An efficient synthesis of dibenzocycloocta-4a,6a,-diene-5,11-diyne and its precursors

Article abstract of DOI:10.1055/s-2002-32542

Two efficient syntheses of dibenzocyclooctadienediyne 1 were developed employing known reactions, which utilize commercially available reagents. Both methods are an improvement on known syntheses resulting in 41% and 43% overall yields. The latter method also offers an efficient synthesis of dibenzocyclooctatetraene 9, which is one of the key reagents now commercially unavailable.

Full text of DOI:10.1055/s-2002-32542

PAPER
1191
An Efficient Synthesis of Dibenzocycloocta-4a,6a,-diene-5,11-diyne and its
Precursors
Synthesis of
D
t
ibenzo
e
cycloocta-4
r
a,6a-die
l
ne-5, 11
i
-diyneng Chaffins, Michael Brettreich, Fred Wudl*
Exotic Materials Institute and Department of Chemistry and Biochemistry, University of California, Los Angeles, CA 90095, USA
Fax +1(310)8250767; E-mail: wudl@chem.ucla.edu
Received 27 April 2002
Fieser obtained a dicyano derivative in 47% yield in a sin-
gle step.5 Two more steps afforded the tetraene together
with dibenzodihydropentalene. Another promising effort
was the work done by Jung and Kagan in which phenylac-
Abstract: Two efficient syntheses of dibenzocyclooctadienediyne
1 were developed employing known reactions, which utilize com-
mercially available reagents. Both methods are an improvement on
known syntheses resulting in 41% and 43% overall yields. The lat-
ter method also offers an efficient synthesis of dibenzocyclooctatet- etaldehyde could be cyclized to Kagan’s ether by double
ortho Friedel–Crafts alkylation in 27% yield.6,7 The low
yield of the initial cyclization step of phenylacetaldehyde
raene 9, which is one of the key reagents now commercially un-
available.
Key words: dibenzocyclooctadienediyne, elimination, lithium,
sonication
by the Kagan–Jung procedure as well as the multistep
synthesis of dimethoxyphenylacetaldehyde also made this
approach to the target compound undesirable and forced
the development of new syntheses of the dienediyne 1 and
Due to its high angle strain, dibenzocyclooctadienediyne the tetraene 9.
1 has long been recognized as a reactive tricyclic diyne
Beginning with commercially available dibenzosuber-
(Figure). This strain is based on the planarity of the mole-
cule in which the crystal structure shows that the bond an-
gle between the aromatic rings and acetylene units is
distorted from 180 to 153°.1 In the past, advantage was
taken of the enhanced dienophilic reactivity of these triple
bonds to build extended aromatic systems2 and much
more recently we have employed the same reactivity in a
very efficient synthesis of a complex cyclophane3 deriva-
tive. These examples clearly show the diversity of 1 as a
starting material. Further applications of 1 in organic ma-
terials, such as high-performance polymers and conjugat-
ed high-performance polymers, is hampered by the
relatively lengthy and inefficient literature synthesis.4 It is
therefore necessary to develop an efficient synthesis to
fully exploit the potential of dibenzocyclooctadienediyne
1 as a starting material.
enone (2), a ring expansion was carried out to give 70% of
the trieneone 3.8 Base-treatment of the latter afforded an
enolate that was converted to the enol triflate with triflic
anhydride, affording a tetraene derivative 4 in 60% yield.9
The base used in this step, 2,6-di-tert-butyl-4-methylpyri-
dine, was expensive but could be recycled. Unfortunately,
other bases used were less effective. Bromination and
subsequent base-induced elimination furnished the diyne
1 in high yield (Scheme 1). The overall yield from diben-
zosuberenone was 41%, a much improved approach from
previously known methods (10.4%10, <1%2a).
Interestingly, 4 could be eliminated to the eneyne 6 in
80% yield (Scheme 2), providing 6 in 42% yield from
dibenzosuberenone, as well as an efficient entry to the
more reactive eneyne. Under many solvent conditions, 6
was found to readily decompose.
Figure The structures of tricyclic dienes and diynes 1, 6, and 9
Due to the recent unavailability of dibenzocyclooctate-
traene 9, where a bromination-dehydrobromination se-
quence affords the desired diyne 1,4 a new route from
other commercially available starting materials was nec-
essary. Previous work to obtain 9 was accomplished by
numerous multi-step and inefficient routes. By condensa-
tion of o-phthalaldehyde and o-phenylenediacetonitrile,
Synthesis 2002, No. 9, Print: 01 07 2002.
Art Id.1437-210X,E;2002,0,09,1191,1194,ftx,en;C12202SS.pdf.
© Georg Thieme Verlag Stuttgart · New York
ISSN 0039-7881
Scheme 1

1192
S. Chaffins et al.
PAPER
Scheme 2
Although the synthesis in Scheme 1 appeared to be a dra-
matic improvement over known methods, the cost of re-
agents such as dibenzosuberenone, trimethylsilyl
diazomethane, and 2,6-di-tert-butyl-4-methylpyridine
was considered too steep. Therefore, a more cost effec-
tive, yet still efficient synthetic protocol to produce diben-
zocyclooctadienediyne 1 was sought. Based on Cope’s
procedure,11 the saturated precursor 7 was expected to be
a relatively inexpensive target since the efficient conver-
sion from dibenzocyclooctatetraene 9 to 1 had already
been established.4 Cope’s work afforded only moderate
yields (30–46%) of dibenzocyclooctadiene 7 under high
dilution conditions using sodium to couple , -dibromo-
o-xylene. Boudjouk et al.,12 improved this step using lith-
ium powder with sonication to achieve a yield of 80%
yield on a 2 gram scale. Unfortunately, lithium powder
has also become commercially unavailable, and other
forms of lithium, such as wire, granule, or dispersion, are
much less effective in coupling , -dibromo-o-xylene.
Other research groups reported preparing their own highly
reactive lithium powder using specialized grinding mills
under inert conditions. The lithium sand we were able to
produce by stirring other commercial forms of lithium,
such as wire, at high temperatures in silicon oil was ade-
quate to afford 65% of 7 from , -dibromo-o-xylene by
sonication. These results were also reproducible when
scaling up to 50 g scale. NBS bromination of the saturated
ring to the dibromo derivative 8 could also be done in near
100 g scale, maintaining >90% yield in all attempts. Elim-
ination to the tetraene 9 worked best at smaller scales, giv-
ing 80% under optimized conditions; larger scale
elimination reactions gave, on average, 60% yield. Bromi-
nation of the tetraene 9 to the tetrabromo derivative 10
gave a mixture of isomers, which without purification,
was converted to the diyne 1 in 87% yield from the tet-
Scheme 3
diate, we are poised to develop very robust polymeric or-
ganic materials.
All chemicals were used as purchased from Aldrich unless other-
wise noted. , -Dibromo-o-xylene (‘98%’) was recrystallized from
petroleum ether (bp 35–60 °C) to afford white needles. CCl4 was
distilled from CaCl2 and THF was distilled from a purple sodium
and benzophenone solution. NMR data were obtained with ARX
400 and Avance 500 instruments. High resolution EI mass spectra
(70 eV) were recorded on a triple sector instrument (VG/Micromass
Autospec) tuned to 8,000 static resolution (M/ M, 10% valley) us-
ing perfluorinated kerosine (formula weight 705, Lancaster Synthe-
sis Inc., NH) as internal calibrant.
Lithium Sand
Silicon oil (400 mL) and Li wire (15 g) were rapidly stirred at 210
°C using a mechanical stirrer in a 1000 mL Morton flask. Once the
Li had been broken up to the desired particle size, the mixture was
cooled in an ice bath while continuously stirring. The cooled Li sus-
pension was filtered, washed with hexanes, and stored under inert
conditions.
6H-Dibenzo[a,e]cyclooctatrien-5-one (3)
To a stirred solution of dibenzosuberenone (2.9 g, 14.06 mmol) and
BF3 OEt2 (2.66 mL, 21.0 mmol) in CH2Cl2 (20 mL) was added
dropwise a solution of trimethylsilyl diazomethane (10.5 mL, 21.0
raene (Scheme 3). Once again, the dehydrobromination mmol) in CH2Cl2 (25 mL) at –10 °C over 1 h. The mixture was
stirred at –10 °C for 2 h, and then poured into ice-water. The aque-
ous layer was extracted with CH2Cl2 (3 100 mL) and the organic
yield of 10 decreased on scale up. Possible improved
scale-up conditions are being examined. Using this im-
layers were combined. The combined organic layers were washed
proved methodology, we were able to obtain dibenzocy-
with brine and dried (MgSO4). Removal of solvent afforded a yel-
clooctadienediyne 1 in 43% overall yield.
low powder which was recrystallized from hexanes to give 2.2 g
(70%) of product.
In summary, two improved, efficient syntheses of 1 have
1H NMR (CDCl3): = 8.26 (1 H, q, J = 1.4, 6.6 Hz), 7.35 (7 H, m),
been developed. The target was reached in much higher
yields than previously reported. The first approach 7.05 (2 H, q, J = 12.9, 3.8 Hz), 4.06 (2 H, s).
13C NMR (CDCl3): = 196.6, 136.9, 136.3, 135.4, 133.8, 133.1,
132.4, 132.4, 131.4, 130.6, 129.3, 128.8, 128.0, 127.3, 126.9,
48.4.13
(Scheme 2) afforded the product in only four steps and an
isolated yield of 41%. In addition, the more reactive eney-
ne 6 was obtained in 42% yield. Also, triflate elimination
was discovered to proceed more cleanly than dehydrobro-
mination. The second method (Scheme 3) provided 1 in
an overall yield of 43% in five steps. The same scheme
shows that the tetraene 9 can be produced efficiently.
With moderately large amounts of this important interme-
Dibenzo[a,e]cyclooctatetraene-5-trifluoromethanesulfonate (4)
To a CH2Cl2 (25 mL) solution of triflic anhydride (0.82 mL, 4.60
mmol) and 2,6-di-tert-butyl-4-methylpyridine (1.19 g, 5.80 mmol)
was slowly added a solution of 3 in CH2Cl2 (25 mL) at r.t. After stir-
ring overnight at r.t., the solution slowly turned to a reddish color.
Synthesis 2002, No. 9, 1191–1194 ISSN 0039-7881 © Thieme Stuttgart · New York

PAPER
Synthesis of Dibenzocycloocta-4a,6a-diene-5,11-diyne
1193
By TLC analysis in hexanes–EtOAc (9:1), a less polar spot was de-
tected as the major product. Upon complete consumption of starting
material, the reaction was quenched with pentane (300 mL). The
latter was washed with a 5% aq HCl solution and brine, followed by
drying the organic layer (MgSO4). After evaporation of solvent,
1.19 g (60%) of product was recovered.
1H NMR (CDCl3): = 7.31 (5 H, m), 7.14 (3 H, m), 6.98 (1 H, s),
6.88 (2 H, s).
13C NMR (CDCl3): = 147.2, 138.2, 137.0, 133.3, 132.5, 132.0,
131.5, 129.9, 129.4, 129.4, 129.1, 128.7, 128.2, 127.5, 127.4, 124.9,
118.6 (q, J = 321 Hz).
1H NMR (CDCl3): = 7.09 (m, 6 H), 6.97 (dd, J = 7.2, 1.2 Hz, 2 H),
5.34 (dd, J = 4.2, 2.8 Hz, 2 H), 4.29 (dd, J = 4.2, 2.8 Hz, 2 H), 3.66
(q, J = 8.4 Hz, 2 H).13
Dibenzo[a,e]cyclooctatetraene (9)
A solution of 8 (0.95 g, 2.6 mmol) in THF (25 mL) was slowly add-
ed to an excess of 1 M KOBu-t (42 mL, 16 equiv) solution in THF
(16 mL) at 0 °C, and the reaction was allowed to warm to r.t. with
continued stirring overnight. By TLC analysis (silica gel, hexanes)
indicated the formation of a less polar spot, corresponding to the
formation of product. The reaction was carefully quenched with
H2O and the product was extracted with CHCl3. The organic layer
was evaporated, dried and purified by column chromatography (sil-
ica gel, hexanes) to afford 0.42 g (79%) of product.
HRMS: m/z calcd 392.0381; found 392.0377.
1H NMR (CDCl3): = 7.17 (q, J = 3.4 Hz, 4 H), 7.08 (q, 3.4 Hz, 4
11,12-Dibromo-11,12-dihydrodibenzo[a,e]cyclooctatriene-5-
trifluoromethanesulfonate (5)
H), 6.78 (s, 4 H).4
To a solution of 4 (100 mg, 0.28 mmol) in CCl4 (3 mL) was slowly
added over 45 min a solution of Br2 (14.5 L, 0.28 mmol) in CCl4
(1.65 mL). Purification of the product by column chromatography
in hexanes on silica gel gave a yellow oil (136 mg, 95%), which
eventually solidified.
1H NMR (CDCl3): = 7.82 (1 H, br), 7.24 (8 H, m), 6.09 (1 H, br),
5.64 (1 H, br).
5,6,11,12-Tetrabromo-5,6,11,12-tetrahydrodibenzo[a,e]cy-
clooctadiene (10)
To a solution of 9 (0.42 g, 2 mmol) in CCl4 (10 mL) was added drop-
wise a solution of Br2 (0.21 mL, 4 mmol) in CCl4 (10 mL). The re-
action was monitored by TLC (silica gel, hexanes) for the
appearance of two more polar spots. The volume of the reaction
mixture was then reduced and passed through a short silica gel col-
umn and eluted with CH2Cl2 to give a pale white solid (0.87 g),
which was isolated as a mixture of isomers.
1H NMR (CDCl3): = 7.74 (dd, J = 7.8, 0.9 Hz, 2 H), 7.17 (td,
J = 7.7, 1.3 Hz, 2 H), 7.02 (td, J = 7.7, 1.3 Hz, 2 H), 6.81 (dd,
J = 7.8, 1.0 Hz, 2 H), 6.28 (d, J = 9.8 Hz, 2 H), 5.29 (d, J = 9.8 Hz,
2 H).
HRMS: m/z calcd 509.8747; found 509.8726.
sym-Dibenzo-1,3,5-cyclooctatrien-7-yne (6)
A solution of 4 (0.215 g, 0.61 mmol) in THF (20 mL) was added
dropwise to a stirred solution of t-BuOK in THF (1 M, 50 mL) at 0
°C under N2. After the addition was complete, the cold solution was
further diluted with hexanes and washed with H2O (3 100 mL) in
a separatory funnel. Subsequent washings with brine and drying
(MgSO4) gave a yellow solution, which upon removal of solvent
gave the crude product. Further purification by column chromatog-
raphy on silica gel afforded the pure product as an orange solid
(0.098 g, 80%).
13C NMR (CDCl3): = 136.6, 135.2, 130.2, 129.8, 129.2, 129.1,
128.3, 60.8, 56.2.4
Dibenzocycloocta-4a,6a-diene-5,11-diyne (1)
A mixture of isomers of 10 (0.87 g, 1.6 mmol) in THF (50 mL) was
added to an excess of 1 M KOBu-t in THF (20 mL) at 0 °C. The re-
action was monitored by TLC (silica gel, hexanes) for the formation
of a less polar spot, which turned red on the silica gel plate. The re-
action was quenched with H2O and extracted with hexanes, which
upon removal of solvent gave an orange solid. Purification through
a short alumina column in hexanes produced a yellow solid (0.36 g,
87%). Further purification could be accomplished by recrystalliza-
tion from hexanes to give crystallographic quality prisms.
1H NMR (CDCl3): = 6.93 (q, J = 3.3 Hz, 4 H), 6.74 (q, J = 3.3 Hz,
4 H).
13C NMR (CDCl3): = 132.9, 129.1, 126.9, 109.3.4
1H NMR (CDCl3): = 6.56 (td, J = 7.47, 1.71 Hz, 2 H), 6.43 (m, 6
H), 5.40 (s, 2 H).
5,6,11,12-Tetrahydrodibenzo[a,e]cyclooctadiene (7)
CAUTION: Large scale reactions require special care due to the
exothermic formation of the dibromide.
, -Dibromo-o-xylene (50 g, 189 mmol) was added at 0 °C to a
suspension of Li sand (7.74 g, 1.1 mol) in anhyd THF (225 mL) in
a 1 L round bottom flask, equipped with a reflux condenser. The
mixture was degassed with argon at r.t., and then sonicated for 3–6
h. The reaction was complete once a blood red solution was ob-
served in the flask, though unreacted Li was still present. The mix-
ture was filtered, taking great care in handling Li, and then
concentrated to a gummy solid. Purification using a short silica gel
column with hexanes gave a white crystalline product (12.52 g,
63%) after solvent evaporation.
Acknowledgments
We are grateful to the National Science Foundation for support
through Grant DMR 9796302, FRG 0073046 and NSF equipment
grant no. CHE-9974928 for the Avance 500 spectrometer.
1H NMR (CDCl3): = 7.12 (s, 8 H), 3.18 (s, 8 H).14
5,11-Dibromo-5,6,11,12-tetrahydrodibenzo[a,e]cyclooctadiene
(8)
References
A solution of 7 (6.05 g, 29 mmol) in anhyd CCl4 was brought to re-
flux after the addition of NBS (10.35 g, 58 mmol). After 1.0–1.5 h,
the reaction was complete by TLC analysis (silica gel, hexanes, for-
mation of a more polar spot). Upon cooling to r.t., the solid succin-
imide was removed by filtration and washed with CCl4. The
washings were added to the mother liquor and the combined organic
solution, upon evaporation, afforded a solid mass which was
washed with H2O and dried to give the crude product as a white sol-
id (10.16 g, 95%).
(1) Destro, R.; Pilati, T.; Simonetta, M. J. Am. Chem. Soc. 1975,
97, 658.
(2) (a) Man, Y. M.; Mak, T. C. W.; Wong, H. N. C. J. Org.
Chem. 1990, 55, 3214. (b) Wong, H. N. C. Acc. Chem. Res.
1989, 22, 145. (c) Muller, M.; Iyer, V. S.; Kubel, C.;
Enkelmann, V.; Mullen, K. Angew. Chem., Int. Ed. Engl.
1997, 36, 1607.
(3) Brettreich M.; Bendikov, M.; Chaffins, S.; Duong, H.;
Perepichka, D. F.; Helgeson, R.; Wudl, F., in press.
Synthesis 2002, No. 9, 1191–1194 ISSN 0039-7881 © Thieme Stuttgart · New York

1194
S. Chaffins et al.
PAPER
(6) (a) Jung, M. E.; Mossman, A. B.; Lyster, M. A. J. Org.
(4) (a) Wawzonek, S. J. Am. Chem. Soc. 1940, 62, 745.
(b) Bachman, G. B.; Hoaglin, R. I. J. Org. Chem. 1943, 8,
300. (c) Wittig, G. Angew. Chem. 1951, 63, 15.
Chem. 1978, 43, 3698. (b) Kagan, J.; Chen, S. Y.; Agdeppa,
D. A.; Watson, W. H.; Zabel, V. Tetrahedron Lett. 1977,
4469.
(d) Bergmann, E. D.; Pelchowicz, Z. J. Am. Chem. Soc.
1953, 75, 4281. (e) Cava, M. P.; Pohlke, R.; Erickson, B.
W.; Rose, J. C.; Fraenkel, G. Tetrahedron 1962, 18, 1005.
(f) Griffin, C. E.; Peters, J. A. J. Org. Chem. 1963, 28, 1715.
(g) Vogel, E.; Frass, W.; Wolpers, J. Angew. Chem. 1963,
75, 979. (h) Avram, M.; Dinulescu, I. G.; Dinu, D.;
Mateensen, G.; Nenitzesen, C. D. Tetrahedron 1963, 19,
309. (i) Wong, H. N. C.; Garratt, P. J.; Sondheimer, F. J. Am.
Chem. Soc. 1974, 96, 5604. (j) Chan, T. L.; Huang, N. Z.;
Sondheimer, F. Tetrahedron 1983, 39, 427. (k) Dürr, H.;
Klank, G.; Peters, K.; von Schnering, H. G. Angew. Chem.
1983, 95, 321. (l) Barton, J. W.; Lee, D. V.; Shepherd, M. K.
J. Chem. Soc., Perkin Trans. 1 1985, 1407.
(7) Jung, M. E.; Miller, S. J. J. Am. Chem. Soc. 1981, 103, 1984.
(8) (a) Hashimoto, N.; Aoyama, T.; Shioiri, T. Tetrahedron Lett.
1980, 21, 4619. (b) Boehshar, M.; Heydt, H.; Regitz, M.
Tetrahedron 1986, 42, 1815.
(9) Jigajinni, V. B.; Wightman, R. H. Tetrahedron Lett. 1982,
23, 117.
(10) Chan, T. L.; Mak, T. C. W.; Poon, C. D.; Wong, H. N. C.;
Jia, J. H.; Wang, L. L. Tetrahedron 1986, 42, 655.
(11) Cope, A. C.; Fenton, S. W. J. Am. Chem. Soc. 1951, 73,
1668.
(12) Boudjouk, P.; Sooriyakumaran, R.; Han, B. H. J. Org. Chem.
1986, 51, 2818.
(5) Fieser, L. F.; Pechet, M. M. J. Am. Chem. Soc. 1946, 68,
2577.
(13) Moore, J. A.; Mitchell, T. D. Org. Prep. Proced. Int. 1984,
16, 411.
(14) Jimeno, M. L.; Alkorta, I.; Eleguro, J.; Anderson, J. E.;
Claramunt, R. M.; Lavandera, J. L. New J. Chem. 1998,
1079.
Synthesis 2002, No. 9, 1191–1194 ISSN 0039-7881 © Thieme Stuttgart · New York