A bimetallic aluminum(salen) complex for the synthesis of 1,3-oxathiolane-2-thiones and 1,3-dithiolane-2-thiones

Article abstract of DOI:10.1021/jo101121h

Figure presented. The combined use of the bimetallic aluminum(salen) complex [Al(salen)]₂O and tetrabutylammonium bromide (or tributylamine) is found to catalyze the reaction between epoxides and carbon disulfide. In most cases, at 50 °C, the reaction produces 1,3-oxathiolane-2-thiones, while at 90 °C, 1,3-dithiolane-2-thiones are the main product. The structure and stereochemistry of three of the 1,3-dithiolane-2-thiones is unambiguously determined by X-ray crystallographic analysis, and this is used to correct errors in the literature concerning the synthesis of cyclic di- and trithiocarbonates. The kinetics of 1,3-oxathiolane-2-thione synthesis are determined, and the resulting rate equation, along with a stereochemical analysis of the reaction and catalyst modification studies, is used to determine a mechanism for the synthesis of 1,3-oxathiolane-2-thiones which contrasts with the mechanism previously determined for cyclic carbonate synthesis using the same bimetallic aluminum(salen) complex.

Full text of DOI:10.1021/jo101121h

pubs.acs.org/joc
A Bimetallic Aluminum(salen) Complex for the Synthesis
of 1,3-Oxathiolane-2-thiones and 1,3-Dithiolane-2-thiones
William Clegg, Ross W. Harrington, Michael North,* and Pedro Villuendas
School of Chemistry and University Research Centre in Catalysis and Intensified Processing,
Bedson Building, University of Newcastle, Newcastle upon Tyne NE1 7RU, U.K.
michael.north@ncl.ac.uk
Received June 21, 2010
The combined use of the bimetallic aluminum(salen) complex [Al(salen)]2O and tetrabutylammo-
nium bromide (or tributylamine) is found to catalyze the reaction between epoxides and carbon
disulfide. In most cases, at 50 °C, the reaction produces 1,3-oxathiolane-2-thiones, while at 90 °C,
1,3-dithiolane-2-thiones are the main product. The structure and stereochemistry of three of the
1,3-dithiolane-2-thiones is unambiguously determined by X-ray crystallographic analysis, and this is
used to correct errors in the literature concerning the synthesis of cyclic di- and trithiocarbonates. The
kinetics of 1,3-oxathiolane-2-thione synthesis are determined, and the resulting rate equation, along
with a stereochemical analysis of the reaction and catalyst modification studies, is used to determine a
mechanism for the synthesis of 1,3-oxathiolane-2-thiones which contrasts with the mechanism
previously determined for cyclic carbonate synthesis using the same bimetallic aluminum(salen)
complex.
Introduction
earlier work7 involving bimetallic titanium(salen) complexes
[Ti(salen)O]2, we showed that the combined use of complex 1 as
a Lewis acid and triphenylphosphine oxide as a Lewis base
would catalyze the asymmetric addition of trimethylsilyl
cyanide to aldehydes.8,9 In addition, we have shown that a
combination of complex 1 and tetrabutylammonium bromide
forms an exceptionally active catalyst for the synthesis of cyclic
carbonates from terminal epoxides and carbon dioxide
(Scheme 1), allowing cyclic carbonate synthesis to be carried
In 2003, Jacobsen first reported the use of bimetallic
aluminum(salen) complex 1 as a catalyst for the asymmetric
Michael addition of doubly stabilized carbanions to R,β-
unsaturated imides.1,2 Michael additions of azide,2 nitro-
alkanes,2 cyanide,3 oximes,4 and heterocycles5 were also found
to be catalyzed by complex 1. Subsequently, Zhu showed that
complex 1 would catalyze Passerini-type reactions between
aldehydes, isocyanides, and hydrogen azide.6 Building on our
(7) (a) Belokon, Y. N.; Caveda-Cepas, S.; Green, B.; Ikonnikov, N. S.;
Khrustalev, V. N.; Larichev, V. S.; Moscalenko, M. A.; North, M.; Orizu, C.;
Tararov, V. I.; Tasinazzo, M.; Timofeeva, G. I.; Yashkina, L. V. J. Am.
Chem. Soc. 1999, 121, 3968–3973. (b) Belokon, Y. N.; Green, B.; Ikonnikov,
N. S.; Larichev, V. S.; Lokshin, B. V.; Moscalenko, M. A.; North, M.; Orizu,
C.; Peregudov, A. S.; Timofeeva, G. I. Eur. J. Org. Chem. 2000, 2655–2661.
(c) Belokon, Y. N.; Ishibashi, E.; Nomura, H.; North, M. Chem. Commun.
2006, 1775–1777. (d) Belokon, Y. N.; Clegg, W.; Harrington, R. W.; Young,
C.; North, M. Tetrahedron 2007, 63, 5287–5299. (e) Belokon, Y. N.; Clegg,
W.; Harrington, R. W.; North, M.; Young, C. Inorg. Chem. 2008, 47, 3801–
3814.
(1) (a) Taylor, M. S.; Jacobsen, E. N. J. Am. Chem. Soc. 2003, 125, 11204–
11205. (b) Raheem, I. T.; Goodman, S. N.; Jacobsen, E. N. J. Am. Chem. Soc.
2004, 126, 706–707. (c) Balskus, E. P.; Jacobsen, E. N. J. Am. Chem. Soc.
2006, 128, 6810–6812.
(2) Taylor, M. S.; Zaltan, D. N.; Lerchner, A. M.; Jacobsen, E. N.
J. Am. Chem. Soc. 2005, 127, 1313–1317.
(3) Sammis, G. M.; Danjo, H.; Jacobsen, E. N. J. Am. Chem. Soc. 2004,
126, 9928–9929.
(4) Vanderwal, C. D.; Jacobsen, E. N. J. Am. Chem. Soc. 2004, 126,
14724–14725.
(5) Gandelman, M.; Jacobsen, E. N. Angew. Chem., Int. Ed. 2005, 44,
2393–2397.
(6) Yue, T.; Wang, M.-X.; Wang, D.-X.; Zhu, J. Angew. Chem., Int. Ed.
2008, 47, 9454–9457.
(8) North, M.; Williamson, C. Tetrahedron: Lett. 2009, 50, 3249–3252.
(9) For a comprehensive review of Lewis acid catalyzed asymmetric
cyanohydrin synthesis, see: North, M.; Usanov, D. L.; Young, C. Chem.
Rev. 2008, 108, 5146–5226.
DOI: 10.1021/jo101121h
r
Published on Web 08/23/2010
J. Org. Chem. 2010, 75, 6201–6207 6201
2010 American Chemical Society

Clegg et al.
JOCArticle
SCHEME 1. Synthesis of Cyclic Carbonates
SCHEME 2. Reaction between Epoxides and Carbon Disulfide
and DMAP at 120 °C,14b or triethylamine or DMAP in
water14g selectively produced 1,3-oxathiolane-2-thiones, and
potassium hydroxide or alkoxides,18b-f (2-propanolato)-
titanatrane,18g or timethylamine (under high pressure)18a
formed compounds 4 selectively. It is also known that com-
pounds 2 can be isomerized to compounds 3 by potassium
iodide15 or by protic or Lewis acids,19 that compounds 2 react
with excess carbon disulfide to give trithiocarbonates 4,18g and
that compounds 4 can be converted into compounds 3 by
treatment with lead tetraacetate,17b diaryltellurium species,17c
or benzeneseleninic anhydride.17d
1,3-Oxathiolane-2-thiones 2 and 1,3-dithiolane-2-thiones 4
have both been shown to possess radioprotective activity.20
Dithiocarbonates 2 have also been used in polymer synthe-
ses,14c,19,21 while trithiocarbonates 4 have been found to possess
insecticidal activity.22 In this paper, we give full details of the
synthesis of compounds 2 and 4 from epoxides and carbon
disulfide catalyzed by complex 1 and tetrabutylammonium
bromide,23 present the structures of three of the 1,3-dithiolane-
2-thiones determined by X-ray crystallography, correct erro-
neous structural assignments in the literature, and report a
kinetic analysis of the synthesis of dithiocarbonates 2 which
allows a reaction mechanism to be determined.
out at atmospheric pressure and room temperature10 and to be
integrated with oxyfuel combustion in a combined energy and
chemical production unit.11 Immobilized versions of complex 1
subsequently allowed cyclic carbonate synthesis to be carried
out in a gas-phase flow reactor.12
In view of the high catalytic activity shown by complex 1 for
the synthesis of cyclic carbonates, we decided to investigate
its use in the related reaction between epoxides and carbon
disulfide. This is known to be a more complex reaction as a
variety of products can be formed13 depending on the catalyst
and reaction conditions, including 1,3-oxathiolane-2-thiones14,15 2,
1,3-dithiolane-2-ones16,17 3, 1,3-dithiolane-2-thiones18 4,
1,3-oxathiolane-2-ones13b 5, and thiiranes14a 6 as shown in
Scheme 2. Previous work has shown that use of triethylamine
(under high pressure)13 as catalyst gave mixtures of products,
while lithium bromide,14a hydrotalcite,14e sodium methoxide,14f
tributylphosphine and lithium perchlorate,14c,d 4-methoxyphenol
(10) (a) Melendez, J.; North, M.; Pasquale, R. Eur. J. Inorg. Chem. 2007,
3323–3326. (b) North, M.; Pasquale, R. Angew. Chem., Int. Ed. 2009, 48,
2946–2948. (c) Clegg, W.; Harrington, R. W.; North, M.; Pasquale, R.
Chem.;Eur. J. 2010, 16, 6828–6843.
(11) Metcalfe, I. S.; North, M.; Pasquale, R.; Thursfield, A. Energy
Environ. Sci. 2010, 3, 212–215.
Results and Discussion
Preliminary attempts to catalyze the reaction between
epoxides and carbon disulfide using complex 1 and/or tetra-
butylammonium bromide at atmospheric pressure and room
temperature, the conditions used for the synthesis of cyclic
carbonates10 (Scheme 1), were unsuccessful. Therefore, reac-
tions were carried out using propylene oxide 7a as substrate with
up to 7 equiv of carbon disulfide in a sealed Young’s tube for
16 h under solvent-free conditions (Scheme 3). The results of
ꢀ
(12) (a) Melendez, J.; North, M.; Villuendas, P. Chem. Commun. 2009,
2577–2579. (b) North, M.; Villuendas, P.; Young, C. Chem.;Eur. J. 2009,
11454–11457.
(13) For reactions leading to mixtures of compounds 2-6, see: (a) Taguchi,
Y.; Yanagiya, K.; Shibuya, I.; Suhara, Y. Bull. Chem. Soc. Jpn. 1988, 61, 921–
925. (b) Taguchi, Y.; Yasumoto, M.; Shibuya, I.; Suhara, Y. Bull. Chem. Soc.
Jpn. 1989, 62, 474–478.
(14) (a) Kihara, N.; Nakawaki, Y.; Endo, T. J. Org. Chem. 1995, 60, 473–
475. (b) Shen, Y.-M.; Duan, W.-L.; Shi, M. Eur. J. Org. Chem. 2004, 3080–
3089. (c) Motokucho, S.; Itagaki, Y.; Sudo, A.; Endo, T. J. Polym. Sci., Part
A: Polym. Chem. 2005, 43, 3711–3717. (d) Wu, J.-Y.; Luo, Z.-B.; Dai, L.-X.;
Hou, X.-L. J. Org. Chem. 2008, 9137–9139. (e) Maggi, R.; Malmassari, C.;
Oro, Ch.; Pela, R.; Sartori, G.; Soldi, L. Synthesis 2008, 53–56. (f) Yavari, I.;
Ghazanfarpour-Darjani, M.; Hossaini, Z.; Sabbaghan, M.; Hosseini, N.
Synlett 2008, 889–891. (g) Halimehjani, A. Z.; Ebrahimi, F.; Azizi, N.; Saidi,
M. R. J. Heterocycl. Chem. 2009, 46, 347–350.
SCHEME 3. Synthesis of Propylene Di- And Trithiocarbonates
2a and 4a
(15) For the synthesis of 1,3-oxathiolane-2-thiones using thiophosgene,
see: Jones, F. N.; Andreades, S. J. Org. Chem. 1969, 34, 3011–3014.
(16) For a related synthesis of 1,3-dithiolane-2-ones by reaction of
epoxides with trithiocarbonates, see: Barbero, M.; Degani, I.; Dughera,
S.; Fochi, R.; Piscopo, L. J. Chem. Soc., Perkin Trans. 1 1996, 289–294.
(17) For other syntheses of 1,3-dithiolane-2-ones, see: (a) Corey, E. J.;
Mitra, R. B. J. Am. Chem. Soc. 1962, 84, 2938–2941. (b) Adley, T. J.;
Anisuzzaman, A. K. M.; Owen, L. N. J. Chem. Soc. C 1967, 807–812. (c) Ley,
S. V.; Meerholz, C. A.; Barton, D. H. R. Tetrahedron Lett. 1980, 21, 1785–
1788. (d) Cussans, N. J.; Ley, S. V.; Barton, D. H. R. J. Chem. Soc., Perkin
Trans. 1 1980, 1650–1653. (e) Yasuda, N.; Yatami, T.; Ohnuki, T.; Okutsu,
M. J. Heterocycl. Chem. 1984, 21, 1845–1848.
(18) (a) Durden, J. A., Jr.; Stansbury, H. A., Jr.; Catlette, W. H. J. Am.
Chem. Soc. 1960, 82, 3082–3084. (b) Overberger, C. G.; Drucker, A. J. Org.
Chem. 1964, 29, 360–366. (c) Kyaw, M.; Owen, L. N. J. Chem. Soc. 1965,
1298–1305. (d) Hayashi, S.; Furukawa, M.; Fujino, Y.; Nakao, T.; Nagato,
K. Chem. Pharm. Bull. 1971, 19, 1594–1597. (e) McCasland, G. E.; Zanlungo,
A. B.; Durham, L. J. J. Org. Chem. 1974, 39, 1462–1466. (f) Jesudason, M. V.;
Owen, L. N. J. Chem. Soc., Perkin Trans. 1 1974, 1443–1446. (g) Motokucho,
S.; Takeuchi, D.; Sanda, F.; Endo, T. Tetrahedron 2001, 57, 7149–7152.
(19) (a) Choi, W.; Sanda, F.; Endo, T. Macromolecules 1998, 31, 2454–
2460. (b) Choi, W.; Sanda, F.; Endo, T. Heterocycles 2000, 52, 125–128.
(20) Robbe, Y.; Fernandez, J.-P.; Dubief, R.; Chapat, J.-P.;
Sentenac-Roumanou, H.; Fatome, M.; Laval, J.-D.; Subra, G. Eur. J.
Med. Chem.-Chem. Ther. 1982, 17, 235–243.
(21) (a) Choi, W.; Sanda, F.; Kihara, N.; Endo, T. J. Polym. Sci., Part A:
Polym. Chem. 1997, 35, 3853–3856. (b) Motokucho, S.; Sudo, A.; Sanda, F.;
Endo, T. Chem. Commun. 2002, 1946–1947.
(22) Runge, F.; El-Heweki, Z.; Renner, H. J.; Taeger, E. J. Prakt. Chem.
1960, 11, 284–308.
(23) For a preliminary report of some of this work, see: North, M.;
Villuendas, P. Synlett 2010, 623–627.
6202 J. Org. Chem. Vol. 75, No. 18, 2010

Clegg et al.
JOCArticle
TABLE 1. Synthesis of Di- and Trithiopropylene Carbonates 2a and 4aa
SCHEME 4. Synthesis of Dithiocarbonates Catalyzed by 1 and
Bu4NBr or Bu3N
1
(mol %)
Bu4NBr
(mol %)
CS2
(equiv)
T
(°C)
yield
(%)
entry
2a/4a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0
5
0
5
0
5
2.5
2.5
2.5
5
5
5
5
2.5
5
0
0
5
0
5
5
2.5
2.5
2.5
5
5
5
5
5
2.5
5
2.5
1.8
7
7
1.8
1.8
1.8
7
7
7
7
3.5
1.8
1.8
1.8
1.8
1.8
1.8
90
50
50
90
90
90
90
70
45
40
50
50
30
50
50
90
90
0
0
0
<1
2
95
59
47
27
67
94
97
26
51
91
78
96
0
0
10:90
5:95
71:29
69:31
93:7
87:13
84:16
89:11
81:19
75:25
91:9
of 1,3-oxathiolane-2-thiones derived from epoxides 7a-k was
investigated (Scheme 4). The results are summarized in Table 2,
and in most cases, good selectivities in favor of 1,3-oxathiolane-
2-thiones 2a-i were obtained. Terminal, aliphatic epoxides
7a-e all gave predominantly the 1,3-oxathiolane-2-thione
2a-e, though the 2/4 ratio decreased as the length of the alkyl
chain increased (entries 1-5). Styrene oxide 7f was exceptional
in that even at 50 °C it gave 1,3-dithiolane-2-thione 4f as the
only product (entry 6). Functionalized epoxides 7g,h both gave
good results, with phenoxypropylene oxide 7h giving a 99%
yield of a 97:3 ratio of 2h/4h (entries 7 and 8).
A notable difference between reactions involving carbon
disulfide and those we have previously reported using carbon
dioxide is that vicinally disubstituted cyclohexene oxide 7i
was converted into an almost 1:1 ratio of 1,3-oxathiolane-
2-thione 2i and 1,3-dithiolane-2-thione 4i in moderate yield
(entry 9), whereas vicinally disubstituted epoxides were inert to
reaction with carbon dioxide.10a Compounds 2i and 4i were
both obtained exclusively as the trans-isomer, corresponding to
inversion of configuration having occurred during the synthesis
of 2i and 4i from cis-cyclohexene oxide. Attempts to extend the
chemistry to geminally disubstituted substrate 7j or trisubsti-
tuted epoxide 7k were unsuccessful (entries 10 and 11).
Determination of the structure of compounds 2a-i and 4a-i
was far from straightforward as the spectroscopic data pre-
sent in the literature contained numerous errors and incon-
sistencies. A full tabulation of all of the literature data for
compounds 2a-i and 4a-i is given in the Supporting Informa-
tion. However, as an example, the 1H and 13C NMR spectra we
recorded for the crystalline product obtained from styrene
oxide matched the literature data for 4f24 but also matched
the data in two papers claiming to have prepared 2f.14f,g
A third paper19b on the synthesis of 2f, however, had spectro-
scopic data which differed from those we recorded. The high-
resolution mass spectrum of our material was entirely consis-
tent with structure 4f, but to confirm this unambiguously, a
single-crystal X-ray analysis of our product was undertaken,
which clearly showed that the product was indeed trithiocar-
bonate 4f (Figure 1). This, combined with numerous other
spectroscopic inconsistencies detailed in the Supporting Infor-
mation, suggests that some of the previously reported14f,g
syntheses of 1,3-oxathiolane-2-thiones 2 actually synthesized
1,3-dithiolane-2-thiones 4. This will be discussed later in this
manuscript.
2.5
5
26:74
13:87
aAll reactions were carried out for 16 h.
this optimization study are shown in Table 1. Entry 1 shows
that, at 90 °C, there is no significant uncatalyzed background
reaction between epoxide 7a and carbon disulfide. Entries 2 and 3
illustrate that neither complex 1 nor tetrabutylammonium bro-
mide alone is capable of inducing the reaction at 50 °C, and even at
90 °C they have negligible catalytic activity (entries 4 and 5).
However, the simultaneous use of complex 1 and tetra-
butylammonium bromide was capable of catalyzing the
formation of di- and trithiocarbonates 2a and 4a (entries
6-13). At 90 °C, using less than 2 equiv of carbon disulfide,
the reaction gave 2a/4a in a 1:14 ratio, while use of 7 equiv of
carbon disulfide gave a 2.4:1 ratio of 2a/4a (entries 6 and 7).
In the latter case, reducing the reaction temperature increased the
2a/4a ratio to >13:1 at 45 °C, albeit at the expense of a lowered
chemical yield (entries 7-9). By increasing the amount of both
catalysts to 5 mol %, the yield was significantly improved, and
under these conditions, the amount of carbon disulfide used
could be reduced to less than 2 equiv (entries 10-13) while
retaining the preference for formation of product 2a. A reaction
temperature of 50 °C was found to be optimal in terms of both
total yield and 2a/4a ratio (entry 12), since further lowering of the
reaction temperature significantly reduced the chemical yield and
also slightly reduced the product ratio (entry 13).
Entries 14-17 of Table 1 show that the amount of catalyst
1 used rather than the amount of tetrabutylammonium
bromide principally determined the yield and chemoselecti-
vity of reactions carried out at both 50 and 90 °C. Thus, at
50 °C, halving the amount of catalyst 1 significantly reduced
both the yield and the 2a/4a ratio (compare entries 12 and 14),
while halving the amount of tetrabutylammonium bro-
mide used only slightly reduced the chemical yield and had
no significant effect on the 2a/4a ratio (compare entries 12
and 15). Similarly, at 90 °C, halving the amount of catalyst 1
significantly reduced both the yield and the 4a/2a ratio
(compare entries 6 and 16), while halving the amount of
tetrabutylammonium bromide used only slightly reduced the
4a/2a ratio and had no effect on the chemical yield (compare
entries 6 and 17).
Products 2i/4i obtained from cyclohexene oxide 7i were
separable by careful column chromatography. The structure
and trans-stereochemistry of compound 2i have previously
been determined by Endo using X-ray analysis.14a Again,
however, some,13a,14g but not all,14f of the literature data
Since the conditions of Table 1, entry 12, gave the best
chemical yield and a high 2a/4a ratio, they were adopted as
the standard conditions for the synthesis of 1,3-oxathiolane-
2-thiones, and the applicability of the chemistry to the synthesis
(24) (a) Taguchi, Y.; Yanagiya, K.; Shibuya, I.; Suhara, Y. Bull. Chem.
Soc. Jpn. 1987, 60, 727–730. (b) Sugawara, A.; Nakamura, A.; Araki, A.;
Sato, R. Bull. Chem. Soc. Jpn. 1989, 62, 2739–2741.
J. Org. Chem. Vol. 75, No. 18, 2010 6203

Clegg et al.
JOCArticle
TABLE 2. Synthesis of Dithiocarbonates 2a-i and Trithiocarbonates 4a-ia
50 °C (Bu4NBr)
90 °C (Bu4NBr)
yield (%) 2/4
50 °C (Bu3N)
entry
R1
CH3
C2H5
C4H9
C6H13
C8H17
Ph
CH2Cl
CH2OPh
R2
R3
products 2, 4
yield (%) 2/4
yield (%)
2/4
1
2
3
4
5
6
7
8
9
H
H
H
H
H
H
H
H
H
a
b
c
d
e
f
g
h
i
97
54
56
61
49
62
76
97
35
0
89:11
85:15
66:34
70:30
56:44
0:100
85:15
97:3
94
90
87
81
85
91
84
87
76
0
5:95
47:53
43:57
39:61
45:55
0:100
36:64
71:29
19:81
89
40
38
41
35
71
66
80:20
78:22
58:42
87:13
91:9
0:100
85:15
98:2
H
H
H
H
H
H
H
100
25
-(CH2)4-
H
58:42
66:33
10
11
12
CH3
CH3
H
CH3
H
CH3
CH3
H
j
k
a
0
94
0
CH3 (R)
82:18
aAll reactions carried out for 16 h.
FIGURE 1. Molecular structure of compound 4f with 40% prob-
ability ellipsoids.
were consistent with those reported by Endo. Our spectro-
scopic data were also consistent with those of Endo, and
1
analysis of the H NMR spectrum allowed the three-bond
FIGURE 2. Molecular structure of compound 4i with 40% prob-
ability ellipsoids.
coupling constant between the two methine hydrogen atoms
to be determined as 11.9 Hz which is also consistent with a
trans-ring fusion. The symmetrical nature of trithiocarbo-
nate 4i meant that its stereochemistry could not be unambi-
guously determined from its spectroscopic data, though
our data were consistent with those previously reported.24a
Therefore, the single-crystal X-ray analysis of compound 4i
was undertaken, and this confirmed the trans nature of the
ring fusion (Figure 2). The analysis was complicated by the
extremely small size of the crystals, necessitating the use of
synchrotron radiation, and by the fact that the molecule is
disordered across a crystallographic mirror plane passing
through atoms S(1) and C(1) and bisecting the two fused
rings; this disorder has been satisfactorily resolved with the
aid of some refinement restraints, clearly indicating a trans
ring fusion, essentially planar heterocycle, chair conforma-
tion for the cyclohexane ring, and a racemic mixture of the
R,R and S,S enantiomers.
From the results shown in Table 1 (especially comparing
entries 6 and 12), it appeared that, while reactions carried out
at 50 °C produced predominantly 1,3-oxathiolane-2-thiones 2,
reactions conducted at 90 °C should produce 1,3-dithiolane-
2-thiones 4 as the major product. Therefore, the synthesis of
compounds 4 from epoxides 7 was investigated (Scheme 4)
by carrying out the reaction between the epoxide and carbon
disulfide at 90 °C under the conditions of Table 1, entry 6.
The results of this study are also shown in Table 2.
Even at 90 °C, geminally substituted epoxides 7j,k failed to
undergo any reaction (entries 12 and 13) while styrene oxide
7f, which had given trithiocarbonate 4f as the sole product
at 50 °C, also gave trithiocarbonate 4f as the only isolated
product at 90 °C (entry 6). At 90 °C, monosubstituted epox-
ides 7a-e,g did produce trithiocarbonates 4a-e,g, though,
with the exception of propylene oxide 7a, there was only a
slight preference for formation of trithiocarbonate 4 over
dithiocarbonate 2 (entries 1-5 and 7). Cyclohexene oxide 7i
gave the trans-diastereomers of 2i and 4i in a 1:4 ratio (entry 9).
Epoxide 7h was unique among the terminal epoxides in that,
at 90 °C, the major product was still the dithiocarbonate 2h,
though the 2h/4h selectivity was reduced from that obtained
at 50 °C (entry 8). To confirm this unexpected result, since
there were only melting point and infrared data for com-
pound 4h in the literature,18d and the literature data for
compound 2h contained inconsistencies,14a,f,g compounds 2h
and 4h were separated by column chromatography. The
major product had high-resolution mass spectral data con-
sistent with structure 2h. The minor product was a crystalline
To further investigate the stereochemistry of this reaction
and to extend the chemistry to the synthesis of nonracemic
dithiocarbonates, the use of enantiomerically pure (R)-7a as
substrate was investigated (Table 2, entry 12). The yield and
2a/4a ratio were similar to those obtained with the racemic
epoxide (compare Table 2, entries 1 and 12). Compounds 2a
and 4a were separable by column chromatography, and both
were found to be optically active ((R)-2a: [R]20 -29.2
D
(c = 2.5, CHCl3); (R)-4a: [R]20D -35.9 (c = 2.5, CHCl3)). Un-
fortunately, the enantiomers of 2a and 4a were not separable
by chiral HPLC (ADH or ASH columns), and attempted
analysis by chiral GC resulted in thermal decomposition to
mixtures of compounds 2a-6a.
6204 J. Org. Chem. Vol. 75, No. 18, 2010

Clegg et al.
JOCArticle
SCHEME 6. Catalytic Cycle for 1,3-Oxathiolane-2-Thione
Synthesis
FIGURE 3. Molecular structure of compound 4h with 40% prob-
ability ellipsoids.
SCHEME 5. Catalytic Cycle for Cyclic Carbonate Synthesis
cyclization to form the cyclic carbonate and regenerate the
catalytic species.
However, this mechanism is not consistent with the stereo-
chemical results obtained in the reaction of cyclohexene
oxide 7i with carbon disulfide. In particular, the mechanism
shown in Scheme 5 involves double inversion at one end of
the epoxide (formation of the bromide intermediate followed
by its displacement during the cyclic carbonate forming step)
and so overall retention of stereochemistry should be ob-
served. However, the conversion of epoxide 7i into 1,3-
oxathiolane-2-thione 2i was shown to involve inversion of
stereochemistry. The same inversion of stereochemistry was
previously observed by Endo et al.14a and was explained on
the basis of a mechanism in which the epoxide ring is opened
by a sulfur-based nucleophile. The mechanism shown in
Scheme 5 can similarly be modified to the rather simpler
catalytic cycle shown in Scheme 6, which is consistent with
the observed stereochemical outcome.
In the mechanism shown in Scheme 6, the tetrabutylam-
monium bromide still decomposes in situ to form butyl
bromide and tributylamine. The tributylamine reacts as
a Lewis base with carbon disulfide to form a dithiocarba-
mate,13,18a,24a which acts as a nucleophile and ring opens
the aluminum-coordinated epoxide. Subsequent cyclization
directly forms the 1,3-oxathiolane-2-thione and regenerates
the Lewis acid and Lewis base catalysts. This suggests that
the only role for the tetrabutylammonium bromide is to pro-
vide an in situ source of tributylamine and that bromide is not
involved in ring-opening the epoxide. An experiment carried
out at 50 °C under the conditions of Table 2 with epoxide 7h as
substrate but using tributylamine instead of tetrabutylammo-
nium bromide supported this hypothesis as a quantitative yield
of 2h/4h in a 98:2 ratio was obtained. In view of this success, this
procedure was extended to all epoxides 7a-i, and the results are
included in Table 2. In each case, epoxide 7a-i was converted
into products 2/4a-i, though in most cases, the yields and/or
2/4 ratios were lower than those obtained using tetrabutyl-
ammonium bromide as an in situ source of tributylamine,
suggesting that too high a concentration of amine may inhibit
the catalytic activity of aluminum complex 1. The exceptions
were epoxides 7d,e,i which all contain long alkyl groups
attached to the epoxide and which all gave better 2/4 ratios
with tributylamine than with tetrabutylammonium bromide,
though with lower chemical yields, and styrene oxide 7f which
gave a higher chemical yield (exclusively of trithiocarbonate 4f)
with tributylamine as cocatalyst.
solid, and its high-resolution mass spectral data and single-
crystal X-ray structure (Figure 3) confirmed that it was indeed
the trithiocarbonate. Compound 4h also gave very small cry-
stals which required synchrotron radiation to allow the X-ray
structure to be determined. In addition, the crystals were
found to be twinned, though this was resolved satisfactorily.
The NMR data for the major isomer matched the data in two
of the previous reports of the synthesis of compound 2h.14a,g
However, a third paper14f claiming the synthesis of compound
2h contained different spectroscopic data, and these were
foundtobe identical tothose we obtainedfor trithiocarbonate
4h. Thus, we were able to confirm unambiguously the struc-
ture of products 2h and 4h and correct the literature data.
We have previously proposed10b the catalytic cycle
shown in Scheme 5 for the synthesis of cyclic carbonates
from epoxides and carbon dioxide catalyzed by complex 1
and tetrabutylammonium bromide. The key features of
this mechanism are that complex 1 acts as a Lewis acid,
activating the epoxide toward ring-opening by bromide. A
second tetrabutylammonium bromide molecule decom-
poses in situ by a reverse Menschutkin reaction25 to form
butyl bromide and tributylamine. The latter species can
act as a Lewis base and activate the carbon dioxide as a
carbamate. The bimetallic nature of complex 1 allows this
also to coordinate to an aluminum ion to give key inter-
mediate 8. Formation of the aluminum-coordinated carbo-
nate can then occur intramolecularly and is followed by
(25) (a) Hofman, A. W. Proc. R. Soc. London 1860, 10, 594–596.
(b) Collie, N.; Schryver, S. B. J. Chem. Soc. 1890, 767–782. (c) Zaki, A.;
Fahim, H. J. Chem. Soc. 1942, 270–272. (d) Gordon, J. E. J. Org. Chem.
1965, 30, 2760–2763.
J. Org. Chem. Vol. 75, No. 18, 2010 6205

Clegg et al.
JOCArticle
TABLE 3. Synthesis of Compounds 2h and 4h Using Monometallic
Aluminum(salen) Complexes
catalyst
yield (%)
ratio of 2h/4h
9
10
11
93
80
98
91:9
78:22
90:10
To provide additional evidence to support the catalytic
cycle shown in Scheme 6, the reaction kinetics were moni-
tored. These reactions were carried out using epoxide 7h as
substrate so that dithiocarbonate 2h was formed essentially
as the sole product, without coformation of trithiocarbonate
4h complicating the kinetic analysis. It was found that the
kinetics could be monitored by 1H NMR spectroscopy using
CDCl3 assolventat50°C. The concentration of each component
of the reaction was varied, allowing the rate equation (eq 1) to be
derived. This can be compared with the rate equation previously
determined10b,c for cyclic carbonate synthesis (eq 2).
The mechanism by which 1,3-dithiolane-2-thiones 4 are
formed has previously been discussed on the basis that
they are formed from 1,3-oxathiolane-2-thiones 2. The
accepted mechanism13,18a,b,d (Scheme 7) involves reaction of
a nucleophile (bromide or dithiocarbamate 12 in our case)
with the thiocarbonyl bond of 1,3-oxathiolane-2-thione 2.
SCHEME 7. Mechanism for the Synthesis of 1,3-Dithiolane-2-
thiones 4
rate ¼ k½1ꢀ½7hꢀ and zero order in both ½CS2ꢀ and ½Bu4NBrꢀ ð1Þ
2
rate ¼ k½1ꢀ½epoxideꢀ½CO2ꢀ½Bu4NBrꢀ
ð2Þ
Clearly, the two processes follow very different rate equa-
tions, which is consistent with them occurring by different
mechanisms. The zero-order dependence of the rate of dithio-
carbonate synthesis on the concentrations of both carbon
disulfide and tetrabutylammonium bromide is particularly
notable. This suggests that these species are only involved in
the mechanism after the rate-determining step and is consistent
with the mechanism in Scheme 6 provided formation of the
catalyst-epoxide complex is rate-limiting. This is reasonable
since the high nucleophilicity of sulfur will ensure that the
subsequent epoxide ring-opening is rapid and the final step of
the catalytic cycle is intramolecular (5-exo-trig cyclization).
The difference in the mechanisms for carbonate and dithio-
carbonate formation is explained on the basis of sulfur being a
better nucleophile than bromide, which is, however, a better
nucleophile than oxygen. This also explains why cyclohexene
oxide is a substrate for the reaction with carbon disulfide but not
for reaction with carbon dioxide. Only the most reactive sulfur-
based nucleophile is able to react with the vicinally disubstituted
epoxide under the reaction conditions. It is also notable that the
mechanism shown in Scheme 5 exploits the bimetallic nature of
complex 1 to preorganize both the ring-opened epoxide and
carbamate for intramolecular reaction. In contrast, the mechan-
ism in Scheme 6 involves no such preorganization, which may
explain why reactions involving carbon disulfide require a
higher reaction temperature (50-90 °C) than those involving
carbon dioxide (20 °C).
Since the mechanism shown in Scheme 6 involves only one
aluminum atom in the catalytic cycle, the catalytic activity of
complexes 9-11 was investigated. These complexes had all
previously been shown to be inactive for cyclic carbonate
synthesis,10c but in the presence of tetrabutylammonium bro-
mide and under the conditions of Table 1, entry 12, all were
found to catalyze the reaction between epoxide 7h and carbon
disulfide to form dithiocarbonate 2h and trithiocarbonate 4h,
again with dithiocarbonate 7h as the major product (Table 3).
This clearly demonstrates that the bimetallic nature of complex
1 is not essential for its catalytic activity and provides further
evidence to support the mechanism shown in Scheme 6.
Subsequent cleavage of the carbon-sulfur bond within the
five-membered ring establishes intermediate 13 in which the
oxygen is part of a good leaving group and the thiolate is a
good nucleophile. Thus, intermediate 13 can cyclize to
thiirane 6, which can then be ring-opened by dithiocarba-
mate 12 to give intermediate 14 which can cyclize to 1,3-
dithiolane-2-thione 4. The fact that (R)-2a is converted into
optically active 4a implies that, for this substrate at least, the
formation of thiirane 6 occurs by an SN2 reaction. However,
the extremely facile formation of 1,3-dithiolane-2-thione 4f is
consistent with an SN1-type mechanism for the synthesis of
thiirane 6, which in this case would be facilitated by the
benzylic nature of the intermediate carbenium ion.
To resolve some of the inconsistencies in the literature, we
repeated the synthesis of some of 1,3-oxathiolane-2-thiones
2a-i using the literature procedures. Yavari et al.14f have
reported a method for the synthesis of oxathiolane-2-thiones
2a,b,f,h,i using sodium methoxide (generated in situ from
sodium hydride) in methanol. We repeated this work (using
commercial sodium methoxide) and found that it did give
oxathiolane-2-thiones 2a,b,f,h,i as the major product, but some
of the corresponding 1,3-dithiolane-2-thione 4a,b,f,h,i was also
formed as a minor product. The spectroscopic data of the
reaction mixtures (see Supporting Information) clearly showed
that the major product had spectra which matched those we
report for 2a,b,f,h,i, while the minor product gave the spectro-
scopic data reported in the literature. Saidi et al. reported a
synthesis of oxathiolane-2-thiones 2a,b,e-i using triethylamine
in water.14g We repeated their synthesis using epoxides 7a,b,f-i
and again found that the oxathiolane-2-thiones 2a,b,f-iwere the
major product in each case but that the spectroscopic data
again matched our data rather than those reported by the
6206 J. Org. Chem. Vol. 75, No. 18, 2010

Clegg et al.
JOCArticle
authors. In some cases, a mixture of oxathiolane-2-thione 2
and 1,3-dithiolane-2-thione 4 was again formed. Thus, styrene
oxide 7f gave a 9:1 ratio of 2f/4f, 3-chloropropylene oxide 7g
gave a 98:2 ratio of 2g/4g, and cyclohexene oxide 7i gave a
3.5:1 ratio of 2i/4i.
In those cases where compounds 2 and 4 could be separated,
full spectroscopic data (IR, 1H NMR, 13C NMR, low- and high-
resolution mass spectra, and X-ray data for 4f,h,i) are given in
1
the Supporting Information. Copies of the H and 13C NMR
spectra of all pure compounds and/or the mixture of compounds
2 and 4 are also given in the Supporting Information. Spectro-
scopic data for new compounds are given below.
Conclusions
5-Butyl-1,3-oxathiolane-2-thione 2c: νmax 1191 cm-1; δH (CDCl3)
0.87 (3H, t J = 6.7 Hz), 1.2-1.4 (3H, m), 1.4-1.5 (1H, m), 1.7-1.8
(1H, m), 1.9-2.0 (1H, m), 3.35 (1H, dd J = 9.5, 6.2 Hz), 3.52 (1H,
dd J = 11.0, 6.5 Hz), 5.0-5.1 (1H, m); δC (CDCl3) 13.8, 22.3, 27.4,
33.4, 39.3, 91.8, 212.1; m/z (ESI) 199 (M þ Naþ) found 177.0405,
(C7H13S2O, MHþ) requires 177.0408.
In conclusion, we have shown that bimetallic aluminum-
(salen) complex 1 in conjunction with tetrabutylammonium
bromide is an effective catalyst system for the formation of
1,3-oxathiolane-2-thiones and 1,3-dithiolane-2-thiones. The
synthesis of 1,3-oxathiolane-2-thiones 2 is a stereospecific
process which proceeds with inversion of configuration. The
structures and stereochemistry of three of the 1,3-dithiolane-
2-thiones has been unambiguously determined by single-
crystal X-ray analysis, and this has allowed structural assign-
ments in the literature to be corrected. The stereochemical and
reactivity differences between cyclic carbonate and cyclic
dithiocarbonate formation can be explained on the basis of the
established mechanism for cyclic carbonate synthesis and the
greater nucleophilicity of sulfur-based nucleophiles than their
oxygen counterparts. This mechanistic analysis is supported by
an analysis of the reaction kinetics and by modification of the
catalyst structure.
4-Butyl-1,3-dithiolane-2-thione 4c: νmax 1061 cm-1;δH (CDCl3)
0.86(3H,tJ=6.7Hz),1.2-1.4(4H,m),1.7-2.0 (2H, m), 3.64 (1H,
dd J = 11.9, 8.0 Hz), 3.90 (1H, dd J = 11.9, 5.4 Hz), 4.2-4.4 (1H,
m); δC (CDCl3) 13.8, 22.3, 30.3, 33.2, 48.2, 60.9, 227.9; m/z(ESI) 215
(M þ Naþ) found 192.0095, (C7H13S3, Mþ) requires 192.0101.
4-Octyl-1,3-dithiolane-2-thione 4e: νmax 1068 cm-1; δH (CDCl3)
0.82 (3H, t J = 6.8 Hz), 1.1-1.4 (12H, m), 1.8-2.0 (2H, m), 3.64
(1H, dd J = 10.8, 6.1 Hz), 3.90 (1H, dd J = 11.0, 6.4 Hz), 4.2-4.4
(1H, m); δC (CDCl3) 14.0, 22.6, 28.3, 29.1, 29.2, 29.3, 31.7, 33.5, 48.2,
61.0, 227.9; m/z (ESI) 271 (M þ Naþ) found 249.0793, (C11H21S3,
MHþ) requires 249.0805.
Reaction Kinetics. Complex 1, tetrabutylammonium bromide,
and epoxide 7h were dissolved in CDCl3 (0.75 mL) to give the
concentrations required for a particular experiment. The solution
was added to an NMR tube followed by addition of the
required amount of carbon disulfide. The tube was sealed and
inserted into the NMR spectrometer with the probe heated to
50 °C. A 1H NMR spectrum was collected every ca. 30 min for
ca. 10 h.
Experimental Section
Synthesis of 1,3-Oxathiolane-2-thiones and 1,3-Dithiolane-
2-thiones. Carbon disulfide (0.18 mL, 2.98 mmol), an epoxide
(1.67 mmol), catalyst 1 (74 mg, 0.083 mmol), and tetrabutyl-
ammonium bromide (28 mg, 0.083 mmol) or tributylamine
(15 mg, 0.083 mmol) were placed in a sealed Young flask and
stirred at 50 or 90 °C for 16 h. The solution was evaporated and the
residue purified by column chromatography (chloroform/hexane 1/1)
to give compounds 2a-i and/or 4a-i. Compounds 2a,c-e
and 4g could not be separated from the other reaction product in
any solvent system tried. In each case, however, the other
reaction product (4a,c-e and 2g) was obtained pure, and
spectra of compounds 2a,c-e and 4g showed that no other
impurities were present. A sample of compounds 4f, 4i, and 4h
was dissolved in MeOH (2 mL) and left in a refrigerator for 48
h to produce crystals suitable for X-ray analysis.
Acknowledgment. We thank EPSRC for funding for the
National Crystallography Service and STFC for access to
synchrotron facilities at Diamond Light Source.
Supporting Information Available: 1H and 13C NMR spec-
tra of compounds 2,4a-i; a tabulated comparison of all of the
literature data for compounds 2,4a-i; crystallographic infor-
mation files for compounds 4f,h,i; details of all kinetic experi-
ments. This material is available free of charge via the Internet at
http://pubs.acs.org.
J. Org. Chem. Vol. 75, No. 18, 2010 6207