Wavelength-dependent selectivity in [4 + 4]-cycloaddition reactions

Article abstract of DOI:10.1021/jo060838q

Selective excitation of a variety of anthracene derivatives in the presence of anthracene led to the formation of only the cross-cyclomer products, while irradiating these mixtures at wavelengths where anthracene also absorbs led to both dianthracene and

Full text of DOI:10.1021/jo060838q

Wavelength-Dependent Selectivity in
[4 + 4]-Cycloaddition Reactions
David Bailey and Vance E. Williams*
Department of Chemistry, Simon Fraser UniVersity, 8888
UniVersity DriVe, Burnaby, British Columbia, Canada V5A 1S6
Vancew@sfu.ca
ReceiVed April 21, 2006
FIGURE 1. UV-vis absorption spectra of anthracene (solid line) and
TPA (dashed line) in acetonitile at 25 °C. Inset: enlarged view of these
spectra in the range 310-420 nm.
Selective excitation of a variety of anthracene derivatives in
the presence of anthracene led to the formation of only the
cross-cyclomer products, while irradiating these mixtures at
wavelengths where anthracene also absorbs led to both
dianthracene and the mixed cyclomers being formed. This
method was shown to be general, so long as the chromophore
was inert toward forming its homodimer and could be
irradiated at wavelengths at which anthracene does not
absorb.
exciting only one component in a binary mixture of two
chromophores, X and Y.3 Because the dimer di-Y can only form
from the excited state Y*, this photoproduct should not be
observed when the solution is irradiated at wavelengths where
only X absorbs. We posited that the cross cyclomer XY could
be obtained as the exclusive product if we could identify an
anthracene derivative X that (i) could be selectively excited in
the presence of Y and (ii) was limited by steric constraints from
forming its homodimer di-X, yet was still able to form XY.
It has previously been reported that 2,3,6,7-tetraphenylan-
thracene (TPA) is not converted into its homodimer (di-TPA)
even after prolonged irradiation in a Rayonet with 300 nm light,
yet does form cross-cyclomers with derivatives that lack bulky
substituents on the outer rings of the anthracene.3a As such, this
molecule satisfies our second criteria for selectivity. The UV-
vis absorption spectrum of this compound is shown in Figure
1, along with that of anthracene (A). TPA exhibits characteristic
anthracene-like absorption bands between 325 and 400 nm; these
are red-shifted by approximately 25 nm relative to those of
anthracene (Figure 1, inset). This compound also possesses a
strong absorption band at shorter wavelengths (λmax ) 301 nm)
where the absorption of anthracene is negligible. The presence
of two ranges of wavelengths at which TPA absorbs while
anthracene does not, coupled with the reluctance of TPA to
form homodimers, suggested that TPA and anthracene could
constitute a complementary pair under conditions where TPA
is selectively excited.
The concept of complementarity is one of the cornerstones
of modern organic chemistry, with reactions between acids and
bases, donors and acceptors, dienes and dienophiles, or nucleo-
philes and electrophiles facilitating the efficient construction
of a myriad of complex structures. Unfortunately, examples of
complementary pairs of molecules are notably lacking in the
context of [4 + 4]-photocycloaddition reactions. Thus, irradiat-
ing two anthracene derivatives, X and Y, generally yields both
the di-X and di-Y photodimers in addition to the cross-cyclomer
XY.1 For this reason, while these bimolecular photochromic
reactions provide unique opportunities for the construction of
dynamic materials, their use has largely been restricted to the
creation of relatively simple structures created via the ho-
modimerization of identical chromophores.2
We therefore decided to explore the possibility of increasing
the selectivity of anthracene photocycloaddition reactions by
Our initial photochemical experiments on these compounds
were carried out in a Rayonet using either 300 or 350 nm lamps.
When a 1:1 mixture of TPA and A was irradiated with light
centered at 300 nm, two photoproducts were formed in an
approximately 1:9 ratio. These products were identified as
dianthracene (di-A) and the cross-cyclomer TPA-A, respec-
tively.
(1) (a) Bouas-Laurent, H.; Castellan, A.; Desvergne, J. P.; Lapouyade,
R. Chem. Soc. ReV. 2001, 30, 43 and references therein. (b) Bouas-Laurent,
H.; Castellan, A.; Desvergne, J. P.; Lapouyade, R. Chem. Soc. ReV. 2000,
29, 43-55.
(2) (a) Ihara, T.; Fujii, T.; Mukae, M.; Kitamura, Y.; Jyo, A. J. Am.
Chem. Soc. 2004, 126, 8880. (b) de Schryver, F. C.; Anand, L.; Smets, G.;
Switten, J. Polym. Lett. 1971, 9, 777. (c) Jones, J. R.; Liotta, C. L.; Collard,
D. M.; Schiraldi, D. A. Macromolecules 2000, 33, 1640. (d) Mitsuishi, M.;
Tanuma, T.; Matsui, J.; Chen, J.; Miyashita, T. Langmuir 2001, 17, 7449.
(e) McSkimming, G.; Tucker, J. H. R.; Bouas-Laurent, H.; Desvergne, J.
P.; Coles, S. J.; Hursthouse, M. B.; Light, M. E. Chem.sEur. J. 2002, 8,
3331. (f) Nakatsuji, S.; Ojima, T.; Akutsu, H.; Yamada, J. J. Org. Chem.
2002, 67, 916. (g) Ojima, T.; Akutsu, H.; Yamada, J.; Nakatsuji, S.
Polyhedron 2001, 20, 1335. (h) Takaguchi, Y.; Tajima, T.; Yanagimoto,
Y.; Tsuboi, S.; Ohta, K.; Motoyoshiya, J.; Aoyama, H. Org. Lett. 2003, 5,
10, 1677. (i) Molard, Y.; Bassani, D. M.; Desvergne, J. P.; Horton, P. H.;
Hursthouse, M. B.; Tucker, J. H. R. Angew. Chem., Int. Ed. 2005, 44, 1072.
When the TPA/anthracene mixture was irradiated with 350
nm light, both dianthracene and TPA-A were again formed,
(3) For other examples of approaches to improving the selectivity of
these reactions, see: (a) Bailey, D.; Williams, V. E. Chem. Commun. 2005,
2569. (b) Bouas-Laurent, H.; Lapouyade, R. Chem. Commun. 1969, 817.
(c) Lapouyade, R.; Castellan, A.; Bouas-Laurent, H. C. R. Acad. Sc. Paris
1969, 268, 217.
10.1021/jo060838q CCC: $33.50 © 2006 American Chemical Society
Published on Web 07/06/2006
5778
J. Org. Chem. 2006, 71, 5778-5780

SCHEME 1
SCHEME 2a
a Reagents and conditions: (a) (i) LiAlH4, AlCl3, THF, reflux 22 h; (ii)
10% Pd/C, xylenes, reflux, 5 days (82% yield/two steps); (b) (i) MeLi,
Et2O; (ii) SnCl2, aqueous HCl (23% yield/two steps).
but in a 1:3 ratio rather than the 1:9 ratio observed at 300 nm.
After long irradiation periods (>140 min) at this wavelength,
trace quantities (∼3%) of a third photoproduct also started to
appear. This minor product was identified as di-TPA, the
homodimer of TPA, a compound that had not previously been
observed. The formation of this dimer is quite slow; irradiation
of a benzene solution of TPA at 350 nm for 2 h led to only
25% of the starting material being converted into this product.
Under the same conditions, anthracene was quantitatively
converted to dianthracene in under 2 h. It is unclear at this time
why this product is not observed when 300 nm light is
employed, although photofragmentation of the dimer at shorter
wavelengths may be a contributing factor.
FIGURE 2. Structure of photocyclomers DPA-A, di-DPA, and
DMDPA.
The greater selectivity for the cyclomer TPA-A over di-A
when light centered at 300 nm was employed is likely due to
the preferential excitation of TPA at these wavelengths.
However, because the light sources employed in these initial
experiments deliver light over a relatively broad spectral range,
it is difficult to exclusively excite only one chromophore of a
binary mixture. To explore the effects of selective excitation,
we decided to repeat the experiments described above using a
fluorimeter as the light source, hence enabling irradiation over
a much narrower range of wavelengths.4 Three solutions of 10
mM TPA and 10 mM anthracene in 1 mL of CDCl3 were
prepared; these solutions were irradiated at 301, 376, or 399
nm. The sample irradiated at 301 nm exhibited slow formation
of a single photoproduct, which was identified as the cross-
While four peripheral phenyl substituents greatly inhibit
homodimer formation, two of these groups appear to be much
less effective at blocking this reaction. DPA was found to
undergo facile homodimer formation when illuminated in a
Rayonet; after 130 min of irradiation at 300 nm, a 5 mM
acetonitrile solution of this compound was converted to a single
photoproduct in 50% yield. This product was identified as one
of the two possible photodimers, (ht)-di-DPA or (hh)-di-DPA.
Although the available information does not allow us to
distinguish between these two photoproducts, steric consider-
ations suggest that the head-to-tail product (ht)-di-DPA is likely
to be favored. DMDPA, like 9,10-dimethylanthracene,3b fails
to undergo any observable photodimer formation under the same
conditions, presumably due to the presence of the methyl groups
at the 9- and 10-positions.
The UV-vis spectra of DPA and DMDPA are similar to
that of TPA. Both compounds absorb strongly between 270 and
300 nm and exhibit a 10-30 nm red-shift in their S0fS1
absorption bands relative to those of anthracene. Selective
irradiation experiments on mixtures of DMDPA and anthracene
yield similar results to those carried out with TPA. Irradiation
at 290 nm, where only DMDPA absorbs, led to the exclusive
formation of the cross-cyclomer DMDPA-A, whereas excitation
with 372 nm light yields dianthracene and the cross-coupled
product in a 2:9 ratio. The outcome was somewhat more
complex when DPA and anthracene mixtures were illuminated,
because DPA is capable of forming its homodimer. Selectively
exciting DPA with 281 nm light afforded both di-DPA and
DPA-A in a 1:2 ratio, while excitation of both chromophores
at 372 nm led to the formation of all three possible photoprod-
ucts (Figure 2).
1
cyclomer TPA-A by H NMR. Similar results were observed
with the sample irradiated at 399 nm; again, no di-A formation
was observed. This is consistent with our expectations, since at
these wavelengths only TPA should be excited. In contrast,
irradiation of the third mixture with 376 nm light, a wavelength
where both chromophores absorb, causes the formation of both
di-A and TPA-A in approximately a 1:4 ratio. Trace amounts
of di-TPA were detected in the samples excited at 376 and 399
nm, while no homodimer was formed when the sample was
irradiated at 301 nm (Scheme 1).
These results suggest that the phenyl groups attached to the
anthracene core provide enough of a perturbation to the
absorption spectrum to allow for selective excitation. We were
interested in determining whether the same strategy could be
carried out using analogues that contained fewer pendant phenyl
substituents. Two such derivatives, 2,3-diphenylanthracene
(DPA) and 9,10-dimethyl-2,3-diphenylanthracene (DMDPA)
were therefore prepared from 2,3-diphenyl-9,10-anthraquinone
(Scheme 2).5
Using selective irradiation to restrict the outcome of these
reactions to the creation of cross-cyclomers is not limited to
phenyl-substituted anthracenes such as TPA and DMDPA. The
UV-vis absorption spectrum of 9,10-dimethylanthracene (DMA)
is similar to that of anthracene, but with peaks that are red-
shifted by approximately 20 nm relative to those of the latter.
It should therefore be possible to selectively excite DMA in
(4) It should be noted that while the use of a fluorimeter necessarily
limited the scale of the reactions studied herein, it offered the advantage of
allowing us to easily change the wavelength of excitation. In principle,
selective irradiation can be accomplished at larger scales using higher
intensity light sources (e.g., Rayonets or immersion well systems) fitted
with appropriate cutoff filters.
(5) Bailey, D.; Williams, V. E. Tetrahedron Lett. 2004, 45, 2511-2513.
J. Org. Chem, Vol. 71, No. 15, 2006 5779

specified. In a typical experiment, a 2.5 mL benzene solution
containing 4 mM of each chromophore was subjected to three
freeze-pump-thaw cycles in order to ensure the exclusion of
oxygen from the reaction mixtures. After irradiation for 2 h, the
solvent was removed in vacuo and the ratio of the products was
the presence of anthracene. Indeed, irradiation of a mixture of
DMA and anthracene at 398 nm led to the formation a single
observed photoproduct that was identified as DMA-A.6 When
this mixture was irradiated at 376 nm, both dianthracene and
the mixed cyclomer were formed, as expected. The same
outcome is observed over a range of concentrations, although,
as expected, the reaction is much slower in dilute solution than
in more concentrated samples. For example, when both chro-
mophores were present at initial concentrations of 4 mM, only
about 30% of the DMA was consumed after 2 h, while
approximately 80% of this starting material had reacted after
the same period from a solution containing 10 mM of each
reagent.
In the examples described above, the initial concentrations
of both chromophores were equal. To determine whether the
same selectivity would be observed if one of the two chro-
mophores was present in large excess, we irradiated a chloro-
form solution containing 4 mM DMA and 11 mM anthracene
with 398 nm light. During the initial stages of this experiment,
the cross-cyclomer was formed as the exclusive product. Despite
the low concentration of DMA, its conversion to DMA-A was
complete in under 2 h, presumably due to the large excess of
anthracene available for it to react with. Only after DMA had
been completely consumed was dianthracene observed to slowly
form in trace quantities.
This strategy can also be extended to other substituted
derivatives, such as 9,10-dimethoxyanthracene, which does not
homodimerize, yet is known to form cross-cyclomers with
anthracene.7 Selective excitation of this chromophore with 400
nm light in the presence of anthracene led to the formation of
the cross-cyclomer; again, no dianthracene product was ob-
served.
In conclusion, we have demonstrated the viability of a new
strategy for enforcing formation of the cross-cyclomers in the
[4 + 4]-photocycloaddition reaction of anthracene derivatives
by exciting only one component of a binary mixture. This
selectivity rests on two key features of the molecule being
excited: (a) it must absorb in a range where the other
chromophore does not, and (b) it must be relatively inert toward
homodimer formation. These studies indicate that this strategy
is fairly general, and works either when the excitation wave-
length is blue- or red-shifted relative to S0fS1 absorption bands
of anthracene. This approach is much less restrictive than
previous examples of selective cross-cycloadditions, which
either suffered from poor product stabilities3a,b or required that
both chromophores be unreactive toward homodimerization.3a,c
This added flexibility should facilitate the design of modular
reversible materials whose structures can be altered using light.
1
determined by H NMR.
Samples of the photocyclomers prepared for the purposes of
characterization were obtained by irradiating deoxygenated mixtures
of the appropriate monomers with 350 nm lamps. Reagent ratios
and irradiation times for these preparative experiments were
optimized to maximize the formation of the desired products. Details
of these experiments and the characterization of the photoproducts
can be found in the Supporting Information.
Selective irradiation experiments were carried out in deoxygen-
ated CDCl3 solutions in quartz NMR tubes and using a spectro-
flourimeter as the light source (slit width ) 5 nm). These
experiments were typically carried out using 1 mL solutions
containing 10 mM of each chromophore and were monitored by
1H NMR.
9,10-Dimethyl-2,3-diphenylanthracene (DMDPA). In oven-
dried glassware, 2,3-diphenylanthraquinone (0.30 g, 0.83 mmol)
was dissolved in 30 mL dry diethyl ether and cooled to 0 °C, at
which point 1.4 M MeLi (2.65 mL 3.7 mmol) was slowly added.
After addition, the solution was warmed to room temperature. After
90 min, 30 mL of 10% HCl solution saturated with SnCl2 was added
and allowed to stir for a further 60 min. This solution was then
diluted with water (75 mL) and extracted with diethyl ether (3 ×
75 mL). After the ethereal layer was dried with magnesium sulfate,
the solution was concentrated in vacuo. This crude product was
then directly applied to a silica column and eluted with a mixture
of hexanes and toluene (1:1) to afford 0.07 g (20 mmol, 23%) of
1
the desired product as a yellow solid: mp 204-205 °C; H NMR
(400 MHz, CDCl3) 8.36 s (2H), 8.34 dd (2H, J ) 3.3 Hz, J ) 6.9
Hz), 7.53 dd (2H, J ) 3.3 Hz, J ) 6.9 Hz), 7.33-7.25 m (10H),
3.14 s (6H); 13C NMR (125 MHz, CDCl3) 142.1, 138.2, 130.5,
130.3, 129.5, 128.7, 128.1, 127.3, 126.8, 125.6, 125.1, 14.4; FT-
IR: (cm-1) 3077, 3050, 3017, 2920, 1598, 1490, 1443, 1386, 1021,
876, 766, 742, 699, 638, 595, 544; EI-MS 358.2 (M+ 100), Anal.
Calcd for C28H22: C, 93.81; H, 6.19. Found: C, 93.57; H, 6.30.
2,3-Diphenylanthracene (DPA). In an oven-dried round-bottom
flask, 2,3-diphenylanthraquinone (0.200 g, 0.56 mmol) was dis-
solved in 30 mL of dry THF; to this was carefully added 210 mg
(5.6 mmol) of LiAlH4 and 0.37 g (2.8 mmol) of AlCl3. The resulting
mixture was refluxed for 22 h, cooled, poured over diethyl ether
(50 mL), and carefully quenched with water (50 mL). This mixture
was extracted with diethyl ether (3 × 50 mL). The combined
organic extracts were dried (MgSO4) and concentrated in vacuo.
The crude product was dissolved in 20 mL of xylenes, and 0.20 g
of Pd/C was added to the solution, which was then refluxed for 5
days. The product was purified by column chromatography (silica
gel, eluent: gradient 100% hexanes to 1:1 hexanes/toluene) to afford
0.150 g (82%) of the desired product as a yellow solid: mp 98-
1
100 °C; H NMR (500 MHz, CDCl3) 8.47 s (2H), 8.06 s (2H),
Experimental Section
8.02 dd (2H, J ) 3.3 Hz, J ) 6.4 Hz), 7.48 dd (2H, J ) 3.2 Hz,
J ) 6.5 Hz), 7.28-7.23 m (10H); 13C NMR (100 MHz, CDCl3)
141.5, 139.0, 132.1, 130.0, 129.6, 128.3, 127.9, 126.6, 126.2, 125.5;
FT-IR (cm-1) 3054, 3017, 2916, 1598, 1490, 1427, 1068, 1020,
957, 903, 772, 739, 699, 564, 467; EI-MS 330.1 (M+, 100). Anal.
Calcd for C26H18: C, 94.51; H, 5.49. Found: C, 94.17; H, 5.65.
Nonselective irradiation experiments were performed under a
nitrogen atmosphere in glass Schlenk tubes. These experiments were
carried out in a Rayonet fitted with ten 300 or 350 nm lamps, as
(6) Although no other photoproducts were observed from this reaction,
we did find that more DMA was consumed than anthracene, suggesting
that the former decomposes via an additional, unidentified pathway. This
observation may be associated with the pronounced tendency of DMA to
form a 9,10-endoperoxide when irradiated in the presence of even trace
quantities of oxygen. A similar observation was made for experiments
involving 9,10-dimethoxyanthracene. See: Schmidt, R.; Schaffner, K.;.
Trost, W.; Brauer, H.-D. J. Phys. Chem. 1984, 88, 956.
Supporting Information Available: 1H and 13C NMR spectra
of all compounds; UV-vis absporption spectra of compounds
DMDPA, DMA, and DPA. Synthetic and analytical details for the
preparation of the photocyclomers. This material is available free
of charge via the Internet at http://pubs.acs.org.
(7) Bouas-Laurent, H.; Lapouyade, R. C. R. Acad. Sci., Ser. C 1967, 12,
1061.
JO060838Q
5780 J. Org. Chem., Vol. 71, No. 15, 2006