Synthesis and structure of novel sulfur bridged cyclic di- and tetraalkynes

Article abstract of DOI:10.1016/S0040-4039(01)01554-4

Some novel sulfur bridged 13- to 30-membered cyclic di- and tetraalkynes derived from 1,2-, 1,3- and 1,4-dihydroxybenzene and 1,2-bis(bromomethyl)benzene were synthesized and their structures confirmed by X-ray analysis. The unexpected formation of 2,6-divinyl-1,4-dithiin during Na₂S/alumina induced cyclization was also observed and the reaction mechanism is discussed.

Full text of DOI:10.1016/S0040-4039(01)01554-4

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 7485–7488
Synthesis and structure of novel sulfur bridged cyclic
di- and tetraalkynes
S. Braverman,a,* M. Cherkinsky,a M. L. Birsa,a S. Tichmana and I. Goldbergb
aDepartment of Chemistry, Bar-Ilan University, Ramat-Gan 52900, Israel
bSchool of Chemistry, Tel Aviv University, Ramat Aviv 69978, Israel
Received 5 July 2001; revised 9 August 2001; accepted 23 August 2001
Abstract—Some novel sulfur bridged 13- to 30-membered cyclic di- and tetraalkynes derived from 1,2-, 1,3- and 1,4-dihydroxyben-
zene and 1,2-bis(bromomethyl)benzene were synthesized and their structures confirmed by X-ray analysis. The unexpected
formation of 2,6-divinyl-1,4-dithiin during Na2S/alumina induced cyclization was also observed and the reaction mechanism is
discussed. © 2001 Elsevier Science Ltd. All rights reserved.
The so-called enediyne antibiotics such as calicheamicin
or esperamicin1 are amongst the most potent antitumor
agents known to date. However, due to the toxicity and
instability of the enediyne natural products on the one
hand, and their difficult synthesis on the other, a great
variety of enediyne models have been designed and
tested for their biological activity during the last
decade. This remarkable activity which resulted in the
publication of scores of articles was initiated by K. C.
Nikolaou’s report on pH-dependent DNA-cleavage by
propargylic and allenic sulfones,2 which, in turn, is
based on the cyclization of diallenic sulfones, a reaction
discovered by us more than two decades ago and
demonstrated to involve a diradical intermediate.3
Recently, we have investigated the behavior of some
novel acyclic p-conjugated bis-propargylic sulfides, sul-
foxides and sulfones.4 These compounds have been
found to undergo facile isomerization to the corre-
sponding diallenes followed by a tandem cyclization
and aromatization via a diradical intermediate in the
presence of various bases at room temperature.4 In
continuation, we decided to prepare some novel cyclic
sulfur bridged di- and tetrapropargylic systems (1, 2)
(Fig. 1) and investigate their tandem isomerization and
cyclization under basic conditions in order to compare
their reactivity with the reactivity of the acyclic sulfur
bridged propargylic systems.4 We expect that the pres-
ence of oxygen atoms in the new macrocycles would
Figure 1.
Keywords: cyclic dipropargylic sulfide; cyclic tetrapropargylic sulfide; 1,4-dithiin.
* Corresponding author. Tel. +972-3-5318322; fax: +972-3-5351250; e-mail: bravers@mail.biu.ac.il
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01554-4

7486
S. Bra6erman et al. / Tetrahedron Letters 42 (2001) 7485–7488
provide further enhancement to their chemical and
potential biological activity due to the well-known
metal-ion acceleration.5 Furthermore, we wanted to
study the effect of the nature of the bridge as well as
the cycle size on the reactivity of these macrocycles.
Here we wish to report some preliminary results about
synthesis and structure of the above-mentioned
compounds.
1,4-Dithiins are well-studied and documented com-
pounds.10 However, we are not aware of any previously
reported 2,6-divinyl-1,4-dithiin derivatives, e.g. (4).
Vinyl derivatives of 1,4-dithiins, which can potentially
act both as dienes as well as dienophiles, may be
subjected to various cycloaddition reactions and thus
open a route to some new polycyclic systems.
In view of the above results and in order to prevent the
alumina-catalyzed propargyl-allenyl isomerization, we
ran the reaction of dibromide 3a with Na2S·9H2O
under high dilution conditions (Scheme 2). These
resulted in formation of the 30-membered cyclic tetra-
alkyne 2a (vide supra).
1,4-, 1,3- and 1,2-Bis(4%-bromobut-2%-ynyloxy)benzenes
3a–c required for the synthesis of the corresponding
sulfur-bridged cyclic alkynes 1a–c and 2a–c were pre-
pared by two alternative ways ‘a’ and ‘b’ according to
Scheme 1, starting from 1,4-, 1,3- and 1,2-dihydroxy-
benzene, respectively.
Interestingly, unlike 1,4-derivative 3a, dibromide 3b
reacts either with Na2S·Al2O3 or under high dilution
with sodium sulfide nonahydrate in an intramolecular
fashion with the formation of the dipropargylic 13-
membered cyclic sulfide 1b11 only. The corresponding
26-membered cyclic tetrapropargylic bis-sulfide 2b was
obtained by the base-induced reaction of bis-thioacetate
7b with bis-mesylate 6b (Scheme 3).
However, during our attempts to prepare the bis-
propargylic sulfide 1a, we found that the reaction of the
corresponding dibromide 3a with an excess of alumina-
supported sodium sulfide6 in boiling THF, resulted in
formation of 2,6-divinyl-1,4-dithiin (4) instead of the
expected product (Scheme 2). Keeping in mind the high
reactivity of allenes towards sulfur nucleophiles7 and
the good leaving group ability of phenolate anions, we
suggest the mechanism shown in Scheme 2. Following
the formation of the cyclic sulfide 1a and its alumina-
catalyzed rearrangement to the corresponding diallenyl
sulfide 5, attack of the sulfide dianion on the latter
leads to fragmentation of 1a and formation of the
dithiin 4.8 Analogous rearrangement of a cyclic bis-
(propargylic)sulfide to the corresponding bis-allene
induced by the alumina-supported sodium sulfide
reagent was reported by Kerwin.9
Concerning the resorcinol derivatives 1c and 2c, 1,3-
bis(4%-bromobut-2%-ynyloxy)benzene 3c, available only
by pathway ‘b’ (Scheme 1) reacts with Na2S·9H2O
under high dilution with the formation of a mixture of
both types of ring systems in the ratio 1c/2c 1:5 and in
a total yield of 60%. The products were separated by
fractional recrystallization from acetone. Here, we may
see the effect of the substitution in the starting di-
hydroxybenzenes on the competitive intra- versus inter-
Scheme 1. Reagents and conditions: (i) propargyl bromide, K2CO3, DMF, rt, 48 h, 80–90%; (ii) n-BuLi, (CH2O)n, THF, −78°C,
4 h, 80–85%; (iii) 4-bromobut-2-yn-1-ol, K2CO3, DMF, rt, 48 h, 85–90%; (iv) CH3SO2Cl, Et3N, Et2O, 0°C, 6 h, 70–75%; (v) NaBr,
CH3CN, rt, 48–72 h, 80–85%.
Scheme 2.

S. Bra6erman et al. / Tetrahedron Letters 42 (2001) 7485–7488
7487
Scheme 3. Reagents and conditions: (i) AcSK, MeOH, rt, 2 h, 80%; (ii) KOH, MeOH/THF, HD, rt, 1 h, 35%.
bromine by iodine. In any case, the high yields of
cyclization by the use of the latter, especially for the
formation of 2c, are rather remarkable.
molecular cyclization. The instability of 1a as compared
to 1b and 1c may be explained in terms of excessive ring
strain in the former. Formation of the less strained 1c
as a minor product may serve as indirect evidence for
the intermediacy of the macrocycle 1a in the formation
of 2,6-divinyl-1,4-dithiin (4). Interestingly, another fac-
tor which has an influence on the mode of cyclization is
the nature of halogen in the bis(4%-halobut-2%-yny-
loxy)benzenes. Since, the reaction of the corresponding
dimesylates with NaBr takes 2–3 days (Scheme 1) we
used NaI to accelerate the formation of the required
dihalogenides. However, when we ran the reaction of
the diiodides with Na2S·9H2O the output of the cycliza-
tion changed. For the resorcinol derivatives we found
that both rings are formed in a total yield of 90% and
a 1c/2c ratio of 1:10. For the catechol derivatives this
effect was even more striking. Using 1,2-bis(4%-iodobut-
2%-ynyloxy)benzene, we obtained instead of 1b as the
only product, a mixture of 1b and 2b in the ratio of 2:1
with the total yield of 75%. We assume that these
results reflect the increase in leaving group ability and
steric hindrance associated with the substitution of
1,2 - Bis(4% - bromobut - 2% - ynyloxymethyl)benzene, re-
quired for the synthesis of products 1d and 2d, was
obtained according to Scheme 1 (pathway ‘a’) except
that corresponding diol was prepared by the reaction of
1,2-bis(bromomethyl)benzene with propargyl alcohol
and NaH in THF. Reaction of 1,2-bis(4%-bromobut-2%-
ynyloxymethyl)benzene with Na2S·9H2O under high
dilution gave the dipropargylic cycle 1d with the yield
of 80% (Scheme 1), whereas tetrapropargylic cycle 2d
was obtained by the base induced reaction of the
corresponding bis-mesylate with bis-thioacetate (24%
yield) according to Scheme 3. All the cyclic sulfides
1b–d and 2 a–d obtained12 were oxidized by MCPBA to
the corresponding sulfones.
For the compounds 2a and 2b we were able to grow
single crystals (recrystallization from chloroform/hex-
ane), so the molecular structure of both compounds
,
Figure 2. Crystal structure of compounds 2a and 2b. Selected bond lengths (A) and angles (°): 2a: S1ꢀC2 1.82, C2ꢀC3 1.46, C3ꢀC4
1.19, C5ꢀO6 1.44, O6ꢀC7 1.38; C34ꢀS1ꢀC2 98.18, C3ꢀC2ꢀS1 112.13, C4ꢀC3ꢀC2 175.56, C3ꢀC4ꢀC5 178.36, C7ꢀO6ꢀC5 117.22,
O6ꢀC7ꢀC12 124.73, O6ꢀC7ꢀC8 115.43, O6ꢀC5ꢀC4 112.95. 2b: S1ꢀC2 1.819, C2ꢀC3 1.465, C3ꢀC4 1.189, C4ꢀC5 1.463, C5ꢀO6
1.439, O6ꢀC7 1.372; C17ꢀS1ꢀC2 98.66, C3ꢀC2ꢀS1 112.10, C4ꢀC3ꢀC2 177.3, C3ꢀC4ꢀC5 173.3, C7ꢀO6ꢀC5 115.22, O6ꢀC7ꢀC12
115.39, O6ꢀC7ꢀC8 125.0, O6ꢀC5ꢀC4 108.82.

7488
S. Bra6erman et al. / Tetrahedron Letters 42 (2001) 7485–7488
was unambiguously elucidated by means of X-ray analy-
sis (Fig. 2). In the case of tetrapropargylic cycle 2a the
structure contains two crystallographically independent
molecules, each residing on a center of inversion. The
two molecules differ slightly in their conformation at
atoms C(14) and C(31), which represent the most
flexible part of the molecular structure.
8. 2,6-Divinyl-1,4-dithiin (3): yellow–orange oil, yield 45%
(column chromatography on silica gel with hexane as
1
eluent); H NMR (CDCl3, 300 MHz) l 5.15 (1H, d, 10),
5.67 (1H, d, 17), 6.31 (1H, d, 0.5), 6.54 (1H, ddd, 0.5, 10,
17); 13C NMR (CDCl3, 75 MHz) l 115.19 (CH2), 122.05
(CH), 133.07 (CH), 134.34 (Cq); MS-EI m/z 168 (M+,
100), 153 (9.71), 141(8.2), 135 (24.9), 124 (22.74); HRMS
calcd. for C8H8S2 168.006744; found 168.006321.
9. Kerwin, S. M. Tetrahedron Lett. 1994, 35, 1023.
10. (a) Klar, G. In Houben-Weyl Methods of Organic Chem-
istry; Schaumann, E., Ed.; G. Thieme Verlag: Stuttgart,
1997; Vol. E9a, pp. 250–407; (b) Cook, M. J. In Compre-
hensive Heterocyclic Chemistry; Katritzky, A. R.; Rees,
Ch. W., Eds.; Pergamon Press: Oxford, 1984; Vol. 3, pp.
943–994.
Reactivity of the above di- and tetrapropargylic sulfur
bridged cycles under basic conditions is now under
investigation and the results will be published soon.
Acknowledgements
11. Basak, A.; Khamrai, U. K. Tetrahedron Lett. 1995, 43,
7913.
The financial support of this study by the Israel Science
Foundations is gratefully acknowledged. One of us
(M.L.B.) is also grateful for the award of a Vatat
postdoctoral fellowship.
12. All new compounds prepared gave satisfactory analytical
and spectral data in accordance with their structures.
Selected data are as follows: 2a: mp 216°C (white crys-
tals, from CHCl3), yield 35% (flash chromatography,
1
CH2Cl2/hexane 3:1); H NMR (acetone-d6, 300 MHz) l
References
3.37 (4H, t, 2.2), 4.79 (4H, t, 2.2), 6.99 (4H, s); 13C NMR
(DMSO-d6, 75 MHz) l 18.24 (CH2S), 56.19 (CH2O),
78.71 and 82.65 (CꢁC), 116.12 (CHarom), 151.57 (C-ipso);
MS-CI (NH3) m/z 506 (MNH4+, 100), 489 (MH+, 6.5),
374 (8), 212 (7), 161 (7.2); HRMS calcd for C28H24O4S2
488.111603; found 488.111585. 1b: mp 93–94°C (white
1. For reviews, see: (a) Nicolaou, K. C.; Smith, A. L. In
Modern Acetylene Chemistry; Stang, P. J.; Diedrich, F.,
Eds.; VCH: Veinheim, 1995; Chapter 7; (b) Grissom, J.
W.; Gunawardena, G. U.; Klinberg, D.; Huang, D. Tet-
rahedron 1996, 52, 6453; (c) Wang, K. K. Chem. Rev.
1996, 96, 207; (d) Maier, M. E. Synlett 1995, 13.
2. Nicolaou, K. C.; Skokotas, G.; Maligres, P.; Zuccarello,
G.; Schweiger, E. J.; Toshima, K.; Wenderborn, S.
Angew. Chem., Int. Ed. Engl. 1989, 28, 1272.
1
crystals, from CHCl3/hexane), yield 30%; H NMR (ace-
tone-d6, 300 MHz) l 3.39 (4H, t, 2.0), 4.91 (4H, t, 2.0),
7.01 (4H, m); 13C NMR (acetone-d6, 75 MHz) l 21.17
(CH2S), 57.55 (CH2O), 77.68 and 84.40 (CꢁC), 118.62
and 122.59 (CHarom), 148.56 (C-ipso); MS-EI: m/z 244
(M+, 100), 198.08 (27.10), 160.04 (25.61), 136.04 (51.45),
103.05 (35.31), 91.06 (86.02); HRMS for C14H12O2S calcd
244.055802; found 244.054000; 1d: yield 80% (flash chro-
3. Braverman, S.; Segev, D. J. Am. Chem. Soc. 1974, 96,
1245.
4. Braverman, S.; Zafrani, Y.; Gottlieb, H. Tetrahedron
1
Lett. 2000, 41, 2675.
matography on silica gel, eluent CH2Cl2/hexane 2:1); H
5. Konig, B. Eur. J. Org. Chem. 2000, 381.
NMR (CDCl3, 300 MHz) l 3.61 (4H, t, 2), 4.29 (4H, t,
2), 4.75 (4H, s), 7.36 (4H, m); 13C NMR (CDCl3, 75
MHz) l 19.30 (CH2S), 57.63 (OCH2Cꢁ), 68.07 (PhCH2),
80.24 and 81.94 (CꢁC), 128.17 and 129.36 (CHarom),
136.00 (C-ipso); MS-CI(CH4): m/z 272.1 (M+, 6.76),
225.1 (10.09), 179.1 (18.10), 171.1 (59.29), 153 (17.08),
151 (21.01), 135 (21.28), 119.1 (48.40), 104 (81.52), 91
(100); HRMS calcd. for C16H16O2S 272.087102; found
272.086319.
6. Tan, L. C.; Pagni, R. M.; Kabalka, G. W.; Hillmyer, M.;
Woosley, J. Tetrahedron Lett. 1992, 33, 7709.
7. (a) Stirling, C. J. M. J. Chem. Soc. 1964, 5856; (b)
Mueller, W. H.; Griesbaum, K. J. Org. Chem. 1967, 32,
856; (c) Reisman, D. Ph.D. Thesis, Bar-Ilan University,
1985; (d) Xu, Y. Y.; Jin, F. Q.; Huang, W. Y. J. Fluor.
Chem. 1995, 70, 5; (e) Naruse, Y.; Kakita, S.; Tsunekawa,
A. Synlett 1995, 711.