Rapid Bergman cyclization of 1,2-diethynylheteroarenes

Article abstract of DOI:10.1021/jo980879p

The synthesis and cyclization of acyclic quinoxaline, pyridine, and pyrimidine enediynes (1-3) are described. These compounds were prepared using palladium(0) coupling of trimethylsilyl acetylene to o-dihalo- or o- halotriflic heteroarenes. All compounds were prepared in modest to good yields. The enediynes prepared were shown to undergo Bergman cyclization. Kinetics over a minimum of 3 half-lives were used to construct ]Arrhenius plots. Pyrimidine 3 was found to have an activation energy of 16.1 kcal/mol. Cyclization of the closest known aromatic analogue, o-diethynylbenzene (15), has E(a) = 25.1 kcal/mol (Grissom, J. W.; Calkins, T. L.; McMillen, H. A.; Jiang, Y. J. Org. Chem. 1994, 59, 5833-5835). Pyridine 2 and quinoxaline 1 gave activation energies of 21.5 and 33.6 kcal/mol, respectively. The results illustrate that heteroarenes can be used to activate Bergman cyclization. We expect these compounds to play an important role in furthering the understanding of Bergman cyclization and in aiding the development of new biologically significant enediynes.

Full text of DOI:10.1021/jo980879p

J . Org. Chem. 1998, 63, 8229-8234
8229
Ra p id Ber gm a n Cycliza tion of 1,2-Dieth yn ylh eter oa r en es
Chang-Sik Kim and K. C. Russell*
Department of Chemistry, University of Miami, 1301 Memorial Drive, Coral Gables, Florida 33124
Received May 8, 1998
The synthesis and cyclization of acyclic quinoxaline, pyridine, and pyrimidine enediynes (1-3) are
described. These compounds were prepared using palladium(0) coupling of trimethylsilyl acetylene
to o-dihalo- or o-halotriflic heteroarenes. All compounds were prepared in modest to good yields.
The enediynes prepared were shown to undergo Bergman cyclization. Kinetics over a minimum of
3 half-lives were used to construct Arrhenius plots. Pyrimidine 3 was found to have an activation
energy of 16.1 kcal/mol. Cyclization of the closest known aromatic analogue, o-diethynylbenzene
(15), has Ea ) 25.1 kcal/mol (Grissom, J . W.; Calkins, T. L.; McMillen, H. A.; J iang, Y. J . Org.
Chem. 1994, 59, 5833-5835). Pyridine 2 and quinoxaline 1 gave activation energies of 21.5 and
33.6 kcal/mol, respectively. The results illustrate that heteroarenes can be used to activate Bergman
cyclization. We expect these compounds to play an important role in furthering the understanding
of Bergman cyclization and in aiding the development of new biologically significant enediynes.
In tr od u ction
Although originally studied in the early 1970s,1 Berg-
man cyclization received little attention until nearly a
decade and a half later. That changed when a number
of natural products incorporating a (Z)-3-ene-1,5-diyne
and possessing antibiotic anticancer activity were re-
ported.2 Since that time enormous effort has been poured
into elucidating the biological modes of action of these
novel natural products.3 The synthesis of the natural
enediynes4 and the development of simplified synthetic
analogues5 have also been pursued in earnest. However,
many factors which govern this enormously important
reaction remain unresolved.
allow investigation of both aromatic character and elec-
tronic effects but also provide synthetic entrance to
biologically interesting compounds. Herein, we give an
account of the synthesis of the first of these compounds,
novel arenediynes 1-3. The activation energies of these
compounds toward Bergman cyclization, including the
remarkable reactivity of 3, are described.
Bergman cyclization is the orbital-symmetry-allowed
rearrangement of a (Z)-3-ene-1,5-diyne to a 1,4-didehy-
drobenzene diradical (5, Scheme 1). Depending on its
stability, intermediate 6 may undergo dearomatization
to form either starting material (5) or a rearranged
enediyne (7). Recently, this has been elegantly demon-
strated for the aza-Bergman cyclization of C,N-dialky-
nylimines.7 Alternatively, in the presence of a suitable
radical trapping agent, the intermediate can be irrevers-
ibly quenched to maintain the new aromatic moiety (8).
Studies have shown that the ring size and strain
energy of cyclic enediynes and electronic factors in all
enediynes can influence the rate of diradical formation.8-11
Reports which directly address the electronic nature of
simple enediynes include the work of Schmittel and
Maier (Table 1). Schmittel demonstrated that electron-
withdrawing groups attached to the triple-bond termini
(9-11) lower the activation energy of cycloaromatization.
Although not proven, it was suggested that this was the
result of decreased steric repulsion between the cyclizing
Our group has recently begun a program to more
closely examine how aromaticity and electronic factors
influence the rate of Bergman cyclization. We have
started our work with an exciting new class of simple
enediynes, in which the double bond is adjacent to one
or more heteroatoms (e.g., 1-4). Such targets not only
* To whom correspondence should be addressed. Phone: (305) 284-
6856. Fax: (305) 284-1880. E-mail: krussell@umiami.ir.miami.edu.
Internet: http://gilligan.cox.miami.edu.
(1) (a) Bergman, R. G. Acc. Chem. Res. 1973, 6, 25-31. (b) J ones,
R. R.; Bergman, R. G. J . Am. Chem. Soc. 1972, 94, 660-661.
(2) (a) Edo, K.; Mizugaki, M.; Koide, Y.; Seto, H.; Furihata, K.;
Otake, N. Tetrahedron Lett. 1985, 26, 331-334. (b) Lee, M. D.; Dunne,
T. S.; Siegel, M. M.; Chang, C. C.; Morton, G. O.; Borders, D. B. J .
Am. Chem. Soc. 1987, 109, 3464-3466. (c) Golik, J .; Clardy, J .; Dubay,
G.; Groenewold, G.; Kawaguchi, H.; Konishi, M.; Krishman, B.;
Ohkuma, H.; Saitoh, K.; Doyle, T. W. J . Am. Chem. Soc. 1987, 109,
3461-3462. (d) Golik, J .; Dubay, G.; Groenewold, G.; Kawaguchi, H.;
Konishi, M.; Krishman, B.; Ohkuma, H.; Saitoh, K.; Doyle, T. W. J .
Am. Chem. Soc. 1987, 109, 3462-3464. (e) Masataka, K.; Ohkuma,
H.; Matsumoto, K.; Takashi, T.; Kamei, H.; Miyaki, T.; Oki, T.;
Kawaguchi, H. J . Antibiot. 1989, 42, 1449-52.
(3) For a review, see: Nicolaou, K. C.; Dai, W.-M. Angew. Chem.,
Int. Ed. Engl. 1991, 30, 1387-1416.
(6) (a) Hoffner, J .; Schottelius, M. J .; Feichtinger, D.; Chen, P. J .
Am. Chem. Soc. 1998, 120, 376-385. (b) David, W. M.; Kerwin, S. M.
J . Am. Chem. Soc. 1997, 119, 1464-1465.
(7) Nicolaou, K. C.; Zuccarello, G.; Ogawa, Y.; Schweiger, E. J .;
Kumazawa, T. J . Am. Chem. Soc. 1988, 110, 4866-4868.
(8) Magnus, P.; Fortt, S.; Pitterna, T.; Snyder, J . P. J . Am. Chem.
Soc. 1990, 112, 4986-4987.
(9) Magnus, P.; Carter, P. A. J . Am. Chem. Soc. 1988, 110, 1626-
1628.
(10) Magnus, P.; Annoura, H.; Harling, J . J . Org. Chem. 1990, 55,
1709-1711.
(11) Schmittel, M.; Kiau, S. Chem. Lett. 1995, 953-954.
I
(4) For the total synthesis of calicheamicin γ1 , see: (a) Nicolaou,
K. C. Angew. Chem., Int. Ed. Engl. 1993, 32, 1377-1385. (b) Hitchcock,
S. A.; Boyer, S. H.; Chu-Moyer, M. Y.; Olson, S. H.; Danishefsky, S. D.
Angew. Chem., Int. Ed. Engl. 1994, 33, 858-862. For the total
synthesis of dynemicin A, see: (c) Meyers, A. G.; Fraley, M. E.; Tom,
N. J .; Cohen, S. B.; Madar, D. J . Chem. Biol. 1995, 2, 33-43. (d) Shair,
M. D.; Yoon, T. Y.; Mosny, K. K.; Chou, T. C.; Danishefsky, S. J . J .
Am. Chem. Soc. 1996, 118, 9509-9525.
(5) For reviews, see: (a) Maier, M. E. Synlett 1995, 13-26. (b)
Nicolaou, K. C.; Smith, A. L.; Yue, E. W. Proc. Natl. Acad. Sci., U.S.A.
1993, 90, 5881-5888.
10.1021/jo980879p CCC: $15.00 © 1998 American Chemical Society
Published on Web 10/27/1998

8230 J . Org. Chem., Vol. 63, No. 23, 1998
Kim and Russell
Ta ble 1. c-d Dista n ces a n d Kin etic Da ta for Ber gm a n Cycliza tion of Selected En ed iyn es
Sch em e 1. Ber gm a n Cycloa r om a tiza tion
The activation energy determined by Grissom et al. for
the cyclization of acyclic aromatic enediyne 15 was found
to be lower than that for nonaromatic 14, suggesting that
an aromatic ring in place of the enediyne double bond
has no significant influence on the rate of cyclization.1,15
The situation for cyclic enediynes does not appear to
parallel the acyclic case. Ten-membered ring compounds
exhibit significantly different rates of cyclization, with
the aromatic analogues being slower. For example,
nonaromatic 16 cyclizes with a reported half-life of 18 h
at 37 °C whereas aromatic 17 cyclizes with a half-life of
24 h at 84 °C.13,16 The activation energies for 16 and 17
have not been measured.
Semmelhack and Nicolaou have prepared a series of
quinone/dihydroquinone enediyne pairs (18-23).13,17 In
each case, the quinone was found to react significantly
faster than the dihydroquinone analogue. Even when the
quinone is removed from the enediyne by an aromatic
ring (23), the cyclization rate was faster. This substanti-
ates the position that electron-withdrawing groups as-
sociated with the double bond can facilitate Bergman
cyclization. Anthraquinone 24 suggests that the situa-
tion is actually more complicated.18 The enediyne is
attached to an electron-rich furan, which itself is attached
to the electron-deficient anthraquinone. This molecule
shows unexpected reactivity, cyclizing nearly twice as fast
in-plane π orbitals.14 Maier hypothesized that the electron-
donating substituent in 12 inhibited cyclization relative
to 13 by stabilizing the ground state of the starting
material more than destabilizing the transition state.12
It has been proposed that the amount of double-bond
character in the double bond along with the distance
between the diyne termini (c-d , Scheme 1) determines
the activation energy required for cyclization.13 This was
based on the fact that arenediynes generally gave longer
half-lives than their nonaromatic counterparts.14 Al-
though the link between c-d distance and cyclization
has been firmly established, the evidence supporting the
double-bond character is less clear.
The activation energies and half-lives for some aro-
matic and nonaromatic enediynes are given in Table 1.
(15) Grissom, J . W.; Calkins, T. L.; McMillen, H. A.; J iang, Y. J .
Org. Chem. 1994, 59, 5833-5835.
(16) Nicolaou, K. C.; Zuccarello, G.; Riemer, C.; Estevez, V. A.; Dai,
W.-M. J . Am. Chem. Soc. 1992, 114, 7360-7371.
(17) Nicolaou, K. C.; Liu, A.; Zeng, Z.; McComb, S. J . Am. Chem.
Soc. 1992, 114, 9279-9282.
(18) Meyers, A. G.; Dragovich, P. S. J . Am. Chem. Soc. 1992, 114,
5859-5860.
(12) Maier, M. E.; Greiner, B. Liebigs Ann. Chem. 1992, 855-861.
(13) Semmelhack, M. F.; Neu, T.; Foubelo, F. J . Org. Chem. 1994,
59, 5038-5047.
(14) Semmelhack, M. F.; Neu, T.; Foubelo, F. Tetrahedron Lett. 1992,
33, 3277-3280.

Rapid Bergman Cyclization
J . Org. Chem., Vol. 63, No. 23, 1998 8231
Ta ble 2. Syn th esis of P r otected a n d Dep r otected Heter oa r en ed iyn es
coupling prod
(% yield)
deprotection prod
(% yield)
entry
reactant
catalysta
solvent
temp, °C
time, h
1
2
3
4
5
6
7
8
9
25a
26a
26a
26a
26a
26b
26b
27c
27c
27c
27c
27c
A
A
A
B
B
A
B
A
A
A
B
B
Et3N
Et3N
Et3N
100
170
197
140
170
100
100
117
130
152
130
120
10
48
27
1.5
15
not isolated
26d (trace)
26d (20)
26d (67)
26e (34)
26e (19)
26e (47)
27d (32)
27d (9)
27d (6.7)
27e (49)
27d (7.9)b
27e (61)b
27d (2.4)b
27e (35)b
1 (71)
i-Pr2NH
i-Pr2NH
i-Pr2NH
i-Pr2NH
i-Pr2NH
i-Pr2NH
i-Pr2NH
i-Pr2NH
i-Pr2NH
26f (40)
2 (48)
16
16
2 (50)
4.3
4.3
4.3
4.2
5.3
27f (78)c
27f (78)c
27f (78)c
3 (77)d
10
11
12
3 (77)d
13
27c
B
i-Pr2NH
152
4.3
3 (77)d
A: 10% w/w Pd(PPh3)2Cl2, 10% w/w CuI. B: 10% w/w Pd(PPh3)4. Separated prior to deprotection. c Entries 8-10 were combined for
a
b
d
deprotection. Entries 11-13 were combined for deprotection.
as quinone 21, which contains no intervening aromatic
moiety. Certainly, more studies are needed to ferret out
the causes of the apparent anomalous behavior of these
compounds.
Sch em e 2
To explore how both electronics and aromaticity can
influence Bergman cyclization, we have turned our
attention to the heteroarenediynes in this study. The
fusion of heteroaromatics with enediynes also has the
advantage of providing a convenient handle for further
modification through the use of the heteroatom lone
pairs. Furthermore, these compounds give easy access
to biological enediyne chimeras, which may find utility
as biological probes or pharmaceuticals. Although re-
ports of heteroarenediynes have begun to surface,18,19 no
accounts of the activation energies of these compounds
have been described.
Pyridine 2 was prepared by two separate routes
(entries 2-7). Coupling dichloride 26a with Pd(PPh3)2-
Cl2/CuI catalyst in Et3N at 170 °C gave only a trace of
monoacetylene 26d after 2 days of reaction (entry 2).
Elevating the temperature to 197 °C resulted in a 20%
yield of 26d in 27 h (entry 3). Changing to the Pd(PPh3)4
catalyst and i-Pr2NH solvent resulted in a good yield of
26d , with no detectable diacetylene adduct (entry 4).
Notably, the Pd(PPh3)4 catalyst required both lower
temperatures (140 °C) and shorter reaction times (1.5 h)
than did the Pd(PPh3)2Cl2 catalyst. Increasing the tem-
perature to 170 °C with the Pd(PPh3)4 catalyst furnished
2 in a disappointing 16% overall yield after cleavage of
the TMS groups (entry 5). The lack of reactivity of 26a
relative to 25a is not unexpected. It has been established
that positions ortho or para to the ring nitrogen of
pyridine are usually reactive enough with chlorine to give
acceptable yields of coupled products whereas meta
positions require bromine, iodine, or triflates for sufficient
reactivity.22 Although some arenediynes have been
reported to decompose before they cyclize,13,14,19 one cause
for the low yields may be the reactivity of 26e and 2
toward Bergman cyclization at the coupling temperature.
Although no efforts were made to identify cyclization
products from the coupling reactions, the presence of
substantial amounts of insoluble black material is in
agreement with the polymers formed from cyclization.23
Resu lts a n d Discu ssion
Syn th esis. Targets 1-3 were selected so that they
could be directly synthesized from commercially available
starting materials via palladium-mediated coupling strat-
egies. Although the reactivity of halogens toward Stille
or Suzuki couplings is known to be I > Br . Cl,20 we
were somewhat limited by the lack of commercially
available o-dihaloheteroarenes, with the exception of
some dichloro compounds. As an alternative, we turned
to triflates, which have also enjoyed success in palladium-
catalyzed acetylene couplings,21 and these allowed access
to our enediynes via the heteroaryl alcohols. The results
of our coupling reactions are summarized in Scheme 2
and Table 2.
The synthesis of 1 proceeded smoothly from com-
mercially available 2,3-dichloroquinoxaline (25a ) through
couplings at moderate temperatures (entry 1). No
monoacetylene products were observed under these
conditions. The final enediyne was obtained without
isolation of the TMS intermediate in 71% yield. Because
the yield of 1 was acceptable, no further efforts were
made to improve the synthesis either by changing the
reaction conditions or by using other halogens or triflate
couplings. Recently, an alternate synthesis of 1 was
reported using a condensation strategy.19
(19) Faust, R.; Weber, C.; Fiandanese, V.; Marchese, G.; Punzi, A.
Tetrahedron 1997, 53, 14655-14670.
(20) (a) Stille, J . K. Angew. Chem., Int. Ed. Engl. 1986, 25, 508-
524. (b) Stille, J . K. Pure Appl. Chem. 1985, 57, 1171-1180.
(21) Ritter, K. Synthesis 1993, 735-762.
(22) Sakamoto, T.; Shiraiwa, M.; Kondo, Y.; Yamanaka, H. Synthesis
1983, 312-314.
(23) J ohn, J . A.; Tour, J . M. J . Am. Chem. Soc. 1994, 116, 5011-
5012.

8232 J . Org. Chem., Vol. 63, No. 23, 1998
Kim and Russell
Ta ble 3. Cycliza tion Ra tes (kobs) a n d Ha lf-Lives (τ1/2) for
Heter oa r en ed iyn es 1-3 in CCl4
Sch em e 3
compd
temp, °C
kobs, s-1
τ
1/2, min
R
1a
140
152
162
175
1.7 × 10-5
7.8 × 10-5
1.6 × 10-4
4.5 × 10-4
2.5 × 10-4
3.8 × 10-4
8.9 × 10-4
2.0 × 10-3
680
148
72
0.994
0.994
0.992
0.994
26
2b
135
145
155
175
46
30
13
6
0.991
0.993
0.996
0.997
3c
139
154
173
9.2 × 10-4
1.6 × 10-3
4.1 × 10-3
13
7
3
0.993
0.981
0.992
Reaction of isolated monoacetylene 26d with excess
TMS-acetylene under the same conditions as in entry 5
also resulted in a poor yield of 26e (not shown) with no
recovered starting material. An attempt was also made
to couple a second acetylene to 26d after protection of
the triple bond as its cobalt carbonyl complex. Unfortu-
nately, coupling at 140 °C for 10 h with Pd(PPh3)4
catalyst and i-Pr2NH only resulted in recovery of uncom-
plexed starting material.
Alternatively, 2-bromo-3-pyridinol was converted to
triflate 26b and coupled to an excess of TMS-acetylene
(entries 6 and 7). As with 26a , the coupling was catalyst
dependent, with Pd(PPh3)2Cl2-CuI being inferior to Pd-
(PPh3)4, the yield of diacetylene 26e at 100 °C over 10 h
was 19% and 47%, respectively, for the two catalysts.
After TMS deprotection with NaF-HF buffer, pure 3 was
obtained in 50% yield (24% for the two steps).
a
b
Ea ) 33.6 kcal/mol (R ) 0.994). Ea ) 21.5 kcal/mol (R )
0.994). c Ea ) 16.1 kcal/mol (R ) 0.993).
The synthesis of pyrimidine analogue 3 was accom-
plished by iodination of 6-chloro-2,4-dimethoxypyrimidine
with N-iodosuccinimide in trifluoroacetic acid24 to yield
the dihalopyrimidine (27c) in 98% yield. Coupling reac-
tions were carried out using the conditions shown in
entries 8-13. The use of the Pd(PPh3)2Cl2-CuI catalyst
gave selective monosubstitution for the iodine (27d ), and
no desired diacetylene adducts were detected (entries
8-10). Elevation of the temperature did not improve the
situation: only lower yields of 27d were obtained. The
observed selectivity is likely based on the reactivity of
iodine vs chlorine in these reactions. Literature prece-
dent has shown there to be little difference in reactivity
among 2, 4, and 5 positions of substituted pyrimidines
toward palladium-catalyzed couplings.25 This is contrary
to those observations relating activating and deactivating
positions in pyridines. The most efficient coupling condi-
tions used the Pd(PPh3)4 catalyst and i-Pr2NH as the
solvent. At 120 °C, 27e was generated in 61% yield along
with 8% 27d (entry 12). At 130 °C, the yield of 27e was
lower, with a smaller amount of monosubstituted mate-
rial present (entry 11). At 152 °C, the yields of 27d and
27e were even lower (entry 13). This may be indicative
of a competition between coupling and cyclization or of
decomposition reactions. Again, no starting material was
left at the end of the coupling reactions. Pure 27g was
isolated after separation of the TMS products and depro-
tection.
F igu r e 1. Plot of the rate of disappearance of 1 vs time in
CCl4 at various temperatures.
rophenazine (1a ). No other characterizable products
were isolated. However, the presence of 2-chlorophena-
zine was detected by GC-MS. The cyclization product
of pyridine 2 was troublesome to isolate. The reaction
led to large amounts of dark, insoluble materials. These
may be polyquinoline-type compounds, as previously
discussed. The identity of 2a was only confirmed by GC-
MS and was the only identifiable product. The major
cyclization product of pyrimidine 3 was substituted
quinazoline 3a , isolated in 15% yield.
Kin etics. The kinetics of 1-3 were measured, and
their activation energies were calculated from the Ar-
rhenius relationship as has been done for other arene-
diynes.15 All cyclization reactions showed first-order
kinetics over a minimum of 3 half-lives and a tempera-
ture range of 35-40 °C. The data are summarized in
Table 3. Figure 1 shows the rates of cyclization of
quinoxaline 1, which was the slowest of the three
compounds examined. The activation energy calculated
from the Arrhenius plot gave a value of 33.6 kcal/mol.
That is less reactive than o-diethynylbenzene (15, Ea )
25.1 kcal/mol) and more comparable to the energies
reported for terminally substituted acyclic enediynes.1 In
this case, it appears that the acceleration from the
heteroatoms is more than compensated for by the pres-
ence of a second aromatic ring. Pyridine 2 showed
enhanced reactivity (Ea ) 21.5 kcal/mol) which was
somewhat faster than 15. Introduction of a second
nitrogen into the arene, 3, further accelerated cyclization,
Cycliza tion . Enediynes 1-3 were examined for their
propensity to undergo Bergman cyclization (Scheme 3).
Cyclizations were conducted by heating pure samples of
the enediynes in CCl4 at 120 °C for 20 h. The products
were isolated by column chromatography or HPLC.
Cyclization of quinoxaline 1 yielded 81% of 2,5-dichlo-
(24) Bhatt, R. S.; Kundu, N. G.; Chwang, T. L.; Heidelberger, C. J .
Heterocycl. Chem. 1981, 18, 771-774.
(25) Edo, K.; Sakamoto, T.; Yamanaka, H. Chem. Pharm. Bull. 1978,
26, 3843-3850.

Rapid Bergman Cyclization
J . Org. Chem., Vol. 63, No. 23, 1998 8233
Gen er a l Meth od for TMS Dep r otection . The product
from the coupling reaction was dissolved in ethanol, and excess
hydrogen fluoride-sodium fluoride buffer solution (pH 5.5) was
added. The solution was stirred at room temperature over-
night. The product was extracted with ethyl acetate, and the
organic extracts were dried over anhydrous magnesium sul-
fate. The drying agent was filtered, and the volatiles were
removed by rotary evaporation under reduced pressure. The
products were isolated by flash chromatography as described
in the individual experiments.
even with the presence of two donating oxygen atoms (Ea
) 16.1 kcal/mol). This compound is the most reactive
acyclic arenediyne reported to date. The lowest activa-
tion energy reported for a Bergman cyclization is 12.3
kcal/mol.26 This was for an enediyne activated by metal
coordination and resulted in severe c-d contraction.
Con clu sion
2,3-Dieth yn ylqu in oxa lin e (1). The title compound was
prepared as described in the general procedure on the following
scale. Dichloroquinoxaline (25a ; 1.3 g, 6.6 mmol), dichlorobis-
(triphenylphosphine)palladium (130 mg), copper(I) iodide (130
mg), and trimethylsilylacetylene (2.3 mL, 17 mmol, 2.5 equiv)
were reacted for 13 h at 80 °C. The crude product was filtered
through a plug of silica gel using EtOAc/Hex (1:9, v/v) and
directly deprotected. The product was isolated by flash
chromatography using the same solvent as above. The title
compound was isolated as a tan solid (830 mg, 71% yield) after
We have reported the synthesis of novel heteroarene-
diynes 1-3 in reasonable yields. All of these molecules
were shown to undergo thermal Bergman cyclization. The
rate of cyclization was shown to be accelerated relative
to the rates of simple arenediynes by the incorporation
of heteroatoms into the aromatic ring. Pyrimidine 3, in
particular, showed remarkable reactivity. This com-
pound has the greatest intrinsic reactivity of any reported
acyclic arenediyne. This raises the possibility for the
preparation of acyclic enediynes with reactivity compa-
rable to those which use c-d contraction to lower the
activation energy of cyclization. These investigations
open the door to rate modulation of enediynes by pH
dependence or Lewis acid complexation. Conversion of
these enediynes into the more reactive 10-membered ring
analogues should give these compounds reasonable reac-
tion rates under physiological conditions. Furthermore,
even greater reactivity should be possible by the hydroly-
sis of 3 to afford uracil 4, which may have a number of
important biological applications. The synthesis and
kinetics of 4, as well as the ability of 1-3 to cleave DNA,
will be reported elsewhere.
1
two columns: mp 140 °C (dec); H NMR (CDCl3) δ 8.05 (dd,
2H, J ) 3.6, 6.4 Hz), 7.79 (dd, 2H, J ) 3.6, 6.4 Hz), 3.56 (s,
2H); 13C NMR (CDCl3) δ 140.6, 139.6, 131.4, 129.0, 83.4, 80.0;
FTIR (KBr) νmax 3266, 2104 cm-1
. EIMS calcd: 178.05.
Found: 178. Anal. Calcd for C12H6N2: C, 80.89; H, 3.39; N,
15.72. Found: C, 80.74; H, 3.41; N, 15.61.
2-Br om o-3-tr iflic P yr id in a te (26b). 2-Bromo-3-pyridinol
(2.00 g, 11.6 mmol) was dissolved in dry dichloromethane (30
mL) in a two-neck round-bottom flask fitted with a rubber
septum and a gas inlet. The solution was cooled to 0 °C under
an atmosphere of Ar, and triethylamine (1.8 mL) was added,
followed by the slow addition of triflic anhydride (2.3 mL) via
syringe. The mixture was allowed to warm to room temper-
ature and to stir overnight. The volatiles were removed by
rotary evaporation under reduced pressure, and the black, oily
solution was subjected to flash chromatography (EtOAc/Hex/
1:10, v/v), to give a clear, yellowish oil (3.3 g, 93%): 1H NMR
(CDCl3) δ 8.36 (d, 1H, J ) 4.4 Hz), 7.64 (d, 1H, J ) 8.4 Hz),
7.36 (dd, 1H, J ) 8.4, 4.4 Hz); 13C NMR (CDCl3) δ 149.0, 144.6,
135.6, 130.2, 124.0, 118.4 (q, C-F coupling; J ) 320 Hz); FTIR
Exp er im en ta l Section
Gen er a l. All commercial chemicals were purchased from
Aldrich, were ACS-certified grade, and were used without
further purification unless otherwise noted. Triethylamine
and N,N-diisopropylamine were distilled from phosphorus
pentoxide prior to use. Carbon tetrachloride was purified by
filtration through basic alumina. 1H and 13C NMR spectra
were recorded at 400 and 100 MHz, respectively. Chemical
shifts are recorded in parts per million on the δ scale
referenced to the solvent peak as an internal standard. Thin-
layer chromatography was conducted on Merck F254 silica gel
TLC plates with fluorescent indicator. Normal-phase HPLC
was performed with a spheri-5 silica column (5 µ, 250 × 46
mm) for purification and kinetic studies. Flash chromatog-
raphy was performed with Baker silica gel (40 µm). Melting
points are uncorrected. All palladium coupling reactions were
performed in Ace pressure tubes with the added protection of
a blast shield.
Gen er a l Exp er im en ta l P r oced u r e for P a lla d iu m Cou -
p lin g Rea ction s. Triethylamine or N,N′-diisopropylamine
(13 mL) was placed in a 15-mL Ace pressure tube and degassed
by bubbling with an Ar stream for 10 min. The corresponding
dihalide or halotriflate (3.0 mmol), bis(triphenylphosphine)-
palladium chloride (10% w/w) and copper(I) iodide (10% w/w),
or tetrakis(triphenylphosphine)palladium(0) (10% w/w) were
added to the solvent, and the slurry was stirred and bubbled
with Ar for an additional 1 or 2 min. Trimethylsilylacetylene
(7.5 mmol, 2.5 equiv) was added to the solution. The pressure
tube was sealed and placed in a preheated oil bath behind a
blast shield. The reaction mixture was stirred at the temper-
atures and for the times listed in Table 2. After being cooled
to room temperature, the product was isolated for character-
ization by flash chromatography or filtered through a plug of
silica gel using EtOAc/Hex (1:9, v/v) prior to deprotection.
(KBr) νmax 3072, 1575, 1411 cm-1
Found: 305.
. GC-MS calcd: 304.9.
2,3-Dieth yn ylp yr id in e (2). Meth od A. 2,3-Dichloropy-
ridine (26a ; 500 mg, 3.4 mmol), trimethylsilylacetylene (1.2
mL, 8.5 mmol), and tetrakis(triphenylphosphine)palladium (50
mg) were prepared in diisopropylamine (13 mL) as described
in the general procedure. The reaction mixture was stirred
at 170 °C for 13 h. The TMS derivative (26d ; 313 mg, 1.2
mmol, 34%) was isolated by flash chromatography using
EtOAc/Hex (1:9, v/v) as the eluent. The product was subjected
to deprotection as described above and purified by flash
chromatography using EtOAc/Hex (1:9, v/v) as the eluent to
give the title compound as a white solid (70 mg, 48% for
deprotection, 16% overall).
Meth od B. 2-Bromo-3-triflic pyridinate (26b; 100 mg, 0.3
mmol), trimethylsilylacetylene (0.24 mL, 1.7 mmol), and Pd-
(PPh3)4 (10 mg) in diisopropylamine (4 mL) were stirred at
100 °C for 13 h to give 26e (42 mg, 0.16 mmol, 47%) as a brown
oil. After deprotection and isolation as described in method
A, 2 was isolated as a white solid (10 mg, 50% yield): mp 128-
130 °C (dec); 1H NMR (CD2Cl2-d2) δ 8.53 (dd, 1H, J ) 8.1, 1.6
Hz), 7.81 (dd, 1H, J ) 4.8, 1.6 Hz), 7.27 (dd, 1H, J ) 8.0, 4.8
Hz), 3.51 (s, 1H), 3.42 (s, 1H); 13C NMR (acetone-d6) δ 150.3,
145.2, 140.6, 123.9, 123.0, 85.9, 82.5, 82.3, 80.4; FTIR (CCl4)
νmax 3310, 2125 cm-1. GC-MS calcd: 127.04. Found: 127.
Anal. Calcd for C9H5N: C, 85.02; H, 3.96; N, 11.02. Found:
C, 84.76; H, 4.02; N, 10.84.
6-Ch lor o-2,4-dim eth oxy-5-iodopyr im idin e (27c). A mix-
ture of 6-chloro-2,4-dimethoxypyrimidine (100 mg, 0.57 mmol),
trifluoroacetic acid (5.0 mL), and trifluoroacetic anhydride (1.0
mL) was refluxed for 30 min. N-Iodosuccinimide (128 mg,
0.572 mmol) was added, and the reaction mixture was refluxed
overnight. The solution was cooled to room temperature, and
the solvent was evaporated under reduced pressure. Ethanol
(20 mL) was added and evaporated under reduced pressure.
(26) Warner, B. P.; Millar, S. P.; Broene, R. D.; Buchwald, S. L.
Science 1995, 269, 814-816.

8234 J . Org. Chem., Vol. 63, No. 23, 1998
Kim and Russell
The residue was dissolved in chloroform (20 mL) and washed
with a saturated sodium bicarbonate solution (100 mL). The
organic phase was dried over anhydrous magnesium sulfate
and filtered, and the filtrate was evaporated under reduced
pressure. The title compound, a white solid, was collected and
dried under high vacuum (167.5 mg, 98%): mp 126 °C; 1H
NMR (CDCl3) δ 4.02 (s, 3H), 3.98 (s, 3H); 13C NMR (CDCl3) δ
177.6, 170.7, 164.6, 164.5, 55.9, 55.5; IR (CCl4) νmax 1554, 1538,
161.9, 150.4, 133.0, 129.7, 126.4, 55.12, 55.09.27 GC-MS
calcd: 258.0. Found: 258.
P r oced u r e for th e Ra te Deter m in a tion for En ed iyn es
1-3. Meth od A: HP LC . Stock solutions of 6.0 mM 1 (10.6
mg, 0.06 mmol) and 6.0 mM 2 (7.6 mg, 0.060 mmol) in CDCl3
containing naphthalene (22.9 mg, 0.18 mmol for 1; 7.7 mg, 0.06
mmol for 2) as an internal standard were prepared in 10.00-
mL volumetric flasks. Aliquots of the solutions (50 µL each)
were transferred into melting-point tubes and sealed under
vacuum. Twenty sample tubes were prepared for each of four
temperatures (140 ( 1, 152 ( 1, 162 ( 1, and 175 ( 1 °C for
1; 2: 140 ( 1, 145 ( 1, 155 ( 1, and 175 ( 1 °C for 2). Tubes
were removed from the oil baths at regular intervals and
quickly cooled in water to room temperature. The reaction
course was followed by HPLC using EtOAc/Hex (25:75, v/v)
and a flow rate of 1.5 mL/min on a spheri-5 silica gel column.
The disappearance of starting material versus the internal
standard was measured at wavelengths of 254 and 261 nm
for 1 and 2, respectively.
Meth od B: NMR. A 6.0 mM stock solution of 3 (11.2 mg,
0.06 mmol) was prepared in CDCl3 with 0.05 M CH3CN in a
10.00-mL volumetric flask. Aliquots of the stock solution (1.0
mL) were transferred to NMR tubes, and the tubes were flame-
sealed under vacuum. Three samples were heated in oil baths
at each of three temperatures (139 ( 1, 154 ( 1, and 173 ( 1
°C). At regular intervals the samples were removed from the
oil baths and quickly cooled in water to room temperature.
The reaction course was followed by 1H NMR spectroscopy,
comparing the integration of terminal acetylene protons rela-
tive to that of the acetonitrile standard.
1267, 1037 cm-1. FABMS calcd: 299.91. Found: 301(M+
+
1). Anal. Calcd for C6H6N2ClI: C, 23.98; H, 2.01; N, 9.32.
Found: C, 24.13; H, 1.97; N, 9.19.
5,6-Dieth yn yl-2,4-d im eth oxyp yr im id in e (3). The title
compound was prepared as described in the general procedure
on the following scale. 27c (1.0 g, 3.3 mmol), trimethylsily-
lacetylene (1.4 mL, 10 mmol), and tetrakis(triphenylphos-
phine)palladium(0) (100 mg) in diisopropylamine (13 mL) were
reacted for 5 h at 135 °C to give 27e (651 mg, 61%). After
deprotection of the product (620 mg, 1.9 mmol), 3 was obtained
as a white solid (274 mg, 77%): mp 135-136 °C (dec); 1H NMR
(CDCl3) δ 4.05 (s, 3H), 4.01 (s, 3H), 3.57 (s, 1H), 3.50 (s, 1H);
13C NMR (CDCl3) δ 171.6, 163.6, 153.1, 102.1, 87.1, 84.1, 79.7,
74.7, 55.4, 55.1; FTIR (CCl4) νmax 3272, 3233, 2116 cm-1
.
FABMS calcd: 189.06. Found: 189 (M+).
5-Ch lor o-6-eth yn yl-2,4-d im eth oxyp yr im id in e (27f): mp
150-151 °C; 1H NMR (CDCl3) δ 4.04 (s, 3H), 3.99 (s, 3H), 3.57
(s, 1H); 13C NMR (CDCl3) δ 132.2, 131.5, 128.5, 128.4, 87.5,
73.5, 55.6, 55.4; FTIR (CCl4) νmax 2219, 1581 cm-1. FABMS
calcd: 198.02. Found: 198 (M+). Anal. Calcd for C8H7N2O2-
Cl: C, 48.38; H, 3.55; N, 14.10. Found: C, 48.48; H, 3.62; N,
14.07.
Gen er a l P r oced u r e for Bu lk Th er m ocycliza tion . Ene-
diynes 1-3 (0.1 mmol) were dissolved in carbon tetrachloride
(20 mL), and the solution was heated in a pressure tube at
165 °C until TLC indicated that no more starting material
remained. The solvent was then removed by rotary evapora-
tion, and the products were isolated by flash column chroma-
tography, HPLC, or both.
Th er m ocyclization of Qu in oxalin e 1. 2,5-Dichlorophena-
zine (1a ) was isolated as an off-white solid in 81% yield by
column chromatography using EtOAc/Hex (1:9, v/v) as the
eluent: 1H NMR (CDCl3) δ 8.39 (dd, J ) 3.6, 3.2 Hz), 7.93
(dd, J ) 3.6, 3.2 Hz), 7.88 (s, 2H); 13C NMR (CDCl3) δ 143.6,
140.5, 132.3, 132.0, 130.0, 129.1; FTIR (CCl4) νmax 2926, 2854,
1559, 1554 cm-1. GC-MS calcd: 249.0. Found: 249.
Ack n ow led gm en t. The donors of the Petroleum
Research Fund, administered by the American Chemi-
cal Society, are recognized for partial support of this
project. We are also grateful for the financial assistance
provided by the American Cancer Society, Florida
Division, Inc. We also thank Ms. W. Calyptratus for
helpful discussions.
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra,
graphs of kinetics data, and Arrhenius plots for 1-3 at various
temperatures (18 pages). This material is contained in librar-
ies on microfiche, immediately follows this article in the
microfilm version of the journal, and can be ordered from the
ACS; see any current masthead page for ordering information.
Th er m ocycliza tion of P yr id in e 2. 5,8-Dichloroquinoline
(2a ) was only identified by GC-MS (calcd, 196.98; found, 197).
J O980879P
Th er m ocycliza tion of P yr im id in e 3. 5,8-Dichloro-2,4-
dimethoxyquinazoline (3a ) was isolated as a white solid by
normal-phase HPLC using EtOAc/Hex (2:8, v/v) as the eluent
(15% yield): 1H NMR (CDCl3) δ 7.67 (d, J ) 4.0 Hz), 7.28 (d,
J ) 4.0 Hz), 4.13 (s, 1H), 4.17 (s, 1H); 13C NMR (CDCl3) δ 169.1,
(27) Two of the quaternary carbons in the 13C NMR spectrum were
not visible. The eight visible signals in the 13C NMR spectrum and all
other spectral data are consistent with the assigned structure.