Synthesis and photochromic properties of molecules containing [e]-annelated dihydropyrenes. Two and three way π-switches based on the dimethyldihydropyrene-metacyclophanediene valence isomerization

Article abstract of DOI:10.1021/ja0288136

The syntheses of several new simple negative, a simple positive, and multiple negative photochromes containing the dihydropyrene-cyclophanediene photochromic system are described. The photo-openings of the negative photochromes, the [e]-annelated benzo (7

Full text of DOI:10.1021/ja0288136

Published on Web 02/19/2003
Synthesis and Photochromic Properties of Molecules
Containing [e]-Annelated Dihydropyrenes.
Two and Three Way π-Switches Based on the
Dimethyldihydropyrene-Metacyclophanediene
Valence Isomerization
Reginald H. Mitchell,*,† Timothy R. Ward,† Yongsheng Chen,† Yunxia Wang,†
S. Ananda Weerawarna,† Peter W. Dibble,‡ Michael J. Marsella,§
Adah Almutairi,§ and Zhi-Qiang Wang§
Contribution from the Department of Chemistry, UniVersity of Victoria,
P.O. Box 3065, Victoria, British Columbia, Canada V8W 3V6,
Department of Chemistry, UniVersity of Lethbridge, Alberta, Canada T1K 3M4, and
Department of Chemistry, UniVersity of California, RiVerside, California 92521-0403
Received October 4, 2002 ; E-mail: regmitch@uvic.ca
Abstract: The syntheses of several new simple negative, a simple positive, and multiple negative
photochromes containing the dihydropyrene-cyclophanediene photochromic system are described. The
photo-openings of the negative photochromes, the [e]-annelated benzo (7), naphtho (9), anthro (11), furano
(19), and triphenyleno (15) derivatives of the parent 2,7-di-tert-butyl-trans-10b,10c-dimethyl-dihydropyrene
(5), as well as its 4,5-dibromo derivative (13), are described to give the corresponding cyclophanedienes,
as well as their photoclosures and thermal closures back to the dihydropyrenes. These are compared to
the results obtained for the positive photochrome dibenzo[e,l]dihydropyrene (21) and to the bis-
(dihydropyreno)chrysene (44) and the (dihydropyrenobenzo)(benzo)metacyclophanediene (47) photo-
chromes, which have more than one photochromic switch present and thus have more than a simple “on-
off” state. Thermodynamic data are obtained for the thermal closing reactions. The anthrodihydropyrene
(12) has the fastest thermal closing (τ1/2 ) 20 min), while the furanodihydropyrene (19′) has the slowest
(τ1/2 ) 63 h) at 46 °C. An electrochemical readout of the state of the switch is demonstrated for the
benzodihydropyrene (7).
Introduction
to the benzene rings and thus only shows benzenoid absorption
with an extended tail.5 The effective size of the π-system thus
A recent issue of Chemical ReViews is devoted to the topic
“Photochromism: Memories and Switches”1 and indicates the
considerable current interest in such molecules, which include
diarylethenes, fulgides, and spiropyrans. Much less investigation
of dihydropyrenes has been made, even though they are also
photochromic2 and are thus potential photoswitches3 which may
have some future application. The dihydropyrene (DHP)-
cyclophanediene (CPD) system is however especially interesting
because it is an example of the less common negative photo-
chromes,4 in which the thermally stable form is colored and
which on irradiation becomes colorless.
changes from 14π in 1 to 6π in 2. Irradiation of the parent green
dihydropyrene form 1 with visible light from an ordinary
tungsten household lightbulb partially converts it to the colorless
cyclophanediene form 2. Unfortunately, the reverse reaction
occurs not only photochemically with UV light but also
thermally. This is surprising, since the photochemical processes
involve allowed conrotatory opening and closing of the central
bond and are examples of the more general cis-stilbene
dihydrophenanthrene interconversion.6 The thermal back reac-
The green dimethyldihydropyrene 1 has a rigidly planar
extended π-system and has absorptions out to 600 nm. The
colorless cyclophanediene form 2 is a stepped molecule which
is also rigid but has the olefinic bridges almost at right angles
† University of Victoria.
‡ University of Lethbridge.
§ University of California.
(5) Blattmann, H. R.; Meuche, D.; Heilbronner, E.; Molyneux, R. J.; Boekel-
heide, V. J. Am. Chem. Soc. 1965, 87, 130-131. Blattmann, H. R.; Schmidt,
W. Tetrahedron 1970, 26, 5885-5899.
(6) Muszkat, K. A.; Gegiou, D.; Fischer, E. Chem. Commun. 1965, 447-448.
Muszkat, K. A.; Eisenstein, M.; Fischer, E.; Wagner, A.; Ittah, Y.; Luttke,
W. J. Am. Chem. Soc. 1997, 119, 9351-9360.
(1) Irie, M. (Guest Editor) Chem. ReV. 2000, 100, 1683-1890.
(2) Mitchell, R. H. Eur. J. Org. Chem. 1999, 2695-2703.
(3) Marsella, M. J.; Wang, Z. Q.; Mitchell, R. H. Org. Lett. 2000, 2, 2979-
2982.
(4) Bouas-Laurent, H.; Du¨rr, H. Pure Appl. Chem. 2001, 73, 639-665.
9
2974
J. AM. CHEM. SOC. 2003, 125, 2974-2988
10.1021/ja0288136 CCC: $25.00 © 2003 American Chemical Society

Dihydropyrene−Cyclophanediene System
A R T I C L E S
tion of 2 to 1 on the other hand, since it gives the same product
as the photochemical reaction, must break Woodward-Hoff-
mann rules if it is a concerted process.7 The thermal reaction
has an Eact of about 23 kcal/mol, and so to achieve high
conversions of 1 to 2, the reaction mixture first needs to be
cooled to <0°C. Second, the quantum yield of the back reaction,
2 to 1, is approximately 50 times greater than that for the
forward reaction, and so for the best conversion of 1 to 2,
irradiation into the tail of 2 must be avoided; thus, a filter is
required to restrict the wavelength of the light used to >500
nm.2
When benzenoid systems are fused to dihydropyrenes, the
extended π-system continues to be approximately planar, and
the principle absorptions are red shifted. In their open CPD
forms however, the staircase geometry interrupts the π-system,
and benzenoid absorption persists. Benzenoid annelation of
dihydropyrenes thus modifies the properties of the switch but
does not change its nature.8
This paper examines the changes in photochromic properties
that occur when various benzenoid systems are annelated onto
1 and the properties of switches with three states which are
formed when two dihydropyrene systems are fused onto a
benzenoid.9
Figure 1. Partial 1H NMR spectrum of CPD 8 (top) after irradiation (see
text) of the DHP 7 (bottom). The tert-butyl signals are the larger signals in
each spectrum.
starting material, the 2,7-di-tert-butyl derivative of 1 () 5) (the
Supporting Information gives a large scale preparation of 5)
has led11b to a variety of [e]-annelated dihydropyrenes suitable
for photochemical study.
Results and Discussion
The disadvantages of the parent system mentioned previously
are much reduced in the benzo[e]annelated system 3 to 4. First,
Eact for the thermal return reaction 4 to 3 is higher by 2 kcal/
mol (25 kcal/mol), and so, the half-life of 4 is extended to about
a week at room temperature.10
Second the quantum yield for the conversion of 3 to 4 appears
to be much better than that for 1 to 2. Simply holding a solution
of DHP 3 in front of an ordinary tungsten bulb causes complete
bleaching to the colorless CPD 4. As noted previously, during
the photochromic conversion, there is a large change in the
effective π-system: Red 3 has 18 interacting p-orbitals, while
colorless 4 has only isolated 6π-systems, because of the
orthogonality of the orbitals in the CPD form.
Simple Arene [e]-Annelated Derivatives. Although we first
reported 3 some two decades ago,10 its lengthy multistep
synthesis from benzene precursors prevented a full exploration
of its properties. Discovery of a route11a to a series of
[a]-annelated dihydropyrenes from a common dihydropyrene
“aryne” reactive intermediate opened up syntheses in this area.
Coupling this discovery with using the more easily synthesized
Photochemistry of 5-12. The reported12 quantum yield for
the conversion of the di-t-Bu substituted 5 to 6 was 0.002, an
order of magnitude smaller than that for conversion of the
unsubstituted parent 1 to 2, so we were unsure as to whether 7
would be as good at switching as was 3. In the event, 7 is readily
bleached with visible light. Placing a reddish-purple solution
of 7 in cyclohexane on the surface of an ordinary overhead
transparency projector bleaches it in a couple of minutes, at a
rate not noticeably different from that of the parent 3. The
conversion of colored 7 to colorless 8 was essentially quantita-
tive by NMR. When an NMR sample was held about 8 cm
from a 250 W tungsten lamp with ice cooling (Pyrex glass), no
peaks of the starting isomer 7 were observable after a few
minutes of irradiation. This can be seen clearly in Figure 1.
This is important, because many photochromic systems give
photo stationary states and complete conversion from one isomer
to the other is not possible.1 It is not possible to achieve this on
the parent 5, unless a 500 nm cutoff filter is used with ice
(7) Schmidt, W. HelV. Chim. Acta 1971, 54, 862-868. Schmidt, W. Tetrahe-
dron Lett. 1972, 13, 581-584.
(8) Mitchell, R. H.; Iyer, V. S.; Mahadevan, R.; Venugopalan, S.; Zhou, P. J.
Org. Chem. 1996, 61, 5116-5120.
(9) Mitchell, R. H.; Ward, T. R.; Wang, Y.; Dibble, P. W. J. Am. Chem. Soc.
1999, 121, 2601-2602.
(10) Mitchell, R. H.; Yan, J. S. H.; Dingle, T. W. J. Am. Chem. Soc. 1982, 104,
2551-2559.
(11) (a) 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. (b) Mitchell, R. H.; Ward, T. R. Tetrahedron, 2001, 57, 3689-3695.
(12) Murakam, S.; Tsutsui, T.; Saito, S.; Yamato, T.; Tashiro, M. Nippon Kagaku
Kaishi 1988, 221-229.
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 2975

A R T I C L E S
Mitchell et al.
length has to undergo considerable change. For the annelated
derivatives, this change would be smaller, because the annelated
rings cause partial bond fixation in each other. This might
reduce Eact for the thermal reactions of 8/10/12 relative to 6.
Another factor is strain. In the planar annelated DHP derivatives
7/9/11, there are “phenanthrene bay type” steric interactions
between the periprotons of the annelating ring and the protons
of the DHP ring (protons 1/12 and 8/9 in 7 for example). These
are absent in the twisted CPD isomers. Thus the thermal reaction
8 f 7 has to overcome this strain, while the parent 6 f 5 does
not. This should raise Eact for the former relative to the latter.
The summation of these effects, the stabilization of the
“intermediate”, the change in bridge bond length, and the change
in strain may then produce the observed order, in which the
benzo system 8 to 7 has the greatest value for Eact (slowest
thermal return).
Figure 2. UV-vis spectrum of CPD 8 and DHP 7 in cyclohexane.
Table 1. Half-Lives (τ1/2, h, 46 °C), Energies of Activation (Eact
,
kcal/mol), and AM1 Calculated Differences in Heats of Formation
(∆Hf(DHP-CPD), kcal/mol) for the Thermal Reactions Shown
Two additional data points were easily accessible making use
of the dibromide pair 13/14 and the triphenylene annelated pair
15/16.
reaction
τ1/2
Eact
∆H
f
6 f 5
8 f 7
10 f 9
12 f 11
1.88
5.75
1.15
0.33
23.0
24.5
22.1
19.1
2.27
1.47
1.12
0.96
cooling. Holding a hand-held UV chromatography lamp in front
of the NMR tube containing bleached sample 8 quickly and
quantitatively converted it back to 7.
The corresponding UV-visible spectra using cyclohexane as
solvent are shown in Figure 2. The long wavelength absorptions
of purple 7 are removed on visible irradiation, as colorless 8
forms. Almost identical spectra are obtained in the solid state,
if the DHP 7 is dissolved in polystyrene-dichloromethane (20:
80) (1% 7 by weight of the polystyrene) and spin coated onto
a film on glass. Photoisomerization also takes place in the crystal
state. Small crystals bleach in color on exposure to visible light
and return to purple on exposure to UV light.
The naphtho analogue 9 bleached to 10 even faster than did
7 to 8; however, bleaching of the anthro analogue 11 was slower
than that of 7. The major differences between these three
analogues was in the thermal return reactions.
Thermal Return Reactions. Previously, we have observed8
that when ∆Hf(DHP-CPD) (AM1) is large, the thermal return
of the CPD to the DHP form is fast. For the [e]-annelated
compounds here however, the ∆Hf values are all small. The
thermal return of 12 to 11 is approximately 3 to 4 times faster
than that of 10 to 9, which is 4 times faster than that of 8 to 7,
which is 3 times slower than the parent 6 to 5. Table 1 shows
the relevant data.
Thus, for a practical application, although under the same
irradiation conditions, the naphtho derivative 9 photochemically
opens and closes faster than the benzo derivative 7, a slow
thermal return reaction (CPD to DHP) is important, and so the
benzannelated derivative 8/7 is the best.
Why the thermal return of 8 to 7 is slowest is not immediately
obvious. Since this reaction is “Woodward-Hoffmann forbid-
den”, perhaps the transition state has radical or ionic character,
which the anthro and naphtho rings would certainly stabilize
better, lowering Eact. This would not explain why Eact for 6 to
5 is not higher than that for 8 to 7. A number of factors may
contribute, one of which may be the change in length of the
bridge bond during the photoreaction. In the cyclophanediene
6, the bridge bond is a pure double bond, 1.34 D. In the annulene
5, all bonds are equal at 1.40 D. During the transition, this bond
Bromination of 7 with NBS/DMF/CH2Cl213 at 20 °C gave a
48% yield of dibromide 13. The τ1/2 value found at 46 °C for
the thermal return reaction 14 to 13 was 9.50 h, almost double
that of 8 to 7, with Eact increasing to 25.8 kcal/mol. Indeed,
PCMODEL14 calculations indicate an additional 1 kcal/mol
difference in strain energies between the DHP pair (7/13) and
the CPD pair (8/14). In the dibromide, the DHP ring hydrogens
sterically interact with the bromine atoms in the planar DHP
form more than they do in the twisted CPD form (see 2). This
is suggestive that increasing strain in the DHP form relative to
the CPD form slows the thermal reaction. However, Lai15 has
shown that too much strain destabilizes the DHP form too much
and then prevents full equilibration between the photoisomers;
for example, 17 and 18 will not fully interconvert. Bromine
may also introduce an electronic effect.
Reaction of the furan 1911b with 9-bromophenanthrene and
NaNH2/t-BuOK in THF gave an 85% yield of adduct 20, which
on deoxygenation with Fe2(CO)9 in refluxing benzene gave an
86% yield of the triphenylene 15, as a stable purple compound,
(13) Mitchell, R. H.; Lai, Y. H.; Williams, R. V. J. Org. Chem. 1979, 44,
4733-4735. Mitchell, R. H.; Chen, Y.; Zhang, J. Org. Prep. Proced. Int.
1997, 29, 715-719.
(14) PCMODEL, version 7.0; Serena Software: Bloomington, IN 47402-3076.
(15) Lai, Y. H.; Chen, P.; Peck, T. G. Pure Appl. Chem. 1993, 65, 81-87.
9
2976 J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003

Dihydropyrene−Cyclophanediene System
A R T I C L E S
mp 188-193 °C (dec). The τ1/2 value found at 46 °C for 16 to
15 was 3.6 h (Eact ) 23.4 kcal/mol), 1.6 times faster than that
for 8 to 7.
o-dibromobenzene failed to provide any cyclophane 21 (RdH).
Bis-alkylation of the dianion of the sulfone 24 with either o-bis-
(bromomethyl)benzene or o-phthalaldehyde, followed by SO2
expulsion, was also fruitless.
Calculations14 suggest there is little change in the steric
interactions (<0.2 kcal/mol) between the DHP ring hydrogens
with the adjacent aromatic hydrogens on going from 7 to 15.
However, a charged or radical intermediate would be expected
to be additionally stabilized, and thus, the thermal reaction of
16 to 15 might be expected to be faster than that of 8 to 7.
Both pairs, 13/14 and 15/16, photochemically opened and
closed easily; however, of the simple [e]-annelated dihydropy-
renes, the benzo compound 7 appears to be the best photoswitch
so far, offering a reasonable compromise among ease of
synthesis, efficiency or ease of the photoconversions, and
relatively slow thermal return reaction.
In principle, the corresponding arene[a]annelated derivatives
should also photoswitch. Laser flash studies, similar to that of
the parent,16 suggest they do, but the thermal return reactions
are extremely fast, which does not make them very convenient
to study. AM1 calculations show they have much larger values
of ∆Hf(DHP-CPD) with the DHP being the more stable
isomer.8
Diarene Annelated Derivatives. The situation for the
diarene annelated derivative 21/22 is quite different from that
of either the [a]- or the [e]-annelated derivatives. In this case,
the CPD 21 is calculated to be the thermally stable isomer,
∆Hf(CPD-DHP) (AM1)17 ) 18.4 kcal/mol (18.2 kcal/mol when
RdH), and so, we would expect the DHP form 22 to revert
thermally quite rapidly to CPD 21 when the latter is irradiated
with UV light.
Our first19 successful synthesis of the o-m-o-m-cyclophane
system of 21 utilized the dialdehyde 25.20 McMurry coupling21
of 25 using Zn/Cu and TiCl3 in refluxing dimethoxyethane
yielded 75-85% yields of the pinacol 26, mp 215-218 °C.
Swern oxidation22 of diol 26 using oxalyl chloride-DMSO gave
75-80% yields of the hydroxyketone 27, which on repeat
treatment gave 75-90% yields of yellow diketone 28, mp 186-
187 °C. Direct oxidation of 26 to 28 gave a lower yield (20%).
The internal methyl proton chemical shifts for the three
cyclophanes, 26 (δ 0.61), 27 (δ 0.56 and 0.84), and 28 (δ 0.89),
were all shielded as expected. Attempted Wittig reaction on the
diketone 28, or reaction with vinyl or ethynyl Grignard reagents,
caused cleavage of the cyclophane ring, as evidenced by the
loss of these upfield methyl protons. However, a derivative of
our target 21 could be obtained by reaction of 28 with
o-phenylenediamine which gave 55-60% of cyclophane 29, as
yellow crystals, mp 211-213 °C. These showed the character-
istic upfield cyclophane internal methyl protons at δ 1.20 and
gave a correct mass spectrum molecular ion at m/z 384 and
elemental analysis. Likewise, from dialdehyde 30,19 the dione
31 gave the cyclophane 32.
Irradiation of benzene or THF solutions of 29 or 32 at 254
nm rapidly produces unstable reddish-purple solutions. We
believe these contain the dihydropyrenes 33, just as we observed
for the monoannelated cyclophanedienes 8 and 10. However,
such symmetrically annelated annulenes have the possibility of
biradicaloid structures contributing to the resonance hybrid (see
structure 33) which reduces the bond fixation in the strongly
aromatic annelating arenes. This has been demonstrated in [a,h]-
fused dihydropyrenes.23 Such radical contributors prevent an
unchallengeable structure proof from being obtained.19 Unam-
Syntheses. Despite considerable effort, synthesis of the
dibenzannulene 22 was not achieved until almost 20 years after
our synthesis of 3!18 Coupling the bis-Grignard reagent 23 with
(18) (a) Mitchell, R. H.; Yan, J. S. Can. J. Chem. 1977, 55, 3347-3348. (b)
Mitchell, R. H.; Chen, Y. Tetrahedron. Lett. 1996, 37, 5239-5242.
(19) Mitchell, R. H.; Weerawarna, S. A. Tetrahedron. Lett. 1986, 27, 453-456.
(20) Mitchell, R. H.; Yan, J. S. H. Can. J. Chem. 1980, 58, 2584-2587.
(21) McMurry, J. E.; Kees, K. L. J. Org. Chem. 1977, 42, 2655-2656.
(22) Mancuso, A. J.; Huang, S. L.; Swern, D. J. Org. Chem. 1978, 43,
2480-2482.
(16) Murphy, R. S.; Chen, Y. S.; Ward, T. R.; Mitchell, R. H.; Bohne, C. Chem.
Commun. 1999, 2097-2098.
(17) Spartan 02 for Windows, version 1.0; Wavefunction, Inc.: Irvine, CA 92612.
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 2977

A R T I C L E S
Mitchell et al.
biguous formation of [e]-position dibenzannelated dihydropy-
renes had to await the discovery of the “aryne route”.24
removed from the bath, and placed in front of a 300 nm source
for one minute, the sample turned green. As the sample warmed,
the green color was lost. This process could be repeated many
times. If the cold green sample was wiped free of frost and
quickly transferred to an NMR spectrometer at -50 °C, the
spectrum of the green dihydropyrene 22 could easily be seen,
with the internal methyl protons highly shielded at δ -3.41.
Integration suggested a 35:65 ratio of 22/21; warming the
solution resulted in complete reversion to cyclophane 21. The
thermal return reaction, 22 to 21, is essentially stopped, so long
as the green solution is held at -50 °C or below. Irradiation of
this green solution with visible light from an ordinary tungsten
lamp quickly (<1 min) converts 22 to 21. The photoreaction
thus does proceed in both directions.
Thermal Reaction of 22 to 21. As we anticipated, thermal
conversion of 22 to 21 is fast at room temperature. Using NMR
integrations, Eact for the thermal reaction was estimated (see
later) at 20 kcal/mol, giving DHP 22 a half-life of 400 min at
-40 °C but only 1-2 s at 20 °C.
We thought that this thermal reaction might be slowed in
the solid state. Evaporation of a dichloromethane solution of
polystyrene mixed with CPD 21 on a glass plate gave a film,
which when irradiated with 300 nm light at room temperature
became green. When the plate was removed from the light, the
green color faded after 2-3 s. The process could be repeated
at least 15 times, even after storing the sample for a year. The
photoisomerization was also observed when silica gel or alumina
was used instead of polystyrene; however, the sample suffers
some decomposition on these supports. Clearly the thermal
reaction is slowed somewhat in the solid state.
Initially18b we utilized dibromide 34 with NaNH2/t-BuOK in
furan-THF to yield the adducts 35a,b (1:1 mixture of isomers)
in 60% yield. Subsequently, we found that use of the tetra-
bromide 36 with n-BuLi as base in furan/THF improved the
yield of 35 to 85%.
Dinaphtho Annelated Derivative 38. Reaction of the dibro-
mide 34 with isobenzofuran (generated in situ from 1-methoxy-
phthalan) and NaNH2 gave 33% of the mixed isomers of 37
which when deoxygenated with Fe2(CO)9 as previously yielded
95% of colorless cyclophane 38, mp 340-342 °C. Its internal
methyl protons appeared at δ 1.06, similar to those of 21 (δ
1.07). However, irradiation of CPD 38 with 300 nm light at
-90 °C produced no colored species. We suspect, based on
the behavior of 10 relative to 8, that the thermal reaction is just
too fast to see any buildup of the DHP derived from CPD 38.
The anti isomer 35a, with the two oxygens anti to each other,
was less soluble, mp 236-238 °C, and showed only one internal
methyl proton peak at δ -3.99, while the more soluble syn
isomer 35b showed two peaks at δ -3.70 and -4.30. Deoxy-
genation of the mixed isomers 35 with Fe2(CO)9 in refluxing
benzene yielded 68% of the colorless cyclophane 21, mp 231-
232 °C. The internal methyl protons of 21 appeared at δ 1.07,
and the UV spectrum was typical of that of a cyclophane rather
than that of the highly colored dihydropyrenes.11
Photochemistry of 21. Irradiation of 21 with UV light (254
or 300 nm) at room temperature produced no apparent change;
to prove that the valence isomerization does take place but the
thermal return reaction is rapid, the irradiation was carried out
at low temperature. Indeed, when a sample of 21 in d8-THF in
an NMR tube was cooled to about -90 °C (pentane-N2(l) bath),
Synthesis of the elusive dibenzo annelated pair 21/22 has
finally been achieved. Photoswitching between 21 and 22 is
possible, but the very fast thermal return reaction has to be taken
into account in any switching applications.
Molecules Containing More Than One Switch. Because
of the stepped nature of the CPDs (see 2), the π-orbitals of the
benzenoid rings in 8 are prevented from overlapping to make a
contiguous π-system, as is present in the DHP 7. Thus, when 7
switches to 8, an 18π-system switches to what is essentially a
6π-system. This is a rather dramatic change of the π-system
and leads to the very different properties of 8 versus 7. Multiple
(23) Mitchell, R. H.; Dingle, T. W.; West, P. R.; Williams, R. V.; Thompson,
R. C. J. Org. Chem. 1982, 47, 5210-5214.
(24) Mitchell, R. H.; Zhou, P. Tetrahedron. Lett. 1990, 31, 5277-5280.
9
2978 J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003

Dihydropyrene−Cyclophanediene System
A R T I C L E S
for dihydropyrenes, between δ -3 and -4. This mixture of
isomers was directly deoxygenated with Fe2(CO)9 as previously
to yield 82% of the bis(dihydropyreno)chrysene 44a,b. These
two isomers did not separate on TLC or column chromatogra-
phy. In CDCl3, the internal methyl protons appear together as
a somewhat broadened singlet at δ -0.73, while the t-Bu protons
appear in two sets at δ 1.57 and 1.52. In C6D6, the methyl
protons separate into two peaks at δ -0.28 and -0.29 in an
approximate 1:3 ratio, probably derived from two methyl signals
from 44a and one from 44b. The two t-Bu signals remain at δ
1.45 and 1.44. The bay protons were also unresolved slightly
broadened singlets.
The DHP-DHP 44 was intense red, with the most intense
UV-vis band at 427 nm (ꢀ ) 132 000 dm3 mol-1 cm-1),
strongly red shifted from that of 1 (338 nm).26 The DHP 44
also shows longer wavelength bands at 527 and 576 nm (ꢀ ≈
10 000) and a tail to 700 nm. Irradiation of a 5 mg sample of
44 with a light of a wavelength > 600 nm (from a 250W
tungsten filament bulb and a 600 nm cutoff filter) in an ice-
water chilled d8-toluene solution enabled the photoswitching
to be followed conveniently by NMR, Figure 4.
Conversion of DHP-DHP 44 (Figure 4, left, top) to CPD-
CPD 45 (Figure 4, left, bottom) was complete after about 75
min of irradiation. No evidence by NMR (or UV, see later) could
be found for CPD-DHP 46 (Figure 4, right) at intermediate
irradiation times. As can be seen from Figure 4, the internal
methyl protons of 44 at δ -0.22 are replaced by those of 45
(two isomers) at δ 1.71-1.75, and likewise, the tert-butyl signals
become less deshielded, moving from ∼δ 1.6 to 1.45. The other
isomer, CPD-DHP 46, could however be observed in both the
thermal and UV induced reverse reactions. It showed a new
NMR signal for the internal methyl protons at δ -0.27 (those
of 44 were at δ -0.22 in d8-toluene) and a new t-Bu signal at
δ 1.62 (those of 44 were at δ 1.66 and 1.64). Likewise, when
a cyclohexane solution of 44 was similarly irradiated, the 427
nm absorption of 44 disappeared into the baseline and only the
chrysene absorption at 313 nm remained.27 The presence of an
isosbestic point suggested also that 46 was not involved. In
cyclohexane, on UV (∼350 nm) irradiation, new peaks first
appeared at 392 and 413 nm, which then reduced and were
replaced by those of 44 at 406 and 427 nm.27
Figure 3.
state switches can in principle be obtained, by linking two or
more dihydropyrenes together via conjugated spacers as in
Figure 3.
In the closed-closed state, the π-system has 12 + spacer +
12 ) 24 + s (s ) number of spacer π-electrons) contiguous
planar p-orbitals; in the closed-open state, it has 12 + s planar
p-orbitals; and in the open-open state, s p-orbitals. In the latter
case, if s is less than 6, then the cyclophane benzene rings would
dominate. Thus, if s ) 6, that is, the spacer is a benzene ring,
the switch goes from a 30π to essentially an 18π to a 6π-system
as it changes its states. Such dramatic changes should lead to
significantly different properties.
Our first7 test of this idea utilizes chrysene as the spacer.
The main reason for this choice was synthetic. The bis-furan
39 is accessible25 easily from its precursor 40 and reacts with
2.7 equiv of bromide 41 in the presence of NaNH2/t-BuOK in
THF at room temperature to generate aryne 42, which reacts
twice in situ with the bis-furan to give a surprisingly good yield,
81%, of the adduct 43, as mainly an equal mixture of two sets
of two isomers (oxygen’s syn/anti, methyl’s syn/anti of trans).
Thus, although the closed-closed and open-open states of
the switch could be attained fairly easily, the intermediate
position, open-closed (isomer 46), could not be obtained as a
single species, only mixed with 45 early in the return reaction
or with 44 late in the reaction. Clearly, future work on switches
with DHPs at both ends needs to find a way of distinguishing
one end from the other, so that the intermediate switch position
can be more clearly attained.
Relatively few systems appear to have been reported where
several photochromes are linked, and as far as we know, none,
where a rigid conjugated spacer is used. Interestingly, in the
most closely connected analogues, the dithienylethylenes of Irie
and Branda, four papers28 report dithienylcyclopentenes linked
by conjugated but not rigid spacers, and in all of these papers,
it appears that both photochromes do not close. It is not obvious
to us why the dihydropyrenes do.
(25) Yu, D. W.; Preuss, K. E.; Cassis, P. R.; Dejikhangsar, T. D.; Dibble, P.
W. Tetrahedron Lett. 1996, 37, 8845-8848.
(26) Boekelheide, V.; Phillips, J. B. J. Am. Chem. Soc. 1967, 89, 1695-1704.
(27) See Figure 2 in ref 6.
This mixture gave a correct mass spectrum at m/z 893 for MH+
and showed the internal methyl protons at the characteristic shift
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 2979

A R T I C L E S
Mitchell et al.
1
Figure 4. Partial H NMR spectrum of DHP-DHP 44 (top left) before irradiation with visible light and after irradiation to form CPD-CPD 45 (bottom
left) and then (right) after thermal or UV closing to form some CPD-DHP 46 and then DHP-DHP 44.
A Clear Three Position Photoswitch. One method of
distinguishing one DHP from the other is to dibenzannelate one.
As we observed previously, a DHP with a benzene fused at
both [e]-positions has a thermal state that is open, that is, 21.
Having only one benzene [e]-fusion gives a thermal state that
is closed, that is, 7. Thus, compound 47 should be the thermally
stable state of the three way switch 48/47/49, which on
irradiation with UV should close to 48 or on irradiation with
visible light should open to 49. Isomers 48 and 49 should both
thermally return to 47 or by irradiation with visible or UV light,
respectively.
four distinct internal methyl proton signals at δ -4.18, -3.60
(DHP) and -1.99, and -1.04 (benzo-DHP), as well as four
distinct tert-butyl signals. Mass spectroscopy (LSIMS and exact
mass) confirmed the overall structure.
Synthesis and Structure of 47. Reaction of dibromoben-
zopyrene 13 with excess isofuran 19 in toluene at -40 °C with
n-butyllithium for just 15 min gave about 12% of a 1:1 mixture
of the adducts 50a and 50b. Despite much experimentation, this
yield could not be improved. This mixture showed the expected
1
eight internal methyl proton signals in its H NMR spectrum
Deoxygenation of the mixed isomers with Fe2(CO)9 gave
about 60% yield of two isomers of 47, namely 47a and 47b,
while deoxygenation of the chromatographed single isomer of
50 gave a single isomer of 47, in which the DHP internal methyl
protons appeared at δ -1.37 (DHP). From the spectra of the
mixed isomers, the corresponding signals for the methyl protons
of the other isomer were at δ -1.67. The CPD methyl protons
and was used in the next step directly. However, chromatog-
raphy yielded one isomer pure, mp 222-225 °C, which showed
(28) Peters, A.; Branda, N. R. AdV. Mater. Opt. Electron. 2000, 10, 245. Kawai,
T.; Sasaki, T.; Irie, M. Chem. Commun. 2001, 711-712. Osuka, A.;
Fujikane, D.; Shinmori, H.; Kobatake, S.; Irie, M. J. Org. Chem. 2001, 66,
3913-3923. Kaieda, T.; Kobatake, S.; Miyasaka, H.; Murakami, M.; Iwai,
N.; Nagat, Y.; Irie, M. J. Am. Chem. Soc. 2002, 124, 2015-2025.
9
2980 J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003

Dihydropyrene−Cyclophanediene System
A R T I C L E S
Table 2. Thermal Return Reaction (CPD to DHP) Reaction Rates
(min-1 at 46 °C), Half-Lives (τ1/2, H, 46 °C), and Energies of
Activation (Eact, kcal/mol) for the Thermal Reactions Showna
were at δ 1.18 in both isomers. The overall structure was
confirmed again by mass spectroscopy (EI and exact mass), but
in the absence of an X-ray structure, it is difficult to be certain
which isomer is which. We have previously prepared29 the
reaction
k
τ1/2
Eact
1
12 f 11
10 f 9
6 f 5
45 f 44
16 f 15
49 f 47
8 f 7
0.0344(9)
0.0101(3)
0.006 10(12)
0.0057(2)
0.003 07(7)
0.002 24(6)
0.002 00(5)
0.001 24(3)
0.000 183(6)
0.33
1.15
1.88
2.0
3.76
5.16
5.75
9.50
19.1
22.1
24.0
24
23.9
24.1
24.5
25.8
26.1
furans 51a and 51b, and by analysis of their H NMR spectra
and AM1 calculated structures, we suggested that 51b was the
less distorted and had the most shielded internal methyl DHP
protons (δ -1.66) and most deshielded aromatic proton (H-4,
δ 9.10); isomer 51a signals were at δ -1.37 and 8.95,
respectively. Since the isomer we separated had internal methyl
DHP protons at δ -1.37 and the most deshielded proton at δ
8.94 (the other isomer was δ -1.67 and 9.09), this is strongly
suggestive that the single isomer of 47 that we were able to
obtain pure was 47a.
14 f 13
19′ f 19
63.3
a Note: Since 45 to 44 has intermediate 46, the value for Eact is the
overall reaction apparent value. 19′ is the CPD form of 19. Values in
parentheses after the rates are 3σ errors, based on integration error estimates.
half-life of about 0.1 s at 20 °C. Based on an Eact of about 20
kcal/mol, the parent system 22 has a half-life of about 2 s at 20
°C. There are more steric interactions present in 48 than in 22,
and so, the former would be expected to revert thermally to 47
faster than 22 does to 21, consistent with our observations. Even
after hundreds of laser flashes, in the absence of oxygen and in
cyclohexane, the UV spectrum of 47 was returned unchanged
at the end of the experiment.
Thus, it is possible to prepare a three way switch where the
other two states (48, 49) of the switch thermally revert to the
central position (47) and where different signals (vis, UV) do
the switching.
In an attempt to improve the yield, we also tried an alternate
method of adduct formation, namely a reaction of furan 5229
with aryne 42 derived from bromide 41. However, despite
several tries, no product could be isolated. The more condensed
adducts 50 decompose slowly on standing in solution, need
protection from air and light, and are substantially less robust
than the simple adducts such as 35 which lead to the simple
benzo-derivatives. As well, the more condensed dihydropyrenes
such as 47 are also more sensitive to air and light than their
simpler counterparts, and for example, NMR solutions on
standing give spectra with broader lines.
Thermal Return of CPDs to DHPs. Because the thermal
return reactions have Eact values around 24 kcal/mol, the
1
reactions can conveniently be followed by H NMR spectros-
copy, using the sharp methyl or tert-butyl signal integrations
to estimate the amounts of CPD and DHP present. The thermal
return reactions follow first-order kinetics,4 and so for com-
parison, the reaction rates, and more conveniently for discussion,
the half-lives at 46 °C arranged in increasing value order, and
Eact values are shown in Table 2.
Not surprisingly, the three way switch 49 has similar values
as the benzo system 8 and the chrysene containing switch 45
behaves between the naphtho switch 10 and the benzo switch
8.
Because of the thermal return reaction, dihydropyrenes are
both T and P types.4 If the thermal return reaction could be
shut down, they would then be pure P-type. Clearly converting
the benzo ring of 7/8 into a different arene increases the thermal
return rate. The exception is the furan 19. In the dithienylethenes
of Irie,30 replacement of the benzene rings of cis-stibene with
less aromatic thiophene rings results in less of a loss of aromatic
stabilization energy of the closed versus the open form.
However, in the dihydropyrene 7, it is the ethene bridge of
stilbene (metacyclophanediene) that has been annelated. In our
cases, if the forbidden thermal reaction involves some inter-
mediate with partial ionic character, then the annelating aromatic
with the largest resonance energy should stabilize the charge
better, and so in our cases, we would expect the thermal rates
to be in the order anthracene > naphthalene > benzene > furan,
which indeed (Table 2) is exactly the order found. Unfortunately,
the furan 19, despite its nice slow thermal return rate, is too
unstable to use as a photoswitch. Although it opens readily and
cleanly with visible light, photoclosing with UV causes some
Photochemistry of 47. Despite these longer term stability
problems, the DHP unit in 47 easily opens on irradiation either
with broad band (household tungsten lamp) white light or with
a 598 nm cut off filter to give the colorless bis-CPD 49. The
intense red of 47 thus bleaches with visible light, the main λmax
of 49 at 268 nm,2 but can be returned on irradiation at 300 nm
or thermally (see later). The photoswitching of 47/49 can also
1
be followed by H NMR but, because of stability problems,
not in a halogenated solvent. In d8-toluene, the internal methyl
protons of the single isomer 47a were found at δ -1.01 (DHP)
and 1.64 (CPD). On photoswitching, this solution gave peaks
at δ 1.78 and 1.58 corresponding to the bis-CPD 49. Again,
this reversed with 300 nm light. There were similar changes in
the other protons; for example, the most deshielded bay protons
of 47 at δ 9.7 and 8.9 disappear on irradiation and reform
thermally, while the olefinic protons of 49 at δ 6.6 form on
irradiation and then disappear thermally.
Accessing the other photoisomer 48 was expected to be more
difficult, because of the extremely easy thermal return (see 21).
Laser flash photolysis of 47 at 355 nm (10 ns pulse) in deaerated
cyclohexane solution produced a transient with a new λmax at
486 nm, corresponding to the larger π-system of 48, while that
due to 47 at 406 nm decreased in intensity.
Thermal Reaction of 48. In the thermal reaction, the peak
at 486 nm decreased, and the peak at 406 nm reformed, with a
(29) Mitchell, R. H.; Ward, T. R.; Wang, Y. Heterocycles 2001, 54, 249-257.
(30) Irie, M. Chem. ReV. 2000, 100, 1685-1716.
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 2981

A R T I C L E S
Mitchell et al.
decomposition, and as well, it is somewhat air sensitive. It does
suggest that it would be worth investigating other heteroarenes,
for example, the thiophene fused species in the future. The fact
that steric bulk at the 4,5-positions of 7/8, as in the dibromide
13/14, also slows the thermal return reaction suggests that,
instead of switches with fused arene spacers (as in 44 and 48),
perhaps sterically bulky substituted spacers might be worth
pursuing. We shall report on such attempts in the future.
Best Switch So Far, the Benzo Switch 7/8. The best
“switch” in the dihydropyrene series is the benzo switch 7/8. It
has a reasonably slow thermal return yet relatively good photo-
opening and -closing. We need to examine in detail though how
substituents change the quantum yields and the thermal return
rate.
It would be helpful if such an investigation could be guided
by calculations. We have noted previously that AM1 type
calculations give quite reasonable results for ∆Hf calculations,
but they fail miserably however at predicting reliable bond
lengths and at predicting the S0 - S1 transition (the band gap
Eg). DFT calculations give reliable bond length data for these
annulenes31 but do not solve the band gap problem; for example,
simple estimation of Eg for benzo-DHP 7 by subtracting EHOMO
from ELUMO yields the highly overestimated values of 1117 and
1127 nm for AM1 and DFT calculations, respectively, while
the experimental value is 584 nm. We were pleased to find that
time dependent (TD-DFT) calculations32 yielded a value for 7
of 580 nm and likewise for 8 a value of 380 nm (experimental,
404 nm).
Electrochemical Readout of the Benzo Switch. We have
noted previously that the CPD isomers of the photochromic
compounds discussed in this paper are essentially isolated 6π-
systems, while the DHP isomers have contiguous large π-sys-
tems. In the case of 7, there are 18 contiguous π-orbitals. The
change in the π-system, ∆π, on photoswitching, is thus rather
dramatic. Electrochemically, the 6π-system, benzene, is rela-
tively inert. The 18π-system, 7, should be relatively easy to
oxidize electrochemically. Electrochemical readout of the two
states of the photochrome 7/8 should thus be relatively easy
and clear. Indeed, when a solution of benzopyrene 7 in dry
degassed dichloromethane containing TBA PF6 was placed in
a quartz cell using Pt button electrodes and a Ag0 reference
electrode, using a scan rate of 100 mV/sec over a potential
window from 0 to 1.00 V (ferrocene Ep,a ) 0.55 V), the peak
anodic current (Ep,a) was centered at 0.75 V. The cell was then
irradiated with a 500 W household tungsten lamp affixed with
a 400 nm cutoff filter for a few seconds, which completely
opened 7 to CPD 8, and then across the same potential window,
this state was redox silent. The cyclic voltammogram is shown
in Figure 5. Switching the light source to a 4 W 365 nm
ultraviolet light rapidly promoted the back reaction and gave
the same redox features as before. The results appeared to be
stable over several cycles, though no quantitative degradation
studies were made at this time. The cyclic voltammogram clearly
demonstrates that the electrochemical readout of each state is
unequivocally discernible.
Figure 5. Cyclic voltammogram of DHP 7 (bold line, redox active) and
CPD 8 (fine line, redox silent).
Clearly, study of future photochromic systems using elec-
trochemistry will also be worthwhile.
Summary and Conlusions
Simple arene [e]-fused dihydropyrenes are negative photo-
chromes, where the thermally stable form is colored. We
conclude that the benzo compound 7/8 is the best photochrome
in this series and has the best combination of desirable
properties, namely 100% switching, reasonably high quantum
yield, good stability, and slow thermal return. We have also
prepared the positive photochrome, the dibenzo derivative 21/
22, in which the colorless form is the thermally stable one. This
can also be photochemically switched but has a very fast thermal
return reaction. By fusing two dihydropyrenes onto a common
chrysene spacer, 44, a switch was obtained with more than two
states. Although the fully open 45 and fully closed 44 states of
the switch were easily accessible, it was not possible to access
the third position, 46, completely. The three way switch 47/
48/49, which contains both mono- and dibenzannelated dihy-
dropyrenes, has a cleanly accessible middle switch position but
difficult to obtain fully closed position, 48. Switches containing
dihydropyrenes joined by fused conjugated spacers also have
limited stability. Overall this suggests that future efforts should
be made by joining benzo systems with nonfused arene spacers,
to improve stability and give optimal switching properties. Bulky
substituents at the 4,5-positions may also serve to slow the
thermal return reaction. Electrochemical readout of the state of
the switch is demonstrated for the benzo system as a potential
useful alternative to optical readouts.
Experimental Section
Melting points were determined on a Reichert 7905 melting point
apparatus integrated to an Omega Engineering model 199 chromel-
alumel thermocouple. Infrared spectra, calibrated with polystyrene, were
recorded as KBr disks on a Bruker IFS25 FTIR spectrometer and only
the major fingerprint bands are reported. UV-visible spectra were
recorded on a Cary 5 UV-vis-NIR spectrometer using cyclohexane
as solvent unless otherwise stated. Proton NMR spectra were recorded,
unless otherwise specified, on a Bruker AMX 360 spectrometer at 360
MHz in CDCl3, using the solvent residual peak for calibration (7.240
ppm for CHCl3). Where stated, 250 MHz spectra were recorded on a
Bruker WM 250. Carbon NMR spectra were recorded at 90.6 MHz
(or where stated 62.9 MHz), using the solvent (CDCl3) peak at 77.0
ppm for calibration. Where peaks within the same sample are very close
(31) Mitchell, R. H.; Chen, Y.; Iyer, V. S.; Lau, D. Y.; Baldridge, K. K.;
Siegel, J. S. J. Am. Chem. Soc. 1996, 118, 2907-2911. Williams,
R. V.; Armantrout, J. R.; Twamley, B.; Mitchell, R. H.; Ward, T. R.;
Bandyopadhyay, S. J. Am. Chem. Soc. 2002, 124, 13495-13505.
(32) For example, see: Houk, K. N.; Lee, P. S. Nendel, M. J. J. Org. Chem.
2001, 66, 5517-5521. Performed using Gaussian 98W (revision A.9),
Gaussian, Inc, Pittsburgh, PA, 1998.
9
2982 J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003

Dihydropyrene−Cyclophanediene System
A R T I C L E S
in chemical shift, a third decimal place is given. Mass spectra were
recorded on a Finnigan 3300 gas chromatography-mass spectroscopy
system using methane as a carrier gas for chemical ionization. Exact
mass measurements were obtained on a Kratos Concept-H instrument
using perfluorokerosene as the standard. Elemental analyses were
performed by Canadian Microanalytical Services Ltd., Vancouver, B.
C. All evaporations were carried out under reduced pressure on a rotary
evaporator or by using an oil pump and dry ice condenser. SiGel refers
to Merck silica gel, 60-200 mesh. Alumina refers to Aldrich aluminum
oxide, activated, neutral, Brockmann I, standard grade, ∼150 mesh.
Solutions of oxygen sensitive compounds for NMR spectroscopy were
prepared by dissolving the compound in the minimum amount of
deuterated solvent, filtration through alumina deactivated with 5% water,
rinsing the alumina with sufficient deuterated solvent to make up the
correct volume for the spectrometer, and bubbling argon through the
filtrate in the NMR tube for 5-10 min. The tubes were then carefully
capped under argon, and the spectra obtained. NMR spectra of
photoisomerizable compounds were obtained using 5 mm amber NMR
tubes. Where NMR assignments are made, these were on the basis of
2D COSY/NOESY experiments for 1H and HETCORR/HMQCB
experiments for 13C. H-1,2 means H-1 and H-2; H-1/2 means H-1 or
H-2. Expanded data sets were used to obtain coupling constant data.
To clarify the NMR assignments, numbering of the relevant parent
structures is shown here in the order they appear in the experiment
(compound numbers in parentheses):
1.3 Hz, 2H, H-3,6), 7.65-7.60 (AA′XX′, 2H, H-10,11), 1.49 (s, 18H,
C(CH3)3), -1.30 (s, 6H, CH3); 13C NMR δ 148.67, 136.16, 135.71,
129.15, 126.67 (C-10,11), 124.69 (C-9,12), 120.15 (C-3,6), 117.51 (C-
1,8), 117.04, 39.13, 35.87, 30.37 (C(CH3)3), 17.34 (C-12c,d); IR 1616,
1602, 1475, 1445, 1367, 1361, 1318, 1255, 1201, 1121, 1000, 959,
913, 866, 800, 754, 637, 610 cm-1; CIMS m/z 551, 553, 555 (MH+)
(Br2 pattern). Anal. Calcd for C30H32 Br2: C, 65.23; H, 5.84. Found:
C, 65.33; H, 5.89.
2,7-Di-tert-butyl-9,18-epoxy-18c,18d-dimethyl-9,18,18c,18d-tet-
rahydrotetrabenzo[de,j,l,qr]naphthacene 20. NaNH2 (570 mg, 14.6
mmol) and t-BuOK (2 mg) were added to a stirred solution of the
isofuran 1911b (84 mg, 0.22 mmol) and 9-bromophenanthrene (282 mg,
1.10 mmol) in dried THF (25 mL) under argon. The mixture was stirred
overnight, and then methanol (5 mL) and hexanes (25 mL) were added.
The organic phase was washed, dried, and evaporated. The solid was
chromatographed over deactivated SiGel using benzene/hexanes (1:1)
as eluant. The green band was collected and gave 104 mg (85%) of
the adduct, which on recrystallization from toluene gave dark green
crystals: mp 221-222 °C; 1H NMR δ 8.59 (d, J ) 8.3 Hz, 2H,
H-13,14) 8.43, 8.40 (d, J ≈ 1.2 Hz, 1H each, H-1/8), 8.41, 8.27 (dd,
J ) 8.1 Hz, 1.1 Hz, 1H each, H-10/17), 8.13, 8.10 (d, J ) 1.2 Hz, 1H
each, H-3/6), 8.07 (s, 2H, H-4,5), 7.63, 7.61 (s, 1H each, H-9/18), 7.72-
7.64, 7.59-7.52 (m, 2H each, H-11,16/12,15), 1.66, 1.63 (s, 9H each,
C(CH3)3), -3.14, -4.12 (s, 3H each, CH3); 13C NMR δ 145.52, 145.30,
145.01, 144.66, 137.28, 136.22, 136.18, 135.76, 129.61, 129.47, 127.81,
127.22, 126.78 and 126.75 (C-11,16), 126.67, 126.62, 126.04 and
125.97 (C-12,15), 124.47 and 124.36(C-4,5), 124.23 and 124.03 (C-
10,17), 123.57 (C-13,14), 121.23 (C-3,6), 115.47 and 114.70 (C-1,8),
81.72 and 81.14 (C-9,18), 35.74, 35.69, 34.21, 33.32, 31.54 and 31.51
(C(CH3)3), 16.50 and 14.25 (C-18c,d); IR (KBr) 1616, 1448, 1363,
1343, 1260, 883, 865, 777, 755, 745, 721, 716, 671, 632 cm-1; CIMS
m/z 561 (MH+). Anal. Calcd for C42H40O: C, 89.96; H, 7.19. Found:
C, 89.66; H, 7.34.
2,7-Di-tert-butyl-trans-18c,18d-dimethyl-18c,18d-dihydrotetrabenzo-
[de,j,l,qr]naphthacene 15. A mixture of the epoxide 20 (88 mg, 0.16
mmol) and Fe2(CO)9 (88 mg, 0.24 mmol) in benzene (30 mL) was
refluxed under argon in the dark for 5 h. After the mixture was cooled
to room temperature, it was filtered through a short column (5 cm) of
deactivated SiGel using benzene as eluant. The solvent was then
removed, and the dark purple solid product was carefully washed with
pentane until the initial brownish color disappeared and gave 73 mg
(86%) of the product 15. Slow evaporation from hexanes/CH2Cl2
1
yielded dark purple crystals: mp (dec) 188-193 °C; H NMR δ 9.92
(s, 2H, H-9,18), 8.92 (dd, J ) 8.2 Hz, 1.3 Hz, 2H, H-10,17), 8.67 (dd,
J ) 8.2 Hz, 1.2 Hz, 2H, H-13,14), 8.40 (d, J ) 1.3 Hz, 2H, H-1,8),
7.78-7.74 (m, 2H, H-11,16), 7.71-7.67 (m, 2H, H-12,15), 7.23 (d, J
) 1.2 Hz, 2H, H-3,6), 6.99 (s, 2H, H-4,5), 1.55 (s, 18H, C(CH3)3),
-1.07 (s, 6H, CH3); 13C NMR δ 144.51, 138.81, 135.19, 130.15,
130.06, 128.59, 127.54, 127.46 (C-11,16), 127.37 (C-12,15), 123.53
(C-13,14), 123.33 (C-10,17), 121.26 (C-4,5), 120.15 (C-3,6), 118.57
(C8,19), 117.43 (C-1,8), 36.47, 35.39, 30.63 (C(CH3)3), 18.31 (C-18c,d);
UV (CHCl3) λmax (ꢀmax) nm 284 (43 200), 337 (38 600), 374 sh (22 200),
392 (46 200), 412 (72 400), 529 (5700), 644(600); IR (KBr) 1627, 1605,
1490, 1434, 1360, 1227, 886, 872, 754, 719, 673 cm-1; CIMS m/z 545
(MH+). Anal. Calcd for C42H40: C, 92.60; H, 7.40. Found: C, 92.58;
H, 7.64.
4,5-Dibromo-2,7-di-tert-butyl-trans-12c,12d-dihydro-12c,12d-di-
methylbenzo[e]pyrene 13. A solution of vacuum-dried benzopyrene
711b (285 mg, 0.721 mmol) in dry CH2Cl2 (200 mL) was prepared under
argon in the dark and cooled in an ice bath. NBS (vacuum-dried, 257
mg, 1.44 mmol) in dry DMF (40 mL) was then added slowly with
stirring. After the addition, the reaction mixture was allowed to warm
to room temperature and let stir for 2 h. The reaction mixture was then
poured into hexanes (280 mL) and was washed with water (8 × 200
mL). The organic phase was separated, dried, and evaporated. The
residue was purified by repeated (4×) chromatography (to remove
fluorescent close moving compounds) on SiGel (deactivated with 5%
water), using hexanes/CH2Cl2 (9:1) as eluant and yielded 191 mg (48%)
of purple crystals of the dibromide 13, which on recrystallization from
cyclohexane had mp (dec) 187-189 °C; 1H NMR δ 8.69-8.64
(AA′XX′, 2H, H-9,12), 8.21 (d, J ) 1.3 Hz, 2H, H-1,8), 7.73 (d, J )
anti-8,16-Dimethyl-9,10-benzo[2.2]metacyclophane-trans-1,2-di-
ol 26. Zn-Cu couple33 (13.6 g, 202 mmol) was added to TiCl3 (15.7
g, 102 mmol) in dry 1,2-dimethoxyethane (400 mL) under N2, and the
solution was stirred vigorously under reflux for 15 h, which yielded a
black slurry. The dialdehyde 2520 (1.46 g, 4.46 mmol) in dimethoxy-
ethane (1500 mL) was then added slowly to the refluxing slurry over
85 h. After the mixture was cooled in ice, saturated aq K2CO3 (300
mL) was added over 1 h. The reaction mixture was then filtered through
(33) Gribble, G. W.; Leese, R. M. Synthesis 1977, 172-176.
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 2983

A R T I C L E S
Mitchell et al.
a short pad of florisil, dried, and evaporated. The crude product was
then chromatographed over florisil (15 cm × 1.5 cm) using hexanes/
THF (1:4) as eluant to yield 1.1 g (78%) of white crystals of diol 26:
mp 215-218 °C; 1H NMR (250 MHz, d8-THF) δ 7.71-7.65 (m, 4H),
7.44-7.41 (m, 2H), 7.07-7.04 (m, 4H), 4.98 (bs, 2H, -OH), 4.24 (s,
2H, H-1,2), 0.61 (s, 6H, CH3); 13C NMR (62.9 MHz, d8-THF) δ 143.06,
139.27, 138.66, 138.02, 130.73, 129.45, 129.04, 125.65, 124.91, 81.29,
17.22; IR 3510, 3365, 1446, 1418, 1050, 1032, 1012, 752, 738, 714
cm-1; EIMS at m/z 316 (M+). Anal. Calcd for C22H20O2: C, 83.51; H,
6.37. Found: C, 83.48; H, 6.32.
(25 mL), dried, and evaporated. The residue was chromatographed on
florisil (12 cm × 1.5 cm) using dichloromethane/pentane (9:1) as eluant
and gave 0.15 g (20%) of the diketone 28, identical to the previous
sample.
Benzoquinoxalinocyclophane 29. The diketone 28 (0.25 g, 0.80
mmol) and o-phenylenediamine (0.09 g, 0.83 mmol) were refluxed in
methanol (25 mL) under nitrogen for 2 h with stirring. The reaction
mixture was then cooled to room temperature, and the yellow deposit
of 29 was filtered. A second crop of 29 was obtained by dropwise
addition of water to the filtrate. The combined crops were recrystallized
from benzene/methanol to give 0.19 g (56%) of colorless cyclophane
29: mp 211-213 °C; 1H NMR (250 MHz) δ 8.19-8.12 (m, 2H), 7.78-
7.69 (m, 4H), 7.47-7.39 (m, 2H), 7.36-7.21 (m, 4H), 7.14-7.09 (m,
2H), 1.20 (s, 6H, -CH3); 13C NMR (62.9 MHz) δ 156.43, 142.06,
141.54, 140.68, 139.81, 136.49, 132.98, 132.14, 129.70, 129.22, 128.85,
anti-8,16-Dimethyl-9,10-benzo[2.2]metacyclophan-1-ol-2-one 27.
Dry DMSO (2 mL, 26 mmol) was added to oxalyl chloride (l mL, 11
mmol) in dichloromethane (25 mL) maintained at -30 °C under
nitrogen. The mixture was stirred for 2 min, and then the diol 26 (0.46
g, 1.45 mmol) in THF (5 mL) was added within 5 min. The reaction
mixture was then stirred for 1 h at -30 °C, and then triethylamine (7
mL, 50 mmol) was added; the mixture was stirred again for 30 min.
The reaction mixture was then allowed to warm to ambient temperature
and extracted with diethyl ether (150 mL). The ether layer was washed
with water (100 mL) and 10% aq HCl (100 mL). Each aq layer was
re-extracted with diethyl ether (150 mL), and organic layers were
combined. These were then washed with water, 5% aq Na2CO3, and
water and then were dried and evaporated. The crude product obtained
was chromatographed on a column of florisil (15 cm × 1.5 cm) using
pentane/dichloromethane (1:9) as eluant to give the product, which was
recrystallized from chloroform/hexane to give 0.41 g (90%) of 27 as
white crystals: mp 205-208 °C; 1H NMR (250 MHz, d8-THF) δ 7.77-
7.70 (m, 2H), 7.52-7.44 (m, 3H), 7.23-7.05 (m, 5H), 5.26 (d, J ) 4
Hz, 1H), 5.11 (d, J ) 4 Hz, 1H, -OH), 0.84 and 0.56 (s, 3H each,
-CH3); 13C NMR (62.9 MHz, d8-THF) δ 203.70, 142.01, 141.75,
140.19, 138.52 (×2), 138.01, 134.07, 133.73, 132.81, 131.57, 129.86,
129.69, 129.57, 129.46, 126.72, 126.49, 125.20, 124.37, 82.97, 17.19,
17.10; IR 3462, 1702, 1447, 1410, 1132, 1112, 1052, 940, 789, 759,
748, 732 cm-1; EIMS m/z 314 (M+). Anal. Calcd for C22H1802: C,
84.05; H, 5.77. Found: C, 84.06; H, 5.75.
128.79, 127.63, 19.23; IR 1410, 1302, 1213, 1016, 795, 750, 723 cm-1
;
UV (CH2Cl12) λmax (ꢀmax) nm 256 (117 000), 291 (21 000), 347 (20 800);
EIMS m/z 384 (M+). Anal. Calcd for C28H20N2: C, 87.47; H, 5.24.
Found: C, 87.93; H, 5.36.
Irradiation. When the colorless d8-THF NMR solution of 29 was
irradiated with 254 nm light, it turned reddish-purple and showed
1
new broad absorptions at 350 and 525 nm. The H NMR spectrum
however only gave very broad featureless signals, possibly due to 33
(see text).
anti-8,16-Dimethyl-9,10-(2′,3′-naphtho)[2.2]metacyclophane-1,2-
dione 31. From TiCl3 (14.5 g, 94 mmol) and Zn-Cu couple (12.8 g,
190 mmol) in dimethoxyethane (400 mL) and dialdehyde 3011a (1.2 g,
3.3 mmol) in dimethoxyethane (1200 mL), exactly as described for
26, there was obtained 0.98 g (82%) of anti-8,16-dimethyl-9,10-(2′,3′-
naphtho)[2.2]metacyclophane-trans-1,2-diol as white crystals: mp 244-
1
246 °C; H NMR (250 MHz, d8-THF) δ 8.20 (s, 2H), 7.97-7.90 (m,
2H), 7.72-7.69 (m, 2H), 7.52-7.46 (m, 2H), 7.21-7.18 (m, 2H), 7.11-
7.05 (m, 2H), 4.95 (s, 2H, -OH), 4.26 (s, 2H), 0.64 (s, 6H); 13C NMR
(62.9 MHz, d8-THF) δ 141.77, 139.13, 137.98, 134.68, 131.86, 128.95,
128.56, 128.19, 126.86, 125.53, 125.02, 81.20, 17.19; IR 3492, 3380,
1418, 1210, 1098, 1025, 1010, 988, 785, 718 cm-1; CIMS m/z 367
(MH+). Anal. Calcd for C26H22O2: C, 85.21; H, 6.05. Found: C, 85.26;
H, 6.17.
anti-8,16-Dimethyl-9,10-benzo[2.2]metacyclophane-1,2-dione 28.
Dry DMSO (2 mL, 26 mmol) was added to a stirred solution of oxalyl
chloride (1 mL, 11 mmol) in dichloromethane (25 mL) maintained at
-30 °C under nitrogen. After 2 min, the hydroxyketone 27 (0.52 g,
1.7 mmol) in dichloromethane (5 mL) was added over 5 min, and the
mixture was then stirred for 1 h at -30 °C. Triethylamine (7 mL. 50
mmol) was then added, and the mixture was stirred for a further 30
min at -30 °C. The reaction mixture was then allowed to warm to
ambient temperature and was extracted with dichloromethane (150 mL)
and washed with water and 10% aq HCl. The aq layers were
re-extracted with dichloromethane (150 mL), and the combined organic
layers were then washed with water, 5% aq Na2CO3, and water and
then were dried and evaporated. The crude product was chromato-
graphed on a column of florisil (10 cm × 1.5 cm) using pentane/
dichloromethane (1:9) as eluant to give the diketone product, which
on recrystallization from ether/hexanes gave 0.46 g (89%) of 28 as
From this diol (0.52 g, 1.4 mmol) in THF (5 mL) and DMSO (1.7
mL, 22 mmol) and oxalyl chloride (1 mL, 11 mmol) in dichloromethane
(25 mL) and then triethylamine (7 mL), exactly as described above for
hydroxyketone 27, there was obtained 0.48 g (86%) of anti-8,16-
dimethyl-9,10-(2′,3′-naphtho)[2.2]metacyclophan-1-ol-2-one as white
1
crystals: mp 225-228 °C; H NMR (250 MHz, d8-THF) δ 8.27 (s,
1H), 8.25 (s, 1H), 8.00-7.94 (m, 2H), 7.55-7.50 (m, 3H), 7.42-7.38
(m, 1H), 7.27-7.08 (m, 4H), 5.30 (d, J ) 4 Hz, 1H), 5.11 (d, J ) 4
Hz, 1H, -OH), 0.88 and 0.60 (3, 3H each, -CH3); 13C NMR (62.9
MHz, d8-THF) δ 203.79, 140.46, 140.31, 140.10, 139.17, 138.63,
138.42, 134.84, 134.69, 134.04, 133.75, 133.93, 132.69, 128.76, 128.67
(×2), 128.58, 127.28, 127.22, 126.58, 126.34, 125.25, 124.53, 82.79,
17.18, 17.09; IR 3490, 1710, 1442, 1418, 1378, 1115, 1090, 940, 774,
743, 715 cm-1; EIMS m/z 364 (M+). Anal. Calcd for C26H20O2: C,
85.69; H, 5.53. Found: C, 85.57; H, 5.52.
1
yellow crystals: mp 178-180 °C; H NMR (250 MHz, CD2Cl2) δ
7.80-7.74 (AA′XX′, 2H), 7.58-7.51 (AA′XX′, 2H), 7.49-7.16 (m,
6H), 0.89 (s, 6H); 13C NMR (62.9 MHz, CD2Cl2) δ 191.84, 140.37,
139.34, 139.29, 134.77, 134.28, 129.86, 129.63, 128.08, 127.73, 18.45;
From this hydroxyketone (0.55 g, 1.55 mmol) in dry CH2Cl2 (5 mL)
and dry DMSO (1.7 mL, 22 mmol), oxalyl chloride (1 mL, 11 mmol),
CH2Cl2 (25 mL), and triethylamine (7 mL, 50 mmol), exactly as
described for dione 28, there was obtained 0.48 g (88%) of diketone
IR 1680, 1551, 1403, 1368, 1214, 913, 795, 775, 745, 710 cm-1
;
UV (CH2Cl2) λmax (ꢀmax) nm 249 (49 500), 323 (3260); EIMS m/z 312
(M+). Anal. Calcd for C22H1602: C, 84.59; H, 5.16. Found: C, 84.45;
H, 5.44.
1
31, as yellow crystals: mp 186-187 °C; H NMR (250 MHz, CD2-
Cl2) δ 8.26 (s, 2H), 8.02-7.96 (AA′XX′, 2H), 7.62-7.54 (AA′XX′,
2H), 7.49 (dd, J ) 7.2, 1.8 Hz, 2H), 7.35 (t, J ) 7.2 Hz, 2H), 7.29 (dd,
J ) 7.2, 1.8 Hz, 2H), 0.91 (s, 6H, -CH3); 13C NMR (62.9 MHz, CD2-
Cl2) δ 191.76, 139.88, 139.18, 138.78, 135.82, 134.27 (×2), 128.73,
128.32, 128.22, 127.53, 127.32, 18.44; IR 1695, 1220, 782, 738, 719;
UV (CH2Cl2) λmax (ꢀmax) nm 250 (42 000), 320 (3730); EIMS m/z
362 (M+). Anal. Calcd for C26H18O2: C, 86.16; H, 5.01. Found: C,
86.03; H, 5.25.
Direct Oxidation of Diol 26 to Diketone 28. Sulfur trioxide/pyridine
complex (2.00 g, 12.6 mmol) in DMSO (10 mL) at 20 °C under nitrogen
was added to the diol 26 (0.75 g, 2.4 mmol) in DMSO (10 mL) and
triethylamine (6.12 g, 8.4 mL, 60 mmol) at -20 °C. The reaction
mixture was stirred under nitrogen for 3 h and then was acidified with
10% aq HCl. The product was precipitated by adding water. The
precipitate was collected and then was extracted with dichloromethane
9
2984 J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003

Dihydropyrene−Cyclophanediene System
A R T I C L E S
Naphthoquinoxalinocyclophane 32. From the diketone 31 (0.23
g, 0.63 mmol), o-phenylenediamine (70 mg, 0.64 mmol), and MeOH
(25 mL), exactly as described for cyclophane 29, there was obtained
0.19 g (63%) of cyclophane 32, as colorless crystals from chloroform/
THF: mp 247-250 °C; 1H NMR (250 MHz, d8-THF) δ 8.27 (s, 2H),
8.16-8.09 (m, 2H), 7.98-7.94 (m, 2H), 7.81-7.75 (m, 2H), 7.55-
7.50 (m, 2H), 7.38-7.25 (m, 6H), 1.21 (s, 6H, -CH3); 13C NMR (62.9
MHz, d8-THF) δ 156.89, 142.70, 141.60, 141.09, 140.73, 137.84,
134.88, 134.66, 133.21, 130.32, 130.08, 128.62, 128.52, 128.00, 127.20,
UV Irradiation of Cyclophane 21; Valence Isomerization to
Dibenzodihydropyrene 22. (A) By NMR. An NMR tube containing
the cyclophane 21 (2.9 mg in 0.5 mL of d8-THF) was first cooled to
about -80 °C in a cooling bath of pentane, which was cooled by liquid
nitrogen outside. Then the NMR tube in a small pentane cooling bath
was transferred immediately into a photochemical reactor and irradiated
at 300 nm for ∼1 min. A very heavy frost forms quickly on the cooling
bath and on the NMR tube which is removed by methanol on a tissue.
Then, the NMR tube was transferred into the NMR instrument which
was already set at the desired temperature, and the NMR spectrum was
19.35; IR 1300, 1210, 788, 742, 720 cm-1; UV (CH2Cl2) λmax (ꢀmax
)
1
nm 253 (120 000), 280 (86 300), 342 (26 500), 368 (21 100); EIMS
m/z 434 (M+). Anal. Calcd for C32H22N2: C, 88.45; H, 5.10. Found:
C, 88.39; H, 5.03.
immediately recorded at the desired temperature. H NMR (d8-THF,
-80 °C) δ 9.45 (s, 4H, H-5,7,12,14), 9.38-9.35 (AA′XX′, 4H,
H-1,4,8,11), 7.62-7.59 (AA′XX′, 4H, H-2,3,9,10), 1.79 (s, 18H, t-Bu),
-3.41 (s, 6H, internal -CH3). For reference, the peaks of the unchanged
cyclophane 21 at this temperature were δ 7.74-7.70, 7.47-7.43, 6.98,
1.28, and 1.07, and the ratio 22/21 was about 35:65.
(B) By UV. A solution of cyclophane 21 (1.1 mg in 200 mL of
THF) was cooled to about - 50 °C in a UV cell directly and then was
irradiated with the light of wavelength 300 nm for about 15 s. Then,
the cell was transferred into the UV-vis instrument, and the scanning
started immediately. The green solution containing 22 showed new
peaks (relative intensity) at 327 (0.52), 340 (0.51), 407 (0.44), 431
(1.00), 630 (0.25), 680 (0.25), and 730 (0.23) nm. If conversion were
100%, ꢀ431 ) 54 000 mol-1 cm-1. Irradiation with visible light or
warming returned cyclophane 21.
Thermal Isomerization Studies of 22 to 21. The irradiated NMR
solutions previously mentioned were held in the instrument at the
desired temperature, and the 1H NMR spectra were recorded at 10 min
intervals at -40 °C to 70 min, 6 min intervals at -30 °C to 60 min,
∼1 min intervals at -20 °C to 20 min, and ∼1 min intervals at -10
°C to 12 min. Integrations of both the internal methyl protons and the
aromatic protons were used to obtain the ratio r ) [22]/[21]. Then, if
the initial concentration of 22 is C0, then as the thermal reaction
proceeds, C ) C0[r/(1 + r)] and hence -ln r/(1 + r) ) kt + constant,
from which the rate constant (k) at temperature T is easily determined.
Then, using the standard -ln k versus 1/T plot, we are able to obtain
the activation energy Eact for the process. The values of k (s-1) and half-
6,13-Di-tert-butyl-1,4:8,11-diepoxy-trans-14b,14c-dimethyl-
1,4,8,11,14b,14c-hexahydrodibenzo[fg,op]naphthacene 35. n-BuLi
(0.92 mL, 2.3 mmol, 2.5 M in hexane) in hexane (15 mL) was added
dropwise over 30 min to a stirred solution of the tetrabromide 3634
(0.684 g, 1.04 mmol) in furan (50 mL) and THF (5 mL) at -78 °C
under Ar. The cooling bath was then removed, and the mixture was
stirred for a further 2 h at room temperature. Methanol (2 mL) was
added to quench the mixture, and then ether (100 mL) and ice-water
(100 g) were added. A simple filtration at this stage yielded 0.16 g of
bright green crystals, pure isomer 35a. The filtrate was extracted with
ether (2 × 100 mL). The ether extracts were then washed, dried, and
evaporated to yield a further 0.26 g (total yield, 85%) of the product,
about 20% of isomer 35a and 80% of isomer 35b. For preparative
purposes, these isomers could be combined and used in the next step
or sufficient ether can be added initially to dissolve all the material.
For characterization purposes, isomer 35a could be fractionally
crystallized from hexanes: mp 236-238 °C; 1H NMR δ 8.34 and 8.31
(bs, 2H each, H-5,7,12,14), 7.29 and 7.19 (dd, J ) 5.5 and 1.9 Hz, 2H
each, H-2,3,9,10), 6.674-6.667 and 6.651-6.646 (m, 2H each,
H-1,4,8,11), 1.67 (s, 18H, t-Bu), -3.99 (s, 6H, -CH3); 13C NMR δ
145.07, 141.70 (×2), 139.18, 138.87, 128.41, 128.37, 115.79, 115.21,
82.01, 81.69, 35.97, 31.94, 31.32, 14.19; UV (CHCl3) λmax (ꢀmax) nm
284 (5100), 377 (53 000), 391 (36 100), 446 (6900), 665 (800); IR
1467, 1254, 1090, 1027, 864, 802 cm-1; CIMS m/z 477 (MH+). Anal.
Calcd for C34H36O2: C, 85.94; H, 7.60. Found: C, 85.67; H, 7.61.
lives τ1/2 (min) found were as follows: 233 K, (2.86 ( 0.19) × 10-5
400; 243 K, (2.16 ( 0.12) × 10-4, 54; 253 K, (9.45 ( 0.11) × 10-4
12; 263 K, (4.26 ( 0.08) × 10-3, 2.7; Eact ) 20.1 ( 0.7 kcal/mol.
,
,
1
The main difference in the H NMR spectrum of the more soluble
isomer 35b, was in the chemical shift of the internal methyl protons at
δ -3.70 and -4.30 (35a δ -3.99).
At the end of all thermal isomerizations, only pure cyclophane 21
remained.
When NaNH2 (400 mg, 10 mmol) and t-BuOK (2 mg) were added
to a stirred solution of dibromide 34 (see later) (430 mg, 0.86 mmol)
in dried furan (5 mL) and THF (0.5 mL), and then the mixture was
stirred at 55 °C for 4 days before addition of MeOH as previously.
The residue after extraction was chromatographed over silica gel using
hexanes/ethyl acetate (9:1) as eluant. Eluted first was a small amount
of monoadduct (∼20 mg, 5%), followed by the bis-adducts 35, 241
mg (59%), as a 1:1 mixture of the two isomers 35a and 35b, identical
to those previously mentioned.
anti-[1,2;9,10]-Dibenzo-5,13-di-tert-butyl-8,16-dimethyl[2.2]meta-
cyclophane 21. The mixture of adducts 35 (100 mg, 0.21 mmol) and
Fe2(CO)9 (190 mg, 0.53 mmol) in benzene (10 mL) was stirred under
Ar at reflux for 6 h. The mixture was cooled and preabsorbed onto
SiGel (2 g) and then was chromatographed over SiGel using hexanes
to give 40 mg (68%) of cyclophane 21, as colorless crystals from
hexanes: mp 231-232 °C; 1H NMR (d8-THF) δ 7.71-7.66 (AA′XX′,
4H, H-17,20,21,24), 7.41-7.36 (AA′XX′, 4H, H-18,19,22,23), 6.95 (s,
4H, H-4,6,12,14), 1.29 (s, 18H, t-Bu), 1.07 (s, 6H, -CH3); 13C NMR
(d8-THF) δ 150.88, 144.57, 140.64, 138.94, 129.37, 129.23, 128.97,
34.75, 31.62, 18.70; UV (cyclohexane) λmax (ꢀmax) nm 288 (12 000),
tail to 400 nm; IR 1467, 1417, 1354, 1254, 1228, 877, 730, 638, 588
cm-1; CIMS m/z 445 (MH+). Anal. Calcd for C34H36: C, 91.84; H,
8.16. Found: C, 91.24; H, 8.36.
4,9-Dibromo-2,7-di-tert-butyl-trans-10b,10c-dimethyl-10b,10c-di-
hydropyrene 34. A precooled (-7 °C) solution of NBS (1.10 g, 6.17
mmol) in dry DMF (100 mL) was added slowly over 1 h to a precooled
(-7 °C) solution of dihydropyrene 534 (see also Supporting Information)
(1.01 g, 2.94 mmol) in dry DMF/CCl4 (350 mL/50 mL) at 0 °C, and
then the reaction was stirred without cooling for 2 h. Hexanes (200
mL) were added, and then the mixture was poured into ice-water.
The hexanes extract was washed well with water, dried, and evaporated.
The residue was chromatographed over SiGel using hexanes and gave
1.3 g (88%), from which the less soluble 34 was easily separated from
a small amount of the more soluble 4,10-isomer by fractional
recrystallization from hexanes, as dark green crystals: mp 220-222
1
°C; H NMR δ 8.81 (d, J ) 1.1 Hz, 2H), 8.64 (s, 2H), 8.48 (bs, 2H),
1.68 (s, 18H, t-Bu), -3.83 (s, 6H, -CH3); 13C NMR δ 148.17, 137.22,
132.42, 126.93, 121.94, 121.54, 116.48, 36.32, 32.15, 31.85 (C(CH3)3),
14.26; IR 1329, 1241, 1115, 877, 777, 676 cm-1; UV λmax (ꢀmax) nm
350 (77 000), 389 (40 000), 486 (7000), 654 (1400). CIMS m/z 501
(MH+, correct isotope pattern). HRMS (EI) Calcd (79Br2): 500.0714.
Found: 500.0715. Anal. Calcd for C26H30Br2: C, 62.17; H, 6.02.
Found: C, 62.06, H, 6.01.
If NBS (3.53 g, 19.8 mmol) in CHCl3 (150 mL) is added to the
pyrene 5 (3.1 g, 9.01 mmol) in CHCl3 at 20 °C and the mixture stirred
overnight, 4.5 g (100%) of a 3:2 mixture of the 4,9-dibromo and 4,10-
dibromo isomers is obtained. This mixture is suitable for use in the
(34) Tashiro, M.; Yamato, T. J. Am. Chem. Soc. 1982, 104, 3701-3707.
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 2985

A R T I C L E S
Mitchell et al.
Diels-Alder additions. The spectra of the 4,10-isomer could easily be
obtained by subtraction: 1H NMR δ 9.00 (d, J ) 1.3 Hz, 2H), 8.96 (d,
J ) 1.3 Hz, 2H), 8.84 (d, J ) 1.3 Hz, 2H), 1.70 (s, 9H, t-Bu), 1.67 (s,
9H, t-Bu), -3.69 (s, 3H, -CH3), -3.71 (s, 3H, -CH3); 13C NMR δ
149.49, 137.86, 133.67, 133.52, 132.95, 127.45, 124.56, 124.10, 122.90,
122.05, 117.13, 36.60, 36.34, 34.50, 32.13, 31.76, 26.89, 14.60, 13.93.
Bis-isobenzofuran Adduct 37. LDA (15 mL, 1.5 M in cyclohexane,
22.5 mmol) was added to a solution of 1-methoxyphthalan (3.0 g, 20
mmol) in THF (35 mL) under Ar at room temperature. After 5 min,
the dibromide 34 (or mixture of isomers) (1.0 g, 2.0 mmol) and sodium
amide (400 mg, 10 mmol) were added, and stirring con-
tinued for 4 days. Methanol (2 mL) was then added, and the total
mixture was preabsorbed onto SiGel and chromatographed using first
hexanes as eluant (which gave some dihydropyrene 5 and bromo-5)
and then hexanes-ether (20:1) which eluted some monoadduct and
then was eluted a 1:1 mixture of bis-adduct isomers 37 (0.38 g, 33%).
Recrystallization and chromatography failed to separate these two
isomers completely, though they yielded a mixture enriched in the syn-
isomer of 37 corresponding to 35b, which enabled assignment of their
proton spectra. 1H NMR (d8-THF) 37a (anti-O) δ 8.65 (bs, 4,
H-1,3,10,12), 7.48-7.45 and 7.40-7.37 (m, 2 each set, H-5,8,14,17),
7.03 (s, 4, H-4,9,13,18), 6.86-6.84 (m, 4, H-6,7,15,16), 1.73 (s, 18,
t-Bu), -4.39 (s, 6, -CH3). 37b (syn-O) 8.61 (bs, 4, H-1,3,10,12), 7.41-
7.39 and 7.32-7.30 (m, 2 each set, H-5,8,14,17), 7.02 (s, 4, H-4,9,13,18),
6.78-6.75 (m, 4, H-6,7,15,16), 1.750 and 1.746 (s, 9 each, t-Bu), -3.66
and -5.17 (s, 3 each, -CH3). 13C NMR (mixture of 37a and 37b) δ
149.96, 149.68, 146.40, 146.06, 145.72, 141.47, 141.07, 140.90, 140.78,
129.03, 128.79, 128.49, 128.39, 125.97, 125.90, 125.84, 125.78, 120.37,
120.16, 120.09, 119.91, 117.46, 116.55, 116.49, 83.10, 83.04, 82.65,
82.57, 36.71, 35.02, 32.76, 32.23, 30.88, 16.37, 14.29, 12.11. LSIMS
(Matrix, mNBA) calcd: 577.3 (MH+). Found: 577.3. Anal. Calcd for
C42H40O2: C, 87.45; H, 6.99. Found: C, 86.19; H, 6.95. HRMS (EI)
calcd: 576.3028. Found: 576.3029.
added, and the organic layers were separated, washed with aq NH4Cl,
brine, dried and evaporated. The aqueous layers were further extracted
with ether, dried and evaporated. The combined products were rinsed
with hexanes (4 × 4 mL) and dried to yield 122 mg (81%) of adducts
43 as a green crystalline mixture of approximately equal amounts of
1
two isomers, mp (dec) 205-210 °C; H NMR (C6D6, 300 MHz) δ
8.6-7.5 (m, 16H), 7.4-7.2 (m, 2H), 6.95-6.8 (m, 2H), 1.7-1.6 (at
least 4 overlapping s, 36H), -2.89, -2.95, -3.79, -3.92 (s, 12H total,
internal CH3); LSIMS m/z 893.4 (MH+); HR MS calcd for C66H69O2
(MH+): 893.5297. Found: 893.5294
This mixture of isomers was used directly in the next step.
2,7,14,19-Tetra-tert-butyl-trans-20b,20c-trans-24c,24d-tetramethyl-
20b,20c,24c,24d-tetrahydro-di-(4,5-pyreno)[b,k]chrysene 44. The
mixed isomers of 43 (38 mg, 0.043 mmol) and Fe2(CO)9 (140 mg, 0.4
mmol) were stirred under reflux in benzene (15 mL) under argon in
the dark for 2 h. After cooling, the mixture was filtered through
deactivated SiGel, using an additional 100 mL of CHCl3 to elute the
brown product. The eluate was evaporated, and the residue was purified
by further chromatography on deactivated SiGel under argon using
hexanes/CH2Cl2 (6:1) as eluant. The reddish-brown product band was
collected and evaporated to yield 30 mg (82%) of mixed isomers of
44: mp (dec) >345 °C; 1H NMR (C6D6 {values in CDCl3}) δ
10.35{10.09} (s, 2H, H-9,21), 9.39{9.24} (s, 2H, H-12,24), 9.18{9.06}
(d, J ) 9.0 Hz, 2H, H-10,22), 8.54{8.39} (bs, 2H, H-8,20), 8.41{8.36}
(bs, 2H, H-1,13), 8.17{8.36} (d, J ) 9.0 Hz, 2H, H-11,23), 7.23{7.11},
7.22{7.09} (bs, 4H, H-3,6,15,18), 6.90{6.84} (bs, 4H, H-4,5,16,17),
1.453{1.57} (s, 18H), 1.442{1.52} (s, 18H), -0.284 and -0.295{-0.73}
(bs, 12H total (∼1:3 ratio), -CH3; 13C NMR (C6D6, not all peaks
visible, compound too unstable in CDCl3) δ 144.76, 139.59, 139.36,
136.57, 136.04, 131.05, [129-127 peaks obscured by benzene], 124.37,
122.04, 121.96, 121.77, 120.92, 120.88, 119.31, 118.21, 118.21, 37.96,
35.30, 30.50 and 30.44 (C(CH3)3), 19.28 (C-20b,c;24c,d); UV (cyclo-
hexane) λmax (ꢀmax) nm 290 (51 600), 345 (43 400), 406 (70 900), 427
(132 000), 527 (10 000); LSIMS m/z 860.5 (M+); HRMS Calcd for
C66H68: 860.5321. Found: 860.5323. Note: The peaks in the NMR
spectrum broaden on standing despite the precautions taken described
in the previous general section.
anti-5,13-Di-tert-butyl-8,16-dimethyl[1,2-b;9,10-b]dinaphtho[2.2]-
metacyclophane 38. The mixture of adducts 37 (68 mg, 0.12 mmol)
and Fe2(CO)9 (130 mg, 0.35 mmol) in benzene (30 mL) was stirred
under Ar at reflux for 8 h. The mixture was preabsorbed onto SiGel
and chromatographed over SiGel using hexanes as eluant, which yielded
38 mg (59%) of colorless 38 as needles from cyclohexane: mp 340-
Epoxytetrabenzohexacene 50. n-BuLi (0.64 mmol) in hexane (1.6
mL) was added slowly via syringe to a stirred solution of isofuran 1911b
(720 mg, 1.87 mmol) and dibromide 13 (298 mg, 0.54 mmol) in dry
toluene (15 mL) at -40 °C under argon in the dark. After 16 min,
methanol (3 mL) was added. Then, solvents were removed under
vacuum at ∼10 °C. The residue was chromatographed over neutral
alumina (deactivated with 5% water) using hexanes/benzene (2:1) as
eluant and gave 50 mg (12%) of the adduct 50 as a golden solid as an
equal mixture of two isomers. This mixture could be used directly in
the next step. However, further chromatography and recrystallization
from benzene yielded a single isomer (probably 50A; see text) as dark
1
341 °C (sublimes); H NMR (d8-THF) δ 8.22 (s, 4H, H-17,22,23,28),
7.97-7.94 (AA′XX′, 4H, H-18,21,24,27), 7.51-7.48 (AA′XX′, 4H,
H-19,20,25,26), 7.14 (s, 4H, H-4,6,12,14), 1.31 (s, 18H, t-Bu), 1.06
(s, 6H, CH3-8,16); 13C NMR (d8-THF) δ 150.65, 143.14, 140.32,
138.44, 134.75, 130.54, 128.54, 127.96, 126.84, 34.83, 31.66, 18.56;
UV (THF) λmax nm (ꢀmax) 261 (55 000), 287 (47 000), tail to 380 nm;
IR 890, 870, 750, 740 cm-1; CIMS m/z 545 (MH+). Anal. Calcd for
C42H40: C, 92.59; H, 7.40. Found: C, 92.12; H, 7.60.
Eluted next was unchanged 37, 25 mg (37%).
1
golden crystals: mp 222-225 °C; H NMR (d8-THF) δ 8.82 (d, J )
9,24;12,21-Diepoxy-2,7,14,19-tetra-tert-butyl-trans-20b,20c-trans-
24c,24d-tetramethyl-9,12,20b,20c,21,24,24c,24d-octahydro-di-(4,5-
pyreno)[b,k]chrysene 43. The bis-furan 39 was generated from 40 as
described in the literature:25 bis-acetal 40 (62 mg, 0.23 mmol) was added
to LDA (2 mmol) in THF (20 mL) at -78 °C and then was allowed to
warm to 0 °C for 10 min. The mixture was recooled to -78 °C and
was then quenched with methanol (0.5 mL). Solvents were removed
under vacuum at no more than 7 °C. The product was then taken up in
hexane/ethyl acetate (8:2, 10 mL) and was filtered through a 4 cm
column of alumina deactivated with 5% water, using the same mixed
solvent (90 mL) to elute the colorless product. Solvents were removed
under vacuum at 7 °C, and the product furan 39 was dried for an
additional hour under vacuum at room temperature, to give 35 mg (74%)
of colorless crystalline leaflets. To this was added dry THF (10 mL),
bromopyrene 41 (192 mg, 0.455 mmol), sodium amide (240 mg), and
potassium tert-butoxide (5 mg), and the mixture was stirred under argon
for 24 h. The reaction was then cooled in an ice bath and quenched by
slowly adding saturated aq ammonium chloride (5 mL). Ether was
1.3 Hz, 1H), 8.70 (d, J ) 1.2 Hz, 1H), 8.66-8.62 (m, 2H), 8.45 (bs,
1H), 8.41 (bs, 1H), 8.32 and 8.29 (AB, J ) 8.3 Hz, 2H), 8.16 (d, J )
1.3 Hz, 1H), 8.13 (d, 1H, J ) 1.3 Hz), 7.92 (d, J ) 1.2, 1H), 7.69 (d,
J ) 1.3, 1H), 7.50 (d, J ) 0.6 Hz, 1H), 7.47 (bs, 1H), 7.48-7.43 (m,
2H), 1.76, 1.71, 1.60, 1.53 (s, 9H each, C(CH3)3), -1.04, -1.99, -3.60,
-4.18 (s, 3H each, internal CH3); 13C NMR (d8-THF) δ 146.23, 146.14,
145.34, 145.14, 138.48, 137.80, 137.68, 137.60, 136.65, 136.59, 136.35,
136.16, 130.55, 130.40, 128.46, 127.94, 127.61, 127.21, 126.80, 126.78,
125.88, 125.31, 125.19, 125.01, 124.78, 122.07, 121.85, 117.57, 117.43,
117.12, 116.53, 116.16, 80.47, 80.27, 39.87, 36.90, 36.66, 36.61, 36.16,
36.06, 33.40, 32.16, 31.08, 30.87, 30.77, 30.71, 19.52, 17.53, 15.76,
14.62; IR 1261, 888, 865, 754, 673, 651, 645 cm-1; LSIMS m/z 776.4
(M+); HRMS calcd for C58H64O: 776.4957. Found: 776.4977. Note
in C6D6 the mixed isomers showed internal methyl protons (300 MHz)
at δ -0.63, -0.64, -1.67, -1.70, -3.17, -3.21, -3.87, -3.95.
5,13-Di-tert-butyl-8,16-dimethyl-1,2-[10,11-(2,7-di-tert-butyl-trans-
12c,12d-dihydro-12c,12d-dimethylbenzo[e]pyreno)]-9,10-benzo[2.2]-
9
2986 J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003

Dihydropyrene−Cyclophanediene System
A R T I C L E S
metacyclophane 47. (A) From the Single Isomer of 50A. A mixture
of the single isomer 50A from previously (40 mg, 0.052 mmol) and
Fe2(CO)9 (70 mg, 0.2 mmol) was stirred under reflux in benzene (30
mL) under argon in the dark for 2 h. After the mixture was cooled, the
mixture was filtered under argon through deactivated SiGel using
benzene (50 mL) as eluant. The solvent was removed under argon in
the dark. The residue was then chromatographed on deactivated SiGel
under argon using hexanes/CH2Cl2 (6:1) as eluant. The first ∼70% of
the red product band was collected. Evaporation of the solvent gave
24 mg (60%) of 47A as an intense red solid, which on recrystallization
Dibromobenzo-DHP-13: δ 8.80-8.74 (m), 8.42 (d), 8.20 (d), 7.62-
7.56 (m) 1.56 (s), -1.03 (s) (for assignments, see 13).
Dibromobenzo-CPD-14: δ 7.77 (d, J ) 2 Hz, H-6,12), 7.76-7.69
(m, H-17,20), 7.36-7.29 (m, H-18,19), 7.10 (d, H-4,14), 1.62 (s, CH3),
1.43 (s, (C(CH3)3)).
Phenyleno-DHP-15: δ (CDCl3, insufficiently soluble in d8-toluene)
9.92 (s), 8.92 (dd), 8.67 (dd), 8.40 (d), 7.78-7.74 (m), 7.71-7.67 (m),
7.23 (d), 6.99 (s), 1.55 (s), -1.07 (s) (for assignments, see preparation
from previous text).
Phenyleno-CPD-16: δ (CDCl3) 8.89 (s, H-18,20), 8.76-8.66 (m,
H-15,16), 7.72-7.64 (m, H-14,17), 6.98 (d, J ) 2 Hz, H-6,12), 6.88
(d, J ) 2 Hz, H-4,14), 6.51 (s), 1.51 (s, CH3-8,16), 1.30 (s, t-Bu-5,13).
Furano-DHP-19: δ 7.98 (s, H-9,11), 7.18 (d, J ) 1 Hz, H-1,8),
6.68 (d, J ) 1 Hz, H-3,6), 6.39 (s, H-4,5), 1.40 (s, (C(CH3)3)), 0.61
(s, CH3).
Furano-CPD-19′: δ 7.60 (s, H-17,18), 7.10 (d, J ) 2 Hz,
H-4,14), 7.04 (d, J ) 2 Hz, H-6,12), 6.55 (s, H-9,10), 1.45 (s, CH3),
1.42 (s, (C(CH3)3)).
Chryseno-bis-DHP-44: δ (see preparation from previous text for
aryl signals) 1.657, 1.640 (s, t-Bu) and -0.22 (s, internal-CH3)
Chryseno-bis-CPD-45 (TWO ISOMERS): δ 9.36, 9.35 (s), 8.84,
8.80 (d, J ) 9 Hz), 8.50, 8.47 (s), 8.01, 7.96 (d, J ) 9 Hz), 7.40-7.10
(m), 6.63 (s), 1.748, 1.740, 1.729, 1.716 (s, internal-CH3), 1.472, 1.461,
1.451, 1.440 (s, t-Bu).
Cyclophano-DHP-47: δ 9.73 (s), 8.86 (s), 7.98-7.92 (m), 7.62
(d), 7.61 (d), 7.48-7.42 (m), 7.32 (s), 7.27 (s), 1.638 (s, CH3), 1.630
(s, (C(CH3)3)), 1.37 (s, (C(CH3)3)), -1.01 (s, CH3) (for assignments,
see preparation from previous text).
Cyclophano-CPD-49: δ 8.65 (s, H-17,26), 7.94-7.88 (m, H-27,30),
7.47 (d, H-4,14), 7.46-7.40 (m, H-28,29), 7.40 (d, H-18,25), 7.25 (d,
H-6,12), 7.03 (d, H-20,23), 6.56 (s, H-21,22), 1.78 (s, CH3), 1.58 (s,
CH3), 1.34 (s, (C(CH3)3)), 1.32 (s, (C(CH3)3)) [J values ≈ 2 Hz].
Thermal Return Reactions. The NMR solutions previously men-
tioned were held in the probe of the NMR spectrometer at the
appropriate temperatures, and the reaction followed by integration of
the methyl or tert-butyl (or where appropriate sharp aromatic) signals,
such that -ln[MCD/(MCD + DHP)] could be plotted against time to
obtain the first-order rate constants, as described for 22 to 21 (data
available in Supporting Information). The first-order rate constants were
then used in an Arrhenius plot (ln(k) vs 1/T) to obtain Eact and an Eyring
plot (ln(k/T) vs 1/T to obtain ∆Hq and ∆Sq in the normal way.
1
from cyclohexane gave mp (dec) 170-175 °C; H NMR δ 8.94 (bs,
2H, H-17,26), 8.32 (bs, 2H, H-18,25), 7.79-7.74 (AA′XX′, 2H,
H-27,30), 7.48-7.43 (AA′XX′, 2H, H-28,29), 7.38 (bs, 2H, H-20,23),
7.15 (bs, 2H, H-21,22), 7.08 (d, J ) 2.1 Hz, 2H, H-4,14), 7.00 (d, J )
2.2 Hz, 2H, H-6,12), 1.50 (s, 18H, 19,24-C(CH3) 3), 1.31 (s, 18H, 5,13-
C(CH ) ), 1.18 (s, 6H, 8,16-CH3), -1.37 (s, 6H, 25b,c-CH3) (note
3
3
spectrum broadens on standing); 13C NMR (not satisfactory because
of decomposition); UV (deaerated cyclohexane) λmax (ꢀmax) nm 250
(56 800), 280 (52 300), 354 (19 900), 390 (29 400), 408 (40 500), 520
(4900), 640 (650); LSIMS m/z 760.5 (M+); HRMS calcd for C58H64:
760.5008. Found: 760.5004.
(B) From the Mixed Isomers 50. When the mixture of isomers
was used, almost identical material was obtained. The 1H NMR
spectrum of the second isomer 47B could be obtained by subtraction
of the spectrum of 47A from the spectrum of the mixture of isomers:
1H NMR of 47B (CDCl3, 300 MHz) δ 9.09 (bs, 2H), 8.52 (bs, 2H),
7.80-7.74 (AA′XX′, 2H), 7.49-7.43 (AA′XX′, 2H), 7.52 (bs, 2H),
7.31 (bs, 2H), 7.09 (d, J ) 2 Hz, 2H), 7.01 (d, J ) 2 Hz, 2H), 1.54 (s,
18 H), 1.32 (s, 18 H), 1.18 (s, 6H), -1.67(s, 6H) (see 47A for
assignments). Note: The spectrum broadens on standing despite the
precautions taken described in the previous general section.
Visible Light Irradiations of Dihydropyrenes to Form Cyclo-
phanedienes. A solution of the dihydropyrene (5 mg) in degassed d8-
toluene in an NMR tube, cooled by a small cold water bath, was
irradiated with a 250 W tungsten household lamp (the type used to
light a room during painting) with a square of yellow plastic (absorbs
light <400 nm) between the lamp and the sample, until the sample
solution was colorless (usually a few minutes). The sample was held
at 0 °C while being transferred to the NMR spectrometer (300 MHz),
where the 1H (unless otherwise specified) NMR spectra of the
cyclophanediene form were recorded.
Parent-DHP-5: δ 8.77 (s, H-1,3,6,8), 8.64 (s, H-4,5,9,10), 1.78 (s,
C(CH3)3), -3.61 (s, CH3).
k (min-1) (3σ errors based on 10 integrations)
compd
319 K
333 K
343 K
Parent-CPD-6: δ 6.87 (s, Ar-H), 6.44 (s, vinyl-H), 1.82 (s, CH3),
1.40 (s, C(CH3)3).
12
10
6
45
16
49
8
0.0344(9)
0.0101(3)
0.178(5)
0.042(1)
0.270(7)
0.117(3)
0.086(2)
0.090(2)
0.0431(8)
0.0356(9)
0.0299(8)
0.0212(5)
0.003 22(8)
0.006 10(15)
0.0057(2)
0.003 07(7)
0.002 24(6)
0.001 98(5)
0.001 24(3)
0.000 183(5)
0.0300(8)
0.0297(8)
0.0149(8)
0.0134(3)
0.0095(2)
0.0063(2)
0.001 08(3)
Benzo-DHP-7: δ 9.01-8.94 (m, H-9,12), 8.56 (d, J ) 1 Hz, H-1,8),
7.70-7.63 (m, H-10,11), 7.59 (d, J ) 1 Hz, H-3,6), 1.62 (s, (C(CH3)3)),
-1.11 (s, CH3).
Benzo-CPD-8: δ 7.88-7.80 (m, H-17,20), 7.41-7.33 (m, H-18,19),
7.13 (d, J ) 2 Hz, H-4,14), 6.99 (d, J ) 2 Hz, H-6,12), 6.50 (s, H-9,10),
1.62 (s, CH3), 1.40 (s, C(CH3)3).
14
19′
Eact
kcal/mol
∆Hq
kcal/mol
∆Sq
cal/deg mol
Naphtho-DHP-9: δ 9.21(s, H-9,14), 8.28 (d, J ) 1 Hz, H-1,8),
7.97-7.91 (m, H-10,13), 7.47-7.41 (m, H-11,12), 7.08 (d, J ) 1 Hz,
H-3,6), 6.79 (s, H-4,5), 1.50 (s, (C(CH3)3)), -0.06 (CH3).
Naphtho-CPD-10: δ 8.31 (s, H-17,22), 7.86-7.79 (m, H-18,21),
7.47-7.40 (m, H-19,20), 7.25 (d, J ) 2 Hz, H-4,14), 7.07 (d, J ) 2
Hz, H-6,12), 6.60 (s, H-9,10), 1.66 (s, CH3), 1.45 (s, (C(CH3)3)).
compd
12
10
6
45
16
49
8
19.1
22.1
24.0
24.3
23.9
24.1
24.5
25.8
26.1
18.4
21.4
23.3
22.6
23.3
23.4
23.9
25.0
26.3
-7.4
-0.7
4.3
5.1
2.7
2.9
3.8
6.2
4.4
Anthro-DHP-11: δ 9.30 (s, H-9,16), 8.47 (s, H-10,15), 8.21 (d,
J ) 1 Hz, H-1,8), 8.08-8.03 (m, H-11,14), 7.48-7.42 (m, H-12,13),
6.95 (d, J ) 1 Hz, H-3,6), 6.62 (s, H-4,5), 1.56 (s, (C(CH3)3)), 0.50
(s, CH3).
14
19′
((0.5-0.8)
((0.5-0.8)
((1)
Anthro-CPD-12: δ 8.47 (s, H-17,24), 8.30 (s, H-18,23), 7.99-
7.94 (m, H-19,22), 7.45-7.40 (m, H-20,21), 7.37 (d, J ) 2 Hz, H-4,14),
7.12 (d, J ) 2 Hz, H-6,12), 6.62 (s, H-9,10), 1.63 (s, CH3), 1.45 (s,
(C(CH3)3)).
Laser Flash Studies of 47. A sample of 47 was dissolved in
cyclohexane and adjusted to an absorbance of ∼0.1 at 408 nm using 7
mm fluorescence cells. The sample was deaerated for 10 min with N2
9
J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003 2987

A R T I C L E S
Mitchell et al.
and capped and was held at 20 ( 1 °C. Laser flash photolysis was
conducted using a YAG laser (Spectra Physics GCR-12) at 355 nm
with pulse energies of ∼40 mJ/pulse, with a repetition rate of 1 Hz.
Transient spectra were obtained using a xenon lamp (150 W, Oriel
housing model 66057, PTI power supply, model LPS-220) connected
to a pulser synchronized with the laser pulses. A photomultiplier
(Hamamatsu R446) coupled with a monochromator (CVI Digikrom
240) and a baseline compensation unit was used to monitor the transient
absorption spectra. The operating system is described in detail in ref
34. The transient spectrum showed bleaching of the 406 nm band of
47 and a large increase in absorption at 486 nm for a transient species,
which appears very rapidly, about 10 ns after the laser pulse. The
transient spectra are shown in the Supporting Information). The decay
pattern of the transient was monitored at 486 nm, and it appeared to
be first order, with a half-life of about 0.10 s at 20 °C. The transient
showed an absorbance change of +0.23 at 486 nm, while the 406 nm
band of 47 changed by -0.055. There was no change to the UV-
visible spectrum obtained at the end of the laser flash study, even after
hundreds of flashes.
Acknowledgment. R.H.M. thanks the Natural Sciences and
Engineering Research Council of Canada and the University
of Victoria, and M.J.M. thanks DuPont, UCR, and GAANN
for the support of this work.
Supporting Information Available: An improved large scale
preparation of the dihydropyrene 5, the first-order decay plots
for the thermal isomerizations of the cyclophanedienes, and the
laser flash transient spectra of 47. This material is available free
of charge via the Internet at http://pubs.acs.org.
JA0288136
9
2988 J. AM. CHEM. SOC. VOL. 125, NO. 10, 2003