Controlling the regioselectivity of the di-p-methane rearrangements of 1,2-naphtho-annelated barrelene derivatives-Solution versus solid-state photochemistry

Article abstract of DOI:10.1139/V09-035

The synthesis of 1,2-naphtho-annelated barrelene derivatives, namely dimethyl-7,12-dihydro-7,12-ethenobenzo[a]anthracene-13,14-dicarboxylate (4a) and dimethyl-7,14-dihydro-7,14-ethenodibenzo[a,j]-anthracene-15,16-dicarboxylate (4b), and the investigation

Full text of DOI:10.1139/V09-035

619
Controlling the regioselectivity of the
di-p-methane rearrangements of 1,2-
naphtho-annelated barrelene derivatives —
Solution versus solid-state photochemistry
¨
Jia Luo, Heiko Ihmels, Hans-Jorg Deiseroth, and Marc Schlosser
Abstract: The synthesis of 1,2-naphtho-annelated barrelene derivatives, namely dimethyl-7,12-dihydro-7,12-ethenoben-
zo[a]anthracene-13,14-dicarboxylate (4a) and dimethyl-7,14-dihydro-7,14-ethenodibenzo[a,j]-anthracene-15,16-dicarboxy-
late (4b), and the investigation of their photoreactivity in solution and in the solid state is reported. The irradiation of 4a
and 4b resulted in regioselective di-p-methane rearrangements to give semibullvalene products; however, the product dis-
tribution is inverted upon changing the reaction medium from solution to the solid state. In solution, an a-naphtho–vinyl
bridging predominates as the initial photochemical step because of the selective stabilization of the corresponding transi-
tion state, as discussed on the basis of the structures of the related biradical intermediates, whereas the solid-state photore-
action is significantly influenced by the steric constraints within the crystal lattice that only allow a b-naphtho–vinyl
bridging.
Key words: di-p-methane rearrangement, barrelene derivatives, solid-state photochemistry.
´
´
´
´
`
´
´
´
` `
`
´ ´
Resume : On a realise la synthese de derives condenses 1,2-naphtobarrelenes, a savoir le 7,12-dihydro-7,12-ethenoben-
`
´
´
´
`
zo[a]anthracene-13,14-dicarboxylate de dimethyle (4a) et le 7,14-dihydro-7,14-ethenobenzo[a,j]anthracene-15,16-dicarbo-
´
´
´
´
´
`
´
´
xylate de dimethyle (4b), et on a etudie leur photoreactivite en solution et a l’etat solide. L’irradiation des composes 4a et
`
´
´
´
´
`
`
4b conduit a des rearrangements di-p-methane regioselectifs et a la formation de produits semibullvalenes; toutefois, la
´
´
`
´
distribution des produits est inversee lorsque le milieu reactionnel passe de la solution a l’etat solide. En solution, un pont
´
´
´
´
a-naphtovinyle predomine comme etape photochimique initiale en raison de la stabilisation selective de l’etat de transition
´
´
´
correspondant, tel que discute sur la base des structures des intermediaires biradicalaires apparentes, alors que la photo-
´
`
´
´
´
´
reaction a l’etat solide est fortement influencee par des contraintes steriques dans le reseau cristallin qui ne permettent que
la formation d’un pont b-naphtovinyle.
´
´
´
´
´
` `
`
´
Mots-cles : rearrangement di-p-methane, derives du barrelene, photochimie a l’etat solide.
´
[Traduit par la Redaction]
mation of the dibenzosemibullvalene 2a (Scheme 1), and
this remarkable regioselectivity has been rationalized in
terms of the selective stabilization of one of the four pos-
sible biradicals by the electron-accepting substituent on the
benzene ring. In addition, it has been shown that the nature
and the position of the substituents at the benzene rings of
dibenzobarrelene derivatives 1b–1f have a strong influence
on the regioselectivity of the di-p-methane rearrangement
(Scheme 1).2b,2c Notably, both donor and acceptor substitu-
ents were shown to influence the regioselectivity of the
photorearrangement of dibenzobarrelenes. At the same
time, there are only a few studies with a focus on the in-
fluence of an annelated naphthalene unit on the regioselec-
tivity of the di-p-methane rearrangement of barrelene
derivatives.5 In particular, the competition between vinyl–
naphtho and vinyl–vinyl bridging in the initial step of the
di-p-methane rearrangement of naphthobarrelenes was ex-
amined, and the following order of reactivity was deduced:
a-naphtho–vinyl > vinyl–vinyl > b-naphtho–vinyl > benzo-
vinyl. Nevertheless, no studies have been reported on the
direct competition between photoinduced benzo–vinyl and
a-naphtho–vinyl bridging in annelated barrelene deriva-
tives. To fill this gap, we synthesized the naphtho-annelated
Introduction
The di-p-methane rearrangement1 of benzo-annelated de-
rivatives of barrelene (bicyclo[2.2.2]octatriene)2 and norbor-
nadiene (bicyclo[2.2.1]heptadiene)3 has been shown to be an
appropriate model reaction for the assessment of the sub-
stituent effects on the selectivity of organic photoreactions.
In particular, the structural features that govern the type-
selectivity, the regioselectivity, and the stereoselectivity of
dibenzobarrelene photoreactions in solution1c–1f and in con-
strained media have been evaluated in detail.4 Thus, in a
pivotal paper,2a Zimmerman et al. have shown that the ir-
radiation of dibenzobarrelene 1a leads to the selective for-
Received 18 December 2008. Accepted 29 January 2008.
Published on the NRC Research Press Web site at
canjchem.nrc.ca on 8 April 2009.
J. Luo and H. Ihmels.1 University of Siegen, Organic
Chemistry II, Adolf-Reichwein-Str. 2, 57068 Siegen, Germany.
H.-J. Deiseroth and M. Schlosser. University of Siegen,
Inorganic Chemistry I, Adolf-Reichwein-Str. 2, 57068 Siegen,
Germany.
1Corresponding author (e-mail: ihmels@chemie.uni-siegen.de).
Can. J. Chem. 87: 619–626 (2009)
doi:10.1139/V09-035
Published by NRC Research Press

620
Can. J. Chem. Vol. 87, 2009
Scheme 3. Photoreaction of benzonaphthobarrelene 4a.
Scheme 1. Substituent effects on the di-p-methane rearrangement
of dibenzobarrelene derivatives.
column chromatography, the major photoproduct 5aI was
isolated in 61% yield by crystallization directly from the re-
action mixture. The preferential formation of the photoprod-
uct 5aI in solution was also observed in MeCN and benzene
solution, however, with slightly lower selectivity (Table 1).
Thus, as commonly observed for free-radical reactions, there
is no pronounced solvent effect on the selectivity.8 More-
over in all employed solvents, the product distribution did
not change with varying irradiation wavelength, i.e., from
l > 254 nm (quartz glass) to l > 310 nm (Pyrex glass).
Nevertheless in MeCN and benzene, the reactions were sig-
nificantly slower when the light was filtered through Pyrex
glass, most likely because the Pyrex glass cuts off a signifi-
cant area of the chromophore absorption.
Scheme 2. Synthesis of 1,2-naphthobarrelene derivatives 4a–4c.
The irradiation of crystalline samples of benzonaphthobar-
relene 4a resulted in the formation of the same two semi-
bullvalenes 5aI and 5aII, but the selectivity was reversed
compared with the photoreaction in solution, i.e., the semi-
bullvalene 5aII was formed predominantly upon solid-state
photolysis (Table 1). The direct irradiation of crystals of
barrelene 4a, obtained by crystallization from methanol,
gave the photoproducts 5aI and 5aII in 22:78 ratio at 10%
conversion. The conversion was improved to 41% with only
a slight decrease of selectivity (25:75) by the irradiation of
ground crystals, which were placed between quartz slides,
i.e., by increasing the surface area exposed to the UV light.
Prolonged irradiation times did not increase the conversion
further, probably because of an inner-filter effect that pro-
tects the inside of the solid sample towards the light. On
the other hand, the selectivity was improved to 16:84 by us-
ing ground single crystals of 4a.
barrelene derivatives 4a and 4b and investigated their pho-
tochemical behavior. Herein, we show that the photoin-
duced di-p-methane rearrangements of these compounds
proceed with high regioselectivity in solution and in the
solid state, notably with inverse selectivities.
Results
Synthesis
The benzonaphthobarrelene 4a was obtained in 68% yield
by a Diels–Alder reaction of dimethyl acetylenedicarboxy-
late (DMAD) with benzo[a]anthracene (3a) (Scheme 2).6
For the analogous synthesis of dinaphthobarrelene deriva-
tives 4b and 4c, a mixture of the two isomeric dibenzoan-
thracene derivatives 3b and 3c was employed. The latter
were obtained from the reaction of 1-bromonaphthalene
with N-lithio-2,2,6,6-tetramethylpiperidine in THF in a ratio
The low conversion and small scale of the solid-state pho-
toreactions often limits their potential application in syn-
thetic organic chemistry;7 however, this drawback may be
overcome by the performance of the photoreaction with a
suspension of the finely ground substrate in a solvent in
which it is insoluble.9 To test this approach, thoroughly
stirred suspensions of the benzonaphthobarrelene 4a in hex-
ane or water were irradiated. Thus, on the scale of 60.0 mg
(0.16 mmol) of compound 4a, the photoreaction in hexane
resulted in 90% conversion with a ratio of 33:67 between
the semibullvalenes 5aI and 5aII. In water suspension, the
irradiation proceeded with lower conversion of 67%, but the
selectivity was much better (5aI:5aII = 14:86). From each
reaction mixture, the semibullvalene 5aII was isolated in
yields as high as 37% as colorless cubes, by crystallization
of the photolysate.
1
of 72:28, as determined by H NMR spectroscopic analysis.7
Since the dibenzoanthracene derivatives are air and light
sensitive, the product mixture was immediately made to re-
act with DMAD to give the dinaphthobarrelenes 4b and 4c
in 30% overall yield in a ratio of 71:29 (Scheme 2). At-
tempts to separate the two isomers 4b and 4c by column
chromatography failed; however, the dinaphthobarrelene 4b
was isolated by crystallization in 6% yield. At the same
time, we observed a strong tendency of the two isomers 4b
and 4c to co-crystallize in a 1:1 ratio.
It should be noted that all chiral products are formed
without enantioselectivity. Thus, all presented chiral com-
pounds are actually racemic mixtures.
Photoreactions
1
The product structures were identified on the basis of H
In acetone solution, the naphthobenzobarrelene 4a (long-
wavelength absorption: labs,max = 292 nm) underwent an es-
sentially clean di-p-methane rearrangement upon irradiation
(lex > 254 nm) to give the two semibullvalene derivatives
and 13C NMR spectroscopy and elemental analysis. The an-
nelated semibullvalene structure in compounds 5aI and 5aII
was indicated by the characteristic shifts of the cyclopropane
methine protons (5aI: d 5.21 ppm; 5aII: d 5.59 ppm) and the
proton in the benzylic position (d 4.64 ppm for both 5aI and
5aII). In particular, the unambiguous structure assignment of
1
5aI and 5aII in a 87:13 ratio, as determined by H NMR
spectroscopic analysis (Scheme 3). Although it was impossi-
ble to separate the two semibullvalenes 5aI and 5aII by
Published by NRC Research Press

Luo et al.
621
Table 1. Medium-dependent regioselectivity of the di-p-methane rearrangements of benzo-
naphthobarrelene 4a.
Medium
mba,b (%)
>97
>97
89
97
95
Convna,c (%)
Ratio 5aI/5aII a,d
87:13
80:20
71:29
22:78
25:75
16:84
Acetone solution
Acetonitrile solution
Benzene solution
Bulk crystalsf
Ground crystalsf,g
Single crystalf
>97
>97
>97e
10
41
32
95
Hexane suspension
Water suspension
92
94
90
65h
33:67
14:86
1
aDetermined by H NMR spectroscopic analysis directly on the crude product mixture relative to m-
dimethyl isophthalate as internal standard; error limits: 3% of the given value.
bMass balance.
cConversion.
dIf not stated otherwise, 5aI and 5aII were the only detectable products.
eApprox. 10% of unidentified byproducts were formed.
fCrystallized from methanol solution.
gGround crystals crushed between quartz slides.
hApprox. 3% of unidentified byproducts were formed.
Fig. 1. Structural assignment of the benzonaphthosemibullvalene
derivatives 5aI and 5aII by NOE (arrows indicate NOEs between
the corresponding protons).
Scheme 4. Photoinduced di-p-methane rearrangement of dinaphth-
barrelene 4b (arrows indicate NOEs between the corresponding
protons).
the two isomers 5aI and 5aII was made by NOESY NMR
experiments (Fig. 1). In both cases, the substitution pattern
of the cyclopropane ring, i.e., the position of the ester
groups, was indicated by the NOE between the methine pro-
ton of the cyclopropane and the ortho-proton of the anne-
lated benzene part (arrow a and a’). The assignment of the
benzene protons was confirmed by additional NOE data, in
particular by the NOE between the benzylic proton and the
benzene and naphthalene protons (arrows c,d and b’–d’) .
Moreover, the orientation of the naphthalene unit in com-
pounds 5aI and 5aII was determined by the NOE between
the protons of the methyl substituent and the aromatic pro-
ton (8’-H) of the naphthalene part in isomer 5aI (arrow b).
Scheme 5. Reaction mechanism of the di-p-methane rearrangement
of dibenzobarrlene (1g).
Discussion
The photoinduced di-p-methane rearrangement of anne-
lated barrelenes, e.g., the parent dibenzobarrelene 2g, is ini-
tiated by a carbon–carbon bond formation between the
etheno unit and an ipso carbon of the annelated aromatic
ring to form a 1,4-biradical intermediate BR1-g (Scheme 5),
followed by rearomatization to afford the biradical BR2-g,
and subsequent radical recombination to the semibullvalene
2g.1a,1d Depending on the substitution pattern of the sub-
strate, up to four different regioisomeric products may be
formed, and it was demonstrated that the selectivity of the
initial reaction step determines the selectivity of the overall
reaction (Scheme 1).1c It should be noted, however, that the
selectivity of a chemical reaction, in particular the di-p-
methane rearrangement, is mainly governed by the different
reaction rates, i.e., by the energies of the different transition
states rather than the ones of the corresponding reaction in-
termediates. Nevertheless, it is assumed that factors that sta-
bilize one particular 1,4-biradical intermediate also stabilize
the preceding transition state and thus lead to the preferen-
The photoreactivity of dinaphthobarrelene 4b resembles
the one of barrelene derivative 4a (Scheme 4). In acetone
solution, the irradiation of 4b led to a mixture of semibull-
valenes 5bI and 5bII in a ratio of 87:13, whereas the regio-
selectivity is reversed to 18:82 upon irradiation in the solid
state. The product 5bI was obtained by crystallization in
1
55% yield and fully characterized by H and 13C NMR spec-
troscopy and elemental analysis. Additional NOESY NMR
experiments confirmed the structure of the semibullvalene
derivative 5bI by characteristic NOE signals between an 8’-
H proton of naphthalene and the cyclopropane methine pro-
ton, as well as between the benzyl proton and the 8’-H pro-
ton of naphthalene (Scheme 4). The isomer 5bII could not
be isolated, but its structure was identified by NMR spectro-
scopic analysis directly from the product mixture.
Published by NRC Research Press

622
Can. J. Chem. Vol. 87, 2009
Scheme 6. Reaction pathways of the di-p-methane rearrangement
of barrelene derivative 4a.
same argument may be employed to rationalize the en-
hanced formation of the products 5aI and 5bI resulting from
initial bridging to the C1 carbon upon irradiation of anne-
lated barrelene derivatives 4a and 4b. In contrast to the in-
termediate BR4-aII, the biradicals BR4-aI and BR4-bI
exhibit two resonance structures with a complete benzene
ring providing additional aromatic stabilization. Moreover,
one of the resonance forms, indicated as BR4’, reveals a
particular stabilization as the radical is located both in al-
lylic and benzylic position (Scheme 7).
Notably, the solid-state photoreaction of the annelated
barrelene derivatives 4a and 4b yields the semibullvalene
derivative 5aII and 5bII as the main products. Apparently,
the stereoelectronic factors that determine the regioselectiv-
ity of the di-p-methane rearrangement of 4a and 4b in solu-
tion are no longer effective in the solid state. Since it has
been demonstrated that solid-state photoreactions often pro-
ceed on a reaction pathway that avoids steric interactions
with neighboring molecules,11 it may be assumed that inter-
or intra-molecular steric factors determine the solid-state re-
activity of 4a and 4b rather than electronic effects. To in-
vestigate this aspect, an X-ray diffraction analysis was
performed with single crystals of compound 4a. The benzo-
naphthobarrelene 4a crystallizes in the monoclinic space
˚
˚
group of P21/c (a = 11.3159 A, b = 11.8742 A, c = 14.586
˚
Scheme 7. Resonance stabilization in biradicals BR4-aI and BR4-
bI.
A; b = 104.558), with each unit cell containing two pairs of
enantiomers of the barrelene molecules. Most notably, sig-
nificant close contacts between two molecules were de-
tected. The two different views of the molecular
arrangement in the solid state given in Fig. 2 reveal that the
benzene unit of one molecule (black in Fig. 2) is partially
positioned between the vinyl unit and the naphthalene part
of a neighboring molecule (Fig. 2, white, red, blue, respec-
tively). In this arrangement, the closest contacts between
the two molecules were identified between one carbonyl car-
bon atom of an ester group (Fig. 2, red) and an ortho-hydro-
gen atom of the benzene ring (Fig. 2a, CÁÁÁH: 289 pm), and
between the carbonyl oxygen atom of the same ester func-
tionality and the bridgehead hydrogen atom (Fig. 2a, OÁÁÁH:
262 pm). These distances correspond essentially to the sum
of the van der Waals radii of the respective atoms (CÁÁÁH:
290 pm, OÁÁÁH: 272 pm)12 and may thus result in significant
intermolecular steric repulsion during the solid-state rear-
rangement of compound 4a. In particular, the movement of
one of the ester functionalities, namely the one at the C16
position (Fig. 2, red), in the course of a di-p-methane rear-
rangement appears to be significantly hindered. To achieve
the vinyl–naphtho bridging to give BR4-aI, this ester group
needs to pass the proximate benzene unit of the neighboring
molecule, and the close contacts described above clearly in-
dicate that this pathway is essentially blocked. At the same
time, the other ester group (Fig. 2, blue) is free to move as
there are no significant contacts with the proximate mole-
cule (>350 pm; not shown), so that an almost unhindered vi-
nyl–naphtho bridging is possible in the solid state to afford
biradical BR4-aII. Notably, both vinyl–benzo bridging path-
ways leading to the semibullvalene products 5aIII and 5aIV
are not sterically hindered in the crystal. Thus, in the ab-
sence of steric constraints in the crystal, the solid-state pho-
toreaction is determined by the same electronic factors as in
isotropic solution.
tial formation of the corresponding semibullvalene isomer.
According to these considerations, the regioselectivity ob-
served in the photoreactions of the barrelene derivatives 4a
and 4b may be discussed in terms of the relative stability of
the possible biradical intermediates. Thus, during the di-p-
methane rearrangemenent of the barrelene derivative 4a, the
four regioisomeric biradicals BR4-aI–BR4-aIV may be
formed initially, which lead to the four isomeric benzonaph-
thosemibullvalenes 5aI–5aIV (Scheme 6). Nevertheless, the
products 5aIII and 5aIV are not formed in detectable
amounts, indicating that the intermediates BR4-aIII and
BR4-aIV are only generated to negligible extents. We pro-
pose that the high preference for the photoinduced vinyl–
naphtho bridging is governed by the significantly higher sta-
bilization of the intermediate radicals in the naphthalene
unit, compared with the stabilization in the cyclohexadienyl
radical part in BR4-aIII and BR4-aIV. Moreover, the biradi-
cals BR4-aI and BR4-aII still contain two intact benzene
rings, whose resonance energy contributes more to the over-
all stabilization than the one of the single naphthalene ring
in BR4-aIII and BR4-aIV.10
It was demonstrated that the di-p-methane rearrangeme-
nent of the unsubstituted 1,2-naphthobarrelene is introduced
by a highly selective vinyl–naphtho bridging that takes place
exclusively between a vinyl carbon atom and the C1 position
(a position) of the naphthalene.3 These observations have
been explained by a higher extent of delocalization within
the aromatic part of the corresponding 1,4-biradical. The
Published by NRC Research Press

Luo et al.
623
Fig. 2. Structure of the benzonaphthobarrelene 4a in the solid state (dashed lines indicate closest contacts between two neighboring mole-
cules).
Table 2. Crystallographic data and details of the structure determination for 4a.
Parameter
Value
Molecular formula
Formula mass
Colour
C24H18O4
M = 370.38 g/mol
Colourless
Crystal system
Space group
Monoclinic
P21/c (No. 14)
˚
˚
˚
Unit cell dimensions
Volume
a = 11.3159(9) A, b = 11.8742(7) A, c = 14.586(1) A, b = 104.55(1)8
3
˚
V = 1897.0(3) A
Formula units/unit cell
Calculated density
Diffractometer
Z = 4
r = 1.297 g/cm3
IPDS (STOE)
Wave length
l = 0.710 73 (Mo-Ka)
Monochromator
Crystal size
Measuring temperature
Scan type
Graphite
0.29 mm  0.28 mm  0.17 mm
T = 296(2) K
4 - scan
Measured q-range
Index ranges
2.248 £ q £ 28.148
–15£ h £ 14, –15 £ k £ 15, –19 £ l £ 19
Reflections collected
Independent reflections
Observed reflections
Data averaginga
Completeness to qmax
Absorption coefficient
Extinction coefficient
Structure solution
Structure refinement
Number of parameters
Number of restraints
GooF on F2
20 010
4 614
21 51 [I>2s(I)]
Rint = 0.0664, Rs = 0.0739
99.5%
m = 0.088 mm–1
3 = 0.0079(9)
Direct methods (ref. 12)
Full-matrix least-squares on F2 (ref. 12)
256
0
0.794
F(000)
776
Weighting schemeb
Fig. of merita [I>2s(I)]
Fig. of Merita [all data]
Residual electron density
A = 0.0458, B = 0
R1 = 0.0396, wR2 = 0.0791
R1 = 0.1071, wR2 = 0.0936
3
˚
–0.143 £ r £ 0.194 e/A
aDefinition of R factors see ref.13
bw = 1/[s2(F0 ) + (AP)2 + BP)], P = (F0 + 2Fc2)/3
2
2
Published by NRC Research Press

624
Can. J. Chem. Vol. 87, 2009
8 Hz, ar-CH). 13C NMR (100 MHz, CDCl3) d: 48.2
(CH3), 52.4 (overlapped CH3 and CH), 53.6 (CH), 122.0
(ar-CH), 122.5 (6,11-CH), 123.7 (ar-CH), 123.8 (ar-CH),
125.1 (ar-CH), 125.2 (two ar-CH overlapped), 125.6 (ar-
CH), 126.4 (ar-CH), 128.7 (ar-CH), 129.0 (Cq), 131.7 (Cq),
141.2 (Cq), 142.6 (Cq), 144.0 (Cq), 144.8 (Cq), 146.8 (Cq),
149.0 (Cq), 165.8 (C=O), 166.2 (C=O). EI-MS (70 eV):
370 [M+]. Elemental analysis calcd. (%) for C24H18O4
(370.40): C 77.82, H 4.90; found C 77.62, H 4.86.
Summary and conclusion
In summary, we have shown that the di-p-methane rear-
rangement of 1,2-naphtho annelated barrelene derivatives 4a
and 4b proceeds with high regioselectivity. Most notably, a
drastic change of the selectivity was observed upon chang-
ing from isotropic solution to the solid state. Whereas in sol-
ution, the course of the reaction is mainly governed by the
selective stabilization of the corresponding transition states,
which lead to the primary biradical intermediates BR4, the
solid-state photoreaction is significantly influenced by the
steric constraints within the crystal lattice. To be stressed is
the fact that in the case of 4a, the major products from both
reactions, i.e., in solution and in the solid state, could be iso-
lated in reasonable yields. Considering the importance of the
annelated semibullvalene derivatives as synthons in organic
synthesis,1f the presented systems may be employed along
these lines as a useful source for the selective synthesis of
complex 1,2-naphtho-annelated polycyclic aromatic com-
pounds.1f
Dimethyl-7,14-dihydro-7,14-ethenodibenzo[a,j]-anthracene-
15,16-dicarboxylate (4b)
Under argon protection, a mixture of dibenzo[a,h]- and
dibenzo[a,j]anthracene (28:72, 690 mg, 2.48 mmol), DMAD
(0.50 mL, 4.1 mmol), and AlCl3 (20 mg) in xylene (15 mL)
was heated at 140 8C for 5 days. After cooling RT, the mix-
ture was diluted with EtOAc (70 mL). The organic phase
was extracted with 2N aq. HCl (20 mL) and brine (50 mL),
and subsequently separated, dried with Na2SO4. The solvent
was removed in vacuo and the residue was purified by col-
umn chromatography (SiO2; EtOAc/hexane = 1:4, v/v; Rf =
0.31) to give a mixture of the two dinaphthobarrelenes 4b
and 4c (4b/4c 71:29, 312 mg, 0.74 mmol, 30%). Two-fold
recrystallization from EtOAc/hexane gave pure 4b as color-
less needles (42 mg, 6%), mp 223–225 8C. 1H NMR
(400 MHz, CDCl3) d: 3.79 (s, 3H, CH3), 3.82 (s, 3H, CH3),
5.80 (s, 1H, CH), 7.23 (s, 1H, CH), 7.39 (d, J = 8 Hz, 2H,
ar-CH), 7.55–7.59 (m, 4H, ar-CH), 7.65 (d, J = 8 Hz, 2H,
ar-CH), 7.80 (d, J = 8 Hz, 2H, ar-CH). 13C NMR
(100 MHz, CDCl3) d: 44.0 (CH), 52.5 (CH3), 52.5 (CH3),
54.8 (CH), 122.1 (ar-CH), 122.5 (ar-CH), 125.2 (ar-CH),
125.3 (ar-CH), 126.6 (ar-CH), 128.8 (ar-CH), 129.2 (Cq),
131.6 (Cq), 141.7 (Cq), 144.1 (Cq), 146.8 (Cq), 151.0 (Cq),
165.7 (C=O), 166.4 (C=O). MS (m/z): 420 [M+]. Elemental
analysis calcd. (%) for C28H20O4 (420.46): C 79.98, H 4.79;
found C 79.60, H 4.67.
Experimental section
General
NMR spectra: Bruker Avance 400 (1H NMR: 400 MHz;
13C NMR: 100 MHz) and a Varian NMR System 600
(1H NMR: 600 MHz; 13C NMR: 150 MHz). H NMR chem-
1
ical shifts refer to dTMS 0.0 ppm. 13C NMR chemical shifts
refer to solvent signals (CDCl3: d 77.0 ppm). Melting points:
Bu¨chi 510K apparatus (uncorrected). Mass spectra: Hewlett-
Packard HP 5968 (EI, 70 eV). Elemental analyses: KEKA-
tech EuroEA combustion analyzer by Mr. H. Bodenstedt,
Organic Chemistry I, University of Siegen. Benzo[a]anthra-
cene5 and benzotetracenes6 were prepared according to liter-
ature procedures. All the other reagents and solvents were
obtained from Acros Organics and Sigma-Aldrich and used
as received.
Photoreactions
Dimethyl-7,12-dihydro-7,12-ethenobenzo[a]anthracene-
13,14-dicarboxylate (4a)
Irradiations were performed with a TQ150 middle-pres-
sure mercury lamp (Heraeus, UV-Consulting Peschl), which
was placed in a quartz cooling tube. The sample container
was placed 10–15 cm in front of the lamp. The solution con-
tained one molar equivalent of m-dimethyl isophthalate as
internal standard for the quantitative H NMR spectroscopic
analysis of the photolysate.
Under nitrogen-gas atmosphere, a suspension of ben-
zo[a]anthracene (5.89 g, 25.8 mmol), DMAD (9.18 mL,
75.0 mmol), and AlCl3 (0.2 g) in xylene (15 mL) was heated
at 140 8C for 5 days. The mixture was cooled to room tem-
perature (RT) and diluted with ethyl acetate (100 mL). The
organic phase was extracted with 2N aq. HCl (50 mL) and
brine (50 mL), and subsequently separated, dried with Na2SO4.
The solvent was removed in vacuo and the residue was purified
by column chromatography (SiO2; EtOAc/hexane = 1:2, v/v;
Rf = 0.45) to give an orange oil, which was solidified upon
treatment with methanol to give dimethyl-7,12-dihydro-
7,12-ethenobenzo[a]anthracene-13,14-dicarboxylate (4a) as
1H NMR spectroscopically pure light yellow granules
(6.47 g, 17.5 mmol, 68%). An analytical sample was pre-
pared by recrystallization from methanol as clusters of col-
orless prisms, mp 142–144 8C. UV (CHCl3): lmax (log 3) =
245 (4.40), 270 (3.85), 281 (3.90), 292 (3.92), 329 (3.27).
1
Photolysis in solution
Solutions (10–3 to 10–2 mol/L) were placed in a 200 mL
Pyrex tube (acetone) or a 50 mL quartz tube (other solvents)
and purged with argon for at least 20 min. The stirred solu-
tion was irradiated for 6–13 h until the starting material was
1
fully consumed as determined by H NMR analysis. After
evaporation of the solvent in vacuo, the photolysate was an-
1
alyzed by H NMR spectroscopy, or, in preparative experi-
ments, recrystallized from EtOAc/hexane to obtain the
major photoproduct.
1H NMR (400 MHz, CDCl3) d: 3.79 (s, 3H, OCH3), 3.80
(s, 3H, OCH3), 5.64 (s, 1H, CH), 6.32 (s, 1H, CH), 6.98–
7.01 (m, 2H, ar-CH), 7.39–7.47 (m, 3H, ar-CH), 7.53–7.61
(m, 3H, ar-CH), 7.81 (d, J = 8 Hz, ar-CH), 8.27 (d, J =
Photolysis in crystalline state
15–25 mmol of the crystalline starting material were
placed in an argon-purged and sealed quartz test tube and
irradiated. Afterwards, the stock solution of m-dimethyl iso-
Published by NRC Research Press

Luo et al.
625
phthalate was added as internal standard, and the mixture
was dissolved in dichloromethane. The solvent was removed
and the mixture was analyzed by H NMR analysis.
C24H18O4 (370.40): C 77.82, H 4.90; found C 77.98, H
4.81.
1
Dimethyl-6d,12b-dihydro-cyclopropa[3,4]pentaleno-[1,2-
a;6,5-c’]dinaphthalene-6c,12d-dicarboxylate (5bI)
Alternatively, the crystalline solids were ground to pow-
ders and sandwiched between a pair of quartz slides, which
were sealed in an argon-purged polyethylene bag and irradi-
ated. The photolysate was dissolved in dichloromethane and
the stock solution of m-dimethyl isophthalate was added as
internal standard. After the removal of solvent, the mixture
Prepared by photoreaction in acetone solution. Yield 55%,
white granules, mp 217–218 8C. 1H NMR (600 MHz,
CDCl3) d: 3.78 (s, 3H, CH3), 3.89 (s, 3H, CH3), 5.13 (s,
1H, CH), 5.34 (s, 1H, CH), 7.30–7.38 (m, 4H, ar-CH), 7.41
(d, J = 8 Hz, 1H, ar-CH), 7.46 (dd, J = 7 Hz, 1H, ar-CH),
7.59 (d, J = 8 Hz, 1H, ar-CH), 7.64 (two overlapped dou-
blet, J = 8 Hz, 2H, ar-CH), 7.68–7.71 (m, 2H, ar-CH), 8.10
(d, J = 8 Hz, 1H, ar-CH). 13C NMR (100 MHz, CDCl3) d:
48.5 (CH), 52.4 (CH3), 53.0 (CH3), 56.3 (CH3), 58.8 (Cq),
67.4 (Cq), 120.0 (ar-CH), 120.1 (ar-CH), 123.5 (ar-CH),
124.3 (ar-CH), 125.0 (ar-CH), 125.3 (ar-CH), 126.6 (ar-
CH), 126.6 (ar-CH), 127.5 (Cq), 128.3 (ar-CH), 128.4 (ar-
CH), 128.8 (ar-CH), 129.4 (ar-CH), 129.8 (Cq), 130.5 (Cq),
130.7 (Cq), 132.5 (Cq), 133.1 (Cq), 146.9 (Cq), 148.0 (Cq),
169.6 (C=O), 169.7 (C=O). MS (m/z): 420 [M+]. Elemental
analysis calcd. (%) for C28H20O4 (420.46): C 79.98, H 4.79;
found C 79.47, H 4.71.
1
was analyzed by H NMR analysis.
Photolysis in suspensions
In a 20 mL quartz tube, 0.05–0.16 mmol of starting mate-
rial and 10 mL of hexane or water were added. The mixture
was sonicated for 10 min to destroy the larger crystals and
purged with argon for 20 min. After irradiation, dichlorome-
thane was used to dissolve (hexane) or extract (water) the
photolysate. After evaporation of the solvent in vacuo the
product mixture was subjected to 1H NMR spectroscopic
analysis, or redissolved in acetone/methanol to crystallize
the major photoproduct.
Dimethyl-4b,10b-dihydro-
benzo[5,6]cyclopropa[3,4]pentaleno[1,2-a]naphthalene-
X-ray diffraction analysis of compound 4a
The conditions and parameters adopted for the determina-
tion of crystallographic data of compound 4a are listed in
Table 2.
4c,10d-dicarboxylate (5aI)
Prepared by photoreaction in acetone solution. Yield 61%,
colorless needles, mp 177–179 8C. UV (CHCl3): lmax (log 3) =
246 (4.44), 281 (3.73), 328 (2.97). 1H NMR (400 MHz,
CDCl3) d: 3.76 (s, 3H, CH3), 3.83 (s, 3H, CH3), 4.64 (s,
1H, CH), 5.21 (s, 1H, CH), 7.00–7.08 (m, 2H, ar-CH),
7.17 (d, J = 7 Hz, 1H, ar-CH), 7.33–7.44 (m, 4H, ar-CH),
7.64–7.68 (m, 2H, ar-CH), 7.75 (d, J = 8 Hz, 1H, ar-CH).
13C NMR (100 MHz, CDCl3) d: 49.7 (CH3), 52.4 (CH),
52.9 (CH), 55.5 (CH3), 58.4 (Cq), 66.8 (Cq), 120.2 (ar-
CH), 121.3 (ar-CH), 123.5 (ar-CH), 125.0 (ar-CH), 125.7
(ar-CH), 126.6 (ar-CH), 126.8 (ar-CH), 127.4 (ar-CH),
127.7 (Cq), 128.8 (ar-CH), 129.5 (ar-CH), 130.3 (Cq),
133.1 (Cq), 134.8 (Cq), 147.8 (Cq), 150.0 (Cq), 169.6
(C=O), 169.6 (C=O). MS (m/z): 370 [M+]. Elemental anal-
ysis calcd. (%) for C24H18O4 (370.40): C 77.82, H 4.90;
found C 77.74, H 4.81.
Supplementary data
Supplementary data for this article are available on the
journal Web site (canjchem.nrc.ca) or may be purchased
from the Depository of Unpublished Data, Document Deliv-
ery, CISTI, National Research Council Canada, Ottawa, ON
K1A 0R6, Canada. DUD 3912. For more information on ob-
taining material refer to cisti-icist.nrc-cnrc.gc.ca/cms/
unpub_e.shtml. CCDC 720890 contains the crystallographic
data for this manuscript. These data can be obtained, free of
charge, via www.ccdc.cam.ac.uk/conts/retrieving.html (Or
from the Cambridge Crystallographic Data Centre, 12 Union
Road, Cambridge CB2 1EZ, UK; fax +44 1223 336033; or
deposit@ccdc.cam.ac.uk).
Dimethyl-4b,10c-dihydro-benzo[5,6]cyclopropa[1,2-
Acknowledgement
h]naphthalene-4c,10e-dicarboxylate (5aII)
Prepared by photoreaction in water suspension. Yield
37%, colorless cuboids, mp 173–175 8C. 1H NMR
(600 MHz, CD3Cl) d: 3.75 (s, 3H, CH3), 3.91 (s, 3H, CH3),
4.64 (s, 1H, CH), 5.59 (s, 1H, CH), 6.98–7.04 (m, 2H, ar-
CH), 7.28 (d, J = 8 Hz, 1H, ar-CH), 7.34 (d, J = 8 Hz, 1H,
ar-CH), 7.38 (d, J = 9 Hz, 1H, ar-CH), 7.34 (d, J = 8 Hz,
1H, ar-CH), 7.44 (dd, J = 7 Hz, 1H, ar-CH), 7.59 (dd, J =
7 Hz, 1H, ar-CH), 7.62 (d, J = 8 Hz, 1H, ar-CH), 7.62 (d,
J = 8 Hz, 1H, ar-CH), 7.78 (d, J = 8 Hz, 1H, ar-CH), 8.14
(d, J = 8 Hz, 1H, ar-CH). 13C NMR (150 MHz, CD3Cl) d:
50.1 (CH3), 52.4 (CH), 52.9 (CH), 53.7 (CH3), 67.8 (Cq),
68.7 (Cq), 121.4 (ar-CH), 123.3 (ar-CH), 123.7 (ar-CH),
125.7 (two overlapped ar-CH), 126.7 (ar-CH), 127.0 (ar-
CH), 127.3 (ar-CH), 127.8 (ar-CH), 128.8 (ar-CH), 129.8
(Cq), 133.2 (Cq), 134.9 (Cq), 145.3 (Cq), 148.9 (Cq), 168.6
(C=O), 169.4 (C=O), one Cq signal is overlapped. MS
(m/z): 370 [M+]. Elemental analysis calcd. (%) for
This paper is dedicated to Trevor Linden on the occasion
of his retirement and in appreciation of his exemplary be-
havior on and off the ice. We thank Dr. Thomas Paululat
for carrying out the NOESY experiments and for helpful
discussions on spectra analysis.
References
(1) (a) Zimmerman, H. E. In Rearrangements in ground and ex-
cited states; de Mayo, P., Ed.; Academic Press: New York,
USA, 1980; chap. 16; (b) Liao, C.-C.; Yang, P.-H. In Hand-
book of Photochemistry and Photobiology; Horspool, W. M.,
Song, P.-S., Eds.; CRC Press: Boca Raton, Fla., USA, 1995;
chap. 15; (c) Scheffer, J. R.; Yang, J. In Handbook of Photo-
chemistry and Photobiology; Horspool, W. M., Song, P.-S.,
Eds.; CRC Press: Boca Raton, Fla., USA, 1995; chap. 16;
(d) Zimmerman, H. E.; Armesto, D. Chem. Rev. 1996, 96,
3065 doi:10.1021/cr910109c. PMID:11848853.; (e) Liao, C.-
Published by NRC Research Press

626
Can. J. Chem. Vol. 87, 2009
C.; Peddinti, R. K. In Handbook of Photochemistry and
Photobiology; Horspool, W. M., Lenci, F., Eds.; CRC Press:
Boca Raton, Fla., USA, 2003; chap. 32; (f) Ramaiah, D.; Sa-
jimon, M. C.; Joseph, J.; George, M. V. Chem. Soc. Rev.
2005, 34, 48. doi:10.1039/b300843f. PMID:15643489.
L.; Ding, R. F.; Jenks, W. S.; Armstrong, D. W. Org. Lett.
2005, 7, 335. doi:10.1021/ol047599i. PMID:15646991.
(5) (a) Zimmerman, H. E.; Bender, C. O. J. Am. Chem. Soc.
1969, 91, 7516 doi:10.1021/ja01054a053.; (b) Zimmerman,
H. E.; Bender, C. O. J. Am. Chem. Soc. 1970, 92, 4366.
doi:10.1021/ja00717a038.
(6) Platt, K. L.; Oesch, F. J. Org. Chem. 1981, 46, 2601. doi:10.
1021/jo00325a041.
(7) Photochemistry in organized and constrained media; Rama-
murthy, V., Ed.; VCH: NY, USA, 1991.
(8) Reichardt, C. Solvents and solvent effects on organic chem-
istry; Wiley-VCH: Weinheim, Germany, 2003; chap. 5.
(9) (a) Toda, F. Top. Curr. Chem. 2005, 254, 1 doi:10.1007/
b98357.; (b) Scheffer, J. R.; Wang, K. Synthesis 2001, 8,
1253. doi:10.1055/s-2001-15056.
(10) Dewar, M. I. S.; de Liano, C. J. Am. Chem. Soc. 1969, 91,
789. doi:10.1021/ja01032a001.
(11) Ramamurthy, V.; Venkatesan, K. Chem. Rev. 1987, 87, 433.
doi:10.1021/cr00078a009.
(2) (a) Zimmerman, H. E.; Sulzbach, H. M.; Tollefson, M. B. J.
Am. Chem. Soc. 1993, 115, 6548 doi:10.1021/ja00068a011.;
(b) Ihmels, H.; Mohrschladt, C. J.; Grimme, J. W.; Quast,
H. Synthesis 2001, 8, 1175 doi:10.1055/s-2001-15064.;(c)
Rattray, G.; Yang, J.; Gudmundsdottir, A. D.; Scheffer, J. R.
Tetradedron Lett. 1993, 34, 35. doi:10.1016/S0040-4039(00)
60051-5.
(3) Snow, R. A.; Cottrell, D. M.; Paquette, L. A. J. Am. Chem.
Soc. 1977, 99, 3734. doi:10.1021/ja00453a036.
(4) (a) Garcia-Garibay, M.; Scheffer, J. R.; Watson, D. G. J.
Chem. Soc. Chem. Commun. 1989, 600. doi:10.1039/
c39890000600.; (b) Garcia-Garibay, M.; Omkaram, N.;
Schffer, J. R.; Trotter, J.; Wirenko, F. J. Am. Chem. Soc.
1989, 111, 4985 doi:10.1021/ja00195a066.; (c) Chen, J.;
Scheffer, J. R.; Trotter, J. Tetrahedron 1992, 48, 3251
doi:10.1016/0040-4020(92)85002-V.; (d) Benitez, M.; Bring-
mann, G.; Dreyer, M.; Garcia, H.; Ihmels, H.; Waidelich,
M.; Wissel, K. J. Org. Chem 2005, 70, 2315 doi:10.1021/
jo047878j.; (e) Ding, J.; Desikan, V.; Han, X. X.; Xiao, T.
(12) Bondi, A. J. Phys. Chem. 1964, 68, 441. doi:10.1021/
j100785a001.
(13) Sheldrick, G. M. SHELX-97 [computer program]; University
¨
¨
of Gottingen, Gottingen, Germany, 1997.
Published by NRC Research Press