Multistate π switches: Synthesis and photochemistry of a molecule containing three switchable annelated dihydropyrene units

Article abstract of DOI:10.1021/jo052153g

The synthesis of the multistate photochromic switch 3 is described. This switch contains three dihydropyrene (DHP) units in the most conjugated fully closed form 3-c,c,c. The thermally stable form has the central DHP unit open and is 3-c,o,c. NMR and lase

Full text of DOI:10.1021/jo052153g

Multistate π Switches: Synthesis and Photochemistry of a Molecule
Containing Three Switchable Annelated Dihydropyrene Units
Reginald H. Mitchell,* Cornelia Bohne, Yunxia Wang, Subhajit Bandyopadhyay, and
Chelsea B. Wozniak
Contribution from the Department of Chemistry, UniVersity of Victoria,
PO Box 3065, Victoria, BC, Canada, V8W 3V6
regmitch@uVic.ca
ReceiVed October 14, 2005
The synthesis of the multistate photochromic switch 3 is described. This switch contains three
dihydropyrene (DHP) units in the most conjugated fully closed form 3-c,c,c. The thermally stable form
has the central DHP unit open and is 3-c,o,c. NMR and laser flash photolysis experiments were used to
characterize the multiple states in the photoswitching of 3 by visible and UV light. Three of the possible
five isomeric states of 3-c,o,c were observed. Irradiation of 3-c,o,c by visible light led to the formation
of 3-o,o,o via the isomer 3-c,o,o as an intermediate, which were observed by NMR. Irradiation by UV
light led to the formation of 3-c,c,c, which decays with a lifetime of 7.5 ms.
Introduction
do however, have a disadvantage that the thermal back reaction,
although Woodward-Hofmann forbidden, does occur. In the case
of the thermal reversion, 2-o to 2-c, Eact ) 25 kcal/mol, and so
the open form 2-o has a half-life, τ1/2, of about one week at
room temperature in solution.6 An alternative viewpoint,
especially in examples with lower Eact values, means that these
molecules are also thermochromic!7
Photochromic molecules have become of intense interest
recently because of potential applications to memories and
switches.1 The dimethyldihydropyrene-dimethylmetacyclophane-
diene photochrome 1-c/1-o (c)closed; o)open) belongs in the
diarylethene class of photochromes. They have been much less
studied than the dithienylcyclopentenes of Irie2 and more
recently Branda and others,3 which attract much attention
because of their thermal bistability and fatigue resistance. The
dithienylcyclopentenes are positive photochromes, where the
colorless form on irradiation becomes colored. The dihydropy-
renes such as 1-c/1-o and 2-c/2-o are negative photochromes
where now the thermally stable forms 1-c and 2-c are colored,4-6
but more importantly, when annelated as in the benzo-derivative
2-c/2-o, show a clean 100% conversion6 between the two
photostates, 2-c and 2-o, unlike the dithienylcyclopentene
cousins, which generally tend to form photostationary mixtures
(e.g. 80/20 and 70/30 in refs 3(a) and 3(b)). The dihydropyrenes
(1) Irie, M. Chem. ReV. 2000, 100, 1683-1890.
(2) Irie published 22 papers on dithienylethenes in 2004 alone. Recent
examples: Murakami, M.; Miyasaka, H.; Okada, T.; Kobatake, S.; Irie, M.
J. Am. Chem. Soc. 2004, 126, 14764-14772; Fukaminato, T.; Sasaki, T.;
Kawai, T.; Tamai, N.; Irie, M. J. Am. Chem. Soc. 2004, 126, 14843-14849;
Matsuda, K.; Irie, M. J. Photochem. Photobiol., C 2004, 5, 169-182.
If more than one photochrome is joined together, then
multistate switches are possible, for example with two, the
closed-closed, closed-open, and open-open states might exist.
10.1021/jo052153g CCC: $33.50 © 2006 American Chemical Society
Published on Web 11/30/2005
J. Org. Chem. 2006, 71, 327-336
327

Mitchell et al.
SCHEME 1. Five Possible States of 3 (Only One
Diastereomer of Each State is Shown)
SCHEME 2 a
a Conditions:11 (a) NaNH2/cat. t-BuOK/THF/55 °C; (b) Fe2(CO)9/
benzene/80 °C.
open-closed isomer 3-c,o,c should be the thermally stable
isomer based on the known preference for a dihydropyrene
flanked by two benzene rings to be open, 4, or flanked with
one benzene ring to be closed, 2.6 The all-open isomer, 3-o,o,o
should be accessible by irradiation of 3-c,o,c with visible light
until both closed rings open, while irradiation of 3-c,o,c with
UV light should yield the all closed isomer 3-c,c,c,6 though this
isomer may very rapidly thermally convert back, as does 4. The
isomer 3-c,o,o might be accessible by part-way opening of
3-c,o,c or part-way closing of 3-o,o,o; however the remaining
isomer 3-o,c,o would not be expected to be thermally stable
and thus if formed should have a short lifetime decaying to
3-o,o,o.
Results and Discussion
Syntheses. To prepare the dibenz-annelated compound 4, the
dibromide 5 was reacted with sodium amide and catalytic
t-BuOK in THF, conditions6 which generate arynes, which were
trapped with excess furan to give the bis Diels-Alder adduct
6. The latter was deoxygenated with Fe2(CO)9 to give 4.
We anticipated that 3 could be prepared from 5 via the
analogous adduct 7 derived from addition of iso-furan 8.12 How-
ever, material obtained in this way was very difficult to purify.
It proved superior to first react adduct 6 with 1 equiv of 3,6-
di(2′-pyridyl)-1,2,4,5-tetrazine13 (9) to yield isofuran 10 and
subsequently react this under aryne generating conditions with
bromide 11 to yield adduct 12, which then was subjected again
to tetrazine 9 to yield 13 and again with the aryne from 11 to
yield adduct 7. This adduct showed poor solubility and so was
directly deoxygenated by refluxing in benzene with Fe2(CO)9
to give the tris-DHP product 3. This was then purified by
chromatography and recrystallization from toluene and gave
red crystals, mp 176-181 °C, in 7% overall yield from bis-
adduct 6.
We were the first to report8 fused photochromes where both
photochromes open and close. Dihydropyrenes seem of value
here, since all of the fused or linked ones made so far6,9 have
opened and closed, while in the case of the dithienylethenes,
most have not.10 As more switchable moieties are added in one
molecule, the possibility for more states exists. This paper
reports the synthesis and photochemical study of the “tris-
switch” 3, containing three dihydropyrene units, which in
principle could show 5 distinct states (ignoring diastereomers
of the internal methyl groups, which are always trans within
each dihydropyrene unit), as shown in Scheme 1. The closed-
(3) Recent examples: Gorodetsky, B.; Samachetty, H. D.; Donkers, R.
L.; Workentin, M. S.; Branda, N. R. Angew. Chem., Int. Ed. 2004, 43, 2812-
2815; Wigglesworth, T. J.; Branda, N. R. AdV. Mater. 2004, 16, 123-125;
Myles, A. J.; Wigglesworth, T. J.; Branda, N. R. AdV. Mater. 2003, 15,
745-748; Mulder, A.; Jukovic, A.; Luca, L. N.; van, E. J.; Feringa, B. L.;
Huskens, J.; Reinhoudt, D. N. Chem. Commun. 2002, 2734-2735; Utsumi,
H.; Nagahama, D.; Nakano, H.; Shirota, Y. J. Mater. Chem. 2002, 12, 2612-
2619.
Although this yield was not as great as the direct reaction of
Scheme 3, the product could be obtained pure, which was
essential for the photochemistry studies. The overall structure
was confirmed by mass spectrometry, since MH+ (C82H93)
calculated was 1077.728 and found was 1077.726, and a
satisfactory elemental analysis was obtained as well. In theory,
three diastereomers of 3-c,o,c should be obtained, represented
graphically by 3a, 3b, and 3c in Figure 1, though not necessarily
in equal amounts. The stability of the bis-adducts 7 should
control which isomers of 3 are formed. However, PCMODEL
calculations14 on all of the possible isomers of 7 reveal that
(4) Sheepwash, M. A.; Mitchell, R. H.; Bohne, C. J. Am. Chem. Soc.
2002, 124, 4693-4700.
(5) Sheepwash, M. A. L.; Ward, T. R.; Wang, Y.; Bandyopadhyay, S.;
Mitchell, R. H.; Bohne, C. Photochem. Photobiol. Sci. 2003, 2, 104-112.
(6) Mitchell, R. H.; Ward, T. R.; Chen, Y. Wang, Y.; Weerawarna, S.
A.; Dibble, P. W.; Marsella, M. J.; Almutairi, A.; Wang, Z. Q. J. Am. Chem.
Soc. 2003, 125, 2974-2988.
(7) Mitchell, R. H.; Brkic, Z.; Sauro, V. A.; Berg, D. J. J. Am. Chem.
Soc. 2003, 125, 7581-7585.
(8) Mitchell, R. H.; Ward, T. R.; Wang, Y.; Dibble, P. W. J. Am. Chem.
Soc. 1999, 121, 2601-2602.
(9) Mitchell, R. H. Eur J. Org. Chem. 1999, 2695, 5-2703. Mitchell,
R. H.; Bandyopadhyay, S. Org. Lett. 2004, 6, 1729-1732.
(10) Peters, A.; Branda, N. R. AdV. Mater. Opt. Electron. 2000, 245-
249. Kaieda, T.; Kobatake, S.; Miyasaka, H.; Murakami, M.; Iwai, N.;
Nagata, Y.; Itaya, A.; Irie, M. J. Am. Chem. Soc. 2002, 124, 2015-2024;
but see Kobatake, S.; Irie, M. Tetrahedron 2003, 59, 8359-141 for one
that does.
(11) Mitchell, R. H.; Chen, Y. Tetrahedron. Lett. 1996, 37, 5239-5242.
(12) Mitchell, R. H.; Ward, T. R. Tetrahedron 2001, 57, 3689-3695.
(13) Geldard, J. F.; Lions, F. J. Org. Chem. 1965, 30, 318-319.
328 J. Org. Chem., Vol. 71, No. 1, 2006

Multistate π Switches
SCHEME 3 a
a Conditions: (a) NaNH2/cat. t-BuOK/THF/55 °C; Fe2(CO)9/benzene/
80 °C.
FIGURE 1. Graphical representation of the diastereomers of 3 (u )
up or d ) down refers to the orientation of each of the top internal
methyl groups).
while the syn adducts (both oxygen atoms on the same side)
are about 1 kcal/mol lower in energy than the trans adducts,
within each set there is <0.5 kcal/mol difference in energy
between the various u-d isomers of the methyl groups.
Diastereomers 3a ) 3uuu and 3b ) 3udu have C2h symmetry,
and so the DHP internal methyl protons of each isomer should
be a single peak, while in 3c ) 3uud, there should be two.
Indeed, four NMR signals are seen for the mixture of isomers
at δ -1.360, -1.364, -1.622, and -1.626 (initially the ratio
of the -1.3:-1.6 peaks was ∼1:1, but chromatography and
recrystallization led to fractions richer in the -1.3 peaks). Given
the closeness in energy of the precursors 7, we are unable to
assign which peaks belong to which isomer on these data, though
one might expect that the less symmetrical isomer 3c (3uud)
would have one peak from each group, i.e., at ∼δ -1.36 and
-1.63. The internal methyl protons of the cyclophane rings all
appear as a broad signal at δ 1.28, and the t-Bu methyl signals
appear at δ 1.53 and 1.56 for the DHP rings and 1.360 and
1.365 for the cyclophane rings and are not as well differentiated
as the internal methyl protons. The 13C NMR signals for the
internal methyl carbons of 3 appeared at 18.8, 18.0, and 17.5
ppm, with the peak at δ 18.8 being more intense than the other
two and so also do not permit assignment. It should be noted
that chloroform solutions of 3 readily form small amounts of
cation radicals that broaden the proton spectra; so solutions for
NMR were filtered through alumina immediately prior to
obtaining the spectra. Fortunately, as with other dihydropyrenes,6
the diastereomers (3a,b,c) of 3-o,o,o formed by irradiation with
visible light also thermally close to return the starting isomers
(3-c,o,c) but at slightly different rates. Careful examination using
proton NMR of the growth of the various peaks of the DHPs
formed in this thermal reversal reaction suggests that the internal
FIGURE 2. (a) Sequential UV-vis spectra of 3-c,o,c on irradiation
with visible light (λ > 550 nm) and (b) sequential UV-vis spectra of
3-o,o,o on irradiation with UV (λ ) 350 nm).
methyl proton peaks at δ -1.364 and -1.622 belong to the
same isomer, which must then be 3c (3uud). The peaks at δ
-1.360 and -1.626 thus belong one to each of the other two
isomers.
Simple Photochemistry: Continuous Irradiation.
Irradiation with broadband light sources (λ > 500 nm) of
the thermally stable isomer, 3-c,o,c, is expected6 to give the
fully open isomer, 3-o,o,o. Indeed, irradiation with light above
550 nm led to the disappearance of the visible bands with an
increase in absorption in the UV, as expected (Figure 2a). It is
important to note that a single isosbestic point was not observed
at 300 nm (see SI Figure S1 for an expansion around 300 nm)
but rather one point for the early spectra and one point for the
late spectra, suggesting that 3-c,o,c opens first to 3-o,o,c which
further opens to 3,o,o,o. Likewise irradiation of 3-o,o,o with
350-nm light led to the return of the absorption for 3-c,o,c
(Figure 2b), again with the observation (SI Figure S2) of an
early and a late isosbestic point suggesting sequential closing.
That the fully open compound, 3-o,o,o, is formed, is not in
doubt, since when sequential NMR spectra (Figure 3a) are
obtained (irradiation using >613 nm light in CDCl3), all the
dihydropyrene internal methyl signals disappear with irradiation
of the sample with visible light. The same internal methyl NMR
signals reappear (Figure 3b) when this opened sample is
irradiated at 350 nm. All diastereomers of 3-c,o,c must thus
open to 3-o,o,o and close back to 3-c,o,c, because no residual
NMR peaks corresponding to either 3-c,o,c (in the opening) or
(14) PCMODEL, V8.5; Serena Software: Bloomington, IN, 47402-3076.
J. Org. Chem, Vol. 71, No. 1, 2006 329

Mitchell et al.
one DHP to open at a different rate to the other, and so form
one isomer preferentially in the opening reaction. As can be
seen in Scheme 5, the phenyl-substituted parent 15 was an
intermediate, and the phenyl-substituted benzoDHP 19 was
easily obtained, so that both could act as model compounds in
our interpretation of the behavior of 18. In fact, the phenyl
substituted benzoDHP 19 opens with visible light about four
times faster than benzoDHP 2, and this, on the basis of observed
chemical shifts, turns out to be true also for 18, such that the
spectra for 18-c,o,o are more readily observed, than for
unsubstituted 3-c,o,o. Lai15 has shown that introduction of an
aryl substituent on to a DHP slightly reduces the ring current
in the DHP by resonance conjugation, and thus the internal
methyl protons are slightly deshielded with respect to the
unsubstituted example. Thus in 15 and 19, the internal methyl
protons are at δ -3.84/-3.85 and -1.453/-1.456, respectively,
slightly deshielded from those of the two parents 2,7-di-t-Bu-
1-c and 2-c at δ -4.06 and -1.58, respectively.12 In 3-c,o,c,
the internal methyl protons appear at δ -1.37 and -1.63, while
in 18-c,o,c, there are two sets of each proton type at δ -1.23/
-1.37 and δ -1.49/-1.62. The peaks at δ -1.23 and -1.49
thus belong to the phenyl-substituted side. These more deshield-
ed peaks disappear first on irradiation with visible light (>590
nm in toluene, Figure 4). In fact the internal methyl proton
chemical shifts of the newly formed isomer, 18-c,o,o, are not
exactly the same as for the unsubstituted DHP side of 18-c,o,c,
and so the more shielded pair actually become two close peaks
each, where the peaks due to the original shifts are replaced by
those at the new shifts, i.e., in d8-toluene, δ -1.161 and -1.354
become δ -1.175 and -1.367; in this solvent the peaks at δ
-1.030 and -1.228 have mostly gone in ∼35 min, while
complete disappearance (i.e., conversion to 18-o,o,o) requires
more than 90 min of irradiation. This leaves no doubt that during
visible light irradiation the sequence 18-c,o,c to 18-c,o,o to
18-o,o,o is followed. Sequential UV-visible opening and
closing spectra for 18 are given in the Supporting Information
(Figures S3 and S4) as well as sequential 1H NMR spectra in
the UV closing of 18-o,o,o to 18-c,o,c (Figure S5).
Photochemistry: Transient Studies. Studies with model
compounds were first performed in order to understand the
photochemistry of switch 3. The outside DHP moieties behave
as and thus can be modeled by the benzo-DHP 2-c. The
cyclophanediene (CPD) isomer, 2-o, formed when benzo-DHP
is irradiated with visible light, is known to have a very long
half-life (t1/2 g 5 h, 46 °C).6 In the laser flash photolysis
experiments, the benzo-CPD isomer (2-o) was longer lived than
the longest lifetime that we can measure with the laser flash
photolysis system. The system was reconfigured to measure
lifetimes in the millisecond time domain (see SI for details),
and the upper limit for lifetime measurements is determined by
how fast the transient diffuses from the volume irradiated by
the laser into the adjacent nonirradiated solution. Excitation of
benzo-DHP 2-c at 532 nm led to bleaching of the visible
absorption bands due to the formation of the CPD isomer 2-o.
This signal was used to determine the time it takes for transients
to diffuse out of the irradiation volume (τ ca. 4-6 s). No other
transients were observed on shorter time scales.
FIGURE 3. Sequential NMR spectra in the (a) visible light (λ > 613
nm) opening of 3-c,o,c and (b) UV light (λ ) 350 nm) closing of
3-o,o,o.
3-o,o,o (in the closing) were observed, after the irradiation drove
each reaction to completion.
On both opening and closing, new methyl proton signals
appear (slightly shielded from those of 3-c,o,c) which disappear
again as irradiation is continued. It is important to note that the
chemical shifts observed for the internal methyl peaks of the
transient isomer are the same when 3-c,o,c is opened by
irradiation with visible light and when 3-o,o,o is closed by
irradiation at 350 nm. Therefore the same isomer is formed in
both irradiations. The only two species that can be formed in
the visible irradiation of 3-c,o,c are 3-c,o,o and 3-o,o,o, because
the excited states of the open moieties (which might lead to
3-c,c,c or 3-c,c,o) cannot be populated with irradiation above
550 nm. Therefore, the transient species observed was assigned
to the 3-c,o,o isomer, which is equivalent to 3-o,o,c. Unfortu-
nately, we were not able to separate and purify this isomer. To
obtain clearer evidence, we thought that if we could differentiate
the two DHP sides of 3, such that one of the DHPs opened
much faster than the other, the structure of the intermediate
isomer equivalent to 3-c,o,o, would be easier to observe and
verify.
Our previous work4,5 has shown that substituents can
substantially alter the opening rate of DHPs with visible light.
Unfortunately, 3 immediately turns black under acidic condi-
tions, e.g., BF3‚Et2O, and so acetylation, for example, was not
successful. Eventually we were able to synthesize the 4-phenyl-
substituted derivative of 3. The full synthesis is described in
the Supporting Information but is shown briefly in Scheme 5.
Now, 18-c,o,o is no longer identical to 18-o,o,c and thus since
one DHP is substituted and the other is not, one might expect4,5
The central moiety of 3 corresponds to the disubstituted benzo
compound 4, for which the CPD isomer is the thermodynami-
cally stable isomer.6 Compound 4 absorbs strongly in the UV
(15) Lai, Y. H.; Chen, P.; Dingle, T. W. J. Org. Chem. 1997, 62, 916-
924. Lai, Y. H.; Jiang, J. P. J. Org. Chem. 1997, 62, 4412-4417.
330 J. Org. Chem., Vol. 71, No. 1, 2006

Multistate π Switches
SCHEME 4 a
a Conditions: (a) 0 °C/THF/3 h; (b) NaNH2/cat. t-BuOK/THF/20 °C/5 h; (c) 20 °C/THF/1 h.
SCHEME 5 a
FIGURE 4. Sequential internal methyl proton NMR spectra of visible
light (λ >590 nm) opening of 18-c,o,c.
FIGURE 5. Normalized transient absorption spectra for the irradiation
of 4 with 266 nm obtained with delays after the laser pulse of 27 ms
(O, normalized at 430 nm; ∆A ) 0.43) and 370 ms (b, normalized at
435 nm; ∆A ) 0.056).
of 4 in the laser flash photolysis experiment led to long-lived
transients (Figure 5). The spectrum at long delays was similar
to that observed at short delays, but the absorption at 340 nm
disappeared, while the ratio between the absorption in the 430-
nm region and the absorption in the 650-nm region decreased.
The spectra observed are consistent with the formation of a more
a Conditions: (a) PhB(OH)2/Pd(PPh3)2Cl2/THF/2 M Na2CO3, reflux 2
h; (b) NBS/CH2Cl2-DMF/-78 °C; (c) NaNH2/KOBu-t/THF, RT 2-3 h;
(d) Fe2(CO)9/PhH, reflux 1 h; (e) Furan/NaNH2/KOBu-t/THF, RT 5 h.
region of the spectrum, with a weak shoulder observed between
300 and 350 nm. Irradiation at 266 nm of cyclohexane solutions
J. Org. Chem, Vol. 71, No. 1, 2006 331

Mitchell et al.
FIGURE 7. Normalized transient absorption spectra for the irradiation
of 20 with 532 nm obtained with a delay after the laser pulse of 12 ms
(O, normalized at 480 nm; ∆A ) 0.35) and 220 ms (b, normalized at
480 nm; ∆A ) 0.055).
FIGURE 6. Normalized transient absorption spectra for the irradiation
of 20 with 266 nm obtained with a delay after the laser pulse of 6.8
ms (O, normalized at 480 nm; ∆A ) 0.20) and 440 ms (b, normalized
at 410 nm; ∆A ) 0.016).
conjugated system than the CPD isomer, as expected for the
formation of the DHP isomer. The changes in the transient
spectra with time indicate that more than one transient with
similar absorption properties were present.
The transient kinetics of 4 in cyclohexane were measured at
430 and 660 nm. No decay of the absorptions was observed in
the nano- and microsecond time domains. The kinetics at longer
times did not follow a monoexponential decay. The decays were
fitted to the sum of two exponentials, and the same two lifetimes
(0.29 ( 0.07 s and 0.9 ( 0.3 s) were recovered from the fits at
430 and 600 nm. The longest lifetime value is shorter than the
diffusional lifetime of transients out of the irradiation volume
(ca. 6 s) but could have been shortened somewhat because of
the diffusional effect. The nonexponential decays were observed
in solvents of different polarity (cyclohexane vs acetonitrile),
at various concentrations of 4 (varied by a factor of 5) and at
various energies for the laser pulses. In addition, the same
lifetimes were observed for two independently synthesized
samples of 4. These control experiments suggest that the
nonexponential decay is not due to the presence of an impurity
or due to the aggregation of 4. The transients were not quenched
by oxygen, since the same lifetimes were observed in aerated
solutions and solutions bubbled with nitrogen.
FIGURE 8. Normalized transient absorption spectra for 3 when excited
at 266 nm (A, delays of 14 µs (O) and 3.6 ms (b), normalized at 520
nm) and when excited at 530 nm (B, delays 1.2 ms (O) and 33 ms (b),
normalized at 490 nm). The dashed line at 500 nm was included to
facilitate the comparison of both spectra.
Compound 206 was the second model compound studied. The
thermodynamically stable isomer which is shown, has a DHP
moiety, which corresponds to benzo-DHP 2 and a CPD moiety,
which corresponds to compound 4 and thus is actually 20-c,o.
Excitation at 530-532 nm should lead to the excitation of the
DHP moiety, whereas the photons corresponding to 266 nm
have enough energy to excite the CPD moiety of 20.
whereas at 410 nm a growth was followed by a decay (see
Figure 6 in SI). The lifetime for the fast component at both
wavelengths was 100 ( 8 ms. The long-lived species decayed
with a time constant similar to the diffusion of transients out of
the irradiation volume, and for this reason, the formation of
the long-lived transient is “permanent” (τ g 1 s) on the time
scales accessible by laser flash photolysis. The transients were
not quenched by oxygen.
Excitation of 20 at 532 nm led to the formation of a transient
with maximum absorption at 480 nm (Figure 7).
In the 400-nm region, a bleaching was observed. The same
kinetics was observed at all monitored wavelengths, where the
decays were mono exponential with a lifetime of 110 ( 10 ms.
This transient was not quenched by oxygen. At all wavelengths
above 400 nm, a “permanent” bleaching was observed after the
decay of the transient (see Figure S7 in SI).
The transient absorption spectra for excitation of 3 at 266
nm initially showed a peak at 350 and 520 nm (Figure 8). An
additional absorption grew in at 410 nm at longer delays. The
ratio between the absorptions at 350 and 520 nm did not change
with the delay after the laser pulse. Excitation of 3 at 532 nm
Excitation of 20 in cyclohexane at 266 nm led to a transient
with an absorption maximum at 480 nm, which decayed into a
new transient with an absorption maximum at 410 nm (Figure
6). At 480 nm, the decay kinetics showed two components,
332 J. Org. Chem., Vol. 71, No. 1, 2006

Multistate π Switches
the energy of the laser pulses. Finally, we assumed that the
absorption of one photon led to the breakage/formation of one
bond for the switching between the DHP/CPD systems.
Excitation of 3, 4, and 20 with 266-nm light led to the closure
of the CPD moiety. In the case of 4, the DHP isomer of 4 was
previously detected by NMR at low temperatures, but its lifetime
at room temperature is short (estimated ca. 1-2 s)6 because of
the fast thermal return to 4. Excitation of 4 at 266 nm led to a
transient spectrum with enhanced absorption in the regions
expected for a DHP isomer. The transient absorption above 600
nm with respect to the absorption in the 400 nm region is higher
than that observed in the absorption spectra of stable DHPs.
This difference can be explained by the fact that the transient
spectra correspond to difference spectra between the absorption
of the transient and the absorption of the precursor.
Two transients were observed for 4, which have very similar
absorption spectra. This result suggests that the transients have
a chromophore similar to the conjugated aromatic systems of
the DHP. One of the transients is the DHP isomer of 4, and
this isomer has either a lifetime of 0.3 or 0.9 s. The identity of
the second transient is unclear at this point. The second transient
cannot be a triplet excited state or a radical because both of
these intermediates should have been quenched by oxygen.
Although cis isomers of dihydropyrenes are known,16 no con-
version between the cis series and the trans series has ever been
observed. One possibility is that the biradical 4-DHP-BR is
the second transient. This intermediate is still planar, and
calculations have indicated that radicals of this type could have
low energy17 and for this reason would be much less reactive
toward oxygen. However, we do not have any Supporting
Information for this assignment. In addition, we cannot deter-
mine if the two transients were formed concomitantly within
the laser pulse or if they are formed sequentially. The latter
can only be concluded if a growth of a transient is observed,
but the absence of growth kinetics does not eliminate the
possibility of the sequential formation of the two transients.
FIGURE 9. Transient kinetics at 410 nm for the irradiation at 266
nm of 3 in cyclohexane.
led to an increased absorption at 490 nm. This absorption is
blue shifted when compared to the transient absorption observed
for the excitation of 3 at 266 nm. At longer delays an increased
absorption was observed at 410-415 nm. No increase in
absorption was observed at 350 nm.
The transient absorption signal when 3 was excited at 532
nm did not follow a monoexponential function (see Figure S8
in SI). The decays were fitted to the sum of two exponentials,
and the recovered lifetimes were 28 ( 5 ms and 250 ( 30 ms.
The short lifetime corresponds to the formation of the transient
with enhanced absorption at 410 nm. None of the transients
were quenched by oxygen.
The kinetics for the excitation of 3 at 266 nm were also
biphasic, where at 520 nm only a decay was observed, whereas
at 410 nm a growth (Figure 9) was followed by a decay. The
lifetime for the fast component was 7.5 ( 0.5 ms, whereas the
second component decayed with the time constant for the
diffusion out of the irradiation volume (τ g 1s). None of the
transients were quenched by oxygen.
Discussion
NMR experiments provide valuable information for the
assignment of the transients observed in the laser flash photolysis
experiments. The longest lifetime of an isomer that we can
measure by laser flash photolysis is ca. 1 s. For this reason,
any species that is observable by NMR will appear as a
“permanent” species in the laser flash photolysis experiments.
The most stable isomers of the multichromophoric compounds
3 and 20 contain DHP and CPD moieties within the same
structure. Excitation at 530-540 nm corresponds to energies
significantly lower than required to excite the S1 state of the
CPD chromophore, and for this reason, we assumed that only
the DHP moiety was excited when 3 and 20 were irradiated at
530-540 nm. At 266 nm, the energy of the photons is high
enough to excite the CPD moiety, and at this wavelength, the
molar absorption coefficient is higher for CPD than DHP.4 For
this reason, we assumed that irradiation at 266 nm led mainly
to the excitation of the CPD moieties. This rationale assumes
that the DHP and CPD chromophores can be treated as
independent. This is warranted because molecules containing
at least one CPD unit are not planar and extended conjugation
throughout the whole molecule is hindered. However, a small
degree of extended conjugation has been previously observed
in the fluorescence of arene [e]-annelated CPDs.5 A further
assumption made is that transients are formed from precursors
that absorb one photon. Sequential absorption of photons is
unlikely to lead to the transient phenomena observed because
the transient spectra and lifetimes did not change by varying
The thermodynamically stable isomer of 20 corresponds to
the 20-c,o isomer. It was previously shown6 that irradiation of
20-c,o with visible light above 590 nm led to the formation of
(16) Mitchell, R. H.; Boekelheide, V. J. Chem. Soc., Chem. Commun.
1970, 1555-1557. Mitchell, R. H.; Boekelheide, V. J. Am. Chem. Soc.
1974, 96, 1547-1557. Kamp, D.; Boekelheide, V. J. Org. Chem. 1978,
43, 3475-3477. Mitchell, R. H.; Chaudhary, M.; Kamada, T.; Slowey, P.
D.; Williams, R. V. Tetrahedron 1986, 42, 1741-1744. Mitchell, R. H.;
Bodwell, G. J.; Vinod, T. K.; Weerawarna, K. S. Tetrahedron. Lett. 1988,
29, 3287-3290. Lai, Y. H.; Chew, P. J. Org. Chem. 1989, 54, 4586-
4590. Lai, Y. H.; Zhou, Z. L. J. Org. Chem. 1994, 59, 8275-827. Mitchell,
R. H.; Iyer, V. S.; Khalifa, N.; Mahadevan, R.; Venugopalan, S.;
Weerawarna, S. A.; Zhou, P. J. Am. Chem. Soc. 1995, 117, 1514-1532.
Bodwell, G. J.; Bridson, J. N.; Chen, S. L.; Li, J. Eur. J. Org. Chem. 2002,
243-249.
J. Org. Chem, Vol. 71, No. 1, 2006 333

Mitchell et al.
SCHEME 7. Possible Decay Pathways for 3-c,o,c When
Excited at 266 nm (Isomers Detected by NMR Are Shown in
Squares)
SCHEME 6. Possible Decay Pathways for 20-c,o When
Excited at 266 nm.
than 1 s but short enough for this transient not to be observed
in the NMR experiment.
Excitation of 20-c,o at 532 nm should lead to the formation
of 20-o,o, which was detected by NMR. The permanent
bleaching observed at all wavelengths above 400 nm is con-
sistent with the formation of the 20-o,o isomer, since the
extended conjugation is lost in this isomer. The assignment for
the transient with absorption at 480 nm is not evident. This
transient is not a triplet excited state or a radical because it is
not quenched by oxygen. The transient is also not the 14-c,c
isomer formed by the absorption of two 532 nm photons,
because the transient spectrum obtained is different from that
measured when 20 was excited at 266 nm, and the formation
of the 410 nm transient (20-c,o) was not observed. For this
reason the equal lifetime (100 ms) measured for the transients
when 20 is excited at 266 or 532 nm must be coincidental.
The isomers of 3 that can be detected by NMR are the 3-c,o,c
(stable isomer), 3-o,o,o (fully open isomer), and 3-o,o,c
(equivalent to 3-c,o,o) isomers. Therefore, the possible isomers
with short lifetimes are the 3-c,c,c, 3-c,c,o (equivalent to 3-o,c,c),
and 3-o,c,o isomers. The transient spectra for the irradiation of
3 at 266 nm are consistent with the formation of a more
conjugated isomer, which can only be the 3-c,c,c isomer. If this
transient decayed only to the starting compound monoexpo-
nential kinetics would have been observed, with a decay back
to the baseline at all wavelengths. Since this is not the case,
the isomer 3-c,c,o also has to be formed in the decay of 3-c,c,c.
The lifetime of the 3-c,c,c isomer is 7.5 ms. This lifetime is
shorter by a factor of 13 than the lifetime for the fully closed
form of 20. A slight decrease (ca. 2×) of the lifetime would
have been expected from statistical considerations, because there
are two flanking DHP moieties in 3 when compared to 20.
However, the decrease by an order of magnitude of the lifetime
indicates that a larger degree of conjugation enhances the rate
constant either for the ring opening of the flanking DHPs or
the ring opening of the central dibenzo-DHP moiety.
20-o,o. This isomer was detected by NMR and had a half-life
of 5 h at 46 °C or several days at room temperature. Formation
of 20-o,o led to an increase in the absorption in the UV region
of the spectrum when compared to the absorption of 20-c,o.
Preliminary laser flash photolysis experiments were performed
previously where 20 was irradiated at 355 nm.9a The transient
absorption spectrum recorded was similar to the one observed
in this work when 20 was excited at 266 nm. The growth of
the transient at 410 nm was not observed in the earlier study
because of the limited time range for those laser flash photolysis
experiments.
The excitation of 20 at 266 nm led to a transient with a
lifetime of 100 ms followed by an isomer that had a lifetime
longer than 1 s. In contrast to the kinetics for 4, a growth was
observed at 410 nm, which was followed by a decay. This result
suggests that the transients are sequentially formed. The
excitation of 20-c,o isomer at 266 nm leads to the formation of
the 20-c,c transient. The maximum of the transient at 485 nm
indicates that this transient is more conjugated than the
precursor, as expected for the formation of 20-c,c. This transient
decays with a 100-ms lifetime, but it forms an isomer different
from the precursor (20-c,o) because a second transient with
increased absorption at 410 nm was formed. The 20-c,c isomer
can decay either by (pathway a, Scheme 6) breaking the bond
for the dibenzo-DHP moiety forming 20-c,o (precursor) or by
(pathway b, Scheme 6) breaking the bond of the benzo-DHP
forming the 20-o,c isomer. From the kinetic data available, it
is not possible to determine the partition between the formation
of 20-c,o and 20-o,c from 20-c,c. The 20-o,c isomer has to form
the 20-o,o isomer (pathway c, Scheme 6) before regenerating
the precursor 20-c,o (pathway d, Scheme 6). The transient with
absorption at 410 nm has a lifetime longer than measurable with
the laser flash photolysis system, and it could be either the
20-o,c or the 20-o,o isomer. The transient with the absorption
at 410 nm is assigned to the 20-o,c isomer, because the 20-o,o
formed when 20 is excited at 532 nm does not have an
absorption at 410 nm. The lifetime of 20-o,c has to be longer
The growth of the signal at 410-415 nm indicates the
formation of isomers from 3-c,c,c that are different from the
stable precursor (3-c,o,c). This new transient has a lifetime
longer than 1 s. The decay of 3-c,c,c will lead to the formation
of 3-c,c,o, which can further decay to 3-c,o,o and 3-o,c,o. The
latter will decay to 3-o,o,o. One of the possible isomers
absorbing at 410-415 nm is 3-c,c,o, but the absorption could
also be due to the isomers 3-c,o,o and 3-o,o,o, which are
detectable by NMR.
Excitation of 3-c,o,c with 532-nm light could lead to 3-c,o,o
(which is equivalent to 3-o,o,c). This isomer is stable enough
to be detected by NMR, and for this reason is responsible for
(17) Mitchell, R. H.; Dingle, T. W.; West, P. R.; Williams, R. V.;
Thompson, R. C. J. Org. Chem. 1982, 47, 5210-5214. Mitchell, R. H.;
Williams, R. V.; Dingle, T. W. J. Am. Chem. Soc. 1982, 104, 2560-2571.
334 J. Org. Chem., Vol. 71, No. 1, 2006

Multistate π Switches
(40 mL) under N2, and the mixture was stirred at ∼20 °C for 5 h.
The solvent was evaporated, and the residue was filtered through
alumina using hexanes/benzene (1:1) as eluant and gave 58 mg
(72%) of the product 12 as a green solid as a mixture of four
isomers, LSIMS m/z 793.6 (MH+); HRMS. Calcd for C58H64O2
(MH+): 793.4985. Found: 793.4987. These four isomers on further
chromatography over aluminum oxide using hexanes/benzene (1:
1) as eluant were separated into two portions which were oxygen
bridge syn and anti isomers, each containing two isomers.
the “permanent” spectral changes observed in the laser flash
photolysis experiments. The identity of the transients with 28-
and 250-ms lifetimes is not known. The same arguments as
outlined for 20 are also applicable here. The interesting point
is that two fast-decaying transients are observed for 3 whereas
only one is observed for 20.
Conclusions
Eluted first were anti isomers (I and II in a ∼1:2 ratio): mp
(dec) 183-188 °C; 1H NMR isomer I: δ 8.85 (d, J ) 1.3 Hz, 1H),
8.67 (d, J ) 1.3 Hz, 1H), 8.66 (d, J ) 1.2 Hz, 1H), 8.57 (d, J )
1.3 Hz, 1H), 8.22-8.21 (m, 2H), 8.16-8.15 (m, 2H), 8.084 (s,
2H), 7.97 (d, J ) 0.7 Hz, 1H), 7.88 (d, J ) 0.7 Hz, 1H), 7.22 &
7.09 (dd, J ) 5.5, 1.7 Hz, 1H each), 6.57-6.56 (m, 1H), 6.53-
6.52 (m, 1H), 1.75, 1.72, 1.71, 1.69, 1.68, 1.67, 1.63 (s from both
isomers I and II, ratios: 2:1:1:2:2:2:2, C(CH3)3), -3.28, -3.77,
-4.23, -4.41 (s, 3H each, CH3). Isomer II: δ 8.79 (d, J ) 1.2 Hz,
1H), 8.70 (d, J ) 1.2 Hz, 1H), 8.65 (d, J ) 1.2 Hz, 1H), 8.53 (d,
J ) 1.4 Hz, 1H), 8.25 (d, J ) 1.0 Hz, 1H), 8.20 (d, J ) 1.3 Hz,
1H), 8.17 (d, J ) 1.2 Hz, 1H), 8.13 (d, J ) 1.2 Hz, 1H), 8.080 (s,
2H), 7.93 (d, J ) 0.7 Hz, 1H), 7.90 (d, J ) 0.7 Hz, 1H), 7.21 &
7.08 (dd, J ) 5.5, 1.7 Hz, 1H each), 6.57-6.56 (m, 1H), 6.53-
6.52 (m, 1H), 1.75, 1.72, 1.71, 1.69, 1.68, 1.67, 1.63 (s from both
isomers I and II, ratios: 2:1:1:2:2:2:2, C(CH3)3), -3.25, -3.77,
-4.24, -4.43 (s, 3H each, CH3). 13C NMR Both isomers: δ 145.50,
145.41, 145.21, 144.98, 141.04, 140.95, 140.68, 140.41, 140.27,
137.92, 137.76, 137.35, 137.23, 137.08, 136.98, 136.91, 136.67,
128.56, 128.24, 128.10, 127.88, 127.59, 127.46, 127.20, 126.88,
124.29, 124.26, 121.34, 121.20, 121.08, 117.63, 117.31, 116.53,
116.40, 115.72, 115.30, 115.15, 115.02, 114.83, 81.90, 81.70, 81.62,
81.52, 81.38, 81.19, 35.98, 35.89, 35.83, 35.77, 33.06, 32.91, 32.81,
32.71, 32.28, 32.15, 31.94, 31.91, 31.65, 31.62, 31.25, 31.07, 30.32,
16.39, 16.29, 15.07, 15.00, 14.44, 14.28, 14.14.
One of the challenges in designing multichromophoric
photoswitches is to string together several photochromic units
and to be able to address each one of them individually. The
objective of this work was to establish in a molecule with three
units if each one of them could be switched and if more than
one isomer could be formed. In addition, the molecule chosen
had a mixture of open and closed units in its thermodynamically
stable form (3-c,o,c). The fully closed (3-c,c,c) and fully open
forms (3-o,o,o) can be formed by UV and visible irradiation,
respectively. In addition, the transient formation of 3-c,o,o was
observed by NMR and the formation of 3-c,c,o was inferred
from laser flash photolysis studies. Despite the fact that several
of the lifetimes of the isomers are short, we established that the
units in multichromophoric switches based on dihydropyrenes
can be individually switched. The next challenge will be to tune
the absorption properties of each moiety to be able to address
specifically each unit.
Experimental Section
General conditions and spectral assignment methods are given
in the Supporting Information. The proton numbering scheme for
compounds 10 and 3 is shown below.
Eluted second were the syn isomers III and IV (in about a 2:3
1
ratio): mp (dec) 179-182° C; H NMR of isomer III: δ 8.84 (d,
J ) 1.2 Hz, 1H), 8.69-8.68 (m, 1H), 8.66-8.65 (m, 1H), 8.55 (d,
J ) 1.3 Hz, 1H), 8.21 (d, J ) 1.2 Hz, 2H), 8.14 (d, J ) 1.3 Hz,
1H), 8.13 (d, J ) 1.3 Hz, 1H), 8.07 and 8.06 (AB, J ) 8.5 Hz,
2H), 7.97 (d, J ) 0.6 Hz, 1H), 7.89 (d, J ) 0.5 Hz, 1H), 7.07 (dd,
J ) 5.5, 1.9 Hz, 1H), 6.93 (dd, J ) 5.5, 1.8 Hz, 1H), 6.58 (dd, J
) 1.8, 0.7 Hz, 1H), 6.56 (dd, J ) 1.8, 0.8 Hz, 1H), 1.76, 1.722,
1.719, 1.685, 1.677, 1.65, 1.61 (s from both isomers III and IV,
C(CH3)3), -3.29, -3.47, -4.32, -4.70 (s, 3H each, CH3). Isomer
IV: δ 8.82 (d, J ) 1.1 Hz, 1H), 8.69-8.68 (m, 1H), 8.66-8.65
(m, 1H), 8.51 (d, J ) 1.2 Hz, 1H), 8.27 (d, J ) 1.1 Hz, 1H), 8.20
(d, J ) 1.1 Hz, 1H), 8.16 (d, J ) 1.1 Hz, 1H), 8.11 (d, J ) 1.2 Hz,
1H), 8.06 (s, 2H), 7.93 (d, J ) 0.5 Hz, 1H), 7.90 (d, J ) 0.6 Hz,
1H), 7.09 (dd, J ) 5.5, 1.8 Hz, 1H), 6.94 (dd, J ) 5.5, 1.8 Hz,
1H), 6.53 (dd, J ) 1.9, 0.8 Hz, 2H), 1.76, 1.722, 1.719, 1.685,
1.677, 1.65, 1.61 (s from both isomers III and IV, C(CH3)3), -3.26,
-3.48, -4.31, -4.73 (s, 3H each, CH3). 13C NMR both isomers:
δ 145.47, 145.44, 145.39, 145.32, 145.23, 145.00, 144.82, 141.05,
140.92, 140.84, 140.74, 140.52, 140.45, 140.27, 137.72, 137.62,
137.59, 137.46, 137.13, 137.08, 136.95, 136.88, 136.79, 128.75,
128.16, 127.99, 127.87, 127.83, 127.71, 127.63, 127.43, 127.21,
127.07, 126.76, 124.25, 124.21, 121.32, 121.28, 121.16, 121.07,
117.66, 117.31, 116.57, 116.46, 115.77, 115.32, 115.08, 114.85,
114.77, 114.65, 81.92, 81.69, 81.62, 81.50, 81.39, 81.24, 35.98,
35.91, 35.82, 35.79, 35.74, 34.66, 33.84, 33.77, 33.03, 32.89, 31.96,
31.91, 31.65, 31.60, 31.41, 31.26, 31.21, 31.01, 16.58, 16.45, 16.40,
16.28, 14.22, 14.11, 12.86.
Isofuran Adduct 10. Tetrazine13 9 (51.3 mg, 0.217 mmol) was
added to a cooled (0 °C) solution of the bis adduct6 6 (103.4 mg,
0.217 mmol) in dry THF (50 mL) under N2 and was stirred for 3
h. The solvent was evaporated, and the residue was chromato-
graphed on alumina using hexanes/benzene (1:1) as eluant and gave
28.3 mg (29%) of the product 10 as a red-orange solid, mp (dec)
1
182-183 °C; H NMR δ 8.0183 and 8.0177 (d each, J ) 1.7 Hz,
2H, H-11,12), 6.836 and 6.830 (d each, J ) 1.4 Hz, 2H, H-1,10),
6.675 and 6.569 (dd each, J ) 5.6, 1.8 Hz, 2H, H-5,6), 6.409 and
6.381 (d each, J ) 1.4 Hz, 2H, H-3,8), 5.737 and 5.701 (dd each,
J ) 1.8, 1.4 Hz, 2H, H-4,7), 1.241 and 1.239 (s each, 9H, C(CH3)3),
0.37 and 0.14 (s each, 3H, CH3); 13C NMR δ 144.56, 144.28, 137.01
and 136.99 (C-11,12), 136.69 and 135.45 (C-5,6), 132.12, 131.67,
131.10, 131.04, 128.75, 128.60, 120.72, 120.61, 116.79 and 116.74
(C-1,10), 113.76 and 113.71 (C-3,8), 78.83 and 78.81 (C-4,7), 42.37,
41.64, 34.67, 34.65, 29.56 (C(CH3)3), 22.04, 19.76; CI MS m/z,
451 (MH+); HRMS. Calcd for C32H34O2: 450.2559. Found:
450.2559.
Adduct 13. Tetrazine13 9 (18.3 mg, 0.078 mmol) was added to
a solution of the bis-adduct 12 (56 mg, 0.071 mmol) in dry THF
(20 mL) under N2 and was stirred at 20 °C for 1 h. The solvent
was evaporated, and the residual was filtered through alumina using
hexanes/benzene (1:1) as eluant and gave 51 mg (93%) of the
product 13 (suitable for use in the next step) as a reddish-brown
solid as a mixture of unequal amounts of two isomers (I and II in
Bis-Adduct 12. NaNH2 (400 mg, 10 mmol) and KOtBu (2 mg)
were added to a solution of the isofuran adduct 10 (45.9 mg, 0.102
mmol) and the bromide12 11 (56.1 mg, 0.133 mmol) in dry THF
J. Org. Chem, Vol. 71, No. 1, 2006 335

Mitchell et al.
a ratio of about 3:2), mp (dec) 190 °C, LSIMS m/z 767.5 (MH+);
HRMS. Calcd for C56H62O2 (M+): 766.4750. Found: 766.4737.
Isomer I showed 1H NMR (300 MHz) δ 8.54 (s, 1H), 8.43 (s, 1H),
8.40 (s, 1H), 8.36 (s, 1H), 8.324 (s, 1H), 8.316 (s, 1H), 7.98 (s,
1H), 7.97 (s, 1H), 7.07 (s, 1H), 7.05 (s, 1H), 7.00 (d, J ) 1.6 Hz,
1H), 6.84 (d, J ) 1.6 Hz, 1H), 6.81 (d, J ) 1.6 Hz, 1H), 6.76 (d,
J ) 1.6 Hz, 1H), 1.69, 1.66, 1.35, 1.28 (s, 9H each), 0.45, -0.54,
-3.68, -4.02 (s, 3H each); Isomer II δ 8.58 (s, 1H), 8.48 (s, 1H),
8.40 (2H), 8.33 (2H), 7.96-7.99 (2H), 7.09 (s, 1H), 7.06 (s, 1H),
6.95 (1H), 6.88 (1H), 6.83 (d, 1H), 6.81 (1H), 1.69, 1.68, 1.33,
1.31 (s, 9H each), 0.45, -0.52, -3.72, -4.04 (s, 3H each).
NOTE: isomer II decomposed on chromatography much faster than
isomer I, and hence peaks for each isomer were assigned after much
of isomer II had decomposed.
an NMR tube, which was then chilled with running cold water while
being irradiated by a 500-W tungsten household lamp (available
in hardware stores to light a room during painting), with a ∼550-
nm cutoff filter between the lamp and the sample. After about 3 h
of irradiation, NMR monitoring indicated opening was mostly
complete and that the sample solution was almost colorless. The
1H NMR spectrum (300 MHz) of 3-o,o,o (as mixed isomers 3a,b,c)
then showed δ 8.04 (s), 7.98 (s), 7.96 (s), 7.19-7.00 (m), 6.96 (s),
6.86 (s), 6.49 (s), 1.42 (s), 1.31 (s), 1.30(s) as major peaks. When
this sample was irradiated with 350-nm light (TLC lamp), the
original spectrum was restored after about 1 h irradiation.
(b) Transient Studies. Samples for the laser flash photolysis
experiments were prepared with spectrograde solvents. All experi-
ments were performed at 20 °C. The concentrations were such that
the absorptions at the excitation wavelengths were between 0.1 and
0.7 (l ) 7 mm). Samples were placed in 7 × 7 mm2 Suprasil cells.
Unless otherwise stated, samples were purged with N2 for at least
15 min.
Trisdihydropyrene-dioxide 7. NaNH2 (400 mg, 10 mmol) and
t-BuOK (2 mg) were added to a solution of the adduct 13 (as soon
as possible after it was prepared above) (20 mg, 0.026 mmol) and
the bromide12 11 (18 mg, 0.043 mmol) in dry THF (15 mL) under
N2, and the mixture was stirred at ∼20 °C for 5 h and then was
filtered through a short alumina column (2 cm). For preparative
purposes, the solvent was evaporated and the residual was used
directly in the next reaction step below. The product 7 had poor
solubility and was not very stable in solution, and so purification
tried at this step was not very successful. The crude product which
The laser flash photolysis system previously described18 was
modified for measurements in the millisecond time range (see
Supporting Information for details). Samples were excited using
either a Nd:YAG laser at 266 (e20 mJ/pulse) or 532 nm (e50
mJ/pulse) or by using a tunable Coherent Infinity/OPO system for
excitation at variable wavelengths. Transient decays were recorded
at fixed wavelengths using a Xe-arc lamp. The signal from the
photomultiplier tube was fed into an oscilloscope terminated
externally at 1 kW. Decay traces were measured for samples either
in static or flow cells, whereas transient spectra were collected using
flow cells. The use of flow cells ensures that a fresh sample is
irradiated at each laser pulse. This is important when transients
are formed that have lifetimes longer than seconds and do not decay
back to the precursor between two subsequent laser shots. Static
cells were employed for lifetime measurements to avoid the
diffusion of the transient out of the monitoring volume. When static
cells were used, frequent measurements of the ground-state absorp-
tion of the sample ensured that the conversion of the precursor was
low. The diffusion rate of the transient out of the excitation volume
into the bulk solution determines the longest transient kinetics that
can be measured. The diffusion time varies slightly for different
experiments. For this reason, the diffusion time was measured by
following the absorption of transients known to have known
lifetimes longer than 10 s. For excitation at 266 nm, the formation
of quinone methide from o-hydroxybenzhydrol in acetonitrile:water
9:1 (t > 20 s)19 was used to determine the diffusion time. The
diffusion time for excitation at 532 nm was determined by the
irradiation of benzo-DHP 2-c. The cyclophanediene isomer, 2-o,
formed from this compound has a half-life longer than 5 h at room
temperature.6
1
contained 6 isomers gave H NMR (300 MHz) with peaks in the
following regions: δ 8.86-8.44, 8.08-7.87 (ring protons), 1.87-
1.76 (t-Bu on the middle ring), 1.64-1.55 (t-Bu on the side rings),
-3.36, -3.37, -3.39, -3.50, -4.54, -4.57, -4.58, -4.62, -5.25,
-5.28, -5.30 (internal methyls).
This product could also be obtained by reaction of isofuran12
8
(3 equiv) with dibromide 5 (1 equiv) and NaNH2/t-BuOK (10 equiv)
at 55 °C in THF; however, the product was much more difficult to
purify than using the sequence described above.
Switch 3. Trisdihydropyrene 7 (crude from above) (21 mg, 0.019
mmol) and Fe2(CO)9 (10 mg, 0.027 mmol) were refluxed in benzene
under argon for 1 h, and then the mixture was cooled to ∼20 °C
when the solvent was evaporated. The residue was chromatographed
on deactivated SiGel using hexanes/benzene (6:1) as eluant and
gave 9 mg (44%) of the product 3-c,o,c as a mixture of three
diastereomers. Recrystallization from toluene gave dark-red crystals,
mp (dec) 176-181 °C, which had increased amount of one of the
1
symmetrical isomers, 3a or 3b, which gave H NMR δ 9.00 (s,
4H, H-9,13,22,26), 8.37 (br. s, 4H, H-1,8,14,21), 7.40 (d, J ) 0.9
Hz, 4H, H-3,6,16,19), 7.17 (s, 8H, H-4,5,17,18,10,12,23,25), 1.53
(s, 36H, C(CH3)3 from DHPs), 1.36, 1.35 (s, 36H total, C(CH3)3
from cyclophane), 1.28 (br. s, 6H, CH3 from cyclophane), -1.360
(see text) (s, 12H, CH3 from DHPs). The spectra of the other
isomers were determined by subtraction: δ 9.142 and 9.136 (s,
4H, H-9,13,22,26), 8.55 (br. s, 4H, H-1,8,14,21), 7.52 (br. s, 4H,
H-3,6,16,19), 7.31 (s, 4H, H-4,5,17,18), 7.17 (s, 4H, H-10,12,23,-
25), 1.56 (s, 36H, C(CH3)3 from DHPs), 1.36, 1.35 (s, 36H total,
C(CH3)3 from cyclophane), 1.28 (br. s, 6H, CH3 from cyclophane),
-1.364 and -1.622 (see text, isomer 3c), and -1.626 (s, 12H total,
CH3 from DHPs). Mixed isomers: 13C NMR δ 150.28, 144.74,
144.50, 141.93, 141.49, 140.10, 139.95, 138.63, 138.48, 137.93,
134.92, 134.48, 129.46, 129.25, 129.07, 124.38, 123.42, 121.25,
120.99, 119.87, 117.35, 116.77, 35.62, 35.56, 35.49, 34.99, 34.37,
31.53, 30.91, 30.76, 30.46, 29.71, 18.84, 17.96, 17.46; UV
(chloroform) λmax (ꢀmax) nm 287 (53 400), 310 (47 700), 394
(43 600), 413 (74 200), 513 (9 600), 628 (1 100); EI MS m/z, 1078
(MH+); HRMS. Calcd for C82H93 (MH+): 1077.7277. Found:
1077.7256. Anal. Calcd for C82H92: C, 91.39; H, 8.61. Found: C,
91.13; H, 8.87.
Acknowledgment. We thank the Natural Sciences and
Engineering Research Council of Canada and the University
of Victoria for support of this work.
Supporting Information Available: Synthesis of compounds
in Scheme 5, expansion of Figures 2a and 2b, sequential UV-vis
spectra in photo-opening of 18-c,o,c and sequential UV closing of
18-o,o,o, 1H NMR spectra (internal methyl region) of 18-c,o,c and
in UV closing of 18-o,o,o, relative rate data for visible light opening
of 3-c,o,c and 18-c,o,c with respect to 2-c; changes to the lasher
flash system, transient kinetics at 410 nm for irradiation at 266 nm
of 20, transient kinetics at different monitoring wavelengths for
14, Transient kinetics at 490 nm for irradiation of 3 at 532 nm
(PDF). This material is available free of charge via the Internet at
http://pubs.acs.org.
The syntheses of compounds 14-19 (Scheme 5) are given in
the Supporting Information.
Photochemical Experiments. (a) Continuous Irradiation:
Opening of 3-c,o,c to 3-o,o,o. A solution of dihydropyrene 3-c,o,c
(5 mg) in argon-degassed CDCl3 (which had been filtered through
basic alumina before use) was again filtered through alumina into
JO052153G
(18) Liao, Y.; Bohne, C. J. Phys. Chem. 1996, 100, 734-743.
(19) Diao, L.; Yang, C.; Wan, P. J. Am. Chem. Soc. 1995, 117, 5369-
5370.
336 J. Org. Chem., Vol. 71, No. 1, 2006