A novel photoreversible photochromic system involving a hydrogen transfer/cyclization sequence

Article abstract of DOI:10.1039/b304861f

A novel photoreversible photochromic system, 3-(2-benzyl-benzoyl)-1,2-R,R ¹-4(1H)-quinolinones/12-hydroxy-5-R-5a-R ¹-6-phenyl-5a,6-dihydrobenzo[b]acridin-11(5H)-ones, is described.

Full text of DOI:10.1039/b304861f

A novel photoreversible photochromic system involving a hydrogen
transfer/cyclization sequence†
Vladimir Lokshin,*a Magali Valès,a André Samat,a Gérard Pèpe,a Anatoly Metelitsab and Vladimir
Khodorkovsky*c
a Université de la Méditerranée, UMR CNRS 6114, 13288 Marseille 09, France.
E-mail: lokshin@luminy.univ-mrs.fr; Fax: 33 4 91829301; Tel: 33 4 91829405
b Institute of Physical Organic Chemistry, Rostov University, Rostov-on-Don, 344090, Russia.
E-mail: met@ipoc.rsu.ru; Fax: 7 863 243 46 67; Tel: 7-863 243 34 00
c
Department of Chemistry, Ben Gurion University of the Negev, Beer Sheva 84105, Israel.
E-mail: hodor@bgumail.bgu.ac.il; Fax: 972 8-6472940; Tel: 972-8-6472188
Received (in Cambridge, UK) 7th May 2003, Accepted 27th June 2003
First published as an Advance Article on the web 14th July 2003
A novel photoreversible photochromic system, 3-(2-benzyl-
benzoyl)-1,2-R,R1-4(1H)-quinolinones/12-hydroxy-5-R-5a-
R1-6-phenyl-5a,6-dihydrobenzo[b]acridin-11(5H)-ones, is
described.
(5aS,6R,11aS)-5,5a-dimethyl-6-phenyl-5a,11a-dihydrobenzo-
[b]acridine-11,12(5H,6H)-dione (3)‡ (lmax = 386 nm). In the
presence of catalytic amounts of a base, the equilibrium 3 Ù 2
is established rapidly and irradiation of such solutions at 490 nm
led to the recovery of 1. Both 2 and 3 are fluorescent (lmax
=
Photoreversible photochromic systems are currently the focus
of numerous studies owing to their potential as components of
optoelectronic devices.1 Of the variety of known photochromic
compounds, furylfulgides2 and dithienylethenes3 are the only
thermally irreversible materials which can be used for optical
information storage and processing. The photochromic reaction
in both systems is based on the reversible hexatriene–
cyclohexadiene photocyclization, one of the most intensively
studied photochemical reactions. Another reversible photo-
chemical process, the photoenolization of o-alkylphenylk-
etones,4 has been thoroughly studied during the past four
decades, but these derivatives attracted little interest for
photochromic applications, since the corresponding colored
photoproducts, o-xylylenols formed upon relaxation of the
initial triplet biradical, are unstable and either convert back to
the starting materials or undergo fast irreversible reactions with
dienophiles and oxygen.
557 and 443 nm, quantum yields 0.08 and 0.19 in toluene,
respectively). No transients were detected and two isosbestic
points were observed before the photostationary state was
reached. Similar behavior was observed in acetonitrile, how-
ever, formation of the colored product 2 and its dark conversion
into 3 proceeded more rapidly.
The preparative experiments were carried out in toluene
(xenon lamp, 300–400 nm band pass filter). The reaction
mixture was separated on a silica column to afford derivative 3
as the main product (80% of reacted 1) along with 2 (4%).
In a tentative mechanism, the initial photoenol or a biradical
(formed upon irradiation, which is in fact a triplet state of the
photoenol) can undergo further reversible cyclization owing to
delocalization of the spin over the heterocyclic moiety (Scheme
1). In general, the hydrogen atom can be transferred to either the
quinolinone (route a) or benzoyl (route b) carbonyl groups. The
resulting 1,6- and 1,4-biradicals can, after rotation of the o-
benzylphenyl fragment, cyclize affording a mixture of 2, 5 and
their respective S,S-isomers 4 and 6. The 1,4-biradical can also
relax into the usual o-xylylenol transient 7 (route c).4 It is
noteworthy that isomers 4–7 were not detected either spec-
troscopically or in preparative experiments.
We found that attaching a conjugated moiety to the carbonyl
group of o-xylylenol precursors can provide access to the new
type of photoreversible systems. Here we report on the
photochemical behavior of 3-(2-benzylbenzoyl)-1,2-dimethyl-
4(1H)-quinolinones (1, R
= = Me).‡ Irradiation of
R1
degassed solutions of 1 (lmax = 327 nm) in toluene (Fig. 1) led
to the formation of (5aS,6R)-12-hydroxy-5,5a-dimethyl-
6-phenyl-5a,6-dihydrobenzo[b]acridin-11(5H)-one (2).‡ Upon
irradiation derivative 2 (lmax = 490 nm) can be converted back
into 1. In the dark, derivative 2 is slowly converted into
Exclusive formation of 2 can be rationalized by the results of
quantum mechanical calculations,§ which showed that the bent
conformation of 1, in which hydrogen atom transfer to the
quinolinone carbonyl is possible (C…O distance 3.50 Å, typical
of C–H…O hydrogen bonds,5 Fig. 2), corresponds to the
equilibrium geometry, and is more stable than conformations
enabling hydrogen transfer to the benzoyl carbonyl. The longest
wavelength absorption band of 1 (calc. lmax = 342 nm, f =
† Electronic supplementary information (ESI) available: supporting data for
Fig. 1 Irradiation of a solution of 1 in toluene: 1, 1 min; 2, 3 min; 3, 5 min;
2 and 3. See http://www.rsc.org/suppdata/cc/b3/b304861f/
4, 9 min; 5, 15 min (saturation).
2080
CHEM. COMMUN., 2003, 2080–2081
This journal is © The Royal Society of Chemistry 2003

Scheme 1
0.0032) corresponds to the HOMO–LUMO transition within
Reversible regio- and stereoselective photocyclization of
derivatives 1 affords not only colored but also, unlike
furylfulgides and dithienylethenes,2 fluorescent products,
which is important for non-destructive readout. Another
particularly promising potential application is to utilize the
ability of the photoproducts to form complexes with heavy
metal ions.
the quinolinone moiety and involves the CH2 group (Fig. 2).
The second peak observed at 327 nm should involve the benzoyl
carbonyl group and corresponds to predominantly HOMO21
–
LUMO+1 transition (calc. 323 nm, f = 0.0002). Note also that
the triplet 1,6-biradical is 3.2 kcal mol21 more stable than the
1,4-biradical. In fact to the best of our knowledge formation of
the 1,6-biradicals, in contrast to the 1,4-biradicals, is unprece-
dented, although the possibility of direct 1,7-hydrogen abstrac-
tion giving rise to the 1,5-biradicals has been discussed.6 We
also cannot exclude the possibility that the excited state
intramolecular proton transfer mechanism,7 which should give
rise to the same products, is involved. Mechanistic aspects of
the reaction are currently under study.
Notes and references
‡ 1 was synthesized using a modified procedure;8 colorless crystals, m.p.
161–162 °C, the structure was confirmed by X-ray technique. 2: red oil, 1H
NMR: d (500 MHz, CD2Cl2): 17.19 (1H, s, OH), 8.10 (1H, dd, J1 = 7.7, J2
= 1.5, C10-H), 7.78 (1H, dd, J1 = 7.7, J2 = 1.8, C1-H), 7.45 (2H, m, o-Ph-
H), 7.44 (1H, td, J1 = 7.4, J2 = 1.5, C8-H), 7.36 (1H, td, J1 = 7.5, J2 = 1.3,
C9-H), 7.32 (1H, d, J = 7.5, C7-H), 7.27 (1H, ddd, J1 = 8.6, J2 = 7.1, J3
= 1.7, C3-H), 7.18 (2H, m, m-Ph-H), 7.12 (1H, m, p-Ph-H), 6.69 (1H, ddd,
J1 = 7.8, J2 = 7.1, J3 = 0.9, C2-H), 6.47 (1H, d, J = 8.6), 4.48 (1H, s, C6-
H), 2.76 (3H, s, NMe), 1.51 (3H, s, CMe). 3: yellowish crystals, m.p.
191–192 °C. 1H NMR: d (500 MHz, 213 K, CD2Cl2): 7.96 (1H, d, J = 7.8,
C10-H), 7.76 (1H, dd, J1 = 7.8, J2 = 1.5, C1-H), 7.58 (1H, td, J1 = 7.5, J2
= 1.0, C8-H), 7.39 (2H, m, o-Ph-H), 7.36 (1H, t, J = 6.9, C3-H), 7.37 (1H,
td, J1 = 7.7, J2 = 1.7, C9-H), 7.27 (1H, d, J = 7.7), 7.24 (3H, m, m,p-Ph-
H), 6.73 (1H, t, J = 7.3, C2-H), 6.69 (1H, d, J = 8.7, C4), 4.82 (1H, s, C11a-
H), 3.60 (1H, s, C6-H), 2.76 (3H, s, NMe), 1.11 (3H, s, CMe). NMR spectra
of 2 and 3 are temperature dependent and were assigned using the COSY
and HMBC techniques. These results will be discussed elsewhere. Crystal
data for C25H21NO2 (3): M = 367.43, monoclinic, space group P21/n, a =
7.3820(10), b = 28.831(4), c = 9.2480(10) Å, b = 97.670(10)°, V =
1950.6(4) Å3, Z = 4, m = 0.079 mm21, T = 25 °C. 2253 independent
reflections were used in the refinement. Refinement of 338 parameters gave
wR2 = 0.26 (all data) and R1 = 0.0423 for 952 data with I > 2sI. CCDC
197132. See http://www.rsc.org/suppdata/cc/b3/b304861f/ for crystallo-
graphic data in CIF or other electronic format.
Assuming direct hydrogen transfer to the quinolinone
carbonyl as shown in Fig. 2 can explain not only the absence of
isomer 5, but also the predominant formation of the S,R-isomer
2 over the S,S-isomer 4. Indeed, the triplet biradical, which can
form from this conformation and is the precursor of 2, is
indicated by the calculations to be more stable than the biradical
precursor of 4 by 3.9 kcal mol21 (whose formation requires
additional rotation over the CH–CPh bond). Moreover, the
calculations showed that within the series of 5aS,6R isomers,
derivative 3 is the most stable. Thus, isomers 2 and 5 are 6 and
7 kcal mol21 less stable than 1 and 8–9 kcal mol21 less stable
than 3, which accounts for the thermal conversion of 2 into 3.
The 5aS,6R,11aR isomer of 3 is the least stable (11 kcal mol21
less than 1).
The above experimental and computational results imply that
the outcome of the photoreaction of 1 and its reversibility
should strongly depend on the nature of substituents R and R1,
in particular their bulkiness, since the differences in stability of
isomers 2–6 that stem mostly from steric strain, do not exceed
10 kcal mol21. Indeed, whereas derivative 1 (R = Et, R1 = Me)
behaves similarly to 1 (R = R1 = Me), different chemistry was
observed in cases when R1 = Ph. These results will be
published elsewhere.
§ The equilibrium geometries were found by molecular dynamics simula-
tions (Hyperchem, ver. 5.0) at the PM3 level and further optimized at the
HF/6-31G(d) level. The spectra were calculated using the TD B3LYP/
6-31G(d)//HF/6-31G(d) model chemistry (M. J. Frisch, et al., Gaussian 98,
Revision A.7, Gaussian, Inc., Pittsburgh, PA, 1998).
1 Photoreactive Materials for Ultrahigh Density Optical Memory, ed. M.
Irie, Elsevier, Amsterdam, 1994.
2 J. Whittal, in Photochromism, ed. H. Durr and H. Bouas-Laurent,
Elsevier, Amsterdam, 1990, ch. 9, p. 467.
3 M. Irie, Chem. Rev., 2000, 100, 1685.
4 J. C. Scaiano, Acc. Chem. Res., 1982, 15, 252; P. J. Wagner, M. Sobczak
and B.-S. Park, J. Am. Chem. Soc., 1998, 120, 2488.
5 G. R. Desiraju, Acc. Chem. Res., 1996, 29, 441.
6 P. G. Sammes, Tetrahedron, 1976, 32, 405.
7 S. M. Ormson and R. G. Brown, Prog. React. Kinet., 1994, 19, 211; K.
Kobayashi, M. Iguchi, T. Imakubo, K. Iwata and H. Kamaguchi, Chem.
Commun., 1998, 763.
8 K. R. Huffman, M. Loy and E. F. Ullman, J. Am. Chem. Soc., 1965, 87,
5417.
Fig. 2 Equilibrium geometry and LUMO of 1.
CHEM. COMMUN., 2003, 2080–2081
2081