The Bergman reaction as a synthetic tool: Advantages and restrictions

Article abstract of DOI:10.1016/S0040-4020(01)00247-2

The Bergman cycloaromatization reaction efficiently converts easily prepared acyclic enediynes into aromatic rings. In order to prepare larger, functionalized fused aromatic systems using this reaction, a thorough understanding of how functionalization affects cycloaromatization is necessary. We present here our studies on the influence of substituents at three different functionalization sites on cycloaromatization, and how these functional groups can be tailored to prepare more complex systems.

Full text of DOI:10.1016/S0040-4020(01)00247-2

TETRAHEDRON
Pergamon
Tetrahedron 57 ;2001) 3753±3760
The Bergman reaction as a synthetic tool:
advantages and restrictions
Daniel M. Bowles, Grant J. Palmer, Chad A. Landis, John L. Scott and John E. Anthonyp
Department of Chemistry, University of Kentucky, Lexington, KY 40506-0055, USA
Received 25October 2000; accepted 18 December 2000
AbstractÐThe Bergman cycloaromatization reaction ef®ciently converts easily prepared acyclic enediynes into aromatic rings. In order to
prepare larger, functionalized fused aromatic systems using this reaction, a thorough understanding of how functionalization affects
cycloaromatization is necessary. We present here our studies on the in¯uence of substituents at three different functionalization sites on
cycloaromatization, and how these functional groups can be tailored to prepare more complex systems. q 2001 Elsevier Science Ltd. All
rights reserved.
First reported in the early 1970s, the cycloaromatization that
eventually became known as the Bergman reaction1
received little attention until the early 1990s, when it was
discovered that this reaction formed the DNA-cleaving
`warhead' of a series of marine antitumor antibiotics.2 As
efforts to control the activity of these potential drugs proved
less than successful, a few researchers began to investigate
the Bergman cyclization as a synthetic tool. Recent
examples include the exploitation of the diradical nature
of the p-benzyne, formed through cycloaromatization, for
the synthesis of poly;p-phenylenes),3 as well as the poly-
merization and crosslinking reaction for the formation of a
series of arene-rich polymers.4 A recent application of
cycloaromatization to the synthesis of fused-ring systems
involved tethering alkenes to the enediyne unit, allowing
them to react with the cycloaromatized species to form
additional ;saturated) rings ;Scheme 1).5
We recently began a series of projects applying the
Bergman reaction to the synthesis of fused aromatic
systems. Effective use of this reaction required us to develop
an understanding of not only the functional group tolerances
of this reaction, but also the geometric parameters required
for successful ring closures initiated by the Bergman
radical. Because we are particularly interested in synthetic
applications, rather than detailed mechanistic considera-
tions, our focus in exploring the scope of this reaction has
been to determine how substitution affects the isolated yield
and purity of products of these reactions, rather than how
substituents alter the kinetics or rate of reaction.
Keywords: cycloaromatization; Bergman reaction; functionalization; acenes.
Corresponding author. Tel.: 11-606-264-1146; e-mail: anthony@pop.uky.edu
p
0040±4020/01/$ - see front matter q 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020;01)00247-2

3754
D. M. Bowles et al. / Tetrahedron 57 +2001) 3753±3760
aromatization product even at temperatures as high as
2208C. Electronic factors also appear to play an important
role in this reaction, since the tetrayne 3e ;the alkyne sub-
stituent is no more sterically encumbering across the
enediyne than a halogen) also failed to undergo cyclo-
aromatization.
Figure 1. The three distinct functionalization sites of an arenediyne.
Scheme 2.
The only hydrocarbon-substituted enediynes which have
been shown to cycloaromatize in good yield have been
those with substituents which `trap' the resulting radical.8
Such systems do have rather severe limitations. They
require very high temperatures ;.2208C) to induce cycliza-
tion, and they are restricted to the formation of ®ve-
membered saturated ring systems. Attempts at forming
six-membered rings by this method were not successful.
Numerous efforts to extend this concept to compounds
with a more rigid radical acceptor ;e.g. 5), or to quench
the Bergman radical by hydrogen abstraction rather
than cyclization ;e.g. 6) were unsuccessful in our
hands, with such systems showing no reaction below
2508C ;Fig. 2).
Figure 2. Substituted arenediynes which do not undergo cycloaromatiza-
tion.
The dramatic effect of alkyne substitution on the cyclo-
aromatization reaction has led us to investigate whether
similar functionalization on the ®nal product can be attained
by substitution on the aromatic ring, rather than the alkynes.
Substituents on the positions para to the alkynes have no
steric impact on the cycloaromatization reaction, and are
located far enough from the developing radical centers
that interactions with these reactive species are negligible.
Such characteristics make these the ideal positions for
attachment of large solubilizing groups. Although they are
remote from the enediyne, we have found that substituents
at these positions do have a signi®cant in¯uence on the yield
of the Bergman reaction. While several researchers have
reported that electron-donating groups decrease the rate of
cycloaromatizations,9 we have found that along with lower-
ing the temperature at which cycloaromatization takes
place, such groups also have a salutary effect on the yield
of the reaction ;Scheme 3). Unlike other enediynes we have
studied, dialkoxy arenediynes10 even undergo quantitative
cycloaromatization in the injection port of the GC/MS! A
possible explanation for this result could lie in an increased
reactivity of the p-benzyne unit toward hydrogen-atom
abstraction,11 making it a more competitive alternative to
unproductive intermolecular radical couplings.
An aromatic enediyne can be substituted at three distinct
positions ;Fig. 1): On the positions para to the alkynes
;A), on the positions ortho to the alkynes ;B), or on the
alkynes themselves ;C). Because the chemistry of alkyne
functionalization is both well optimized and diverse,
substitution of the alkynes is particularly straightforward.
Unfortunately, this sort of substitution typically has a detri-
mental impact on the cycloaromatization reaction, with the
majority of disubstituted enediynes showing no cyclo-
aromatization products formed at temperatures in excess
of 2508C.6 A few exceptions to this trend are the dichloro-
dibromo- and diformyl-enediynes 3a±c, which undergo
cycloaromatization in very good yields at much more
synthetically amenable temperatures ;1808C, re¯uxing
o-dichlorobenzene/g-terpinene).7 However, even relatively
subtle changes to these systems appear to shut down the
Bergman reaction ;Scheme 2).
For example, while diformyl derivative ;3c) forms diformyl
naphthalene ;4c) cleanly upon cycloaromatization, the
sterically more demanding diketone 3d shows no cyclo-
Scheme 3.

D. M. Bowles et al. / Tetrahedron 57 +2001) 3753±3760
3755
Scheme 4.
Scheme 5.
The positions ortho to the alkynes reveal both the bene®ts
and disadvantages of the use of cycloaromatization as a
synthetic tool. While substituents at these positions have
little direct impact on the cycloaromatization reaction,
they are close enough to the radical centers generated by
cycloaromatization to react with them. This con¯uence of
attributes can be a double-edged sword: While further
cyclization initiated by the Bergman radical can be a useful
synthetic tool, it can also be a serious competing side-
reaction when further cyclization is not desired.
cycloaromatization of desilylated 13 showed complete
reaction within 2 h. However, GC/MS analysis of the reac-
tion mixture showed seven products in relatively equal
amounts, all of which appear to have arisen from various
side-chain reactions such as hydrogen abstraction, elimina-
tion and fragmentation ;Scheme 5).
In an attempt to avoid such adverse side-reactions we
prepared hexayne 14, which underwent an unprecedented
double cycloaromatization smoothly to yield anthracene 15
in 35% yield. Because double cycloaromatization represents
a new entry into the rapid synthesis of acenes, the low yield
for this cycloaromatization reaction is as disconcerting as it
is puzzling: The alkyl chains are remote enough that hydro-
gen abstraction is unlikely, and the geometry of the system
should preclude reaction of the p-benzyne radicals with the
adjacent alkynes. In order to further investigate the cyclo-
aromatization chemistry of such systems with a simpler
model, triyne 16 was prepared and subjected to cyclo-
aromatization at 1808C. The yield of the expected ethynyl
naphthalene 17 was only 40%; the bulk of the remaining
material ;an additional ,40%) appeared to be an insepar-
able mixture of dimeric species ;a broad peak on the GC
trace, with a mass corresponding to dimerized 17).
Although the actual structure of this by-product could not
We began our investigations of ortho-substituted systems by
preparing two simple arenediynes with small groups in the
ortho position, and discovered that the cycloaromatization
reaction proceeded quite rapidly and in good yield ;Scheme
4).12 Early reports on the cycloaromatization reaction
related ease of cycloaromatization to the distance between
the ends of the two alkynes.2 Calculations on the bromo- and
methyl-substituted enediynes 9 and 11 show that this
Ê
distance is slightly decreased ;from 4.07 to 4.03 A) by the
buttressing ortho substituents.13
Unfortunately, increasing the length of an ortho alkyl chain
beyond methyl leads to cycloaromatization reactions which
produce complicated mixtures. For example the double-
Scheme 6.

3756
D. M. Bowles et al. / Tetrahedron 57 +2001) 3753±3760
Scheme 7.
be determined by further spectroscopic analysis, the absence
of acetylenic carbons in this material led us to believe that
the dimerization occurred Diels±Alder reactions between
the alkyne and the naphthalene ring. In order to prevent
such reactions in future materials, we are currently prepar-
ing derivatives of 14 with signi®cantly bulkier groups on the
solubilizing alkynes ;Scheme 6).
20b yielded characterizable products following cyclo-
aromatization ;Scheme 7).
We next shifted our attention to a slightly longer pendant
group, which would lead to a 6-exo cyclization reaction with
the Bergman radical. While ortho-linked arenediynes
cycloaromatize in parallel to form substituted perylenes,14
whether this reaction occurred by the coupling of two
Bergman radicals ;Scheme 8, Path A) or by a radical
cascade mechanism ;Path B) was not clear. Because the
viability of the radical cascade is critical to the successful
formation of larger fused aromatic systems, we further
investigated this reaction by preparing bromonaphthyl
arenediyne 23. Reaction of this compound with tributyltin
hydride/AIBN generates a radical species, which mimics the
geometry and reactivity of the Bergman radical, and induces
a cycloaromatization reaction with the adjacent enediyne.
While the yield of perylene is not high ;36%), the signi®cant
side product is hydrostannylated enediyneÐan unfortunate
Of course, interactions between the Bergman radical and
functional groups on the aromatic core can also be bene-
®cial, since reactions of the ortho side-chain with the newly
generated radical centers are a rapid and ef®cient way to
construct multiple fused rings. Our ®rst efforts in this vein
focused on the preparation of acenaphthene derivatives.
Triyne 18 underwent cycloaromatization followed by
5-exo cyclization in excellent yield, providing 19 as the
only signi®cant product. This second cyclization step
appears to be highly dependent on the geometry of the
pendant alkyne, since neither the ketone 20a nor the alkene
Scheme 8.

D. M. Bowles et al. / Tetrahedron 57 +2001) 3753±3760
3757
result of the need to use tin-based reagents for this type of
cyclization. Thus the modest yield for this reaction can be
viewed as a minimum yield for such a radical cascade, and
in the absence of tin reagents Bergman-induced radical
cascades should prove effective synthetic tools for the
formation of peri-fused aromatic systems.
nol/THF overnight. The resulting compound was extracted
into hexanes, dried, and the solvent removed. The resulting
dark oil was immediately treated to the cyclization condi-
tions described above ;sealed tube method). Evaporation of
solvents left a dark oil, which was puri®ed by ¯ushing
through a thin pad of silica ;hexanes) to yield ;after evapora-
tion of solvent) 0.95g ;73%) of 2,3-dihexylnaphthalene as a
1
pale yellow oil. H NMR ;CDCl3, 400 MHz) d 0.8 ;t, 6H,
1. Conclusions
Jˆ6.6 Hz), 1.27 ;m, 8H); 1.65;m, 8H), 2.1 ;t, 4H, Jˆ
6.8 Hz), 7.42 ;dd, 2H, Jˆ9.3, 2.5Hz); 7.75 ;s, 2H), 7.82
;dd, 2H, Jˆ9.6, 2.1 Hz);. 13C NMR ;CDCl3, 100 MHz) d
14.07, 22.64, 28.58, 29.59, 31.88, 36.41, 125.23, 126.82,
127.25, 132.18, 138.94 ppm. FTIR ;neat) 3058,
2975cm 21. MS ;EI 70 eV) m/z 296 ;100%, M1).
Methodologies involving cycloaromatization are ef®cient
routes to substituted and fused aromatic systems. This
thermal reaction tolerates a wide range of functional groups,
many of which also increase the yield of the cycloaromati-
zation reaction. The reactive radical generated from
cycloaromatization can be employed to form further rings,
yielding either acenaphthenes or perylene derivatives.
2.3.2. 2,3-Dioctyloxynaphthalene /8c). 1,2-dioctyloxy-4,5-
bis;trimethylsilylethynyl)benzene ;2.0 g, 3.8 mmol)10 was
desilylated by stirring with tetrabutylammonium ¯uoride
in wet THF for 2 h. The compound was extracted into
hexanes, dried, and the solvent removed. The resulting oil
was subjected to cycloaromatization as described above
;sealed tube method, but only 1 h reaction time) to yield
after puri®cation ;silica plug, 3:1 hexanes/methylene
chloride) 1.3 g ;89%) of 8c as a yellow oil. 1H NMR
;CDCl3, 200 MHz) d 0.86 ;t, Jˆ6.6 Hz, 6H), 1.5±1.2
;broad m, 20H), 2.12 ;quintet, Jˆ6.6 Hz, 4H), 4.26 ;t,
Jˆ6.6 Hz, 4H), 7.11 ;s, 2H), 7.46 ;m, 2H), 7.78 ;m, 2H)
ppm. 13C NMR ;CDCl3, 5 0 MHz)d 13.94, 22.55, 25.91,
26.00, 28.98, 29.14, 31.71, 69.18, 108.02, 127.67, 129.33,
130.53, 149.55 ppm. MS ;EI 70 eV) m/z 384 ;25%, M1),
272 ;20%, M12octyl), 160 ;100%).
2. Experimental
2.1. General data
1H and 13C NMR spectra were recorded on Varian ;Gemini
200 MHz/Unity 400 MHz) spectrometers with tetramethyl-
silane as the internal standard. Mass spectral analyses were
performed in the EI-mode on a Varian Saturn GC/MS
spectrometer.
2.2. Techniques and materials
All experiments were carried out under nitrogen in freshly
distilled solvents under anhydrous conditions unless other-
wise noted. Commercial chemicals were used as supplied.
Tetrahydrofuran ;THF) and diethyl ether were puri®ed by
passage through a column of activated alumina. Preparative
column chromatography was performed with silica gel
;230±425mesh) from Fisher Scienti®c Company. Thin
layer chromatography ;TLC) was performed on glass-
backed silica gel 60 F254 plates from Fisher Scienti®c
Company. Radial chromatography was performed on a
Harrison Research Chromatotron, using 4 mm silica plates.
g-terpinene was purchased from Acros Organics. Because
desilylated enediynes are often prone to decomposition,
silylated enediynes were typically desilylated immediately
prior to cycloaromatization. Deprotection is assumed to be
quantitative, unless otherwise noted.
2.3.3. 1-Methylnaphthalene /12). 1.0 g ;3.5mmol) of 2,3-
bis;trimethylsilylethynyl)toluene15 was desilylated by treat-
ment with potassium carbonate in methanol. The resulting
enediyne was extracted into methylene chloride, dried, and
subjected to the cyclization conditions described above
;sealed tube). Chromatography on silica gel ;hexanes) led
to 0.35g of a colorless liquid ;70%). 1H NMR ;CDCl3,
200 MHz) d 3.11 ;s, 3H), 7.70 ;m, 1H), 7.76 ;d,
Jˆ4.2 Hz, 1H), 7.888 ;m, 2H), 8.08 ;d, Jˆ8.0 Hz, 1H),
8.22 ;d, Jˆ8.0 Hz, 1H), 8.36 ;d, Jˆ7.6 Hz, 1H) ppm. 13C
NMR ;CDCl3, 5 0 MHz)d 19.14 124.07, 125.49, 125.53,
125.67, 126.38, 126.55, 128.52, 133.61, 134.19 ppm. GC
retention time and MS fragmentation identical to a commer-
cial sample.
2.3.4. 1,4-Di-hept-1-ynyl-2,3,5,6-tetrabromobenzene. To
a ¯ame dried 100 mL two-neck ¯ask was added a solution
of 1-heptyne ;1.92 g, 20 mmol, 4 equiv.) in 30 mL THF.
After the solution was cooled to 08C, butyllithium
;19.9 mmol, 2.5M in Et 2O), was added dropwise. The
suspension was allowed to warm to 258C and stirred for
1 h, after which bromanil ;2.12 g, 5mmol) was added.
The solution was stirred at this temperature for 12 h, then
quenched with 10 mL saturated NH4Cl. The solution was
extracted with three portions of 100 mL hexanes, and the
organic layer dried and poured onto a thick pad of silica gel.
After ¯ushing the silica with 200 mL hexanes ;to remove
excess heptyne), the desired compounds were ¯ushed off the
silica with 250 mL ethyl acetate. The solvent was removed
to yield 3.0 g of a brown powder ;two spots by TLC in 3:1
hexanes/EtOAc). The two crude enantiomers and 3.79 g
2.3. General conditions used for cycloaromatization
A 0.01 M solution of the enediyne in o-dichlorobenzene/g-
terpinene ;10:1 v/v) was sparged with dry nitrogen for
20 min. For large scale reactions, a condenser was added
to the ¯ask, and the mixture heated to 1808C for 2 h. For
small scale reactions, this mixture was poured into a thick-
walled glass tube which was then capped with a te¯on screw
cap, and heated to 1808C for 2 h. After reaction, the high-
boiling solvents were removed on a rotary evaporator
;connected to a high-vacuum pump), and the residue puri-
®ed by chromatography or recrystallization.
2.3.1. 2,3-Dihexylnaphthalene /8a). 2.0 g of 1,2-dihexyl-
4,5-bis;trimethyl-silylethynyl)benzene ;4.5 mmol)10 was
desilylated by stirring with potassium carbonate in metha-

3758
D. M. Bowles et al. / Tetrahedron 57 +2001) 3753±3760
;20 mmol, 4 equiv.) of SnCl2 were taken up in 25mL aceto-
nitrile and two drops of H2O in a 100 mL ¯ask ®tted with a
re¯ux condenser. The reaction was heated to re¯ux for 12 h,
and after cooling, the precipitate was ®ltered away. The
solvent was removed, and the oily residue passed
through a pad of silica gel ;hexanes) to remove any
unreacted starting material. Evaporation of solvent
yielded 2.8 g of an orange oil ;96% yield). 1H NMR
;CDCl3, 200 MHz) d 1.45;t, Jˆ8 Hz, 3H), 1.87 ;m, 2H),
2.01 ;m, 2H), 2.19 ;m, 2H), 3.15;t, Jˆ8 Hz, 2H). 13C NMR
;CDCl3, 50 MHz): d 13.97, 19.79, 22.49, 28.09, 31.23,
80.99, 103.12, 127.75, 128.96. FTIR ;neat) 2952,
2141 cm21. MS ;EI 70 eV) m/z 582 ;80%, M1), 525
;100%, M12Bu).
give 133 mg ;40%) ethynylnaphthalene 17 as a yellow oil.
1H NMR ;CDCl3, 200 MHz) d 2.49 ;s, 3H), 3.44 ;s, 1H),
7.44 ;s, 1H), 7.62 ;s, 1H), 7.81 ;m, 2H), 8.38 ;m, 2H) ppm.
13C NMR ;CDCl3, 50 MHz) d 21.23, 81.49, 81.82, 119.59,
125.82, 126.03, 126.31, 126.50, 127.62, 128.43, 131.87,
133.35, 134.73 ppm. FTIR ;neat) 3345, 3050, 2920,
2250 cm21. MS ;EI 70 eV) m/z 166 ;80%, M1), 151
;100%, M12CH3).
2.3.8. 3-Bromo-4,5-bis/trimethylsilylethynyl)toluene and
3,4,5-tris/trimethylsilylethynyl) toluene /16). A solution
of 4.23 g ;10.0 mmol) of 3-bromo-4,5-diiodotoluene
;prepared as described in Ref. 14) in 50 mL of piperidine
was placed in a 100 mL screw cap glass tube. The solution
was sparged under N2 for 30 min, and bis[triphenylphos-
phine]palladium;II) chloride ;300 mg, 2 mol%), CuI
;190 mg, 10 mol%), and trimethylsilylacetylene ;3.11 mL,
2.2 equiv.) were added sequentially. The tube was then
sealed, and heated at 608C for 12 h. After cooling, the result-
ing amine salts were removed by ®ltration, and the clear
solution was evaporated to dryness. The resulting dark oil
was taken up in 20 mL hexanes, poured onto a thick pad of
silica gel, and separated into its two components with
hexanes eluent. Evaporation of the solvent from the ®rst-
eluted fraction afforded 2.5g ;70%) of desired diyne as a
yellow oil. 1H NMR ;CDCl3, 200 MHz) d 0.10 ;s, 9H), 0.12
;s, 9H), 0.39 ;s, 1H), 1.98 ;s, 3H), 7.10 ;s, 1H), 7.39 ;s, 1H)
ppm. 13C NMR ;CDCl3, 50 MHz) d 0.19, 0.07, 20.77, 98.82,
101.68, 102.81, 102.85, 124.58, 125.62, 127.15, 131.56,
132.95, 139.12 ppm. FTIR ;neat) 3300, 2923, 2150, 1711,
1592, 1534, 1249, 853 cm21. MS ;EI 70 eV) m/z 362 ;15%,
M1), 267 ;10%), 73 ;100%). Solvent removal from the
second chromatographic fraction gave 0.4 g ;10%) of the
2.3.5. 1,4-Di-hept-1-ynyl-2,3,5,6-tetrakis/trimethylsilyl-
ethynyl)benzene /14). 2.9 g ;5.0 mmol) of 1,4-di-hept-1-
ynyl-2,3,5,6-tetrabromobenzene was dissolved in 20 mL
piperidine and sparged under N2 for 30 min. Bis[triphenyl-
phosphine]palladium;II) chloride ;70 mg, 2 mol%), CuI
;95mg, 10 mol%), and trimethylsilylacetylene ;7.0 g,
50 mmol) were added and the reaction was heated to 508C
for 48 h. After cooling, the solvent was removed under
vacuum and the residue was taken up in 20 mL hexanes,
which puri®ed by silica gel chromatography ;hexane eluent)
to yield 2.8 g ;86%) of hexa-yne as an orange oil. 1H NMR
;CDCl3, 200 MHz) d 1.11 ;m, 36H), 0.88 ;m, 6H), 1.24 ;m,
12H), 1.65;m, 4H) ppm. 13C NMR ;CDCl3, 50 MHz) d
20.11, 13.90, 19.82, 22.15, 28.37, 31.15, 77.80, 100.75,
101.57, 104.13, 127.48, 128.45 ppm. FTIR ;neat) 3050,
2105, 1954, 1758, 1496, 1205 cm21. MS ;EI 70 eV) m/z
650 ;80% M1), 635;100%, M12CH3).
1
2.3.6. 9,10-Di-hept-1-ynylanthracene /15). 0.8 g ;1.2
mmol) of 1,4-di-hept-1-ynyl-2,3,5,6-tetrakis-;trimethylsilyl-
ethynyl)benzene was desilylated by stirring in methanolic
K2CO3 for 1 h. The resulting desilylated material was
extracted into hexanes, dried, evaporated, and immediately
taken up in 50 mL of benzene. After this solution was
sparged with nitrogen for 20 min, 5mL of 1,4-cyclohexa-
diene was added, the solution was transferred under N2 into
a steel reaction vessel and heated to 1808C for 2 h. The
vessel was cooled to room temperature and the solvent
was removed via rotary evaporation to give an orange
greasy solid. Recrystallization ;MeOH) gave 0.15g ;35%)
triyne as a colorless oil. H NMR ;CDCl3, 200 MHz) d
0.285;three overlapping peaks, 27H), 2.27 ;s, 3H), 7.23
;s, 2H). 13C NMR ;CDCl3, 5 0 MHz)d 20.13, 0.01, 20.76,
98.18, 101.74, 101.93, 103.08, 125.35, 126.19, 132.81,
137.66. FTIR ;neat) 3005, 2958, 2210, 2106, 1621, 1490,
1428, 1299, 632 cm21. MS ;EI 70 eV) m/z 380 ;20%, M1),
1
365;100% M 2CH3).
2.3.9. 2,3-Bis/trimethylsilylethynyl)-5-methylbenzalde-
hyde. A ¯ame dried 100 mL two-neck ¯ask under N2 was
cooled to 2788C and charged with 0.88 g ;2.42 mmol) of
3-bromo-4,5-bis-;trimethylsilylethynyl)-toluene in 50 mL
dry THF. The solution was stirred for 5min, after which
3.7 mL sec-BuLi ;4.8 mmol, 1.3 M in Et2O) was added
dropwise, and was stirred for 10 min. N-formyl morpholine
;0.7 g, 6.1 mmol, 2.5equiv.) was then added and the solu-
tion was allowed to stir until no further evidence of starting
material could be detected by TLC ;1:1 hexanes/CH2Cl2).
The reaction was quenched with 10 mL of 10% HCl,
extracted into hexanes, washed with H2O, dried, and passed
through a pad of silica gel ;1:1 hexanes/CH2Cl2). The
solvent was evaporated to yield 0.76 g ;100%) of the
1
of 15 as a waxy yellow±orange solid ;mp 205±2068C). H
NMR ;CDCl3, 200 MHz) d 0.95;t, Jˆ7.4 Hz, 6H), 1.44 ;m,
8H), 1.80 ;m, 4H), 2.73 ;t, Jˆ6.9 Hz, 4H), 7.56 ;t,
Jˆ8.3 Hz, 4H), 8.52 ;d, Jˆ8.3 Hz, 4H) ppm. 13C NMR
;CDCl3, 5 0 MHz)d 13.91, 20.12, 22.18, 28.68, 31.26,
65.78, 103.37, 126.26, 127.31, 132.20, 141.68 ppm.
FTIR ;neat) 3100, 2900, 1590, 1500, 1120, 610,
590 cm21. MS ;EI 70 eV) m/z 366 ;100%, M1), 309
;25%, M12butyl).
2.3.7. 1-Ethynyl-3-methylnaphthalene /17). 0.75g
;2 mmol) of 5-methyl-1,2,3-tris;trimethylsilylethynyl)ben-
zene ;16, obtained as a by-product as described below)
was desilylated by stirring in methanolic KOH overnight.
The compound was extracted into hexanes, then subjected
to cycloaromatization conditions as described above ;sealed
tube method). The solvent was removed, and the compound
puri®ed by chromatography on silica ;hexanes eluent) to
benzaldehyde as a
yellow oil. 1H NMR ;CDCl3,
200 MHz) d 0.02 ;s, 9H), 0.03 ;s, 9H), 1.99 ;s, 3H), 7.02
;s, 1H), 7.21 ;s, 1H), 10.10 ;s, 1H) ppm. 13C NMR ;CDCl3,
50 MHz) d 0.48, 0.40, 98.16, 99.10, 101.86, 105.39, 125.79,
126.85, 126.90, 135.86, 137.43, 138.11 ppm. FTIR ;neat)
3050, 2410, 2350, 1693, 1265, 749 cm21. MS ;EI 70 eV)
m/z 312 ;20%, M1), 297 ;100%, M12CH3), 269 ;25%) 238
;25%), 73 ;50%).

D. M. Bowles et al. / Tetrahedron 57 +2001) 3753±3760
3759
2.3.10. R,S-1-[5-Methyl-2,3-bis/trimethylsilylethynyl)-
phenyl]-prop-2-yn-1-ol. 0.60 g ;1.92 mmol) of 2,3-bis;tri-
methylsilylethynyl)-5-methylbenzaldehyde in 10 mL dry
ether was added to an ethereal solution of ethynyl
Grignard;20 mL of a 0.5M solution) at 0 8C, and stirred at
this temperature for 20 min. The reaction was then
quenched with 10 mL of a saturated aqueous NH4Cl,
extracted into hexanes, washed with water, dried, and
passed through a thin pad of silica gel with CH2Cl2. The
solvent was removed by rotary evaporation to yield 0.64 g
;99%) of a racemic mixture of triyne alcohols as a yellow
solid, mp 70±728C ;MeOH). 1H NMR ;CDCl3, 200 MHz) d
0.13 ;s, 18H), 2.10 ;s, 3H), 2.40 ;d, Jˆ2.2 Hz, 1H), 5.62 ;s,
1H), 6.99 ;s, 1H), 7.21 ;s, 1H) ppm. 13C NMR ;CDCl3,
50 MHz) d 0.36, 0.45, 20.90, 61.99, 73.86, 82.58, 97.66,
100.638, 103.25, 103.65, 120.32, 125.68, 126.77, 132.04,
138.30, 142.56 ppm. FTIR ;neat) 3111, 2995, 2407, 2219,
1355, 810, 755 cm21. MS ;EI 70 eV) m/z 338 ;25%, M1),
323 ;100%, M12CH3).
dry zinc chloride was added, and the solution warmed to
08C. 1-Bromo-8-iodonaphthalene ;1.66 g, 5mmol) and
bis[triphenylphosphine]palladium;II) chloride ;0.35g,
10 mol%) were added, and the solution allowed to warm
to room temperature overnight. The solution was then
poured into hexanes, extracted with water, dried, and passed
through a plug of silica ;hexanes). Recrystallization from
1
methanol led to pure 23 ;1.6 g, 60%), mp 138±1398C. H
NMR ;CDCl3, 200 MHz) d 20.24 ;s, 9H), 0.27 ;s, 9H), 1.32
;s, 12H), 7.4 ;m, 4H), 7.8 ;m, 3H) ppm. 13C NMR ;CDCl3,
5 0 MHz) d 20.64, 20.03, 30.96, 34.64, 97.08, 101.39,
102.63, 104.41, 120.12, 124.35, 124.44, 125.32, 125.86,
127.83 ;two overlapping peaks), 128.82, 129.15, 129.99,
130.85, 133.28, 135.73, 138.89, 146.23, 150.38 ppm.
FTIR ;KBr) 3052.4, 2959.2, 2893.2, 2147.6, 1242.7,
842.7, 753.4 cm21. MS ;EI 70 eV) m/z 532 ;30%, M1),
517 ;100%, M12CH3). The compound was desilylated by
treatment with potassium carbonate in methanol immedi-
ately prior to cyclization.
2.3.14. 2-tert-Butyl-perylene /24). 1.0 g ;2 mmol) of
compound 23 was dissolved in a mixture of 15mL methanol
and 5mL THF, to which a catalytic amount of potassium
carbonate had been added. Upon complete desilylation ;as
determined by GC/MS analysis), the mixture was poured
into hexanes, and extracted with water. Drying and evapora-
tion of the hexanes led to the crude desilylated diyne, which
was immediately dissolved in 200 mL of dry, degassed
benzene. The benzene solution was heated at re¯ux, and a
solution of tributyltin hydride ;1.1 equiv.) and AIBN
;4 equiv.) in benzene ;40 mL) was added by syringe pump
over a period of 6 h. The reaction mixture was allowed to
re¯ux for an additional 10 h and then cooled to room
temperature. The benzene was evaporated under vacuum,
and the resulting residue was dissolved in hexanes. Silica
gel chromatography ;hexanes) gave 24 as a yellow oil in
36% yield. 1H NMR ;CDCl3, 400 MHz) d 1.47 ;s, 9H), 7.45
;m, 4H), 7.65;m, 3H), 8.11 ;d, Jˆ7.2 Hz, 1H), 8.16 ;d,
Jˆ7.6 Hz, 1H), 8.21 ;d, Jˆ6.8 Hz, 1H), 8.29 ;s, 1H) ppm.
13C NMR ;CDCl3, 100 MHz) d 31.59, 35.13, 119.15,
119.79, 120.15, 120.29, 123.60, 126.64, 126.69, 126.75,
127.32, 127.90, 127.93, 128.09, 129.00, 130.99, 131.12,
131.56, 131.76, 135.01, 135.05, 149.21 ppm. FTIR ;KBr)
3048, 2959, 2924, 2850, 1464, 1258, 838, 768 cm21. MS ;EI
70 eV) m/z 308 ;75%, M1), 293 ;100%, M12CH3).
2.3.11.
R,S-1-/5-Methyl-2,3-diethynylphenyl)-prop-2-
yne-1-ol /18). To a racemic mixture of 1-[5-methyl-2,3-
bis;trimethylsilylethynyl)phenyl]-prop-2-yn-1-ol ;0.95g,
2.3 mmol) in 2 mL THF and 10 mL MeOH was added one
KOH pellet. After no evidence of starting material could be
detected by TLC ;5:1 hexanes/CH2Cl2), the reaction
mixture was extracted into 500 mL hexanes, washed with
water, brine, dried, and passed through a pad of silica gel
with 5:1 hexanes/CH2Cl2 eluent. The solvent was removed
to yield 0.43 g ;96%) of a racemic mixture of 18 as a
;marginally stable) yellow oil. 1H NMR ;CDCl3,
200 MHz) d 1.99 ;s, 3H), 2.42 ;s, 1H), 3.15;s, 1H), 3.42
;s, 1H), 4.10 ;broad s, 1H), 5.89 ;s, 1H), 6.97 ;s, 1H), 7.26 ;s,
1H) ppm. 13C NMR ;CDCl3, 50 MHz) d 20.68, 61.38,
74.35, 78.61, 81.16, 81.55, 82.52, 86.32, 119.60, 124.91,
127.16, 132.47, 138.77, 142.46 ppm. FTIR ;neat) 3090,
2955, 2460, 2210, 1290, 700, 605 cm21. MS ;EI 70 eV)
m/z 194 ;100, M1), 169 ;10%).
2.3.12. 7-Methyl-2-methylene-acenaphthen-1-ol /19). 1-
;5-Methyl-2,3-diethynylphenyl)prop-2-yne-1-ol 18 ;0.34 g,
1.74 mmol) in 50 mL benzene was sparged with N2 for 1 h.
1,4-Cyclohexadiene ;5mL, 10%/vol) was added and the
solution was sealed under N2 in a steel pressure apparatus.
The reaction mixture was heated to 1608C for 1.5h. After
cooling, the solvent was removed under vacuum and the
residue was dissolved in 5mL 1:1 hexanes/CH 2Cl2 and
separated into its components via radial chromatography
;2:1 hexanes/CH2Cl2). Compound 19 was isolated as a
yellow solid ;0.23 g, 65%), mp 105±1068C. 1H NMR
;CDCl3, 200 MHz) d 2.35 ;s, 3H), 5.44 ;s, 1H), 5.46
;s,1H), 5.70 ;s,1H), 7.11 ;s, 1H), 7.38 ;m,3H), 7.42 ;m,
1H) ppm. 13C NMR ;CDCl3, 50 MHz) d 22.61, 75.25,
109.01, 115.29, 122.03, 123.64, 124.10, 128.19, 130.96,
135.85, 137.43, 138.23, 142.22, 151.10 ppm. FTIR ;neat)
3350, 3100, 2990, 1750,1410, 600 cm21. MS ;EI 70 eV)
m/z 196 ;100%, M1), 181 ;8%, M12 CH3).
Acknowledgements
This work was supported by the National Science
Foundation ;CAREER Award CHE 9875123), the Of®ce
of Naval Research and the University of Kentucky Research
Foundation. NMR Facilities used in this research were
enhanced by funds from the CRIF program of the National
Science Foundation ;CHE 997841) and from the Research
Challenge Trust Fund of the University of Kentucky.
2.3.13.
1-Bromo-8-/5-tert-butyl-2,3-bis/trimethylsilyl-
ethynyl)phenyl)naphthalene /23). To 2.0 g ;5mmol) of
1-bromo-5-tert-butyl-2,3-bis;trimethylsilylethynyl)benzene13
in 50 mL dry THF at 2788C was slowly added 2.5mL of
n-BuLi ;2.0 M in hexanes). After 20 min, 0.7 g ;5mmol) of
References
1. ;a) Jones, R. R.; Bergman, R. G. J. Am. Chem. Soc. 1972, 94,
660. ;b) Bergman, R. G. Acc. Chem. Res. 1973, 6, 25 .

3760
D. M. Bowles et al. / Tetrahedron 57 +2001) 3753±3760
;c). Darby, N.; Kim, C. U.; SalauÈn, J. A.; Shelton, K. W.;
Takada, S.; Masamune, S. Chem. Commun. 1971, 1516.
Lichtenberger, F.; Shapiro, J.; Russell, K. C. Tetrahedron
Lett. 2000, 41, 6955.
2. Nicolaou, K. C.; Dai, W.-M. Angew. Chem., Int. Ed. Engl.
1991, 30, 1387 ;and references therein).
10. Zhou, Q.; Carroll, P. J.; Swager, T. M. J. Org. Chem. 1994, 59,
1294.
3. ;a) John, J. A.; Tour, J. M. J. Am. Chem. Soc. 1994, 116, 5011.
;b) John, J. A.; Tour, J. M. Tetrahedron 1997, 53, 15,515.
4. Smith Jr, W. D.; Babb, D. A.; Snelgrove, R. V.; Townsend III,
P. H.; Martin, S. J. J. Am. Chem. Soc. 1998, 120, 9078.
5. Grissom, J. W.; Gunawardena, G. U.; Klingberg, D.; Huang,
D. H. Tetrahedron 1996, 52, 6453.
11. ;a) Logan, C. F.; Chen, P. J. Am. Chem. Soc. 1996, 118, 2113.
;b) Schottelius, M. J.; Chen, P. J. Am. Chem. Soc. 1996, 118,
4896.
12. For the synthesis of bromoenediyne 9 and its cycloaromatiza-
tion, see Ref. 14. The yield of methylnaphthalene is actually
considerably higher than 70%, since signi®cant quantities of
the material were found to have distilled off with the high-
boiling solvents.
6. Lockhart, T. P.; Comita, P. B.; Bergman, R. G. J. Am. Chem.
Soc. 1981, 103, 4082.
13. Distances were calculated using the Sybyl force®eld in
MacSpartan by Wave function.
7. Bowles, D. M.; Anthony, J. E. Org. Lett. 2000, 2, 85 .
8. Grissom, J. W.; Calkins, T. L.; Egan, M. J. Am. Chem. Soc.
1993, 115, 1744.
14. Chow, S.-Y.; Palmer, G. J.; Bowles, D. M.; Anthony, J. E.
Org. Lett. 2000, 2, 961.
15. Perry, R. J.; Turner, S. R. J. Org. Chem. 1991, 56, 6573.
9. ;a) Schmittel, M.; Kiau, S. Chem. Lett. 1995, 953. ;b) Choy,
N.; Kim, C.-S.; Ballestero, C.; Artigas, L.; Diez, C.;