢀ
1
955.7, 3050.2, 3381.5 cm ; MS (FAB) m/z (%) 219
+
(
M ); HRMS (FAB) calcd for C H N 219.1048, found
16 13
219.1052. The structure was confirmed by NOEDF, COSY,
NOESY, HMQC, and HMBC.
12. (a) Lewis, F. D.; Reddy, G. D.; Cohen, B. E. Tetrahedron Lett.
1994, 35, 535–538; (b) Pandey, G.; Sridhar, M.; Bhalerao, U. T.
Tetrahedron Lett. 1990, 31, 5373–5376.
Acknowledgements
1
3. (a) Adam, W.; Schneider, K.; Stapper, M.; Steenken, S. J. Am.
Chem. Soc. 1997, 119, 3280–3287; (b) Vermeersch, G.;
Febvay-Garot, N.; Caplain, S.; Couture, A.; Lablache-
Combier, A. Tetrahedron Lett. 1979, 20, 609–612.
We thank Mr. Koji Masuda for his contribution to the initial
stage of this investigation.
1
4. Lepidopterene (2) was reported to be produced from 9-
1
3
15
substituted anthracenes by photochemical or thermal reac-
tions.
15. (a) Vadakkan, J. J.; Mallia, R. R.; Prathapan, S.; Rath, N. P.;
Supplementary data
1
13
The variable temperature H and C NMR of carbamic acid
(Figs. 1 and 2) and the thermodynamic plot of the data in
Figure 1 (Fig. 3). Supplementary associated with this article
can be found in the online version, at doi:10.1016/
j.tet.2007.09.069.
Unnikrishnan, P. A. Tetrahedron Lett. 2005, 46, 5919–5922;
(
2
Yamaguchi, H.; Sakata, Y.; Misumi, S. Bull. Chem. Soc.
Jpn. 1982, 55, 182–187; (d) Felix, G.; Lapouyade, R.;
Bouas-Laurent, H.; Gaultier, J.; Hauw, C. Tetrahedron Lett.
1975, 16, 409–412.
4
b) Fernandez, M.-J.; Gude, L.; Lorente, A. Tetrahedron Lett.
001, 42, 891–893; (c) Higuchi, H.; Otsubo, T.; Ogura, F.;
References and notes
1
6. (a) Schneider, S.; Rehaber, H.; Sch €u ßlbauer, W.; Lewis, F. D.;
Yoon, B. A. J. Photochem. Photobiol., A: Chem. 1996, 99,
103–114; (b) Lewis, F. D.; Cohen, B. E. J. Phys. Chem.
1994, 98, 10591–10597.
17. Hann, R. A. J. Chem. Soc., Perkin Trans. 1 1974, 1379–1380.
18. Rhodes, R. A.; Boykin, D. W. Synth. Commun. 1988, 18, 681–
687.
19. Kumar, C. V.; Punzalan, E. H. A.; Tan, W. B. Tetrahedron 2000,
56, 7027–7040.
20. Gardner, R. A.; Delcros, J.-G.; Konate, F.; Breitbeil, F., III;
Martin, B.; Sigman, M.; Huang, M.; Phanstiel, O., IV.
J. Med. Chem. 2004, 47, 6055–6069.
21. Kumar, C. V.; Asuncion, E. H. J. Am. Chem. Soc. 1993, 115,
8547–8553.
22. Teodori, E.; Dei, S.; Quidu, P.; Budriesi, R.; Chiarini, A.;
Garnier-Suillerot, A.; Gualtieri, F.; Manetti, D.; Romanelli,
M. N.; Scapecchi, S. J. Med. Chem. 1999, 42, 1687–1697.
1
2
3
4
. Dell’Amico, D. B.; Calderazzo, F.; Labella, L.; Marchetti, F.;
Pampaloni, G. Chem. Rev. 2003, 103, 3857–3897.
. Jessop, P. G.; Heldebrant, D. J.; Li, X.; Eckert, C. A.; Liotta,
C. L. Nature 2005, 436, 1102.
. Liu, Y.; Jessop, P. G.; Cunningham, M.; Eckert, C. A.; Liotta,
C. L. Science 2006, 313, 958–960.
. Stastny, V.; Rudkevich, D. M. J. Am. Chem. Soc. 2007, 129,
1018–1019; Xu, H.; Rudkevich, D. M. Org. Lett. 2005, 7,
3223–3226.
5
6
7
. Carretti, E.; Dei, L.; Baglioni, P.; Weiss, R. G. J. Am. Chem.
Soc. 2003, 125, 5121–5129.
. Alauzun, J.; Mehdi, A.; Rey ꢀe , C.; Corriu, J. P. J. Am. Chem.
Soc. 2005, 127, 11204–11205.
. Vaidyanathan, R.; Kalthod, V. G.; Ngo, D. P.; Manley, J. M.;
Lapekas, S. P. J. Org. Chem. 2004, 69, 2565–2568;
Wittmann, K.; Wisniewski, W.; Mynott, R.; Leitner, W.;