Synthesis of dibenzodiynes: Exceptionally easy formation of a chrysene biradical

Article abstract of DOI:10.1055/s-1998-1922

The new dibenzodiynes 1 and 2 have been synthesized using a Nozaki-Hiyama-Kishi reaction as the key step. Treatment of dibenzodiynone 2 with base at -50°C afforded an enyne-cumulene intermediate that underwent a regioselective cycloaromatization via a chrysene biradical.

Full text of DOI:10.1055/s-1998-1922

1282
LETTERS
SYNLETT
Synthesis of Dibenzodiynes: Exceptionally Easy Formation of a Chrysene Biradical
G. Rodríguez, D. Rodríguez, M. López, L. Castedo, D. Domínguez, C. Saá*
Departamento de Química Orgánica y Unidad Asociada al CSIC, Facultad de Química, Universidad de Santiago de Compostela,
15706 Santiago de Compostela, Spain
Received 5 August 1998
Abstract: The new dibenzodiynes 1 and 2 have been synthesized using
a Nozaki-Hiyama-Kishi reaction as the key step. Treatment of
dibenzodiynone 2 with base at –50°C afforded an enyne-cumulene
intermediate that underwent a regioselective cycloaromatization via a
chrysene biradical.
1 as a 5:95 mixture of diastereomers (Scheme 1). Diastereomers 1a and
1b were both stable in CHCl3 or THF solution at r.t.; however, they
slowly decomposed when kept neat.
To determine the relative stereochemistry of the diastereomers, the
major one was silylated to yield the bis-silyloxy derivative 4 (Fig. 1).
The 1H NMR spectrum of 4 in the presence of chiral Eu+3 salts showed
no splitting of signals, indicating 4 to be the meso-compound.
Therefore, we have tentatively assigned anti relative stereochemistry to
the major diastereomer, 1b.12
Current research on the thermal biradical cyclizations of enediynes,
enyne-allenes and enyne-cumulenes focuses on their utility in the
construction of polycyclic ring systems,1 and on the synthesis of models
of the natural enediyne antitumor antibiotics.2 Myers3 has developed
such a model based on the fact that 1,6-didehydro[10]annulene cyclizes
to the 1,5-didehydronaphthalene biradical at -51°C. In the present work,
with the ultimate goal of developing simple and stable DNA-cleaving
agents that can be activated under controlled conditions, we report
preparation of new dibenzodiynes 1 and 24,5 (Scheme 1) using a
Nozaki–Hiyama–Kishi reaction as the key step and the novel
regioselective cycloaromatization of dibenzodiynone 2 to a chrysene
biradical upon treatment with base.
Next, cycloaromatization of 1b was attempted using methyl
thioglycolate as nucleophile in the presence of 1,4-cyclohexadiene and
triethylamine.3,4 Several sets of reaction conditions were tried,13 but no
cycloaromatized products were detected. Activation of the system by in
situ generation of the mesylate derivative of 1b, followed by quenching
with H2O as nucleophile,4b furnished the initial diastereomeric alcohols
1a and 1b in 53% yield and 3:1 ratio (Scheme 1). Regardless of the
mechanism of this reaction (SN1 or SN2), it is noteworthy that the latter
diastereomeric ratio is the opposite of that obtained in the original
intramolecular coupling.
Benzaldehydes 3 were chosen as the immediate synthetic precursors of
1 (Scheme 1). Compound 3a was prepared in 76% yield by coupling of
the acetylide derivative of 2-(2-(ethynylphenyl)-1,3-dioxane with o-
ethynylbenzaldehyde (both prepared from o-bromobenzaldehyde),6
followed by deprotection of the acetal group. Silyl ether 3b was
obtained by treating 3a with TBDMSCl.
To increase the reactivity of the dibenzodiynols 1, they were oxidized to
the alkynyl ketone 2 using Dess-Martin periodinane14 (Scheme 1).15
Alkynyl ketone 2 is rather unstable and so was immediately subjected to
the cycloaromatization reaction. Gratifyingly, treatment of 2 with
methylthioglycolate (2 equiv) in the presence of triethylamine (2 equiv)
and 1,4-cyclohexadiene (40 equiv) in THF at 0°C for 5 min gave
chrysenol 6 in 29% yield as the only cycloaromatized product.16 Careful
control of the reaction conditions showed that a similar result was
obtained at –50°C with triethylamine alone (Scheme 2).
Scheme 2
Formation of 6 in the absence of the nucleophile could be explained by
an enolization of
2 to give [10]annulene 7 (Fig. 1), which
regioselectively cycloaromatizes to chrysenol 6 (route a). Interestingly,
alternative cycloaromatization of annulene 7 to naphthacenol 8 (route b)
has not been observed. In this regard, it is interesting to note that
distances “a” and “b” within structure 7 are determined to be 2.934 and
Attempts to cyclize 3b to dibenzodiyne 1 by intramolecular nucleophilic
addition of the acetylide of 3b to its aldehyde7 [LiHMDS or NaHMDS
(2 equiv) in THF at -78°C] afforded a complex mixture of products,
among which the desired dibenzodiyne 1 was not detected. Next, we
tried treating the trimethylsilylacetylene 3d8 with cesium fluoride [CsF
(3.5 equiv), NaHCO3 (2 equiv), Ac2O (2 equiv)],9 which led
quantitatively to the desilylated acetylene 3b. Finally, the terminal
alkyne was converted into the alkynyl iodide 3c by a standard method
using DBU as a base,10 and ring–closure was achieved by the Nozaki-
Hiyama-Kishi reaction.11 Thus, slow addition of a solution of 3c to a
suspension of CrCl2 (3 equiv) and NiCl2 (0.5 equiv) in THF at -20°C
gave, after separation by flash column chromatography, a 53% yield of
2.998 Å, respectively, by AM1 calculations, suggesting that
a
predisposition exists for cyclization along pathway “a” within ground-
state structure.17
To confirm that 6 (chrysenol) and not 8 (naphthacenol) had been
obtained, HMQC and HMBC NMR experiments were performed on the
symmetric bis-silyloxy derivative 9.18 For the sake of comparison we
also detailed the most significant couplings expected for the putative
isomeric naphthacenol 10 (Table 1).

November 1998
SYNLETT
1283
Grissom, J. W.; Gunawardena, G. U.; Klingberg, D.; Huang, D.
Tetrahedron 1996, 52, 6453.
(2) For reviews, see: (a) ref. 1d. (b) Maier, M. E. Synlett 1995, 13. (c)
Nicolaou, K. C.; Smith, A. L. Modern Acetylene Chemistry;
Stang, P. J.; Diederich, F., Eds.; VCH: Weinheim, 1995; pp 203-
283.
(3) (a) Myers, A. G.; Dragovich, P. S. J. Am. Chem. Soc. 1993, 115,
7021 and references therein.
(4) For closely related simple models, see: (a) Eckhardt, M.;
Brückner, R.; Suffert, J. Tetrahedron Lett. 1995, 36, 5167. (b)
Matsumoto, Y.; Hasegawa, T.; Kuwatani, Y.; Ueda, I.
Tetrahedron Lett. 1995, 36, 5757. (c) Matsumoto, Y.; Kuwatani,
Y.; Ueda, I. Tetrahedron Lett. 1995, 36, 3197.
(5) Dibenzodiynes are structural analogues of both dienediyn-1-ol3
and golfomycin A– a known DNA-cleaving agent with antitumour
activity. Nicolaou, K. C.; Skokotas, G.; Furuya, S.; Suemune, H.;
Nicolaou, D. C. Angew. Chem., Int. Ed. Engl. 1990, 29, 1064.
Figure 1
(6) Padwa, A.; Krumper, K. E.; Weingarten, M. D. J. Org. Chem.
1995, 60, 5595.
From the Table 1, it may be concluded that the observed coupling
between H-4 and quaternary carbon C–4b and the absence of reciprocal
couplings between H–5 and tertiary carbon C–4 and H–4 and tertiary
carbon C–5 confirms a chrysenol–type structure for the bis-silylated
product 9,19 and consequently, structure 6 for the cycloaromatization
product.
(7) (a) Ulibarri, G.; Nadler, W.; Skrydstrup, T.; Audrain, H.;
Chiaroni, A.; Riche, C.; Grierson, D. S. J. Org. Chem. 1995, 60,
2753. (b) Danishefsky, S. J.; Shair, M. D. J. Org. Chem. 1996, 61,
16.
(8) 3d was prepared as described for 3b but using o-
iodobenzaldehyde instead of o-ethynylbenzaldehyde, and then
coupling the resulting aryliodide with trimethylsilylacetylene
using palladium catalyst (62% overall yield).
(9) Wender, P.A.; Beckham, S.; Mohler, D. L. Tetrahedron Lett.
1995, 36, 209.
(10) Nicolaou, K. C.; Liu, A; Zeng, Z.; McComb, S. J. Am. Chem. Soc.
1992, 114, 9279.
(11) Takai, K.; Kuroda, T.; Nakatsukasa, S.; Oshima, K.; Nozaki, H.
Tetrahedron Lett. 1985, 26, 5585.
(12) Diol 5 was highly unstable and difficult to handle, and all attempts
at performing 1H-NMR experiments with this compound in the
presence of chiral Eu+3 salts failed.
1H NMR experiments carried out in deoxygenated THF-d8 at -50°C
provided evidence for the formation of a chrysene biradical upon
treatment of ketone 2 with triethylamine. Specifically, successive
spectra showed slow disappearance of the starting material with
concomitant formation of 6 incorporating 25% deuterium at C–5 and C–
11 (Scheme 2).20
(13) (a) Various solvents (THF, benzene) and conditions (0°C, 20°C,
60°C) were tried using HSCH2CO2Me (2, 9 or 12 equiv) as
nucleophile and Et3N as base (see ref. 4b). In all cases starting
material was recovered as the main product.
(14) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155.
In summary, we synthesized new dibenzodiynes 1 and 2 and found
evidence for the intermediacy of a chrysenic biradical in the novel
regioselective cycloaromatization of dibenzodiynone 2 into chrysenol 6
upon treatment with base. Further work towards the development of
stable DNA–cleaving agents is in progress.
(15) Dibenzodiynone 2: 1H-NMR (CDCl3, 250 MHz): δ= 0.25 (s, 3H),
0.27 (s, 3H), 0.99 (s, 9H), 5.89 (s, 1H), 7.29-7.67 (m, 7H), 8.21 (d,
J= 7.0 Hz, 1H); 13C-NMR (CDCl3, 62.8 MHz): δ= -4,6 (CH3), -
4.1 (CH3), 18.3 (C), 25.8 (3 x CH3), 64.3 (CH), 87.2 (C), 94.5 (C),
94.8 (C), 96.5 (C), 117.8 (C), 123.6 (C), 127.9 (CH), 128.3 (CH),
128.7 (CH), 129.3 (CH), 131.2 (CH), 132.1 (CH), 132.9 (CH),
134.6 (CH), 137.6 (C), 141.7 (C), 175.9 (C=O).
Acknowledgements
(16) Chrysenol 6: 1H-NMR (CD2Cl2, 250 MHz): δ= 0.32 (s, 6H), 1.14
(s, 9H), 7.61-7.70 (m, 4H), 7.94 (s, 1H), 7.98 (s, 1H), 8.33 (d, J=
7.6 Hz, 2H), 8.57 (d, J= 8.0 Hz, 2H); 13C-NMR (CD2Cl2, 62.8
MHz): δ= -3.8 (2 x CH3), 19.0 (C), 26.3 (3 x CH3), 102.9 (CH),
107.4 (CH), 122.8 (CH), 123.6 (CH), 123.7 (2 x CH), 124.8 (C),
125.0 (C), 125.7 (C), 126.6 (2 x CH), 127.2 (CH), 127.5 (CH),
128.9 (C), 131.4 (C), 131.5 (C), 149.3 (C), 149.5 (C).
We thank the Dirección General de Enseñanza Superior, M.E.C.
(Project PB95-0824) and the Xunta de Galicia (Project XUGA 20907–
B96) for financial support. G. Rodríguez and D. Rodríguez thank the
Xunta de Galicia for predoctoral fellowships.
References and Notes:
(1) (a) Schmittel, M.; Steffen, J.-P.; Auer, D.; Maywald, M.
Tetrahedron Lett. 1997, 38, 6177. (b) Wang, K. K.; Wang, Z.;
Tarli, A.; Gannett, P. J. Am. Chem. Soc. 1996, 118, 10783. For
reviews, see: (c) Wang, K. K. Chem. Rev. 1996, 96, 207. (d)
(17) Computer modelling was performed using the AM1
semiempirical method as implemented in MacSpartan Plus 1.1.6,
Wavefunction, 1996. This modelling study is intended to be

1284
LETTERS
SYNLETT
suggestive only since the analysis does not examine the relevant
transition–state energies of the two biradical–forming pathways.
CH3), 19.0 (C), 26.3 (3 x CH3), 107.3 (CH, C–5), 123.6 (CH, C–
4), 123.7 (CH, C–1), 125.1 (C, C–4b), 126.6 (CH, C–2), 127.2
(CH, C–3), 128.9 (C, C–6a), 131.6 (C, C–4a), 149.6 (C, C–6).
(18) Chrysenic ether 9: 1H-NMR (CD2Cl2, 500 MHz): δ= 0.38 (s,
6H, Si(CH3)2), 1.15 (s, 9H, C(CH3)3), 7.62-7.68 (m, 2H, H-2 and
H-3), 7.99 (s, 1H, H-5), 8.34 (d, J= 8.1 Hz, 1H, H-1), 8.56 (d, J=
8.1 Hz, 1H, H-4); 13C-NMR (CD2Cl2, 62.8 MHz): δ= -3.8 (2 x
(19) A two–bond 1H–13C coupling between H–5 and quaternary
carbon C–6 is also observed.
(20) t1/2 of 2 at –50°C ≈10 min.