The complex photochemistry of 2,3-dibenzylidenesuccinates

Article abstract of DOI:10.1139/v04-143

The photochemistry of diethyl E,E-2,3-(3,4,5-trimethoxybenzylidene) succinate (8) is solvent dependent. In both protic and aprotic solvents, there is a photoequilibrium established between 8 and its E,Z-isomer (9). In chloroform at high light intensity, very little 9 is formed and the main product is 1,4-dihydronaphthalene (10), formed via photoinduced intramolecular [1,3]-sigmatropic hydrogen shift within an intermediate 1,8a-dihydronaphthalene (11). In protic solvents, irradiation of either 8 or 9 ultimately gives primarily the cis-1,2-dihydronaphthalene product (13), along with smaller amounts of the trans isomer (14). By using deuterated solvents, it was shown that 13 and 14 are formed by solvent protonation (or deuteration) of the 1,8a-dihydronaphthalene intermediate (11 or 12).

Full text of DOI:10.1139/v04-143

1663
The complex photochemistry of 2,3-
dibenzylidenesuccinates
Tokouré Assoumatine, Brigitte L. Yvon, and James L. Charlton
Abstract: The photochemistry of diethyl E,E-2,3-(3,4,5-trimethoxybenzylidene)succinate (8) is solvent dependent. In
both protic and aprotic solvents, there is a photoequilibrium established between 8 and its E,Z-isomer (9). In chloro-
form at high light intensity, very little 9 is formed and the main product is 1,4-dihydronaphthalene (10), formed via
photoinduced intramolecular [1,3]-sigmatropic hydrogen shift within an intermediate 1,8a-dihydronaphthalene (11). In
protic solvents, irradiation of either 8 or 9 ultimately gives primarily the cis-1,2-dihydronaphthalene product (13), along
with smaller amounts of the trans isomer (14). By using deuterated solvents, it was shown that 13 and 14 are formed
by solvent protonation (or deuteration) of the 1,8a-dihydronaphthalene intermediate (11 or 12).
Key words: 2,3-dibenzylidenesuccinate, photocyclization, dihydronaphthalene, lignan.
Résumé : La photochimie du E,E-2,3-(3,4,5-triméthoxybenzylidène)succinate de diéthyle (8) dépend de la nature du
solvant. Dans les solvants tant protique qu’aprotique, il s’établit un photoéquilibre entre le composé 8 et son isomère
E,Z (9). Dans le chloroforme, à une intensité lumineuse intense, il se forme très peu de composé 9 et le produit princi-
pal est le 1,4-dihydronaphtalène (10) qui se forme, à partir de l’intermédiaire 1,8a-dihydronaphtalène (11), par un dé-
placement photoinduit d’hydrogène [1,3]-sigmatropique. Dans les solvants protiques, l’irradiation des composés 8 ou 9
conduit toujours à des quantités importantes du produit cis-1,2-dihydronaphtalène (13) aux côtés de plus faibles quanti-
tés de l’isomère trans (14). En utilisant des solvants deutérés, il a été possible de démontrer que les composés 13 et 14
se forment par protonation (ou deutération) d’un intermédiaire 1,8a-dihydronaphtalène (11 ou 12) par le solvant.
Mots clés : 2,3-dibenzylidènesuccinate, photocyclisation, dihydronaphtalène, lignane.
[Traduit par la Rédaction] Assoumatine et al. 1667
Introduction
2. The 1,8a-DHNs 3 and 4 are seldom isolated owing to
their thermal and photochemical instability, and the final
isolated products (other than 1 and 2) are more often the 1,4-
DHN 5 and (or) the 1,2-DHNs 6 and 7 (or the corresponding
naphthalenes if oxygen is present). On the basis of tempera-
ture and pH effects, Heller and co-workers (6) tentatively
concluded that 1,4-DHN 5 was formed photochemically
from 3 and 4 (a [1,3]-intramolecular sigmatropic H-shift)
and that 1,2-DHNs 6 and 7 were formed by an acid-
catalysed rearrangement of 3 and 4, respectively, rather than
by a thermal intramolecular [1,5]-sigmatropic shift (Scheme 1)
(6). In previous papers by Heller’s group, it had been as-
sumed that the latter rearrangement involved an intramole-
cular [1,5]-sigmatropic H-shift (17, 18, 20).
While there have been many published studies of the
photochemistry of cyclic derivatives of dibenzylidenesuccin-
ates, such as the fulgides and fulgimides discussed above,
there have been fewer published examples of the study of
the photochemistry of the corresponding acyclic succinate
derivatives. In an earlier paper (23), we reported the first ex-
ample of the photolysis of a dibenzylidenesuccinate diester.
In that preliminary study, we reported that the photolysis of
E,E-2,3-(3,4,5-trimethoxybenzylidene)succinate (8) in etha-
nol acidified with trifluoroacetic acid (TFA), followed by
column chromatography of the photolysis product, gave the
dihydronaphthalene esters 13 (34% yield) and 14 (2% yield)
(Scheme 2) (23). We have now conducted a more detailed
study of the photochemical behaviour of 8 in various sol-
vents and observed that the products depend quite strongly
There is a rich literature on the photochemistry of E,E-
2,3-dibenzylidenesuccinate derivatives (1–23). Although the
majority of the work deals with the photochemical cycli-
zation of fulgides (the cyclic anhydrides of E,E-dibenzyl-
idenesuccinic acids; 1, X = O), there are also examples of
photoreactions of fulgimides (the cyclic imides of E,E-
dibenzylidenesuccinic acids; 1, X = NR), dibenzylidene-
butyrolactones, and acyclic derivatives. Heller and co-
workers (6) have provided a general mechanistic scheme to
explain the photochemistry that has been observed for vari-
ous fulgides and fulgimides (Scheme 1, where R represents
various aromatic substitution patterns).
Although the results of irradiation vary with substituents
R and solvents, photoinduced isomerization between the
E,E- and E,Z-isomers 1 and 2 is generally observed. Isomers
1 and 2 also undergo photochemical conrotatory ring closure
to form the cis- and trans-1,8a-dihydronaphthalenes (1,8a-
DHNs) 3 and 4, respectively. The thermal disrotatory open-
ing of 3 and 4 to give 2 and 1, respectively, is assumed to be
the mechanism for the photoinduced isomerization of 1 and
Received 31 January 2004. Published on the NRC Research
Press Web site at http://canjchem.nrc.ca on 12 January 2005.
T. Assoumatine, B.L. Yvon, and J.L. Charlton.1
Department of Chemistry, University of Manitoba, Winnipeg,
MB R3T 2N2, Canada.
1Corresponding author (e-mail: charltn@cc.umanitoba.ca).
Can. J. Chem. 82: 1663–1667 (2004)
doi: 10.1139/V04-143
© 2004 NRC Canada

1664
Can. J. Chem. Vol. 82, 2004
Scheme 2. Previous study of photolysis of E,E-2,3-(3,4,5-
Scheme 1. Photoreactions of fulgides (X = O) and fulgimides
(X = NH, NR, or NAr).
trimethoxybenzylidene)succinate (8).
Scheme 3. Irradiation of 8 in ethyl acetate.
on the solvent and reaction conditions. By using deuterated
solvents, we have established the mechanisms of the various
prototropic rearrangements.
It is hoped that a better understanding of the photochem-
istry of 2,3-dibenzylidenesuccinates will make their photo-
chemical reactions more useful for the synthesis of lignans,
particularly the aryltetralin and aryldihydronaphthalene
types. This goal has been partially achieved with the recent
publication of an asymmetric synthesis of (+)-lyoniresinol
dimethyl ether from a derivative of compound 8 (24).
Results and discussion
Diethyl E,E-2,3-(3,4,5-trimethoxybenzylidene)succinate (8)
was prepared as previously described (23). Although 8 is
shown in Scheme 2 in an s-cis conformation, it should be
noted that this molecule does not have a planar butadiene
system. Molecular mechanics calculations show that it exists
as an equilibrating mixture of twisted s-cis and s-trans con-
formers. A fuller discussion of its conformation can be
found in our previous publication (24). In the synthesis (23),
8 is formed in an E,E configuration as the sole product. This
product is easily distinguished from the corresponding E,Z
and Z,Z isomers by NMR (17).
Irradiation of 8 in ethyl acetate with a medium-pressure
mercury lamp through Pyrex glass gave only a slow and in-
complete conversion of 8 to the corresponding E,Z-isomer 9
(Scheme 3). Further irradiation of the mixture resulted in the
formation of several products (including 10; see below). It
was possible to separate 9 from 8 by HPLC and to character-
1
ize 9 by H and 13C NMR. A possible mechanism for the
conversion of 8 to 9 is photochemical cyclization to the
1,8a-DHN 11 followed by thermal disrotatory opening to 9.
It is also possible that there is a direct photoisomerization of
the double bond. The irradiation of isolated 9 in ethyl ace-
tate gave the same mixture of 8 and 9, and the conversion of
9 to 8 may pass through intermediate 12.
Surprisingly, irradiation of 8 in chloroform gave the 1,4-
DHN 10 almost exclusively (Scheme 4). The reaction was
most efficient if low concentrations of 8 were irradiated in
tubes placed very close to the lamp (higher light intensity).
© 2004 NRC Canada

Assoumatine et al.
1665
Scheme 4. Irradiation of 8 in chloroform.
Scheme 5. Irradiation of 8 in alcoholic solvents.
On moving the sample further from the lamp (30 cm), it was
possible to detect the concurrent formation of the E,Z-isomer
9, which, although it appeared early in the reaction, also
slowly disappeared to form the 1,4-DHN 10. When 8 was ir-
radiated in deuterochloroform (that had been washed with
D2O to prevent any possibility that traces of water would
protonate an intermediate), there was no incorporation of
deuterium into the 1,4-DHN 10. The lack of any deuterium
incorporation when the reaction was run in D2O-washed
deuterochloroform confirmed that the proposed conversion
of 11 to 10 was an intramolecular 1,3-sigmatropic hydrogen
shift, an allowed photochemical pericyclic reaction. It thus
appears that 11 (or 12) is photochemically converted to 10
and that this allowed photochemical sigmatropic reaction
competes very well with thermal opening of 11 (or 12) to 9
(or 8). The intermediate formation of E,Z-isomer 9 during ir-
radiation in chloroform at lower light intensities, and its ulti-
mate conversion to 10, suggests that this is also a two-step
photochemical conversion via the 1,8a-DHN 12.
Irradiation of 8 in alcoholic solvents (methanol, ethanol,
or 2-propanol), until the disappearance of 8, gave primarily
the cis-1,2-DHN 13 accompanied by the trans-1,2-DHN 14
in a 4:1 ratio (Scheme 5). Similar results were reported in a
preliminary publication (23). However, irradiation for
shorter periods of time showed that the E,Z-isomer 9 was
initially formed along with the two DHN derivatives 13 and
14. When 9 reached about 80% of the concentration of 8, it
ceased to increase and both 8 and 9 subsequently diminished
as 13 and 14 formed. It therefore appears that a photo-
equilibrium between 8 and 9 also takes place in alcoholic
solvents. A similar irradiation of 9 in methanol produced 8
and thereafter a slow conversion to a mixture of 13 and 14 in
a 4:1 ratio.
rearrangement of 8 to 13 involves protonation by solvent. It
seems likely that the 1,8a-DHN 11 (or 12) is thermally
protonated at the 2-position followed by the loss of a proton
at carbon 8a. This mechanism is consistent with that pro-
posed by Heller and co-workers for the acid-catalysed reac-
tions of 1,8a-DHN formed from fulgides (6).
It was not possible to conclusively determine whether
dihydronaphthalenes 13 and 14 were being formed via sol-
vent protonation of both 11 and 12, although it seems likely.
In any event, an earlier suggestion by Heller and co-workers
that the trans-DHN 14 is coming exclusively from irradia-
tion of the E,Z-isomer seems incorrect for this series of com-
pounds (6).
In summary, we have provided the first detailed study of
the photochemistry of an E,E-2,3-dibenzylidenesuccinate
diester and proposed a mechanism for the products formed
in various solvents. In aprotic solvents, there is an equilib-
rium between 8 and 9, in competition with the two-step pho-
tochemical conversion to the 1,4-DHN 10. It is proposed
that the isomerization between the E,E-isomer and the E,Z-
isomer occurs by disrotatory closure to 1,8a-DNH 11 (or 12)
followed by conrotatory thermal opening to 9 (or 8). In
protic solvents, there is also an equilibrium between 8 and 9,
but solvent protonation of 11 (and (or) 12) produces primar-
ily (but not exclusively) the cis-1,2-DHN 13. The possibility
that 13 is formed via an intramolecular thermal [1,5]-
sigmatropic hydrogen shift was ruled out by the fact that
deuterium is incorporated from deuterated solvents.
Experimental
General methods
Proton and 13C-NMR spectra were recorded on a Bruker
AM-300 FT instrument using residual CHCl3 in the CDCl3
as internal standard, unless otherwise specified. Silica gel
(40–63 µm) (Silicycle, Quebec City, Que.) was used for all
chromatography. High-resolution mass spectra (HRMS)
were obtained on a VG Analytical 7070E-HF instrument. Ir-
radiations were conducted with a 450-W medium-pressure
mercury lamp housed in a Pyrex water-cooled jacket or with
Cis-1,2-DHN 13 could be quantitatively converted to
trans-1,2-DHN 14 by warming with a trace of N,N-
dimethyl-4-aminopyridine (DMAP) in ethanol, confirming
that the trans ester 14 is thermodynamically more stable than
13. When the irradiation of 8 was conducted in perdeu-
teromethanol (CD3OD), a mass spectrum revealed that the
1,2-DHN product 13 was monodeuterated. This shows that
© 2004 NRC Canada

1666
Can. J. Chem. Vol. 82, 2004
a circular set of low-pressure mercury lamps (254 nm, 100
W). In all cases, the temperature of the solutions being irra-
diated was 25 5 °C. HPLC was conducted on a reverse-
phase C18 column using methanol (50–80%) in water as
eluent. Literature references for known compounds are: 8
(23), 13 (23), and 14 (25).
MS m/z (relative intensity): 530 (M+, 100), 484 (44), 456
(62), 411 (65), 384 (33), 358 (69); HRMS calcd. for
C28H34O10: 530.2151; found: 530.2139.
1
14. H NMR (300 MHz, CDCl3) δ: 1.19 (t, J = 7.2 Hz,
3H), 1.31 (t, J = 7.1 Hz, 3H), 3.67 (s, 3H), 3.72 (s, 6H), 3.77
(s, 3H), 3.88 (s, 3H), 3.89 (s, 3H), 4.06 (d, J = 1.2 Hz, 1H),
4.12 (2 overlapping quartets, J = 7.2 Hz, 2H), 4.23 (2 over-
lapping quartets, J = 7.1 Hz, 2H), 5.00 (d, J = 1.2 Hz, 1H),
6.28 (s, 2H), 6.71 (s, 1H), 7.62 (s, 1H). 13C NMR (CDCl3,
75 MHz) δ: 14.1, 14.3, 39.4, 46.2, 56.0 (4C), 60.71, 60.77
(one quaternary carbon not observed). EIMS m/z = 530
[M+], 456 (base), 411, 384, 196; HRMS calcd. for
C28H34O10: 530.2152; found: 530.2129.
Diethyl E,Z-2,3-(3,4,5-trimethoxybenzylidene)succinate
(9)
A solution of diethyl E,E-2,3-(3,4,5-trimethoxybenzyl-
idene)succinate (8) (0.020 g, 0.040 mmol) in EtOAc (4 mL)
in a Pyrex test tube was purged with N2 for 5 min. The test
tube was sealed, fixed at 24 cm from the medium-pressure
mercury lamp, and irradiated for 6 h. The solvent was evap-
orated to leave a mixture of the E,E- and E,Z-isomers, 8 and
1
Acknowledgements
9 (approximately 57:43 as determined by H NMR). The
1
E,Z-isomer 9 was separated from the mixture by HPLC. H
We are grateful to the Natural Sciences and Engineering
Research Council of Canada and the University of Manitoba
for financial support of this project.
NMR (CDCl3) δ: 1.17 (t, J = 7.1 Hz, 3H), 1.33 (t, J =
7.1 Hz, 3H), 3.78 (s, 6H), 3.81 (s, 6H), 3.84 (s, 3H), 3.85 (s,
3H), 4.18 (q, J = 7.1 Hz, 2H), 4.29 (q, J = 7.1 Hz, 2H), 6.63
(s, 2H), 6.73 (s, 1H), 6.91 (s, 2H), 7.75 (s, 1H). 13C NMR
(CDCl3) δ: 14.0, 14.3, 56.0 (2C), 56.1 (2C), 60.9, 60.9 (2C),
61.3, 106.5 (2C), 107.8 (2C), 127.8, 129.9, 130.3, 130.7,
140.8, 141.9, 152.8 (3C), 153.0 (3C), 166.9, 167.5.
References
1. S.-T. Chang, S.-Y. Wang, U.C. Su, S.-L. Hunag, and Y.-H.
Kuo. Hozforschung, 53, 142 (1999).
2. A.S.R. Anjaneyulu, D. Santikumar, and U.V. Mallavadhani. In-
dian J. Chem., Sect. B,. 31B, 401 (1992).
3. J.C.A. Boeyens, L. Denner, and G.W. Perold. J. Chem. Soc.,
Perkin Trans. 2, 1749 (1988).
Diethyl 1-(3,4,5-trimethoxyphenyl)-6,7,8-trimethoxy-1,4-
dihydronaphthalene-2,3-dicarboxylate (10)
A solution of 8 (0.138 g, 0.260 mmol) in CHCl3 (26 mL)
in a Pyrex test tube was purged with N2 for 5 min. The test
tube was sealed, fixed at 7 cm from the medium-pressure
Hanovia mercury lamp, and irradiated for 3 h. The solvent
was evaporated, and the residue mixture was purified by
flash chromatography on silica gel using 50% EtOAc in hex-
4. A.S.R. Anjaneyulu, P. Raghu, and K.V.R. Rao. Indian J.
Chem., Sect. B, 19B, 511 (1980).
5. A.S.R. Anjaneyulu, P. Raghu, and K.V.R. Rao. Indian J.
Chem., Sect. B, 18B, 535 (1979).
6. O. Crescente, H.G. Heller, and S. Oliver. J. Chem. Soc., Perkin
1
anes to give the product (0.464 g, 33% yield). H NMR
Trans. 1, 150 (1979).
(CDCl3) δ: 1.18 (t, J = 7.1 Hz, 3H), 1.30 (t, J = 7.1 Hz, 3H),
3.63 (dd, J = 2.6 and 21.5 Hz, 1H), 3.87 (dd, J = 4.2 and
21.5 Hz, partly obscured by OMe signal at 3.85 Hz, 1H),
3.51 (s, 3H), 3.75 (s, 6H), 3.76 (s, 3H), 3.79 (s, 3H), 3.85 (s,
3H), 4.08–4.25 (m, 2H), 4.25 (q, J = 7.1 Hz, 2H), 5.21 (dd,
J = 2.6 and 4.3 Hz, 1H), 6.37 (s, 2H), 6.52 (s, 1H). 13C
NMR (CDCl3) δ: 14.16, 14.22, 31.9, 43.5, 56.1, 56.3 (2C),
60.5, 60.9, 61.0, 61.3, 61.5, 105.8 (2C), 106.3, 123.0, 127.7,
131.5, 137.2, 138.4, 138.6, 141.3, 151.0, 153.0, 153.2 (2C),
167.7 (C=O), 167.8 (C=O). MS m/z (relative intensity): 530
(M+, 64), 484 (60), 456 (35), 411 (56), 380 (27), 362 (26),
317 (100), 289 (66), 245 (12), 181 (28), 168 (21); HRMS
calcd. for C28H34O10: 530.2152; found: 530.2148.
7. T. Momose, T. Nakamura, and K. Kanai. Chem. Pharm. Bull.
26, 3186 (1978).
8. T. Momose, K. Kanai, and K. Hayashi. Chem. Pharm. Bull.
26, 3195 (1978).
9. T. Momose, K. Kanai, and T. Nakamura. Chem. Pharm. Bull.
26, 1592 (1978).
10. T. Momose, K. Kanai, T. Nakamura, and Y. Kuni. Chem.
Pharm. Bull. 25, 2755 (1977).
11. T. Momose, T. Nakamura, and K. Kanai. Heterocycles, 6, 277
(1977).
12. T. Momose, K. Kanai, and T. Nakamura. Heterocycles, 4, 1481
(1976).
13. H.G. Heller and P.J. Strydom. J. Chem. Soc., Chem. Commun.
50 (1976).
14. R.J. Hart, H.G. Heller, R.M. Megit, and M. Szewczyk. J.
Chem. Soc., Perkin Trans. 1, 2227 (1975).
15. J.S. Hastings, H.G. Heller, and K. Salisbury. J. Chem. Soc.,
Perkin Trans. 1, 1995 (1975).
16. H.G. Heller and M. Szewczyk. J. Chem. Soc., Perkin Trans. 1,
1487 (1974).
17. R.J. Hart and H.G. Heller. J. Chem. Soc., Perkin Trans. 1,
1321 (1972).
18. H.G. Heller and K. Salisbury. J. Chem. Soc. C, 873 (1970).
19. M.D. Cohen, H.W. Kaufman, D. Sinnreich, and G.M.J.
Schmidt. J. Chem. Soc. B, 1035 (1970).
Diethyl cis- and trans-1-(3,4,5-trimethoxyphenyl)-6,7,8-
trimethoxy-1,2-dihydronaphthalene-2,3-dicarboxylate
(13 and 14)
The characterizations of 13 and 14 are reported elsewhere
(23, 24) but are included here for completeness.
1
13. H NMR (CDCl3) δ: 0.89 (t, J = 7.2 Hz, 3H), 1.24 (t,
J = 7.2 Hz, 3H), 3.46 (s, 3H), 3.71 (s, 3H), 3.72 (s, 6H),
3.82 (s, 3H), 3.86 (s, 3H), 4.08 (dd, J = 2.8, 9.1 Hz, 1H),
4.17 (m, 2H), 4.76 (d, J = 9.1 Hz, 1H), 6.32 (s, 2H), 6.63 (s,
1H), 7.36 (d, J = 2.8 Hz, 1H). 13C NMR (CDCl3) δ: 13.8,
14.3, 40.5, 48.1, 56.10, 56.13 (2C), 60.4, 60.5, 60.7, 60.8
(2C), 106.4 (2C), 108.0, 123.7, 126.2, 127.1, 135.4, 136.4,
137.2, 144.2, 151.1, 152.6 (2C), 152.9, 167.3, 171.5;
20. R.J. Hart, H.G. Heller, and K. Salisbury. J. Chem. Soc., Chem.
Commun. 1627 (1968).
21. H.G. Heller and B. Swinney. J. Chem. Soc. C, 2452 (1967).
© 2004 NRC Canada

Assoumatine et al.
1667
22. D.C. Ayres, B.G. Carpenter, and R.C. Denney. J. Chem. Soc.
3578 (1965).
23. P.K. Datta, C. Yau, T.S. Hooper, B.L. Yvon, and J.L. Charlton.
J. Org. Chem. 66, 8606 (2001).
24. T. Assoumatine, P.K. Datta, T.S. Hooper, B.L. Yvon, and J.L.
Charlton. J. Org. Chem. 69, 4140 (2004).
25. C. Cow, C. Leung, and J.L. Charlton. Can. J. Chem. 78, 553
(2000).
© 2004 NRC Canada