Chemoselective photorearrangements of diazinobarrelenes. Deuterium labeling study

Article abstract of DOI:10.1021/jo7016795

(Chemical Equation Presented) Chemoselective photorearrangements of pyrazino-, quinoxalino-, and benzoquinoxalinobarrelenes were investigated by deuterium-labeling experiment. Photolysis of pyrazinobarrelene 4 and quinoxalinobarrelene 5 with 300 nm region

Full text of DOI:10.1021/jo7016795

Chemoselective Photorearrangements of Diazinobarrelenes.
Deuterium Labeling Study
†
Ann-Cheng Chen, Gary J. Chuang, Nelson R. Villarante, and Chun-Chen Liao*
Department of Chemistry, National Tsing Hua UniVersity, Hsinchu 300, Taiwan
ccliao@mx.nthu.edu.tw
ReceiVed August 7, 2007
Chemoselective photorearrangements of pyrazino-, quinoxalino-, and benzoquinoxalinobarrelenes were
investigated by deuterium-labeling experiment. Photolysis of pyrazinobarrelene 4 and quinoxalinobarrelene
5
with 300 nm region light under either direct or sensitized conditions afforded semibullvalenes 7 and 8,
respectively; benzoquinoxalinobarrelene 6 was inert to both reaction conditions. Irradiation of deuterated
pyrazinobarrelene 4-d and quinoxalinobarrelene 5-d afforded deuterated semibullvalenes 7-d -A-7-
-F and 8-d -A-8-d -F, respectively. Of the two a priori possibilities of bridging, the deuterium-labeling
experiment has shown that deuterated pyrazinobarrelene 4-d afforded 98% (C ) and 95% (CD CN)
of semibullvalenes generated through aryl-vinyl (A-V) bridging, whereas the quinoxalinobarrelene 5-d
furnished 79% (C ) and 71% (CD
4
4
4
d
4
4
4
4
D
6 6
3
4
6
D
6
3
CN) of semibullvalenes generated through vinyl-vinyl (V-V)
bridging. The contrasting photochemical behavior of heteroarene-fused barrelenes 4-6 was explained
qualitatively in terms of triplet energy minimization and relative stability of diradicaloid intermediates.
Introduction
Bicyclo[2.2.2]octa-2,5,7-triene (1), also named barrelene
1
1a
1
b
because of its concentric π-orbital orientation (Figure 1), is
one of the most fascinating bridged-ring systems in photochemi-
cal reactions.1 Forty years ago, Zimmerman and Grunewald
c,e,f
2
reported a most intriguing photochemical behavior of 1 which
FIGURE 1. Molecular orbital representation of barrelene.
has led to extensive investigations of several barrelene deriva-
3
4
tives and other structural frameworks that contain the required
The initial investigations by Zimmerman et al. on barrelene
1 had shown that the direct photolysis of 1 (full arc, quartz
5
chromophoric groups for di-π-methane (DPM) rearrangement.
*
Corresponding author. Fax: 886-3-5728123.
Current address: Department of Physical Sciences and Mathematics, College
(3) (a) Ramaiah, D.; Sajimon, M. C.; Joseph, J.; George, M. V. Chem.
Soc. ReV. 2005, 34, 48. (b) Liao, C.-C.; Peddinti, R. K. In CRC Handbook
of Organic Photochemistry and Photobiology, 2nd ed.; Horspool, W. M.,
Lenci, F., Eds.; CRC Press: Boca Raton, FL, 2004; p 32/1. (c) Altundas,
R.; Dastan, A.; Unaldi, N. S.; G u¨ ven, K.; Uzun, O.; Balci, M. Eur. J. Org.
Chem. 2002, 3, 526. (d) Sajimon, M. C.; Ramaiah, D.; Kumar, S. A.; Rath,
N. P.; George, M. V. Tetrahedron 2000, 56, 5421. (e) Chen, J.; Pokkuluri,
P. R.; Scheffer, J. R.; Trotter, J. Acta Crystallogr. 1993, B49, 905. (f) Liao,
C.-C.; Lin, S.-Y.; Hsieh, H.-P.; Yang, P.-H. J. Chin. Chem. Soc. 1992, 39,
275. (g) Liao, C.-C.; Yang, P.-H. J. Chem. Soc., Chem. Commun. 1991,
626. (h) Liao, C.-C.; Hsieh, H.-P.; Lin, S.-Y. J. Chem Soc., Chem. Commun.
1990, 545.
†
of Arts and Sciences, University of Philippines, Manila.
1) (a) Zimmerman, H. E.; Paufler, R. M. J. Am. Chem. Soc. 1960, 82,
514. (b) Hassenruck, K.; Martin, H.-D. Chem. ReV. 1989, 89, 1125. (c)
(
1
Gedanken, A.; de Meijere, A. J. Chem. Phys. 1988, 88, 4153. (d) Yamamoto,
S.; Nakata, M.; Fukuyama, T.; Kuchitsu, K.; Hasselmann, D.; Ermer, O. J.
Phys. Chem. 1982, 86, 529. (e) Zimmerman, H. E.; Boettcher, R. J.; Buehler,
N. E.; Keck, G. E. J. Am. Chem. Soc. 1975, 97, 5635. (f) Hoffman, R.;
Heilbronner, E.; Gleiter, R. J. Am. Chem. Soc. 1970, 92, 706.
(2) Zimmerman, H. E.; Grunewald, G. L. J. Am. Chem. Soc. 1966, 88,
1
83.
10.1021/jo7016795 CCC: $37.00 © 2007 American Chemical Society
9
690
J. Org. Chem. 2007, 72, 9690-9697
Published on Web 11/15/2007

ChemoselectiVe Photorearrangements of Diazinobarrelenes
SCHEME 1. Photorearrangement of Barrelene 1
SCHEME 3. Photorearrangement of Deuterated
Benzobarrelene 11-d
SCHEME 2. Photorearrangement of Deuterated Barrelene
1-d
8
theoretical studies. The mechanism of barrelene rearrangement
has become the basis for other workers to predict the pathways
of other rearrangement processes.
tube, methylcyclohexane solution, room temperature) affords
9
,10
cyclooctatetraene 2, as supported by IR, NMR, and VPC
The deuterium-labeling technique was also employed by
Zimmerman et al. in predicting the chemoselective photorear-
rangements of homoarene-fused barrelenes, such as benzobar-
2
analyses (Scheme 1). The reaction was perceived to occur
initially via intramolecular [2 + 2]-cycloaddition of the excited
singlet state of barrelene. On the other hand, the acetone-
sensitized reaction gives semibullvalene 3.
11
12
13
relene, naphthobarrelene, and anthrabarrelene. Compared
with barrelene 1, these barrelene analogues can exhibit two types
of bridging modes, vinyl-vinyl (V-V) and aryl-vinyl (A-
V) bridgings. As shown in Scheme 3, the direct irradiation of
deuterated benzobarrelene 11-d affords cyclooctatetraenes 14-
dA (major) and 14-dB (minor) derived from A-V bridging,
whereas the sensitized reaction gives semibullvalene 15-d
generated from the V-V bridging mode (Scheme 3).11
Zimmerman considered two reaction pathways for the
sensitized irradiation of barrelene 1. The first mechanism
involved the participation of the three vinylic moieties during
the bridging process, and the second involved a stepwise
6
bridging of the two vinylic moieties. They elegantly resolved
the problem of discriminating these two mechanisms by
substituting all the vinylic protons with deuterium atoms and
then determined the distribution of hydrogen in the deuterated
The observed chemoselectivity of 11-d under sensitized
reaction conditions was rationalized by considering the energy
of the triplet state during the initial mode of bridging. Thus,
they hypothesized that if the benzene moiety in 11-d was fused
with another aromatic ring, bridging would probably occur
between the aryl and the vinyl groups as polycyclic aromatic
groups are expected to have lower triplet energies than ethylene.
This prompted them to study the photochemical behavior of
deuterated naphthobarrelene 12-d and anthrabarrelene 13-d.
Surprisingly, these two barrelenes underwent V-V bridging
(Scheme 4). In both reaction conditions (direct and sensitized),
1
products by H NMR and mass spectral analyses. A good
correlation was noticed between the calculated hydrogen
distributions and the experimentally determined values, thus
supporting the second mechanism which was thought to involve
stepwise bridging of the two vinylic moieties to generate the
triplet diradical 9 followed by cyclopropyl carbon fission to
afford the symmetrical allylic diradicals 10 (Scheme 2). Closing
of the diradicals 10 gives semibullvalenes 3-dA and 3-dB. The
considered mechanism was supported by the independent
1
2-d afforded the DPM photoproduct 16-d. However, anthra-
7
barrelene 13-d afforded semibullvalene 17-d under direct
generation of biradicals 10 using azo precursor and through
(
8) (a) Frutos, L. M.; Sancho, U.; Casta n˜ o, O. Org. Lett. 2004, 6, 1229.
(
4) (a) Dura, R. D.; Paquette, L. A. J. Org. Chem. 2006, 71, 2456. (b)
(b) Zimmerman, H. E.; Kutateladze, A. G.; Maekawa, Y.; Mangette, F. J.
Am. Chem. Soc. 1994, 116, 9795. (c) Zimmerman, H. E.; Sulzbach, H. M.;
Tollefson, M. B. J. Am. Chem. Soc. 1993, 115, 6548.
McClure, C. K.; Kiessling, A. J.; Link, J. S. Org. Lett. 2003, 5, 3811. (c)
Janssen, R. A. J.; Hummelen, J. C.; Wudl, F. J. Am. Chem. Soc. 1995, 117,
5
44. (d) Givens, R. S.; Oettle, W. F. J. Am. Chem. Soc. 1971, 93, 3963. (e)
(9) (a) Armesto, D.; Caballero, O.; Ortiz, M. J.; Agarrabeitia, A. R.;
Martin-Fontecha, M.; Torres, M. R. J. Org. Chem. 2003, 68, 6661. (b)
Armesto, D.; Gallego, M. G.; Horspool, W. M.; Agarrabeitia, A. R.
Tetrahedron 1995, 51, 9223. (c) Armesto, D.; Ortiz, M. J.; Ramos, A.;
Horspool, W. M.; Mayoral, E. P. J. Org. Chem. 1994, 59, 8115.
(10) (a) Singh, V. In CRC Handbook of Organic Photochemistry and
Photobiology, 2nd ed.; Horspool, W. M., Lenci, F., Eds.; CRC Press: Boca
Raton, FL, 2004; p 78/1. (b) Liao, C.-C. In CRC Handbook of Organic
Photochemistry and Photobiology; Horspool, W. M., Soon, P. S., Eds.; CRC
Press: Boca Raton, FL, 1995; p 194. (c) Demuth, M. Org. Photochem.
1991, 11, 37.
Prinzbach, H. Pure Appl. Chem. 1968, 16, 17. (f) Ciganek, E. J. Am. Chem.
Soc. 1966, 88, 2882.
(5) (a) Zimmerman, H. E.; Cirkva, V. Org. Lett. 2000, 2, 2365. (b)
Zimmerman, H. E.; Armesto, D. Chem. ReV. 1996, 96, 3065. (c) Armesto,
D. In Handbook of Photochemistry and Photobiology; Horspool, W. M.,
Song, P.-S., Eds.; CRC Press: Boca Raton, FL, 1995; Chapter 73. (d)
Zimmerman, H. E. Org. Photochem. 1991, 11, 1. (e) Zimmerman, H. E. In
Rearrangements in Ground and Excited States; de Mayo, P., Ed.; Academic
Press: New York, 1980; Vol. 3, p 131. (f) Hixson, S. S.; Mariano, P. S.;
Zimmerman, H. E. Chem. ReV. 1973, 73, 531.
(6) (a) Zimmerman, H. E.; Binkley, R. W.; Givens, R. S.; Grunewald,
(11) (a) Zimmerman, H. E.; Givens, R. S.; Pagni, R. M.; J. Am. Chem.
Soc. 1968, 90, 6096. (b) Zimmerman, H. E.; Givens, R. S.; Pagni, R. M. J.
Am. Chem. Soc. 1968, 90, 4191.
G. L.; Sherwin, M. A. J. Am. Chem. Soc. 1969, 91, 3316. (b) Zimmerman,
H. E.; Binkley, R. W.; Givens, R. S.; Sherwin, M. A. J. Am. Chem. Soc.
1
967, 89, 3932.
7) Zimmerman, H. E.; Boettcher, R. J.; Buehler, N. E.; Keck, G. E.;
Steinmetz, M. G. J. Am. Chem. Soc. 1976, 98, 7680.
(12) Zimmerman, H. E.; Bender, C. O. J. Am. Chem. Soc. 1970, 92,
4366.
(13) Zimmerman, H. E.; Amick, D. R. J. Am. Chem. Soc. 1973, 95, 3977.
(
J. Org. Chem, Vol. 72, No. 25, 2007 9691

Chen et al.
SCHEME 5. Synthesis of Heteroarene-Fused Barrelenesa
SCHEME 4. Photorearrangement of Deuterated
Anthrabarrelene 12-d and 13-d
a
Reaction conditions: (a) 220 °C, phenyl vinylsulfide; (b) mCPBA,
78 °C; (c) Et3N, 130 °C, sealed tube (70% over 3 steps); (d) H2SO4,
5 °C (84%).
-
6
SCHEME 6. Photolysis of Heteroarene-Fused Barrelenes 4,
, and 6
5
irradiation, but no photoproduct was observed in the sensitized
reaction conditions.12,13 One interesting aspect about the pho-
tochemistry of arene-fused barrelenes is the preference for V-V
bridging despite a priori possibilities for A-V bridging.
Encouraged by these impressive mechanistic explorations of
6,7,11
barrelene and homoarene-fused barrelenes
by Zimmerman
and his co-workers, we decided to embark on the photochemical
investigation of heteroarene-fused barrelenes 4-6 to find out
if similarities and differences exist among these barrelene
systems in terms of chemoselectivity as compared to that of
the homoarene-fused systems and also to ascertain if the same
type of rearrangement will occur as in the case of barrelene 1.
In principle, the rearrangement products of the title com-
the corresponding 1,2-diamine, heteroarene-fused barrelenes
could be synthesized facilely (Scheme 5). Detailed experimental
1
4
15
procedures for 19, 20, 4-8, and deuterated compounds 19-
d4, 20-d4, 4-d4, and 5-d4 along with spectral data assignment
are available in Supporting Information.
pounds can be accessed by two types of bridging interaction,
A-V and V-V bridging modes. The deuterium-labeling tech-
nique has provided us the opportunity to identify the primary
bridging mode under direct and sensitized reaction conditions.
We report herein the syntheses and photorearrangements of
diazinobarrelenes 4-6.
Photolysis of Heteroarene-Fused Barrelenes 4-6. Irradia-
tion of pyrazinobarrelene 4 either in benzene, acetonitrile, or
acetone solvent afforded semibullvalene 7 in 96-99% yield
(Scheme 6). Under similar reaction conditions, quinoxalinobar-
relene 5 afforded semibullvalene 8 in near quantitative yield;
however, for benzoquinoxalinobarrelene 6, no characterizable
photoproducts were detected; the starting material was recovered
after 2 h of irradiation.
Results
Synthesis of Heteroarene-Fused Barrelenes 4-6. Bicyclic
dienedione 20 was synthesized using masked o-benzoquinone
(
MOB) dimer 18 as the starting material. Employing the method
The structure of 7 was established as semibullvalene based
on spectral analyses (Table 1). The chemical shifts in the H
14,15
1
developed in our laboratory.
Followed by condensations with
NMR spectrum which showed two doublets of doublets at δ
(
14) (a) Chittimalla, S. K.; Shiao, H.-Y.; Liao, C.-C. Org. Biomol. Chem.
5
.65 (J ) 2.4, 5.2 Hz, 1H) and δ 5.59 (J ) 2.4, 5.2 Hz, 1H)
2
8
006, 4, 2267. (b) Liao, C.-C.; Peddinti, R. K. Acc. Chem. Res. 2002, 35,
56. (c) Gao, S.-Y. Ph.D. Thesis, Department of Chemistry, National Tsing-
were assigned to the vinylic protons. The chemical shifts at δ
3.28 (dt, J ) 2.4, 6.4 Hz, 1H), δ 3.67 (q, J ) 6.4 Hz, 1H), and
δ 3.31 (t, J ) 6.4 Hz, 1H) were assigned to the cyclopropyl
protons. The observed multiplicity at δ 3.67 is characteristic of
Hua University, Taiwan, 2005.
15) (a) Scharf, H.-D.; Klar, R. Chem. Ber. 1972, 105, 575. (b) Scharf,
(
H.-D.; Klar, R. Tetrahedron Lett. 1971, 12, 517. (c) McMahon, R. J.; Abelt,
C. J.; Chapman, O. L.; Johnson, J. W. Krei1, C. L.; LeRoux, J.-P.; Mooring,
A. M.; Wede, P. R. J. Am. Chem. Soc. 1987, 109, 2456.
2
,8
16
tricyclo[3.3.0.0 ]octene systems. In the case of quinoxali-
9692 J. Org. Chem., Vol. 72, No. 25, 2007

ChemoselectiVe Photorearrangements of Diazinobarrelenes
irradiation of 4-d4 in perdeuterated acetonitrile; however,
overlapping peaks were observed for the vinylic protons at H-6
and H-7. Irradiation of 5-d4 in perdeuterated benzene showed
spectral profiles similar to Figure 3. One interesting observation
of their spectra is the appearance of proton peaks ascribable to
all the hydrogens of the tricyclooctene moiety.
Discussion
Multiplicities in the Photochemistry of Heteroarene-Fused
Barrelenes. According to the accepted view of DPM photore-
arrangements of bicyclic systems, photoreactions proceed from
FIGURE 2. 1H NMR (400 MHz) spectrum of tricyclooctene moieties
of photoproducts 7-d -A-7-d -F.
5
4
4
the triplet excited states, thus we presume that the same
multiplicity is involved in the rearrangements of barrelenes 4
and 5 since similar rearrangements and product distributions
were observed in both reaction conditions (direct and sensitized).
In addition, the nitrogen-containing aromatic group present in
these heterobarrelene systems which can be excited via n,π*
and π,π* triplet states has greatly influenced the observed
17
multiplicity through spin-orbit interaction. Furthermore, pyra-
1
8b
17d
zine (φISC ) 1.0) and quinoxaline (φISC ) 0.99) moieties,
which are presented in these barrelene systems, were shown to
have very high intersystem crossing quantum yields relative to
FIGURE 3. 1H NMR spectrum of tricyclooctene moieties of photo-
18
their homoarene counterpart. It is important to note that, for
products 8-d
4
-A-8-d
4
-F.
benzoquinoxalinobarrelene 6, no photoproduct was observed in
both reaction conditions.
Bridging Mode of Photochemical Rearrangement. It is of
considerable interest to identify the pathways taken by barrelenes
1
TABLE 1. Pertinent H NMR Data for Nondeuterated
Semibullvalenes 7 and 8
4 and 5 during the photochemical event. As shown in Scheme
7, there are five possible semibullvalenes (7-d4-A ≡ 7-d4-E;
8-d4-A ≡ 8-d4-E) that can be generated from three routes; two
of these routes (I and II) involve A-V bridging, and the other
one (route III) involves V-V bridging. For the purpose of
hydrogen distribution accounting, we label the identical pyrazi-
nosemibullvalenes as 7-d4-A and 7-d4-E, and the quinoxali-
nobarrelenes as 8-d4-A and 8-d4-E. Route I involves primary
aryl-vinyl bridging at C3-C4, generating the diradicaloid
intermediate I-A; breaking of the cyclopropyl moiety generates
the secondary diradicaloid intermediates having the canonical
structures I-B1 and I-B2. Ring closure on both sides of the allylic
radicals gives the photoproducts 7-d4-A and 7-d4-B or 8-d4-A
and 8-d4-B. In the case of route II, initial A-V bridging at
C3-C5 gives rise to intermediates II-A; further rearrangement
generates intermediates of canonical structures II-B1 and II-B2
which ultimately furnish semibullvalenes 7-d4-C and 7-d4-D or
proton chemical shifts in nonaromatic
a
positions (ppm, CDCl3)
semibullvalene
H-1
q
H-2
H-5
H-6
H-7
H-8
t
dd
4.16
dd
dd
5.65
dd
dd
5.59
dd
dt
3.28
dt
7
8
3.67
q
3.31
t
3.51
3.16
4.22
5.65
5.54
2.98
a
Starting material in deuterated benzene was degassed for 1 h prior to
irradiation with 300 nm region light for 2 h.
8
-d4-C and 8-d4-D. Finally, route III commences through V-V
nosemibullvalene 8, the proton spectral features at the fused
cyclopentanoid moiety proved almost directly superimposable
upon that of pyrazinosemibullvalene 7.
bridging at C5-C4 to generate the diradicaloid intermediate
III-A. Breaking of the cyclopropyl ring of III-A at the C5-C8
bond affords intermediate III-B which upon closing of the
diradical furnishes 7-d4-E or 8-d4-E. Bond fission at C4-C8
in III-A affords intermediate III-C, and closure of this diradical
Photolysis of Deuterated Pyrazinobarrelene 4-d4 and
Quinoxalinobarrelene 5-d4. Photoreactions of 4-d4 and 5-d4
1
were monitored by H NMR and were stopped when the starting
1
4 4
furnishes the semibullvalene 7-d -F or 8-d -F.
material had disappeared. The H NMR spectra in Figures 2
and 3 illustrate the spectral profiles of the tricyclic octene
moieties of the deuterated semibullvalenes furnished after
irradiating barrelenes 4-d4 in perdeuterated benzene and 5-d4
in perdeuterated acetonitrile with 300 nm region light. Similar
spectral profiles as that of Figure 2 were obtained for the
(
17) (a) Malkin, Y. N.; Kuz’min, V. A. Russ. Chem. ReV. 1990, 59, 279.
(
b) Leinwand, D. A.; Lefkowitz, S. M.; Brenner, H. C. J. Am. Chem. Soc.
1985, 107, 6179. (c) Yamazaki, I.; Murao, T.; Yamanaka, T.; Yoshihara,
K. Faraday Discuss. Chem. Soc. 1983, 75, 395. (d) Scott, G. W.; Boldridge,
D. W.; Spiglanin, T. A. J. Phys. Chem. 1982, 86, 1976. (e) Madej, S. L.;
Gillispie, G. D.; Lim, E. C. Chem. Phys. 1978, 32, 1. (f) Bent, D. V.; Hayon,
E. P.; Moorthy, N. J. Am. Chem. Soc. 1975, 97, 5065.
(
16) (a) Bender, C. O.; Dolman, D.; Foesier, J. C.; Lawson, S. L.; Preuss,
(18) (a) Pan, Y.; Sheng, Z.; Ye, X.; Ao, Z.; Chu, G.; Dai, J.; Yu, S. J.
Photochem. Photobiol. A: Chem. 2005, 174, 98. (b) Shim, S. C.; Kim, M.
S.; Lee, K. T.; Jeong, B. M.; Lee, B. H. J. Photochem. Photobiol. A: Chem.
1992, 65, 121. (c) Bartocci, G.; Massetti, F.; Mazzucato, U.; Marconi, G.
J. Chem. Soc., Faraday Trans. 2 1984, 80, 1093. (d) Yajima, Y.; Lim, E.
C. Chem. Phys. Lett. 1980, 73, 249. (e) Fischer, G. Chem. Phys. Lett. 1975,
33, 459.
K. E. Can. J. Chem. 2003, 81, 37. (b) Bender, C. O.; Bengton, D. L.;
Dolman, D.; Herle, C. E. L.; O’Shea, S. F. Can. J. Chem. 1982, 60, 1942.
c) Luibrand, R. T.; Broline, B. M.; Charles, K. A.; Drues, R. W. J. Org.
Chem. 1981, 46, 1874. (d) Luibrand, R. T.; Fujinari, E. M. J. Org. Chem.
980, 45, 958. (e) Zimmerman, H. E.; Viriot-Villaume, M.-L. J. Am. Chem.
Soc. 1973, 95, 1274.
(
1
J. Org. Chem, Vol. 72, No. 25, 2007 9693

Chen et al.
4 4
SCHEME 7. Conceivable Bridging Modes and Photoproducts of 4-d and 5-d
Perusal of the spectral profiles of photoproducts (Figures 2
and 3, vide supra) suggests that a number of semibullvalenes
were produced from 4-d4 and 5-d4. This deduction is based on
TABLE 2. Calculated Hydrogen Distributions in Semibullvalenes
7
-d4-A-7-d4-F and 8-d4-A-8-d4-F
calculated hydrogen distributionsa
the following observations: (1) proton peaks appear in all the
semibullvalene
H-1
H-2
H-5
H-6
H-7
H-8
nonaromatic positions of the photoproducts; and (2) from the
mass spectral analysis (see Experimental Section) of the starting
materials, only 6% residual hydrogen was observed at the
7
7
-d4-A or 8-d4-A
-d4-B or 8-d4-B
0.015 0.015 0.015 1.00
0.015 0.015 0.015 0.015 1.00
1.00
0.015
1.00
7-d -C or 8-d -C 1.00
7-d4-D or 8-d4-D
7
7
4
4
1.00
0.015 0.015 0.015 0.015
deuterated sites; however, as depicted in the spectral profiles
of 7-d4 and 8-d4, the amount of hydrogen in most carbons of
the nonaromatic moiety of the photoproducts is relatively high.
For instance, the lowest proton intensity (H-2, 1.31) for the
photoproducts of 4-d4 in benzene (Figure 2, vide supra) when
integrated relative to the other proton peaks is about 13.1% H,
which is way above the 6% residual hydrogen of the deuterated
sites.
1.00
0.015 0.015 0.015 1.00
1.00 0.015 0.015 0.015 0.015 1.00
0.015 1.00
0.015 0.015 0.015
-d4-E or 8-d4-E
-d4-F or 8-d4-F
1.00 0.015
a
Calculated hydrogen distributions were normalized to 2.06 H to account
for the 6% residual hydrogen as determined by mass spectral analysis.
TABLE 3. Observed Hydrogen Distributions in Photoproducts of
4-d4
hydrogen distributions in photoproducts of 4-d4a
A question arises as to which and what ratio of the
photoproducts in Scheme 7 could account for the observed
hydrogen distributions that are depicted in Figures 2 and 3. To
solve this problem, first, we determined the theoretical hydrogen
distribution of each semibullvalene. Considering the 6% residual
hydrogen from the deuterated sites and the two protons from
the nondeuterated sites of the photoproducts, each semi-
bullvalene will have a total of 2.06 H. Neglecting secondary
isotope effect, each deuterated carbon of semibullvalene will
have 0.015 H, whereas the nondeuterated carbon will have 1.0
H. The result of this calculation is summarized in Table 2.
Theoretically, the distribution of hydrogen at the nonaromatic
moiety of pyrazinosemibullvalenes 7-d4-A-7-d4-F and 8-d4-
A-8-d4-F should be the same.
solvent
H-1
H-2
H-5
H-6
H-7
H-8
C6D6
CD3CN
0.511
0.531
0.270
0.268
0.276
0.268
0.260
0.260
0.474
0.472
0.268
0.262
a
Hydrogen distributions calculated from 1H NMR integrations (400
MHz) were normalized to 2.06 H to account for the residual hydrogen as
determined by mass spectral analysis.
residual hydrogen (see Supporting Information). These are
summarized in Tables 3 and 4. As shown in these tables, the
observed hydrogen distributions for the photoproducts 7-d4 and
8-d4 in the two solvents are very close, suggesting that there is
no profound effect of solvent on the observed hydrogen
distribution. If we correlate the observed hydrogen distributions
with that of the calculated values of the individual semi-
bullvalene in Table 1, one would notice a very poor correlation.
Next, we normalized the observed hydrogen distributions in
the photoproducts 7-d4 and 8-d4 to 2.06 to account for the
9694 J. Org. Chem., Vol. 72, No. 25, 2007

ChemoselectiVe Photorearrangements of Diazinobarrelenes
TABLE 4. Observed Hydrogen Distributions in Photoproducts of
derived from A-V bridging are 89 and 92%, whereas the yields
derived from V-V bridging are 2 and 4%. In contrast to that
of 4-d4, the relative yields for the photoproducts of 5-d4 derived
from A-V bridging are 20 and 30%, whereas the yields derived
from V-V bridging are 77 and 75%. Notably, we encountered
a significant experimental error for the values obtained from
5
-d4
Hydrogen distributions in photoproducts of 5-d4a
solvent
H-1
H-2
H-5
H-6
H-7
H-8
C6D6
CD3CN
0.490
0.509
0.076
0.066
0.082
0.082
0.457
0.443
0.494
0.523
0.459
0.437
5
-d in deuterated acetone. This may be attributed to the fact
a
Hydrogen distributions calculated from 1H NMR integrations (400
4
that under sensitized conditions the photoproducts may also
absorb energy within the wavelength region of the sensitizer.
Nevertheless, the A-V/V-V ratios of the photoproducts clearly
show the contrasting bridging preferences of barrelenes 4-d4
and 5-d4.
MHz) were normalized to 2.06 H to account for the residual hydrogen as
determined by mass spectral analysis.
TABLE 5. Bridging Modes and Yields of Semibullvalenes
7
-d4-A-7-d4-F and 8-d4-A-8-d4-F under Direct Irradiation
c
Factors Controlling the Rearrangement. Two striking
results can be noticed regarding the photochemical behavior of
barrelenes 4-6: First, the observation that only semibullvalenes
were furnished in both reaction conditions (direct and sensitized)
in the cases of 4 and 5 and the insensitivity of 6 to photochemi-
cal reaction, and second, the contrasting bridging mode of the
two barrelenes as construed from the deuterium-labeling experi-
ment.
yield (%)
semibullvalene
a,b
(bridging Mode)
C6D6
CD3CN
7
7
7
7
7
7
8
8
8
8
8
8
-d4-A (A-V)b
-d4-B (A-V)
-d4-C (A-V)
-d4-D (A-V)
21.1 (23.2)
21.7 (23.8)
22.9 (25.2)
23.5 (25.8)
0.8 (0.9)
1.0 (1.1)
3.4 (3.5)
3.7 (3.8)
6.2 (6.4)
21.5 (22.3)
21.5 (22.3)
24.4 (25.4)
24.4 (25.4)
2.1 (2.2)
b
-d4-E (V-V)
-d4-F (V-V)
-d4-A (A-V)
-d4-B (A-V)
-d4-C (A-V)
-d4-D (A-V)
-d4-E (V-V)
-d4-F (V-V)
2.3 (2.4)
6.5 (6.2)
In the first case, successful sensitization demonstrates inter-
vention of a triplet reactant, which is in accord with the
rearrangements of several aromatic barrelene derivatives;3
however, we also obtained the same photoproducts with similar
product distributions under direct irradiation. As discussed
previously, this kind of photochemical behavior strongly
8.1 (7.7)
6.9 (6.5)
8.5 (8.1)
38.7 (36.8)
36.5 (34.7)
,11,16
6.8 (7.0)
41.5 (42.9)
35.2 (36.4)
a
A-V ) aryl-vinyl bridging. b V-V ) vinyl-vinyl bridging. c Yields
were calculated based on the relative ratios of the proton intensities of the
deuterated products. The yields in parentheses indicate normalization to
19
indicates that singlet-triplet intersystem crossing of 4 and 5
is very efficient.
1
00%. See Supporting Information for the detailed calculations.
In the second case, the presence of cyano groups in 4 has
effected the preferential aryl-vinyl interaction presumably by
conjugative stabilization of the diradical intermediates IA1-
IA4 as detailed in Scheme 8. This kind of radical stabilization
cannot be observed during the initial V-V bridging (Scheme
TABLE 6. Total Yields of Photoproducts Derived from A-V and
V-V Bridging Modes
3,20
a
total yield (%)
ratio
reactant (solvent)
∑(A-V)b
∑(VsV)c
∑(A-V):∑(V-V)
7
, vide supra) because the radicals are not conjugated to the
4
4
5
5
-d4 (C6D6)
-d4 (CD3CN)
-d4 (C6D6)
89.2 (98.0)
91.8 (95.4)
20.1 (20.7)
30.0 (28.5)
1.8 (2.0)
4.4 (4.6)
76.7 (79.3)
75.2 (71.5)
50:1
21:1
0.26:1
0.4:1
2
sp atoms of the aryl moiety. The strong directional effects of
the nitrile groups, which can stabilize diradicaloid intermediates,
have enhanced the localization of the triplet energy on this part
-d4 (CD3CN)
2
1
of the molecule.
a
Yields in parentheses indicate normalization to 100%. b Sum of
c
The V-V bridging preference of quinoxalinobarrelene 5 is
striking. With the triplet energy of the aromatic moiety (T1
photoproducts derived from A-V bridging. Sum of photoproducts derived
from V-V bridging.
22b
quinoxaline ) 60.6 kcal/mol) relatively lower than the triplet
22a,c
energy of ethylene (T1 ethylene ) 82.1 kcal/mol),
we expect
This suggests that several semibullvalenes were generated from
an initial A-V bonding interaction. To explain this bridging
mode, an alternative consideration would be the triplet energy
of the π-systems during the initial bridging process. Qualita-
tively, we can predict the energy of the reaction course by
4-d4. The proton peak observed at H-1, for example, is therefore
a contribution of all the possible semibullvalenes presented in
Scheme 7 (vide supra). A linear combination of the component
contributions was found which fit the observed distribution (see
Supporting Information). The results of the calculations are
presented in Table 5.
1
1,16e
employing Zimmerman’s bridging hypothesis.
Primary
V-V bridging of 5 leads to a structure that can be approximated
From Table 5, one can easily notice the dominant A-V
bridging mode of 4-d4 and the V-V bridging preference of 5-d4.
The bridging preference of pyrazinobarrelene 4-d4 is opposite
to that of benzobarrelene11 and 2,3-naphthobarrelene; however,
the quinoxalinobarrelene bridging mode is in agreement with
(
19) (a) Kopecky, J. Organic Photochemistry: A Visual Approach;
VCH: Cambridge, U.K., 1992. (b) Turro, N. J. Modern Molecular
Photochemistry; University Science Books: Sausalito, CA, 1991.
(20) (a) Zipse, H. Top. Curr. Chem. 2006, 263, 163. (b) Sajimon, M.
C.; Ramaiah, D.; Muneer, M.; Ajithkumar, E. S.; Rath, N. P.; George, M.
V. J. Org. Chem. 1999, 64, 6347. (c) Yeh-Ku, A.; Paquette, L. A.;
Rozeboom, M. D.; Houk, K. N. J. Am. Chem. Soc. 1979, 101, 5981. (d)
Paquette, L. A.; Cottrell, D. M.; Snow, R. A.; Gifkins, K. B.; Clardy, J. J.
Am. Chem. Soc. 1975, 97, 3275.
12
3
the bridging pattern of several barrelene derivatives, including
the above-mentioned barrelenes. The proximity of the product
distributions to a 1:1 ratio for the different pairs of photoproducts
such as 7-d4-A and 7d4-B and 7-d4-C and 7-d4-D is a good
indication for the absence of an appreciable secondary isotope
(21) (a) Zimmerman, H. E.; Novak, T. J. Org. Chem. 2003, 68, 5056.
(
b) Ito, Y.; Nishimura, H.; Umehara, Y.; Yamada, Y.; Tone, M.; Matsuura,
T. J. Am. Chem. Soc. 1983, 105, 1590. (c) Arnold, D. R.; Bolton, J. R.;
Palmer, G. E.; Prabhu, K. V. Can. J. Chem. 1977, 55, 2728.
6
effect.
(22) (a) Loutfy, R. O. Can. J. Chem. 1976, 54, 1454. (b) Murov, S. L.
In Table 6, we present a summary of the yields of photo-
products derived from the different bridging modes. For the
photoproducts of 4-d4 in the two solvents, the relative yields
Handbook of Photochemistry; Marcel Dekker: New York, 1973; p 3. (c)
Crowley, K. J. Proc. Chem. Soc. 1962, 245. (d) Evans, D. F. J. Chem. Soc.
1960, 1735. (e) Evans, D. F. J. Chem. Soc. 1957, 1351.
J. Org. Chem, Vol. 72, No. 25, 2007 9695

Chen et al.
SCHEME 8. Resonance Structures of Barrelene 4 Generated
through Initial A-V Bridging
state is one factor that has to be considered in predicting the
chemoselectivity and photoreactivity of heteroarene-fused bar-
relenes.
Conclusion
In the course of our investigation on photochemical studies
of heteroarene-fused barrelenes 4-6, we noticed that pyrazi-
nobarrelene 4 preferentially underwent A-V bridging, whereas
quinoxalinobarrelene 5 favored V-V bridging. Benzoquinox-
alinobarrelene 6 was unreactive to both photochemical reaction
conditions. The results clearly demonstrate that heteroarene-
fused barrelenes follow the general mechanism of di-π-methane
rearrangement. The A-V bridging preference of pyrazinobar-
relene 4 and the V-V bridging preference of quinoxalinobar-
relene 5, which are predicted quantitatively through deuterium-
labeling experiment and qualitatively through Zimmerman’s
initial bridging hypothesis and stability of diradicaloid inter-
mediates, suggest that electronic effects and energy minimization
of triplet excited states during the initial bridging modes are
important factors in determining the chemoselectivity of the
process. The inertness of benzoquinoxalinobarrelene 6 is at-
tributed to the minimal excitation energy present in this
molecule. Likewise, similarities and differences between the
1
1-13
homoarene-fused barrelenes
and heteroarene-fused bar-
electronically by a cisoid butadiene of which cyclohexadiene
(
T1 ) 53.5 kcal/mol)2 is an appropriate model. In the case of
2d
relenes 4-6 with regards to the bridging modes can be
rationalized in terms of energy minimization of triplet states.
Finally, the close correlation between the calculated and
observed hydrogen distributions of the photoproducts further
supports the stepwise mechanism of di-π-methane rearrange-
ment.
the quinoxaline moiety, initial bonding with the vinyl group
gives rise to a vinylquinoxaline structure. The triplet energy of
the resulting vinylquinoxaline moiety can be approximated from
2
2b
the triplet energies of napththalene (T1 ) 60.9 kcal/mol),
22b
vinylnaphthalene (T1 ) 60.0 kcal/mol), and quinoxaline (T1
)
60.6 kcal/mol).2 The attachment of the vinyl group in
2a
naphthalene decreases its triplet energy by about one unit.
Assuming that quinoxaline exhibits the same photochemical
behavior as naphthalene (their triplet energies are comparable),
then vinylquinoxaline will have a T1 energy of about 59 kcal/
mol, which is way above the triplet energy of the cisoid
butadiene. In effect, the excitation energy will be more heavily
inclined toward the vinyl group similar to that of naphthobar-
relene.12 If we apply this kind of analysis to pyrazinobarrelene
Experimental Section
3
,3-Dimethoxybicyclo[2.2.2]octa-5,7-dien-2-one (19). To a
1
4a
reaction tube were added 1 equiv of MOB dimer 18 and 4 equiv
of phenyl vinylsulfide in toluene (20 mL) and degassed under liquid
nitrogen for 1 h before sealing. The solution was heated in an oven
at 220 °C for 4 h. After removing the solvent under reduced
pressure, the crude bicyclic sulfide was purified by a silica gel
column (ethyl acetate/hexanes, 1:6). To a solution of the sulfide in
dichloromethane, stirred in a dry ice bath, was gradually added a
solution of mCPBA (1 equiv in CH Cl ). The resulting mixture
4, the dicyanopyrazinovinyl moiety generated through initial
A-V bridging will have a T1 energy of about 43 kcal/mol,
which can be deduced from the triplet energies of benzene (T1
2
2
was heated to room temperature and then quenched with NaHCO
the sulfoxide crude product was extracted with CH Cl and the
3
;
)
(
84.3 kcal/mol),2 styrene (T1 ) 61.7 kcal/mol),
2b,e
22b,e
pyrazine
2
2
22b
solvent removed in vacuo. The crude product in ethyl acetate and
triethylamine (catalytic) was placed in a reaction tube, degassed
before sealing, and heated at 130 °C for 3 h; then the solvent was
stripped off under reduced pressure, and the final product was
purified by column chromatography (ethyl acetate/hexanes, 1:10)
to obtain a light yellow liquid 19 in 70% yield.
T1 ) 76.6 kcal/mol), and dicyanobenzene (T1 ) 72.7 kcal/
22b
mol). This triplet energy (43 kcal/mol) is lower than the triplet
energy of cisoid butadiene, thus A-V bridging is preferred.
The bridging preference of 4 is in contrast to that of benzobar-
relene which favors V-V bridging. A-V bridging in benzobar-
relene will give rise to a homostyrene structure whose triplet
energy is higher than the cisoid butadiene; thus V-V bridging
is observed.1
Bicyclo[2.2.2]octa-5,7-diene-2,3-dione (20). To a dark-colored
reaction flask was added compound 19 (1.80 g, 10 mmol) in 2 N
H SO (250 mL), heated at 65 °C, and stirred for 3 h. The reaction
1,16
2
4
Unlike that of homoarene-fused anthrabarrelene,13 which
mixture was extracted with ether; then the organic layer was washed
with water, dried (MgSO ), and concentrated under reduced
pressure. The crude product was purified by a silica gel column
ethyl acetate/hexanes, 1:2) to obtain a yellow crystalline substance
0 in 84% yield.
,6-Diazatricyclo[6.2.2.0 ]dodeca-2(7),3,5,9,11-pentaene-4,5-
dicarbonitrile (4). Compound 20 (200 mg, 1.5 mmol) in MeOH
15 mL) was mixed with diaminomaleonitrile (21) (178 mg, 1.65
4
gives DPM product during direct irradiation, but is inert under
sensitized reaction, the benzoquinoxalinobarrelene 6 is insensi-
tive to both reaction conditions. The photochemical behavior
of anthrabarrelene is explained in terms of the participation of
the T2 state. In the case of benzoquinoxalinobarrelene 6, the
excitation energy, which is expected to be minimal, is insuf-
ficient to overcome the energy barrier during the initial bridging
due to the aromatic stability of the benzoquinoxalinobarrelene
moiety.12 Hence, energy minimization for the perturbed triplet
(
2
2,7
3
(
mmol) and heated at 50 °C for 8 h. After removing the solvent
under reduced pressure, the residue was chromatographed on a silica
gel column (ethyl acetate/hexanes, 1:1) and yielded upon recrys-
9696 J. Org. Chem., Vol. 72, No. 25, 2007

ChemoselectiVe Photorearrangements of Diazinobarrelenes
tallization from CH
4
3
2
Cl
(mp 183.7-184.1 °C) in 80% yield.
,10-Diazatetracyclo[10.2.2.02,1104,9]hexadeca-2(11),3,5,7,9,13,-
5-heptaene (5). Following the procedure described for 4, con-
2
or ethyl acetate a white crystalline substance
Calculations for Percent Deuterium in Dimer 18-d
ated dimer 18-d was made as the reference compound for
approximating the deuterium content of the photoreactants 4-d and
5-d . The reason being that, in the mass spectra of nondeuterated
barrelenes 4 and 5, we observed [M + H] m/z peaks, so there
was a problem of calculating the deuterium content of the deuterated
8
. Deuter-
8
4
1
4
+
densation of compound 20 (200 mg, 1.5 mmol) with diaminoben-
zene (22, 178 mg, 1.65 mmol) afforded a white crystalline substance
5
1
λ
(mp 144.0-144.6 °C) in 95% yield: IR (neat): νj ) 3076, 2980,
8
barrelenes. Our assumption of using 18-d as the reference com-
-1
575, 1355, 1300, 1101, 902, 770, 754, 602 cm ; UV/vis (MeOH)
pound is reasonable; the deuterated sites in the dimer were exactly
1
1
max (ꢀ) ) 333 (8500), 317 (9600), 251 (14000), 239 (13000); H
located in the same position when we analyzed the H NMR spectra
NMR (400 MHz, CDCl
3
; 25 °C) δ ) 4.97-5.01 (m, 2H; CH),
4 4 4
of the deuterated barrelenes 4-d and 5-d . For example, in 4-d ,
1
6.95-6.98 (m, 4H; dCH), 7.58-7.62 (m, 2H; aromatic), 7.82-
.86 (m, 2H; aromatic); 13C NMR (100 MHz, CDCl
7 ) δ ) 49.5,
128.1, 128.6, 137.6, 137.8, 157.9; MS (EI, 75 eV) m/z (%) 207
the H NMR chemical shift at δ 7.01 was observed which can be
integrated to 2 H, suggesting that the protons in one of the two
vinylic moieties and the bridgehead protons were deuterated. From
3
+
(
100) [M + H], 206 (18), 182 (5), 181 (22), 154 (8), 136 (7), 77
8
LRMS analysis of 18-d , the mass spectrum consisted of a series
+
(6); HRMS (EI) calcd for C14
H
H
10
N
2
[M ] 206.0844; found
(206.08): C, 81.53; H,
of peaks with major contributions from masses 314, 315, and 316;
the normalized mass intensities were 29.61, 15.89, and 100,
respectively. The m/z values of 314, 315, and 316 correspond to
dimers having six, seven, and eight deuteriums, respectively. From
the relative intensities of these m/z values, we obtained about 94%
2
4
06.0841. Anal. Calcd (%) for C14
.89; N, 13.58. Found: C, 81.48; H, 5.62; N, 12.80.
10
N
2
,15 4,13
0
6,11]icosa-2,4,6(11),7,9,-
3
,14-Diazapentacyclo[14.2.2.02
0
12,14,17,19-nonaene (6). Compound 20 (200 mg, 1.5 mmol) in
MeOH was mixed with naphthalenediamine (23, 261 mg, 1.65
mmol) and heated under reflux for 12 h. Following the workup
procedure for 4, compound 6 (mp 204.2-204.7 °C) was obtained
in moderate yield (44%).
8
deuteration of 18-d with the following contributions from the
various masses: m/z 314, 15.26%; m/z 315, 9.56%; m/z 316,
68.73%. Contributions from m/z values with four and five deuterium
were considered negligible so they were not included in the
1
,2,5,6,7,8,11,12-Octadeutero-3,3,10,10-tetramethoxytricyclo-
8
computation. Percent deuterium of 18-d was computed as follows:
2,7
[
6.2.2.0 ]dodeca-5,11-diene-4,9-dione (18-d
containing D O (16 g, 18 mmol) was added SOCl
dropwise. The SO formed during the reaction was removed by
bubbling nitrogen gas before adding guaiacol (24, 5 g, 40 mmol),
then the solution was heated to reflux for 3 days. The crude product
was extracted with ether and concentrated under reduced pressure.
This was done in two runs before cooling (ice bath) the deuterated
8
). To a reaction bottle
Assuming we have 100 molecules for each m/z value, then m/z
of 314 will have 29.61 × 6 D and 29.61 × 2H; m/z of 315 will
have 15.89 × 7 D and 15.89 × 1 H; m/z of 316 will have 100 ×
8 D and 100 × 0 H; therefore, contributions from various masses
will be m/z 314; %D ) 177.66 D/1164 × 100 ) 15.26%; m/z 315,
%D ) 111.23 D/1164 × 100 ) 9.56%; m/z 316, %D ) 800 D/1164
× 100 ) 68.73%.
Acknowledgment. We gratefully acknowledge the financial
support from the National Science Council (NSC) of Taiwan.
N.R.V. thanks the NSC for a postdoctoral fellowship at NTHU.
We also thank Dr. R. K. Peddinti for insightful discussions and
suggestions.
2
2
(10 mL)
2
4
phenol 24-d , which was dissolved in MeOH. Diacetoxyiodoben-
zene (DAIB) (14.2 g, 44 mmol) was gradually added while stirring,
and after 20 min, the ice bath was removed and the product was
monitored by TLC. After removing the solvent using a rotary
evaporator, the crude product was quenched with Na
extracted with ethyl acetate. The organic layer was washed with
water and brine solution, dried (MgSO ), and concentrated under
reduced pressure to obtain a gray-colored powder which can be
recrystallized from ethyl acetate into a white powder, 18-d (6 g,
6%) (mp 172.5-174.0 °C). Percentage deuteration of the MOB
8
dimer 18-d was determined by LRMS which showed m/e peaks
at 314, 315, and 316 with relative intensities of 0.3, 0.15, and 1,
respectively.
2 3
CO (aq) and
4
Supporting Information Available: Experimental procedures
and characterization data for new compounds not reported in the
8
1
13
Experimental Section, copies of H, C NMR, and UV absorption
spectra for new compounds, and detailed sample calculations. This
materialisavailablefreeofchargeviatheInternetathttp://pubs.acs.org.
9
JO7016795
J. Org. Chem, Vol. 72, No. 25, 2007 9697