9-Anthroylacetone and its photodimer

Article abstract of DOI:10.1016/j.tet.2004.09.070

9-Anthroylacetone undergoes a head-to-tail [4π+4π] photo-dimerisation reaction that leads to the formation of 5,11-bis(1,3-diketobutyl)-5,6,11,12- tetrahydro-5,12,6,11-di-o-benzeno-dibenzo[a,e]cyclooctene both in solution and in the solid state when irradiated with different sources (sunlight, tungsten lamp, xenon lamp, UV laser beam 351-364 nm), the reaction being accompanied by a colour variation from bright yellow to colourless. Quantum yields >0.023 mol/Einstein are evaluated for the solid state reaction. Interestingly, the dimer dissociates to give 9-anthroylacetone, both thermally (T>130°C) and photochemically, by short UV wavelength irradiation. The single-crystal X-ray structure of 9-anthroylacetone and its dimer are reported. 9-Anthroylacetone undergoes a head-to-tail [4π+4π] photo-dimerisation reaction both in solution and in the solid state when irradiated with different sources; the dimer reversibly dissociates into 9-anthroylacetone, both thermally and photochemically.

Full text of DOI:10.1016/j.tet.2004.09.070

Tetrahedron 60 (2004) 11959–11968
9
-Anthroylacetone and its photodimer
*
Francesca Cicogna, Giovanni Ingrosso, Fabio Lodato, Fabio Marchetti and
Maurizio Zandomeneghi
Dipartimento di Chimica e Chimica Industriale, Universit a` di Pisa, Via Risorgimento 35, 56126 Pisa, Italy
Received 4 June 2004; revised 26 August 2004; accepted 16 September 2004
Available online 27 October 2004
Abstract—9-Anthroylacetone undergoes a head-to-tail [4pC4p] photo-dimerisation reaction that leads to the formation of 5,11-bis(1,3-
diketobutyl)-5,6,11,12-tetrahydro-5,12,6,11-di-o-benzeno-dibenzo[a,e]cyclooctene both in solution and in the solid state when irradiated
with different sources (sunlight, tungsten lamp, xenon lamp, UV laser beam 351–364 nm), the reaction being accompanied by a colour
variation from bright yellow to colourless. Quantum yields O0.023 mol/Einstein are evaluated for the solid state reaction. Interestingly, the
dimer dissociates to give 9-anthroylacetone, both thermally (TO130 8C) and photochemically, by short UV wavelength irradiation. The
single-crystal X-ray structure of 9-anthroylacetone and its dimer are reported.
q 2004 Elsevier Ltd. All rights reserved.
1. Introduction
1
-Anthroylacetone 1, reported first by Evans in 1961, has
9
been the object of attention since its conjugate base, i.e.
2
-anthroylacetonate 2, gave rise to various boron and rare
earth derivatives showing potentially interesting photo-
physical properties (Scheme 1).
9
3
During the last years, some of us have studied whether and
how the anthrylic fluorophore communicates with different
metal–ligand sub-units, like metal–b-ketoenolato or
metal–cyclopentadienyl systems, under those conditions
(chemical, photochemical, and electrochemical) that can
induce the intramolecular transmission of electronic effects.
Scheme 1.
5
1
In this context, we prepared 1 as reported by Evans with the
More recently, the self assembled compound 3 has been
described in which energy and electron transfer can take
aim to synthesise new rhodium(I) and iridium(I) derivatives
of 2. We observed that freshly prepared bright yellow solid
6
2
3
C
place between Ru(bipy) , the donor sub-unit, and the
samples of 9-anthroylacetone gave rise to a new colourless
compound, when they were exposed to laboratory light in the
presence of air and this drew our attention towards the
photochemistry of 1. We succeeded in isolating the colour-
less compound and found that it has the structure 4 resulting
from the [4pC4p] photo-dimerisation of 1. To our
knowledge, such a reaction of 1 has not yet reported in the
literature. A similar behaviour has instead been documented
ligand 2, the acceptor component, under photo-activation
conditions.
4
Keywords: Anthracene derivatives; Anthracene photo-dimerisation;
Photochromism.
*
Corresponding author. Fax: C39 50 2219410/260;
e-mail: vanni@dcci.unipi.it
7
for 9-benzoylanthracene.
0040–4020/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2004.09.070

11960
F. Cicogna et al. / Tetrahedron 60 (2004) 11959–11968
interconvert between one form and another upon irradiation
with convenient light (photo-chromic molecules)
8b,9b,14
can
be useful, at least in principle, to project various photonic
devices, such as erasable optical memory media and photo-
optical switch components. For this reason it is important
that the direct and inverse photoreactions are not limited to
the solution phase but also occur in the solid state.
All these facts prompted us to study in some detail the
photochemical dimerisation of 1 to 4 and the thermal and
photochemical dissociation of 4, both in solution and in the
solid state. The results of this study are presented herein.
2. Result and discussion
2
.1. The photo-dimerisation of 9-anthroylacetone 1
The [4pC4p] photo-dimerisation of anthracene as well as
of several anthracene derivatives is a very well known
As described in the Section 4, when a d -dichloromethane
2
8
reaction that has been studied extensively since 1950. The
K2
solution of 1 (ca. 10 M) was irradiated for 3 h, under
8
reaction takes place both in solution, in a variety of
,9
dinitrogen atmosphere, with a xenon lamp, about 50% of 1
was converted into 4. Exposure of 1 to sunlight, at the same
concentration, gave 4 in about 28% yield. The photo-
dimerisation took place also when a d -dichloromethane
2
solution of 1 was irradiated with a near-UV laser beam,
although in this case the conversion of 1 into 4 was about
7
,9c,10
organic solvents and in the solid state,
although the
reaction in the solid state is less documented. Photo-
excitation of meso-substituted anthracene derivatives in
solution typically leads to the head-to-tail dimers, though
evidence for the concomitant formation of the thermally
more labile head-to-head dimers has been obtained in some
1
1%. Since each experiment was carried out under very
8
b,9e,10b,11
cases (Scheme 2).
Bimolecular reactions are
different irradiation conditions (irradiation geometries, light
source, energy), it is not possible to draw any conclusion
regarding the relative efficiency of the different irradiation
techniques. Nevertheless, the photo-conversion of 1 into 4
appears a relatively simple clean reaction. A new,
unidentified, species were formed only when the irradiation
is carried out on chloroform solutions.
expected to take place in the solid state only between
nearest neighbours. Thus, the molecular structure of the
dimer is expected to reflect the geometric relationship lying
1
2
between the neighbours in the crystal lattice. As a result of
these restrictions, solid state reactions are generally more
selective then those in solution.
Mixtures of 1 and 4 can be separated by taking advantage of
their different solubility in diethyl ether.
1
The main pattern of the H NMR spectrum of 4 (Table 1)
looks like that reported in the literature for other anthracene
but in this spectrum a peak at dZ14.8 ppm
7
,15
photodimers,
1
is present due to the enolic proton of 4. In the H NMR
spectrum of 1 (Table 1), the enolic peak is observed at dZ
1
4
6.0 ppm, thus showing that 1 is moderately more acid than
. Both compounds are in their enolic form and the keto-
enol equilibrium cannot be shifted by changing the solvent
polarity.
As anticipated in the introduction, solid samples of 1 show
the presence of increasing amounts of 4 when they are
exposed to laboratory light. With the aim to better define the
photo-dimerisation of 1 in the solid state some experiments
were carried out. Irradiation experiments of solids are less
straightforward than experiments on solutions. In particular,
the irradiation geometry (i.e. the relative position and
orientation of the light source and of the sample), the
emission characteristics of the source and the spatial
extension of its light-generating volume, the dimension of
the particles being irradiated and their reflective and
scattering properties and the optics make a precise
evaluation of photochemical quantities difficult. However,
we attempted to estimate the limit values of quantum yields
of the photochemical reactions.
Scheme 2.
One of the most interesting features of the photo-dimerisa-
tion of anthracene derivatives is the reversibility, this being
the prerequisite for the development of new logic devices
9
a,b,13
that operate at the molecular level.
The photo-dimer
can be reverted to the monomer by thermal and photo-
chemical dissociation. While the first process is quantitat-
ive, a steady state mixture of monomer and dimer is
obtained by photochemical dissociation, at least under the
8
a
irradiation conditions used until now. Compounds that

F. Cicogna et al. / Tetrahedron 60 (2004) 11959–11968
11961
1
Table 1. H NMR spectroscopy data for compound 1 and 4
a
The 1 to 4 transformation in the solid state (UV Argon-ion
laser beam) (see Section 4.6, Experiment D) can be
followed by direct analysis using front-surface absorption
1
6
spectroscopy. Figure 1(A), that shows the spectra of pure
micro-crystalline 1, registered before and after irradiation,
as well as the spectrum of pure 4, demonstrates clearly that 1
photo-dimerises, at least partially, giving rise to 4. All the
1
6
spectra show saturation effects at absorbance O0.8 which
are due to the strong absorbing features of pure 1 and 4. To
avoid the saturation effects it is necessary to dilute the
sample with non-absorbing powder species to lower the
concentration of absorbing centres. A comparison between
the spectra reported in Figure 1(A) and those of Figure 1(B),
which were recorded on diluted mixtures of 1 or 4 in BaSO4,
supports the above conclusion, the spectrum of pure 4
looking like that of diluted 4, although, because of
saturation at absorbance O0.8, the absorbance in the
range 350–400 nm seems proportionally higher than that
expected on the base of the absorbance of the band at ca.
Compound
d (ppm), J/Hz
l
16.10 (1H, s, H ); 8.54 (1H, s, H ); 8.20–8.12 (2H, m, H , H );
i
a
h
1
d
e
b
c
f
g
8
5
.06–8.02 (2H, m, H , H ); 7.60–7.50(4H, m, H , H , H , H );
j k
.99 (1H, s, H ); 2.27 (3H, s, H )
300 nm shown by diluted 4. On the other hand, the spectrum
of diluted 1 (Fig. 1(B)) clearly shows the dramatic effect of
dilution, and now it looks similar to that obtained in solution
(see Section 2.2, Fig. 6).
1
2
1
8
4
14.76 (2H, s, H ); 6.98 (4H, d, JZ4 Hz, H , H ); 6.88–6.66
2
3
4
5
6
7
12H, m, H , H , H ,H , H , H ); 5.92 (2H, s, H ); 5.31 (2H, s,
9
(
H
10
11
); 1.94 (6H, s, H
)
a
Spectra recorded in CDCl
3
; sZsinglet; mZmultiplet; dZdoublet.
In order to gain insight into the topochemical aspects of the
photo-dimerisation reaction, single crystals of 1 and 4 were
analysed by X-ray diffraction. The molecular structure of 1
is shown in Figure 2. Bond distances and angles are
provided in Supporting Information. The molecule lies on
two planes, one being defined by the 9-anthryl moiety (max
A variety of irradiation experiments successfully transform
into 4 (see Section 4). Irradiation with a tungsten lamp
24 h), with a xenon arc lamp (4 h) or with a near-UV
Argon-ion laser (2 h) produced the photo-transformation up
to high conversion values: in our experiments, 70% yields
were easily obtained without any detectable formation of
other products. The irradiation of 50 mg of 1 with the
1
(
˚
deviating atom being C(8), 0.06 A) and the other by the 1,3-
diketobutyl moiety (max deviating atom being C(3),
˚
0.002 A). The two planes intersect each other along a line
351–364 nm emission lines of an Argon-ion laser (1/4
conversion, 31%) allowed us to estimate the minimum value
almost coincident with the C(4)–C(5) bond and make a
dihedral angle of about 64.78. The wideness of the
C(2)–C(3)–C(4) angle (121.58) and the presence of only
one hydrogen maximum beside C(3) in the difference
(
F1/4O0.023 mol of 4/Einstein) of the quantum yield of
photo-production of 4. This value was obtained on the
hypothesis that all incident light was absorbed by the
sample, thus ignoring scattering and reflection of light,
which are more and more intense with the increasing
amount of white microcrystals of 4 on the sample surface.
Thus, the above calculated value is a lower limit of the true
quantum yield.
2
Fourier map both suggest the essential sp nature of this
carbon atom, which supports the enolic form for the
1,3-diketobutyl moiety. Anyway, the equal lengths within
the C(2)–C(3)/C(3)–C(4) and C(2)–O(1)/C(4)–O(2) bonds
do not establish reliably which oxygen atom carries the
enolic hydrogen atom, which probably is statistically
Figure 1. Front surface absorbance spectra (absorbance vs l, nm); (A) spectrum of pure micro-crystalline 9-anthroylacetone 1 (
sample after 5 min of irradiation ( ); spectrum of pure 4 ( ); (B) spectrum of 9-anthroylacetone 1 in BaSO (1 mg in 5 g of BaSO
in BaSO (1 mg in 0.5 g of BaSO ) ( ).
); spectrum of the same
) (—); spectrum of pure
4
4
4
4
4

11962
F. Cicogna et al. / Tetrahedron 60 (2004) 11959–11968
Figure 2. View of the molecular structure of 1. Thermal ellipsoids of O
atoms are at 30% probability, those of C atoms have been omitted for
clarity.
bonded to either O(1) or O(2) atoms. However, according to
the finding of a maximum in the difference Fourier map, we
have introduced in our model the hydrogen atom bonded to
O(1). The resulting OH group behaves as a donor in an
Figure 4. View of the molecular structure of 4. Thermal ellipsoids of O
atoms are at 30% probability, those of C atoms have been omitted for
clarity. ZKxC1/2, KyC1/2, KzC1.
0
intramolecular hydrogen bond (O(1)/O(2) distance
2
showing the contents of the unit cell. The molecules are
arranged in pairs, related by the inversion centre placed at
˚
.516 A). Figure 3 suggests the crystal structure of 1 by
wings derived from the bending of the 9-anthryl group make
in 4 a dihedral angle of 133.18, while the 1,3-diketobutyl
moiety maintains its approximately planar conformation,
and, rotating around C(4)–C(5) bond, becomes coplanar
½
other in such a way that C(5) faces C(12 ) and,
, ½, ½. The 9-anthryl moieties of each pair overlap each
0
0
with the C(5)–C(12 ) bond.
0
˚
consequently, C(12) faces C(5 ), at a distance of 3.794 A,
Looking at Figures 3 and 4, the topological relationship
between the pair of molecules in the crystal structure of 1
and the molecular structure of 4 appears evident. Moreover,
it is important to underline that although the distance of
where the primed atoms belong to the second molecule of
the pair.
The molecular structure of 4 is shown in Figure 4. Bond
distances and angles are provided in Supporting Infor-
mation. Compound 4 results from the cycloaddition of two
molecules of 1 that are linked to each other by the new
0 0
˚
.794 A between C(5) and C(12 ) and C(5 ) and C(12) in the
3
crystal structure of 1 should be thought to be too long for the
cycloaddition to take place in the solid state, nevertheless
7
such a reaction indeed occurs in the case of 1, as reported
above. The cycloaddition of 1 to 4 does not follow a
topotactic reaction pathway: in fact, when a single crystal of
0
0
bonds C(5)–C(12 ) and C(12)–C(5 ). As a consequence, the
central rings of the two original 9-anthryl moieties lose their
0
aromaticity and bend along the axes C(5), C(12) and C(5 ),
0
1
was exposed to a near-UV laser the yellow colour
C(12 ), respectively, assuming the shape of flying wings.
The molecule has an inversion centre, which allows us to
confine our description to only one half of compound 4. The
disappeared and 4 was formed, but the crystal flaked off.
This shows that the molecules of 4 cannot move in the
crystal of 1 so to assemble in the packing required by the
new single crystal. The occurrence of any thermal effect
caused by the laser beam was discarded since the crystal was
maintained at room temperature during the irradiation.
2
.2. Thermal and photochemical dissociation of 4
When a solution of 4 in benzene was maintained at 80 8C for
h no appreciable decomposition of the sample was
4
observed. This fact prompted us to heat solid-state samples.
Thus, we observed that when a solid sample of 4 was heated
at a temperature near the melting point of 1 (127–129 8C), it
started to dissociate to give 1, and, at 180 8C, it was
converted almost quantitatively into 1 in about 25 min.
Since at 150 8C the reaction is slower, it can be easily
followed by H NMR spectroscopy. The profile of the
1
thermal dissociation of 4 (containing a small amount of 1,
ca. 1%) as a function of time (Fig. 5) shows that, after an
induction period, the reaction is fast. On the other hand, the
thermal dissociation of a solid sample of 4 which did not
contain any traces of 1 had a completely different course. In
this case, after 200 min of heating at 150 8C, an amount of 1
Figure 3. View of the crystal packing of 1, projected approximately along
the [101] direction. The molecules of the pair are related by the inversion
center at ½, ½, ½. Z1Kx, 1Ky, 1Kz.
0

F. Cicogna et al. / Tetrahedron 60 (2004) 11959–11968
11963
that the molar extinction coefficient at 270 and 280 nm are
equal to four times that of o-xylene.
1
7
One of the most intense emission bands of a medium
1
8
pressure mercury lamp is centred at 254 nm. Exactly at
this wavelength 9-anthroylacetone 1 has an extinction
maximum significantly higher than the extinction of 4 at
the same wavelength. Compound 4 has the extinction
maximum at 290 nm whose value is ca. double with respect
to the extinction exhibited by 1 at the same wavelength.
Finally, at l!240 nm the extinction of 4 becomes much
larger than that of 1. Then, irradiation in the l!240 nm
zone is strongly recommended. Moreover, assuming that the
quantum yield of the photodissociation of 4 is higher than
the quantum yield of dimerisation of 1, as reported for most
8
a
anthracene derivatives in solution, we profited from both
the quantum yield and the favourable absorption features to
shift the photo-equilibrium towards 1. In this connection, it
is important to underline that it is known that several
anthracene photodimers partially photodissociate into two
anthracene nuclei when they are irradiated in the range 250–
Figure 5. Thermal dissociation at 150 8C of 4 as a function of time (min): %
of 1 (—),% of 4 ( ).
(
only 35%), smaller than that predictable based on the data
reported in Figure 5, was obtained. We are inclined to
conclude that small amounts of 1 in solid samples of 4 play
some catalytic role in the thermal dissociation of 4. The fact
that, at 150 8C, 1 is melted while 4 is still solid can favour
the thermal decomposition of 4. At this temperature the
thermal dissociation of crystalline 4 can be kinetically
impeded; while, the solubilisation of 4 into 1 can be
favoured: as a result, a new reaction pathway can be
available.
290 nm. However, at these wavelengths the monomer can
re-dimerise and a steady state mixture is obtained by
8
a,9l
prolonged irradiation,
high amounts of photodimer (up to 94%, under 254 nm
irradiation).
characterised by the presence of
9
l
In our case, we succeeded in shifting the photo-equilibrium
towards 1 taking advantage of the favourable absorption
properties of 4 in the l!240 nm region and by adopting an
unconventional use of the mercury lamp, i.e. by switching
the lamp on/off at regular intervals of time to avoid the
strong increase of Hg vapour pressure which parallels the
warming of the lamp. Indeed, when the cool lamp is just
switched on, it strongly emits at wavelengths shorter than
The photochemical dissociation of 4 to give 1 was followed
by UV–vis absorption spectroscopy and was carried out on
K6
an ethanol solution of pure 4 (4.27!10 M) which was
irradiated by a medium pressure mercury lamp adopting an
unusual technique (see below). Before analysing the results
of these experiments it is necessary to consider the UV–vis
spectra of pure 1 and 4 (Fig. 6). The spectrum of 1 is
characterised by an intense absorption band at 254 nm, a
shoulder at ca. 280 nm and the typical anthracene
254 nm because of the low density of Hg vapour. In
addition, the temperature of the lamp was kept as low as
possible by ventilation. Finally, operating on rigorously
deareated ethanol solutions, the occurrence of the photo-
chemical reaction of anthracene with dioxygen, that leads to
anthraquinone and anthracene endoperoxide,
avoided.
4
,8b
vibrational structure in the range 330–390 nm.
An
4,7,8b,9d,19
was
evident feature of the UV–vis spectrum of 4 is the lack of
any absorption in the spectral zone where the typical
vibrational structure of anthracene is located. In involving
the 9,10-positions, the photo-dimerisation interrupts the
conjugation of anthracene, generating four o-xylene chro-
mophores. This was observed first by Coulson who showed
The UV–vis absorption spectra of ethanol solutions of 4,
registered after 2, 5, 7, 12 and 15 min of irradiation, are
reported in Figure 7. These spectra clearly show the
progressive growth of the absorption bands due to
K6
); the spectra were recorded in 4!10 M ethanol solutions.
Figure 6. Absorption spectra (3 vs l, nm) of: 9-anthroylacetone 1 (—) and 4 (

11964
F. Cicogna et al. / Tetrahedron 60 (2004) 11959–11968
Table 3. IS-MS data of compounds 1 and 4
a
C
C
C
Compound
[MCH]
[MCNH
4
]
[MCNa]
1
4
263 (100)
525 (12.5)
280 (7.5)
542 (70)
285 (20)
547 (100)
a
m/z (% relative intensity of peak).
C
m/zZ263 and 301 are attributable to the ions [M(1)CH]
and [M(1)CK] , respectively, M(1) being the molecular
mass of 9-anthroylacetone 1 (262). The other two peaks at
C
C
m/zZ542 and 547 are associated with [M(4)CNH ] and
4
C
M(4)CNa] , respectively, M(4) being the molecular mass
[
of 4 (524). These results clearly show that 4 photo-
dissociates in solution, and 1 can be obtained in high yields
in few minutes.
The photochemical dissociation in the solid state was
carried out on a solid sample of pure 4 placed in a 1 mm
optical path spectrophotometric quartz cell which was
irradiated with a medium pressure mercury lamp for 30 min.
During the first 15 min the lamp was switched on and off at
regular intervals; whereas, during the last 15 min of
irradiation the lamp was left on. The front surface
absorption spectra registered after 5, 15 and 30 min of
effective exposure to light are reported in Figure 8.
Figure 7. Absorption spectra (absorbance vs l, nm) of 4 in ethanol;
successive spectra taken after 0 ( ), 2 ( ), 5 ( ), 7 ( ), 12 ( ),
and 15 min ( ) of irradiation with a medium pressure mercury lamp.
9
-anthroylacetone 1. Since at 384 nm the photodimer 4
does not absorb and 1 exhibits an absorption maximum
3Z8300), the concentration of 1 in each sample can be
(
For the first 15 min of irradiation the amount of 1 increased
whereas the spectrum registered after 30 min of irradiation
was less intense than that registered after 15 min. This
should be connected to the short-wavelengths emission-
lines of the lamp during the cold period, i.e. the first 15 min
of irradiation, with respect to the emission of the Hg
stationary plasma during the last 15 min. As previously
noted, at 254 and 366 nm the absorption of 1 is much higher
than that of 4. Evidently, for prolonged exposition of the
sample to these wavelengths part of 1, which is formed
during the first 15 min of irradiation, reverts to 4. The
genesis of 1 was confirmed by UV–vis absorption
spectroscopy of the irradiated solid sample upon dissolution
in dichloromethane.
calculated from absorption at this wavelength (Table 2).
Interestingly, the amount of 1 formed for short irradiation
times is proportionally higher than that formed for long
exposure times. Evidently, the photochemical dissociation of
4 tends to a steady state with excellent yields of 1 (O80%)
that can be reached in 15 min of effective irradiation.
Further evidence of the photo-dissociation of 4 to give 1 was
obtained by analysing the irradiated sample by Ion Spray
Mass Spectroscopy (IS-MS). The IS-MS data for pure 1 and
4
are reported in Table 3.
Both the spectra show only the pseudo molecular peaks and
noteworthy in the IS-MS spectrum of 4 no peaks are present
in the range m/zZ263–285, where the spectrum of 1 has its
pseudo molecular peaks. It is interesting to underline that
rarely are the molecular peaks of anthrachene photodimers
The definitive confirmation of the presence of 1 came once
again from the IS-MS analysis of the reaction mixture that
8
b
detected, because generally the photodimer fragments to
anthracene, under electron impact conditions. Only
recently, by FAB(C) analysis and by GC-MS (EI
9
e
9
d
mode), the molecular peak of a photodimer has been
identified. The IS-MS spectrum of the ethanolic solution of
4
, irradiated for 15 min with the medium pressure mercury
lamp, shows four pseudo molecular peaks at m/zZ263
14.3%), 301 (79%), 542 (3%) and 547 (3%). The peaks at
(
Table 2. Concentration of 1 formed by irradiation of an ethanol solution of
4
with a medium pressure mercury lamp
a
a
a
a
a
2
min
5 min
7 min
0.047
5.4!10
12 min
15 min
b
384
c
A
C
1
0.027
0.037
4.5!10
53
0.058
7.0!10
82
0.060
7.2!10
84
K6
K6
K6
K6
K6
1
3.2!10
(%) 37
63
a
Total irradiation time.
Absorbance at 384 nm.
Molar concentration of 1 calculated by Lambert–Beer law 3
b
c
Figure 8. Front surface absorption spectra (absorbance vs l, nm) of a solid
sample of 4; successive spectra taken after 0 ( ), 5 ( ), 15 ( ), and
1
at 384 nmZ
8300.
3
0 min (
) of irradiation by a medium pressure mercury lamp.

F. Cicogna et al. / Tetrahedron 60 (2004) 11959–11968
11965
was dissolved in dichloromethane and diluted with
methanol. The IS-MS spectrum shows one peak at m/zZ
C
63 (42.5%) attributable to [M(1)CH] (M(1) is the molar
2
mass of 1) and two peaks, of the same intensity (47.5%), at
C
m/zZ542 and 547 attributable to [M(4)CNH ] and
4
C
M(4)CNa] (M(4) is the molar mass of 4), respectively.
[
All these results clearly show that 4 photo-dissociates to
give 1 even in the solid state; although, in this case, the
process takes place only at the surface level and then the
overall conversion is not high.
3. Conclusions
For the first time the photodimerisation of 9-anthroylace-
tone 1 to give the photodimer 4 has been observed and
Scheme 3.
1
studied. The H NMR spectra and the X-ray analysis reveal
that both 1 and 4 are in their enolic form in solution as well
as in the solid state.
Compound 4 dissociates to give 1, both thermally and
photochemically (Scheme 3).
1
photodimerises in solution as well as in the solid state by
The thermal dissociation is a well known phenomenon
extensively documented in the literature,
7
,9a,b,g,i
irradiation with different sources, the reaction being
accompanied by a colour variation from bright yellow to
colourless, a manifestation of the deep spectroscopic
differences between 1 and 4. No appreciable amounts of
other products are detected although, in same cases, small
quantities of anthraquinone are present, probably due to the
and 4
behaves like many other anthracene photodimers although
the accelerating effect on thermal dissociation played by the
presence of even small amounts of 1 had not been reported
previously. The photochemical dissociation of 4 is particu-
larly interesting because it can be carried out in solution as
well as in the solid state by irradiation with a medium
pressure mercury lamp. Some examples of this reaction
4
,7,8a,9d,19
presence of oxygen.
9
a,c,d,l
carried out in solution are reported in the literature,
but
The crystal structure of a 9-substituted anthracene deriva-
tive can be classified into three types according to the
geometric relationship between neighbouring molecules
in all cases the samples were irradiated at wavelengths that
seem to favour the photo-recombination of monomers since a
steady state is reached characterised by the prevalent
presence of the photodimer. Instead, under the irradiation
conditions adopted here, O80% yield of 1 can be obtained.
As said, 4 in the solid state can photo-dissociate. It is
noticeable that only a very few examples of photochemical
1
2
(
Fig. 9). In the g type structure the central ring of the two
anthracene units are not superimposed and anthracene
derivatives which crystallise with this structure hardly
photodimerise in the solid state; on the other hand, when
anthracene derivatives crystallise in the a or b type structure
the photodimerisation can take place. The X-ray analysis
of 1 reveals that it crystallises in the a-type structure, this
allowing compound 1 to undergo the photodimerisation to 4
in the solid state.
9
j,10d,20
1
0c
dissociation in this phase are reported in the literature.
4. Experimental
4.1. General
All solvents were used as received. The HPLC grade or
spectrophotometric grade solvents were used in the spectro-
photometric measurements. Microanalyses were performed
by the Laboratorio di Microanalisi, Facolt a` di Farmacia,
Universit a` di Pisa, Italy. 9-Anthroylacetone was prepared as
1
reported.
1
H NMR spectra were run at 200 MHz on a Varian Gemini
200 instrument. Ion Spray Mass Spectra (IS-MS) were
performed on a Perkin–Elmer Sciex API III plus triple
quadrupole mass spectrometer (Sciex Co., Thornhill, ONT,
Canada) equipped with an API ion source and an ionspray
interface. The spectra were obtained under the following
experimental conditions: ionspray voltage, 5.5 kV; orifice
voltage, 35 or 60 V. The samples were dissolved in a very
small amount of dichloromethane and diluted with methanol
to obtain a concentration ca. 10 mM. The irradiation of solid
Figure 9. Main crystalline packing modes found in the case of 9-substituted
anthracene derivatives.

11966
F. Cicogna et al. / Tetrahedron 60 (2004) 11959–11968
samples or solutions of 1 were carried out with a tungsten
lamp (100 W), with a xenon lamp (150 W) which emits at
lO250 nm and with a Argon-Ion laser Coherent Innova 600
the tip of a glass fibre, giving the results summarised in
Table 4. The structure solution was found in the centrosym-
metric C2/c space group. The asymmetric unit resulted to be
done by one half molecule placed beside an inversion
centre. The hydrogen atoms were localised in the difference
Fourier map and refined without constraints. The reliability
factors resulting from the final refinement cycle are listed in
Table 4.
(
solid samples or solutions of 4 were carried out with a
l Z351 nm; l Z364 nm, jointly). The irradiations of
1 2
medium pressure mercury lamp (125 W).
4
.2. Crystal structure determination of 1 and 4
The X-ray diffraction experiments were carried out at room
temperature (TZ293 K) by means of a Bruker P4
diffractometer, operating with a graphite-monochromated
Further details of crystal characterisations and structure
refinements have been deposited in the form of CIF files
with the Cambridge Crystallographic Data Centre. Depo-
sition references: CCDC 237079 and CCDC 237080 for 1
and 4, respectively. Copies of the data can be obtained, free
of charge, on application to the CCDC, 12 Union Rd,
Cambridge CB21EZ UK (fax: C44 1223 336 033); e-mail:
deposit@ccdc.cam.ac.uk. Bond distances and angles for
both 1 and 4 are provided in Supporting Information.
Mo-K radiation. The intensity data collection was carried
a
out with the u/2q scan mode, collecting a redundant set of
data. Three standard reflections were measured every 97
measurements to check sample decay. The intensities were
corrected for Lorentz and polarisation effects and for
2
1
absorption by means of a j-scan method. The absorption
correction was not applied due to the low absorption
coefficient. The structure solutions were obtained by means
4.3. Spectrophotometric measurements
2
2
of the automatic direct methods contained in SHELXS97
programme and the refinements, based on full-matrix least-
The UV–vis absorption spectra were measured at room
temperature in ethanol or dichloromethane solutions using a
Perkin–Elmer UV/VIS LAMDA EZ 201 spectrophotometer
with a spectral bandwidth of 1 nm. Front-surface absor-
bance measurements in the UV–vis spectral region (250–
2
22
squares on F , were done by means of the SHELXL97
programme. Some other utilities contained in the WINGX
2
3
suite were also used. The more relevant crystal parameters
are listed in Table 4.
5
00 nm) were performed on powder samples of 1 and 4
1
6
The crystals of compound 1 are yellow platelets flattened on
(
using the method reported by one of us. Experiments were
carried out on a conventional spectrofluorimeter (ISA
Fluoromax II with photon counting detection) suitably set
to detect light scattered from powder samples. The samples,
ca. 1 mm thick, were enclosed in a cell-holder suitably
designed to avoid specular reflections into the emission
100). One of them, glued at the end of a glass fibre, was
used for intensity data collection. Table 4 lists the essential
collection statistics. The structure solution was obtained in
ꢀ
the centrosymmetric P1 space group. The hydrogen atoms
were observed in the difference Fourier map, they was,
however, introduced in idealised positions and refined
letting them ‘to ride’ on the connected heavy atoms. The
final refinement cycle gave the reliability factors listed in
Table 4.
2
4
monochromator in measurements on powders. Thus, the
incident angle was 358. Powdered BaSO was used as
4
2
5
reference light scatterer. The front-surface absorbance
measurements reported in Figure 1(A) were performed on
powder samples of 1 and 4 obtained by mixing a
The crystals of compound 2 are colourless prisms. The
intensity data collection was done on one of them, glued on
dichloromethane solution of 1 or 4 with BaSO . The slurries
4
so obtained were dried under vacuum, re-powdered by a
mortar and examined. The relative ratios were of 1 mg 1/5 g
of BaSO and 1 mg 4/0.5 g of BaSO , respectively.
Table 4. Crystal data and structure refinements
4
4
Compound
1
4
Empirical formula
Formula weight
Crystal system
Space group
C
18
H
14
O
2
C
36
H
28
O
4
4.4. Purification of 4
262.29
Triclinic
P 1ꢀ (no. 2)
524.58
Monoclinic
C2/c (no. 15)
23.784(3)
9.180(1)
To 200 mg of a mixture of 1 (about 60%) and 4 (about 40%)
was added of diethyl ether (20 mL). The resulting suspen-
sion was stirred for 1 min and decanted. The residual solid
was treated twice with diethyl ether (5 mL). The ethereal
extracts contains 1 with minor amounts of 4 and the solid
residue is constituted by almost pure 4. A further amount of
4 was recovered by drying the ethereal extracts under
reduced pressure and then suspending the resulting solid
residue in 10 mL of diethyl ether. The ethereal solution was
collected and dried under reduced pressure. The solid so
obtained was further extracted with 5 mL of diethyl ether
and, once again, the solution was separated from the solid
and dried. The last sequence was repeated 2 times. All
residual solid phases were dissolved in dichloromethane and
dried under reduced pressure. About 100 mg almost pure 4
was so obtained which can be further purified by two
methods. (A) 1 mL of a benzenic solution of impure 4 was
purified by column (internal diameter, 15 mm; length,
˚
a/A
8.521(2)
8.980(2)
10.079(2)
71.34(1)
68.67(1)
74.24(2)
670.0(2)
2
1.300
0.084
2864
2346 [0.0259]
183
˚
b/A
˚
c/A
16.553(2)
˚
a/A
˚
b/A
132.13(1)
˚
g/A
3
˚
U/A
2680.3(6)
4
1.300
0.084
2186
1749 [0.0168]
237
0.0399, 0.1022
0.961
Z
K3
D
m/mm
calcd/Mg M
K1
No. measured
No. unique [Rint
No. parameters
]
R
1
, wR
2
[IO2s(I)]
Goodness of fit on F
0.0550, 0.1254
1.032
2
P
P
P
P
2
o
2
2
2
o
2
1=2
R Z jjF jKjF jj= F ; wR Z½ ½wðF KF Þ ꢀ= ½wðF Þ ꢀꢀ ; wZ
1
o
c
o
2
c
2
2
o
2
o
2
c
1
=½s ðF ÞCBQꢀ where QZ½MAXðF ; 0ÞC2F ꢀ=3; Goodness-of-fitZ
P
2
o
2
c
2
1=2
½
½wðF KF Þ ꢀ=ðNKPÞꢀ , where N, P are the numbers of observations
and parameters, respectively.

F. Cicogna et al. / Tetrahedron 60 (2004) 11959–11968
11967
2
1
50 mm) chromatography on silica gel 60 (230–400 mesh,
whose spot was about 2 mm , at the centre of the cell
window. During the irradiation the cell was vertically
moved. After 5 min, the portion exposed to laser beam
showed a well defined white strip. The reaction was
monitored by front surface absorption spectroscopy with
illumination of the above strip.
Merck) using benzene as eluant, the first band eluted was
collected and dried. The obtained solid is colourless and
very pure. By this method a small amount of 4 can be
obtained because 4 is slightly soluble in benzene. (B) About
1
00 mg of impure 4 was dissolved in about 25 mL of
dichloromethane. The solution was extracted with portions
of aqueous NaOH 2% (3!25 mL). The organic phase was
discarded. Each aqueous phase was acidified with a few
drops of HCl 35%, which caused the precipitation of a white
solid. The resulting suspensions were extracted (2!20 mL)
with dichloromethane. The organic extracts were then dried
4
.7. Thermal dissociation of 4
1
All tests were run on 2 mg of 4 placed in a H NMR tube
which was heated at various temperatures as shown in
the Section 2. Analysis of the samples, dissolved in
1
d -chloroform, were performed by H NMR.
over anhydrous Na SO , then concentrated to 2 mL under
2
4
1
reduced pressure (17 mmHg), and finally, chromato-
graphated over a column (internal diameter, 5 mm; length,
4.8. Photochemical dissociation of 4 in solution
3
0 mm) of silica gel 60 (230–400 mesh, Merck), using
dichloromethane as eluant. A colourless solution was
collected and dried under reduced pressure (17 mmHg) to
give pure 4. Anal. Calcd for C H O (524) C, 82.4; H,
K6
Three millilitres of an ethanol solution (4.27!10 M) of 4
was placed, under dinitrogen atmosphere, in a quartz
10!10 mm spectrophotometric cuvette and irradiated by
a medium pressure mercury lamp. The sample was placed
about 3 cm far from the lamp in an equatorial position and
oriented so to optimise the interception of the light rays.
Both sample and lamp were cooled by a ventilator. The
lamp was switched on for 30 s, then switched off for at least
3
6 28 4
5
.3%. Found: C, 82.4; H, 5.2%.
4
.5. Photo-dimerisation of 1 in solution
1
Three H NMR tubes, each containing 3 mg of 1 in 0.5 mL
of d -dichloromethane, were irradiated with three different
2
1
min. After 2, 5, 7, 12 and 15 min of effective irradiation,
1
sources and analysed by H NMR. Sample A was irradiated
the sample was analysed by UV–vis absorption spectra. At
the end of the experiment the sample was analysed by IS
MS.
with a xenon lamp for 3 h: the analysis of the sample
indicated the conversion of 50% of 1 to 4. Sample B was
exposed to sun light for 3 h: the conversion of 1 to 4 was
about 28%. Sample C was exposed to the near-UV Argon-
ion laser beam (0.010 W) for 4 h: the conversion of 1 to 4
was about 11%.
4.9. Photochemical dissociation of 4 in the solid state
About 130 mg of pure 4 were placed in a 1 mm thick
spectrophotometric cell and irradiated in the same geo-
metric arrangement as above, with the only difference that
the distance from the plasma source was larger (ca. 6 cm).
The sample was irradiated as before, the lamp being
switched on and off at regular intervals during the first
4
.6. Photo-dimerisation of 1 in the solid state
1
Experiment A. 2 mg of 1 was placed in a H NMR tube
under dinitrogen atmosphere and was irradiated for 24 h by
a 100 W tungsten lamp, at 15 8C, the distance between the
sample and the lamp being about 10 cm. At the end of
15 min of irradiation; finally, during the last 15 min, the
lamp was left on. After 5, 15 and 30 min of irradiation the
sample was analysed by front surface absorption spec-
troscopy. Spectral bandwidths of 0.25 and 4 mm were
employed for the excitation and emission slits, respectively.
The integration time was 0.5 s. 10 mg of the irradiated
irradiation, the solid sample was dissolved in d -chloroform
1
1
and the H NMR spectrum was registered. 4 was formed in a
3
5% yield.
Experiment B. 1 (ca. 5 mg) was placed between two thin
glass slides and irradiated by a 150 W xenon lamp
K3
sample was used to prepare a 3.85!10 M dichloro-
methane solution which was analysed by UV–vis
spectroscopy. The irradiated sample was also analysed by
IS MS.
(collocated at about 15 cm from the sample) for 4 h.
The H NMR analysis of the irradiated sample, dissolved in
d1-chloroform, revealed the presence of 70% of 4.
1
Experiment C. 50 mg of 1, containing ca. 7% of 4, was
placed in a watch glass horizontally disposed and irradiated
with a vertical UV laser beam (0.088 W) that was enlarged
by means of a quartz lens up to 2 cm diameter. The sample
was periodically stirred in order to expose to radiation fresh
portions of solid. During irradiation the well shaped yellow
Acknowledgements
We are grateful to Dr. A. Raffaelli (C.N.R., Pisa, Italy) for
the measurement of the mass spectra. Financial supports
from the M.I.U.R. (Rome, Italy) and from the University of
Pisa are gratefully acknowledged.
crystals of 1 turned opaque white and flaked off. After 2 h of
irradiation, H NMR analysis of the sample in d -chloro-
1
1
form showed the presence of 38% of 4. The temperature of
the sample remained practically unchanged by irradiation.
Supplementary data
Experiment D. About 100 mg of 1, placed in a quartz
spectrophotometric cuvette (1 mm optical path), was
exposed perpendicular to the UV laser beam (0.005 W)
Supplementary data associated with this article can be found
in the online version, at 10.1016/j.tet.2004.09.070

11968
F. Cicogna et al. / Tetrahedron 60 (2004) 11959–11968
References and notes
. Evans, D. F. J. Chem. Soc. 1961, 1987–1993.
J.; Morita, M.; Puza, M. Chem. Phys. Lett. 1976, 42, 288–292.
k) Bradshaw, A. W.; Chapman, O. L. J. Am. Chem. Soc. 1967,
9, 2372–2375. (l) Livingston, R.; Wei, K. S. J. Am. Chem.
(
8
1
2
Soc. 1967, 89, 3098–3100.
0. Ihmels, H.; Leusser, D.; Pfeiffer, M.; Stalke, D. Tetrahedron
. Hartmann, H. J. Prakt. Chem. 1986, 328, 755–762. (b) Ilge,
H.-D.; Birckner, E.; Fassler, D.; Kozmenko, M. V.; Kuz’min,
M. G.; Hartmann, H. J. Photochem. 1986, 32, 177–189.
1
1
2
000, 56, 6867–6875. (b) Ito, Y.; Olovsson, G. J. Chem. Soc.,
Perkin Trans. 1 1997, 127–133. (c) Kaupp, G. Angew. Chem.
Int. Ed. 1992, 31, 595–598. (d) Chandross, E. A.; Ferguson, J.
J. Chem. Phys. 1966, 45, 3564–3567. (e) Ito, Y.; Craig, D. P.;
Sarti-Fantoni, P. Chem. Commun. 1966, 20, 742–743.
(
c) Ilge, H.-D.; Fassler, D.; Hartmann, H. Z. Chem. 1984, 24,
18–219.
. Akutsu, B.; Nakano, Y.; Usui, Y. Anal. Sci. 1990, 6, 827–831.
b) Sato, S.; Wada, M. Bull. Chem. Soc. Jpn 1970, 43,
2
3
(
1. Wu, D.-Y.; Zhang, L.-P.; Wu, L.-Z.; Wang, B.-J.; Tung, C.-H.
Tetrahedron Lett. 2002, 43, 1281–1283. (b) Grimme, S.;
Peyerimhoff, S. D.; Bouas-Laurent, H.; Desvergne, J.-P.;
Becker, H.-D.; Sarge, S. M.; Dreeskamp, H. Phys. Chem.
Chem. Phys. 1999, 1, 2457–2462. (c) Tung, C.-H.; Guan, J.-Q.
J. Org. Chem. 1998, 63, 5857–5862.
1955–1962. (c) Sato, S.; Wada, M.; Seki, T. Jpn. J. Appl. Phys.
1968, 7, 7–13.
4
5
. Kercher, M.; K o¨ nig, B.; Zieg, H.; De Cola, L. J. Am. Chem.
Soc. 2002, 124, 11541–11551.
. Carano, M.; Cicogna, F.; D’Ambra, I.; Gaddi, B.; Ingrosso, G.;
Marcaccio, M.; Paolucci, D.; Paolucci, F.; Pinzino, C.; Roffia,
S. Organometallics 2002, 21, 5583–5593. (b) Carano, M.;
Cicogna, F.; Houben, J. L.; Ingrosso, G.; Marchetti, F.;
Mottier, L.; Paolucci, F.; Pinzino, C.; Roffia, S. Inorg. Chem.
1
2. Cohen, M. D. Angew. Chem., Int. Ed. Engl. 1975, 14, 386–393.
3. Trouts, T. D.; Tyson, D. S.; Pohl, R.; Kozlov, D. V.; Waldron,
A. G.; Castellano, F. N. Adv. Funct. Mater. 2003, 13, 398–402.
1
(
b) Hiraga, H.; Morozumi, T.; Nakamura, H. Tetrahedron Lett.
2002, 41, 3396–3409. (c) Carano, M.; Careri, M.; Cicogna, F.;
D’Ambra, I.; Houben, J. L.; Ingrosso, G.; Marcaccio, M.;
Paolucci, F.; Pinzino, C.; Roffia, S. Organometallics 2001, 20,
2002, 43, 9093–9095. (c) Ojima, T.; Akutsu, H.; Yamada, J.;
Nakatsuji, S. Polyhedron 2001, 20, 1335–1338.
3
478–3490. (d) Cicogna, F.; Colonna, M.; Houben, J. L.;
Ingrosso, G.; Marchetti, F. J. Organomet. Chem. 2000,
93–594, 251–266.
14. Irie, M. Chem. Rev. 2000, 100, 1683–1684. (b) Irie, M. Chem.
Rev. 2000, 100, 1685–1716.
5
15. Becker, H.-D.; Andersson, K. J. Org. Chem. 1983, 48,
4542–4549.
6
7
. Cicogna, F.; Ingrosso, G. Unpublished results.
. Becker, H.-D.; Langer, V.; Becker, H.-C. J. Org. Chem. 1993,
16. Zandomeneghi, M.; Festa, C.; Carbonaro, L.; Galleschi, L.;
Lenzi, A.; Calucci, L. J. Agric. Food Chem. 2000, 48,
2216–2221.
58, 6394–6396.
8
. Bouas-Laurent, H.; Castellan, A.; Desvergne, J.-P.;
Lapouyade, R. Chem. Soc. Rev. 2001, 30, 248–263.
17. Coulson, C. A.; Orgel, L. E.; Taylor, W.; Weiss, J. J. Chem.
Soc. 1955, 2961–2962.
(
b) Bouas-Laurent, H.; Castellan, A.; Desvergne, J.-P.;
Lapouyade, R. Chem. Soc. Rev. 2000, 29, 43–55. (c) Becker,
H.-D. Chem. Rev. 1993, 93, 145–172.
. Nakatsuji, S.; Ojima, T.; Akutsu, H.; Yamada, J. J. Org. Chem.
18. Calvert, J. G.; Pitts, J. N., Jr. Photochemistry; Wiley:
New York, 1966; p. 297.
9
19. Toshima, N.; Sugano, T.; Hirai, H. Can. J. Chem. 1984, 62,
2047–2053.
2
002, 67, 916–921. (b) McSkimming, G.; Tucker, J. H. R.;
Bouas-Laurent, H.; Desvergne, J.-P. Angew. Chem., Int. Ed.
000, 39, 2167–2169. (c) Mori, Y.; Maeda, K. J. Chem. Soc.,
20. Anderson, B. F.; Ferguson, J.; Morita, M.; Robertson, G. B.
J. Am. Chem. Soc. 1979, 101, 1832–1840.
2
Perkin Trans. 2 1996, 113–119. (d) Schottelius, M. J.; Chen, P.
J. Am. Chem. Soc. 1996, 118, 4896–4903. (e) Becker, H.-D.;
Langer, V. J. Org. Chem. 1993, 58, 4703–4708. (f) Becker,
H.-D.; Amin, K. A. J. Org. Chem. 1989, 54, 3182–3188.
21. North, C. T.; Phillips, C.; Mathews, F. S. Acta Crystallogr. A
1968, 24, 351–359.
22. Sheldrick, G. M. SHELXTL-Plus, Rel. 5.1; Brucker A.X.S.
Instruments Inc.: Madison, Wisconsin, USA, 1997.
23. Farrugia, L. J. J. Appl. Crystallogr. 1999, 32, 837–838.
24. Zandomeneghi, M. J. Agric. Food Chem. 1999, 47, 878–882.
25. Hapke, B. In Theory of Reflectance and Emittance Spec-
troscopy; Arvidson, R. E., Rycroft, M. J., Eds.; Cambridge
University Press: Cambridge, UK, 1993; Chapters 10 and 11.
(
g) Becker, H.-D.; Elebring, T.; Sandros, K. J. Org. Chem.
982, 47, 1064–1068. (h) Castellan, A.; Desvergne, J.-P.;
Bouas-Laurent, H. Chem. Phys. Lett. 1980, 76, 390–397.
i) Bergmark, W. R.; Jones, G. II; Reinhardt, T. E.; Halpern,
A. M. J. Am. Chem. Soc. 1978, 100, 6665–6673. (j) Ferguson,
1
(