Benzocyclohex-1-en-3-yne at high temperature

Article abstract of DOI:10.1021/ja981434l

Generation of benzocyclohex-1-en-3-yne in two ways leads to naphthalene and methyleneindene (benzofulvene). Labeling studies favor a multistep mechanism involving an initial electrocyclic ring opening, and carbene- alkene interconversions.

Full text of DOI:10.1021/ja981434l

J. Am. Chem. Soc. 1998, 120, 8315-8318
8315
Benzocyclohex-1-en-3-yne at High Temperature
Peter J. Lu, Weitao Pan, and Maitland Jones, Jr.*
Contribution from the Department of Chemistry, Princeton UniVersity, Princeton, New Jersey 08544
ReceiVed April 27, 1998
Abstract: Generation of benzocyclohex-1-en-3-yne in two ways leads to naphthalene and methyleneindene
(benzofulvene). Labeling studies favor a multistep mechanism involving an initial electrocyclic ring opening,
and carbene-alkene interconversions.
In the phenylcarbene rearrangement (PCR)1 and the inter-
conversion of phenylnitrene and pyridylcarbene,2 an exo-ring
methine or nitrogen becomes incorporated into the ring, as a
ring vertex takes its place in the exo-ring position (Figure 1).
A similar process occurs in carboranylcarbene.3
The graphic device in Figure 2 allows the construction of
any of the carbenes available through these rearrangements. In
each process a formally sp hybridized group enters the ring (or
cage). We wondered if an acetylene could be coaxed into
undergoing a related reaction. In such a process, a divalent
Figure 1.
carbon would be left behind (Figure 3), and it seemed that a
suitable molecule could be constructed that would trap both the
emerging and residual carbenes. Of course, simple acetylenes
are very poor candidates for this reaction, and it seemed prudent
to provide some impetus in the form of ring strain. Reactions
are known in which acetylenes formally act as dicarbenes,4 and
of course, the conversion of alkynes into a different kind of
carbene, vinylidenes, which was pioneered by Roger Brown and
the Australian pyrolysis group,5 has found recent spectacular
application in L. T. Scott’s remarkable synthesis of corannulene.6
Such reasoning, together with perhaps more than a modicum
of chutzpah, led to the idea of examining the chemistry of
Figure 2. In this version of the graph, an old meta carbon becomes
the new divalent center.
benzocyclohex-1-en-3-yne (1) at high temperature. In the end,
Figure 3.
Figure 4.
we were able to show that our imagined reaction was preempted
by another process, but much fascinating chemistry was
uncovered and we report it here.
Two routes to 1, shown below, seemed likely to yield this
intermediate (Figure 4). Indeed, related intermediates with one
(1) For a review, see: Gaspar, P. P.; Hsu, J.-P.; Chari, S.; Jones, M., Jr.
Tetrahedron 1985, 41, 1479.
(2) Wentrup, C. J. Chem. Soc., Chem. Commun. 1969, 1386.
(3) Chari, S.; Agopian, G. K.; Jones, M., Jr. J. Am. Chem. Soc. 1979,
101, 6125.
(4) (a) Shim, S. C. J. Chem. Soc., Chem. Commun. 1996, 2609. (b) Lee,
S. J.; Shim, S. C. Tetrahedron Lett. 1990, 31, 6197. (c) Lee, T. S.; Lee, S.
J.; Shim, S. C. J. Org. Chem. 1990, 55, 4544. Shim, S. C.; Lee, T. S. J.
Org. Chem. 1988, 53, 2410.
(5) (a) Brown, R. F. C. Pyrolytic Methods in Organic Chemistry;
Academic Press: New York, 1980. (b) Brown, R. F. C.; Eastwood, F. W.
Synlett 1993, 9. For a very recent example, and references, see: Sarobe,
M.; Jenneskens, L. W. J. Org. Chem. 1997, 62, 8247.
(6) Scott, L. T.; Cheng, P.-C.; Hashemi, M. M.; Bratcher, M. S.; Meyer,
D. T.; Warren, H. B. J. Am. Chem. Soc. 1997, 119, 10963.
more double bond have been trapped intramolecularly by the
Brown group5b through rearrangement to vinylidenes. The two
required precursors, 2 and 3, were synthesized from an aldol-
like condensation of 1-indanone with Meldrum’s acid and
titanium tetrachloride7a and by the known synthesis of Hershberg
and Fieser,7b respectively (Scheme 1). When 2 was dropped
onto a hot quartz surface at 640 °C and the products were
(7) (a) Baxter, G. J.; Brown, R. F. C. Aust. J. Chem. 1975, 28, 1551. (b)
Hershberg, E. B.; Fieser, L. F. Org. Synth. Collect. Vol. II 1969, 194.
S0002-7863(98)01434-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 08/07/1998

8316 J. Am. Chem. Soc., Vol. 120, No. 33, 1998
Lu et al.
Scheme 4
Figure 5.
Scheme 1
Scheme 5
Scheme 2
Scheme 3
Given all of the evidence for interconversion of vinylidenes
and cycloalkynes,5 and the similarity of the products from 2
and 3, it seemed highly likely to us that 1 was formed.
Cyclohexyne itself is known to undergo an electrocyclic ring-
opening reaction at temperatures above 550 °C to give ethylene
and butatriene.10 Calculations of Johnson and Bradley indicate
that cyclohex-1-en-3-yne would also rearrange exothermically
to vinyl-1,2,3-butatriene.11 If a related reaction were to occur
in 1, o-xylylene 5 would be the product. Surely, given the
exothermicity of rearomatization, 5 would rapidly close to 6.
Strained alkenes are known to rearrange to carbenes at high
temperatures.12 Compound 6 has two possibilities for such a
reaction; one seems to be a cul-de-sac, but the other goes straight
to carbene 7, and then to 4. This mechanism correctly predicts
the fate of the 13C label (dots in Scheme 5).
A second labeling experiment, this time with deuterium, is
consistent with this process but reveals that even more is going
on. It also yields valuable insight into the source of naphthalene.
1-Indanone can be exchanged at the R position by a modification
(D2O/NaOD, not D2O/K2CO3) of the procedure of Lustig and
Ragelis13 to give 2,2-dideuterio-1-indanone. Condensation with
Meldrum’s acid leads to 2-d2. Pyrolysis at 640 °C and 0.005
Torr led to methyleneindene 4-d2 labeled as indicated in Scheme
6. The mechanism shown above would place all of the
deuterium in the exo-methylene position. Therefore, ap-
proximately one-half of the reaction takes another path, not
revealed by the carbon label.
It would appear that some competitive, but slower, process
renders the two methylene positions of 5 and 6 equivalent. The
following Scheme 7 accomplishes that end through the inter-
mediacy of 1,2-dimethylenebenzocyclobutene (8). The carbon
label remains in the experimentally observed positions in this
captured in a trap cooled to -196 °C at 0.005 Torr, napthalene
was formed along with methyleneindene (benzofulvene)8 4 in
the ratio 1:2.6. A similar decomposition of 3 at 640 °C led to
the same two products in the ratio 1:1.2 (Figure 5).
However, a 13C-labeling experiment quickly showed that
napthalene was not formed by our hoped-for mechanism. That
route must produce napthalene with a label in the ring junction
(Scheme 2). In fact, decomposition of 2 with 100% 13C in the
position shown led to naphthalene with the label distributed
between the R and â positions, with no detectable 13C at the
ring junction (Scheme 3).9 The expected5 equilibrium between
1 and the vinylidene would serve to scramble the label, but only
the extended phenylcarbene rearrangement could put it at the
ring junction.
What was most remarkable, at first, was that there was no
label detectable at the methylene position of compound 4.
Clearly, something unusual was going on in the formation of 4
that places one or both of the ring methylenes of 2 in the exo-
ring position of 4. The apparent transfer of hydrogen from the
ring to the divalent carbon cannot be taking place (Scheme 4).
(10) Tseng, J.; McKee, M. L.; Shevlin, P. B. J. Am. Chem. Soc. 1987,
109, 5474.
(11) Johnson, R. P.; Bradley, A. Z. Unpublished work. Bradley, A. Z.
Ph.D. Thesis, University of New Hampshire, 1997.
(12) Liebman, J. F.; Greenberg, A. Strained Organic Molecules; Wiley:
New York, 1978; pp 91-94. Gajewski, J. J. Hydrocarbon Thermal
Isomerizations; Wiley: New York, 1981; pp 22-25.
(13) Lustig, E.; Ragelis, E. P. J. Org. Chem. 1967, 32, 1398.
(8) Neuenschwander, M.; Vo¨geli, R.; Fahrni, H.-P.; Lehmann, H.; Ruder,
J.-P. HelV. Chim. Acta 1977, 60, 1073.
(9) Wilson, N. K.; Stothers, J. B. J. Magn. Reson. 1974, 15, 31.

Benzocyclohex-1-en-3-yne at High Temperature
J. Am. Chem. Soc., Vol. 120, No. 33, 1998 8317
Scheme 6
Scheme 9
Scheme 10
Scheme 7
methylene groups are symmetrized through 8, the naphthalene
skeleton is in place and accessible, as is implied by the known
conversion to naphthalene at 700 °C over Nichrome.15 We do
not know the details, and dicarbene 11 is drawn as a conven-
ience, but any route to naphthalene from 8 must lead to an R,â-
substituted naphthalene (Scheme 10).
Despite these quite remarkable sequences, we have not found
any evidence for our strained acetylene acting in dicarbene-
like fashion. It may be that our choice was foolish, as the
cyclohexyne intermediate 1 had a ready escape route built in,
the reverse Diels-Alder reaction. It remains to be seen if the
carbene-like behavior will appear once this escape route is cut
off. We are pursuing two experiments of this kind and working
out the details of the mechanism of naphthalene formation using
labeled carbenes.
Scheme 8
Experimental Section
General Procedure. Unless otherwise specified, all reagents were
purchased from commercial suppliers (Aldrich or as specified) and used
without further purification. 1H and 13C NMR spectra were recorded
with JEOL GSX-270, GE QE-300, and Varian Unity Inova 500
spectrometers. GC-MS analyses were carried out on a Hewlett-
Packard 5890/5971 gas chromatograph equipped with a mass-selective
detector on a 0.25-mm-i.d., 0.25-mm film thickness, 30-m HP-1701
capillary column. Precise masses were measured on a KRATOS MS
50 RFA high-resolution mass spectrometer. Temperatures were
measured with a thermocouple or a thermometer and are uncorrected.
process (dots). The proposed mechanism finds excellent
precedent in the thermal conversion of 3,4-dimethylenecy-
clobutene to fulvene and benzene at 620 °C in a ratio of 2:1.14
1,2-Dimethylenebenzocyclobutene itself has not been so well
studied but has been reported to rearrange to naphthalene when
heated at 700 °C over a Nichrome wire.15
The naphthalene is labeled with deuterium in both the R and
â positions. 13C also winds up in the R and â positions in the
approximate ratio 1:2. Thus naphthalene cannot be formed from
4, in which the label is equally distributed between two
positions. This surmise is born out by control experiments,
which show that only about 5% of naphthalene is formed from
4 under our reaction conditions. The plethora of potential routes
to naphthalene makes mechanistic speculation difficult, but here
are our thoughts on the likely mechanistic pathways. The
vinylidene intermediate can insert into an adjacent position to
give a fused cyclopropene 9. Opening of the cyclopropene ring
to a vinylcarbene, a common occurence,12 leads to 10 and thence
to napthalene labeled exclusively in the â position (Scheme 8).
Cyclohexyne itself is thought to find its way to 1,3-cyclohexa-
diene by a parallel pathway.10
Some R-substituted naphthalene can be produced by the
rearrangement of 4-d2, but this pathway can account only for
very small amounts as we have shown through control experi-
ments that only about 5% of 4 is converted into naphthalene
under our reaction conditions (Scheme 9).
We suggest that the R-substituted material must come from
the intermediate 8, formed reversibly from 6. Once the two
(14) Henry, T. J.; Bergman, R. G. J. Am. Chem. Soc. 1972, 94, 5103.
Kent, J. E.; Jones, A. J. Aust. J. Chem. 1970, 23, 1059.
(15) Cava, M. P.; Mitchell, M. J.; Pohl, R. J. J. Am. Chem. Soc. 1963,
85, 2080.
2,2-Dimethyl-5-(1′-indanylidene)-1,3-dioxane-4,6-dione (2). The
synthesis follows the general method of Baxter and Brown.7a A solution
of 11 mL of TiCl4 (100 mmol) in 25 mL of CCl4 was added dropwise
to dry tetrahydrofuran (190 mL) at 0 °C. A solution of 6.6 g of
indanone (50 mmol) and 7.2 g of Meldrum’s acid (50 mmol) in dry
tetrahydrofuran (25 mL) was added slowly to the yellow precipitate
and was followed by 16 mL of pyridine (250 mmol) in dry tetrahy-
drofuran (25 mL). The flask was then rinsed with an additional 60
mL of dry tetrahydrofuran. The ice bath surrounding the reaction vessel
was allowed to warm to room temperature. The reaction was then
stirred for 4 days, during which time the solution turned opaque slate
grey. Water and ether were added until the solid dissolved, and the
aqueous layer was removed. The organic layer was washed with H2O
(1 L), saturated NaCl solution (300 mL), saturated NaHCO3 (600 mL),
and H2O again (600 mL). It was dried over NaSO4, and the solvents
were evaporated at the water pump. The crude product was then
recrystallized from ethanol, yielding 4.2 g of white needles, 3 g of
which was redissolved in ether and chromatographed on a silica gel
column with 5:1 pentane:ether solvent. A second recrystallization from
1
ethanol yielded 0.4 g (2.6 mmol, 7.3%): mp 101-108 °C; H NMR
(500 MHz, CDCl3) δ 1.79 (s, 6H), 3.15 (t, J ) 5.2 Hz, 2H), 3.60 (t, J
) 5.2 Hz, 2H), 7.46 (d, J ) 7.5 Hz, 1 H), 7.53 (m, 1H), 8.68 (d, J )
8.2 Hz, 1H); 13C NMR (125 MHz, CDCl3) δ 27.10, 30.54, 36.91,
103.49, 109.61, 125.62, 127.03, 129.76, 134.45, 137.08, 155.75, 162.33,
163.56, 178.29; IR (KBr) cm-1 1720, 1559, 1293, 1201; MS-EI m/z
(relative intensity) 258 (3), 200 (100), 172 (39), 156 (98), 128 (94);
HRMS calcd for C15H14O4 258.0892, found 258.0870.

8318 J. Am. Chem. Soc., Vol. 120, No. 33, 1998
Lu et al.
2,2-Dimethyl-5-(1′-indanylidene)-[5-13C]-1,3-dioxane-4,6-dione. This
compound was synthesized in a fashion similar to that for 2, except
that 2 equiv of 1-indanone was used. The yield was 36.5%: 13C NMR
(67.94 MHz, CDCl3) δ 109.6
least 1 h. The starting material was slowly dropped into the quartz
flask in such a way that the pressure did not exceed 60 × 10-3 Torr.
After the pyrolysis was completed, the products in the cold trap were
immediately dissolved in an appropriate solvent for analysis or further
treatment.
2,2-Dimethyl-5-[5-13C]-1,3-dioxane-4,6-dione. This compound was
synthesized through a modification of the method of Davidson and
Bernhard.16 [2-13C]Malonic acid (Cambridge, 0.34 g, 3.26 mmol) and
isopropenyl acetate (0.38 mL, 3.45 mmol) were placed into a 5-mL
round-bottomed flask equipped with a magnetic stirring bar and cooled
in an ice bath to 0 °C. To the mixture was added one drop of
concentrated sulfuric acid, and stirring was continued at 0 °C for 2 h.
The solution was cooled in the refrigerator overnight and filtered, and
the solid was washed with 2 × 5 mL of cyclohexane. The crude solid
(0.42 g) was dissolved in methylene chloride. Filtration led to 0.13 g
of a solid containing mainly malonic acid and 2,2-dimethyl-[5-13C]-
1,3-dioxane-4,6-dione. Evaporation of the methylene chloride layer
gave 0.32 g of 2,2-dimethyl-[5-13C]-1,3-dioxane-4,6-dione (68%): 1H
NMR (270 MHz, CDCl3) δ 1.79 (s, 6H), 3.63 (d, J ) 133.7 Hz, 2H);
13C NMR (67.94 MHz, CDCl3) δ 36.4 (s) (48 scans of a dilute sample).
2,2-Dideuterio-1-indanone. This compound was synthesized ac-
cording to a modification of the method of Lustig and Ragelis,13 in
which NaH was used as the base. 1-Indanone (2.0 g, 15.0 mmol),
deuterium oxide (10 mL, 500 mmol), NaH (0.20 g, 8.3 mmol), and
dioxane (10 mL) were used in the reaction. A 70% yield of
2,2-dideuterio-1-indanone was obtained: 1H NMR (270 MHz, CDCl3)
δ 2.62 (m, 0.06H), 3.10 (s, 2H). 7.3 (t, 1H), 7.41 (d, 1H), 7.55 (t,
1H), 7.70 (d, 1H).
2,2-Dimethyl-5-(2′,2′-dideuterio-1′-indanylidene)-1,3-dioxane-4,6-
dione. This compound was synthesized in a yield of 35% in a fashion
similar to that for 2, except that deuterium oxide was used instead of
water in the workup: 1H NMR (270 MHz, CDCl3) δ 3.15 (m, 2H),
3.60 (m, 0.5H); other signals are the same as in the unlabeled compound.
Typical Procedure for Pyrolysis. Pyrolyses were carried out in a
500-mL two-necked quartz flask wrapped with heating wire. One of
the necks was connected to a 10-mL round flask into which the
corresponding starting material was placed; the other neck was
connected to a liquid nitrogen trap which was in turn connected to a
vacuum system. Prior to the pyrolysis, the quartz flask was heated to
the desired temperature and the whole system was evacuated for at
Pyrolysis of 2,2-Dimethyl-5-(1′-indanylidene)-1,3-dioxane-4,6-
dione (2). Compound 2 (84 mg, 0.325 mmol) was pyrolyzed to give
the products as a yellow oil. An analysis (GC) of the products revealed
methyleneindene (4) (0.2 mmol, 66%) and naphthalene (0.1 mmol,
26%) as the major products, along with a very small amount of
unidentified compounds whose m/z values were 128 and 116.
Pyrolysis of 2,2-Dimethyl-5-(1′-indanylidene)-[5-13C]-1,3-dioxane-
4,6-dione. The 13C content of the products was measured by comparing
the corresponding signal intensities to those from the products of the
experiment using unlabeled material. The 1H (500 MHz) signal of
unlabeled 4 [δ 6.47 (dd, J ) 5.0, 5.0 Hz, 1H)] was split into a wide
doublet in the 13C-containing material [δ 6.47 (d, J ) 5.0, 170 Hz,
1
0.5H)]. The residual H of unlabeled material appeared as usual at δ
6.47 (0.5H).
Pyrolysis of 2,2-Dimethyl-5-(2′,2′-dideuterio-1′-indanylidene)-1,3-
dioxane-4,6-dione (2). Pyrolysis led to deuterated naphthalene and
methyleneindene. Naphthalene: 1H NMR (500 MHz) δ 7.8 (m, 1H),
2
7.40 (m, 1H); H NMR (500 MHz) δ 7.8 (s), 7.40 (s); GC-MS m/z
(relative intensity) 131 (10), 130 (95), 129 (100). Methyleneindene:
1H NMR (500 MHz) δ 6.47 (m, 0.75H), 6.03, 6.02 (0.5H),17 5.68 (s),
2
5.7 (s)17 (0.5H); H NMR (500 MHz) (relative intensity) δ 6.84 (s,
0.25H), 6.48 (s, 0.25H), 6.0 (s, 1H), 6.03 (s, 1H); GC-MS m/z (relative
intensity) 131 (10), 130 (100), 129 (95).
Pyrolysis of 3,4-Dihydro-1,2-naphthalic Anhydride (3). Com-
pound 3 (60 mg, 0.30 mmol) was pyrolyzed as described above to
give methyleneindene (4) in 20% yield (0.06 mmol) and naphthalene
in 17% yield (0.05 mmol) along with a very small amount of starting
1
material. The determination of yield was by H NMR spectroscopy.
Acknowledgment. We thank the National Science Founda-
tion for support of this work through Grant CHE-9702823. We
also thank Dr. Wei Wang for valuable assistance in obtaining
NMR spectra.
JA981434L
(16) Davidson, D.; Bernhard, S. A. J. Am. Chem. Soc. 1949, 70, 3426.
(17) The multiple signals result from different deuterated isomers of 4.