Synthetic, mechanistic, and structural studies related to 1,2,4-dithiazolidine-3,5-dione

Article abstract of DOI:10.1021/jo960723u

Reaction of 0-ethyl thiocarbamate (4) with (chlorocarbonyl)sulfenyl chloride (5) gives 3-ethoxy-1,2,4-dithiazolin-5-one (2) and 3,5-diethoxy-1,2,4-thiadiazole (3), with the relative amounts of 2 and 3 formed depending very much on the solvent (e.g., diethyl ether favors 2; chloroform favors 3). The effects of added base, order of addition, concentration, and temperature were also studied. Mechanisms for the observed transformations have been proposed and are supported by the characterization of relatively unstable acyclic intermediates, e.g., formimidoyl(chlorocarbonyl)-disulfane 8, symmetrical bis(formimidoyl)disulfane 10, and ethoxythiocarbonyl imidate 11, which are obtained under alternative conditions. Compound 2 is converted with concentrated aqueous hydrochloric acid upon short reflux to 1,2,4-dithiazolidine-3,5-dione (1), rearranges upon prolonged melting to give principally N-ethyl-1,2,4-dithiazolidine-3,5-dione (13), and is desulfurized with various trivalent phosphorus compounds to yield O-ethyl cyanate (15) plus carbonyl sulfide. X-ray crystallographic structures of 1 and 2 have been solved; the planarity and aromatic character of these molecules help to explain some of their reactions.

Full text of DOI:10.1021/jo960723u

J . Org. Chem. 1996, 61, 6639-6645
6639
Syn th etic, Mech a n istic, a n d Str u ctu r a l Stu d ies Rela ted to
1
,2,4-Dith ia zolid in e-3,5-d ion e
Lin Chen,1a Tracy R. Thompson,1a,c Robert P. Hammer,*,1a,b and George Barany*,1a
Department of Chemistry, University of Minnesota, Minneapolis, Minnesota 55455, and Department of
Chemistry, Louisiana State University, Baton Rouge, Louisiana 70803
Received April 19, 1996X
Reaction of O-ethyl thiocarbamate (4) with (chlorocarbonyl)sulfenyl chloride (5) gives 3-ethoxy-
1
,2,4-dithiazolin-5-one (2) and 3,5-diethoxy-1,2,4-thiadiazole (3), with the relative amounts of 2
and 3 formed depending very much on the solvent (e.g., diethyl ether favors 2; chloroform favors
). The effects of added base, order of addition, concentration, and temperature were also studied.
3
Mechanisms for the observed transformations have been proposed and are supported by the
characterization of relatively unstable acyclic intermediates, e.g., formimidoyl(chlorocarbonyl)-
disulfane 8, symmetrical bis(formimidoyl)disulfane 10, and ethoxythiocarbonyl imidate 11, which
are obtained under alternative conditions. Compound 2 is converted with concentrated aqueous
hydrochloric acid upon short reflux to 1,2,4-dithiazolidine-3,5-dione (1), rearranges upon prolonged
melting to give principally N-ethyl-1,2,4-dithiazolidine-3,5-dione (13), and is desulfurized with
various trivalent phosphorus compounds to yield O-ethyl cyanate (15) plus carbonyl sulfide. X-ray
crystallographic structures of 1 and 2 have been solved; the planarity and aromatic character of
these molecules help to explain some of their reactions.
2
3,5-diethoxy-1,2,4-thiadiazole (3).9 In addition, X-ray
crystallographic analyses of 1 and 2 confirm the planar
aromatic character of each heterocycle and facilitate the
understanding of some reactions of the title heterocycles
Recent work from our laboratory has shown that 1,2,4-
dithiazolidine-3,5-dione (1), a compound first prepared
by German scientists several decades ago,3 is a highly
effective sulfurization reagent that provides access to
phosphorothioate analogues of DNA and phosphopep-
tides. In addition, 1 represents a protected form6 of
ammonia and may be useful for mild homologation of the
classical Gabriel synthesis.7 The present contribution
reports an efficient, optimized two-step procedure for the
preparation of 1, via the key intermediate 3-ethoxy-1,2,4-
-5
1
and 2.
Resu lts a n d Discu ssion
R ea ct ion of O-E t h yl Th ioca r b a m a t e (4) w it h
(
Ch lor oca r bon yl)su lfen yl Ch lor id e (5). With the goal
8
to prepare alkyl-free heterocyclic disulfide 1, reaction of
dithiazolin-5-one (2), and also describes how the reaction
designed to give 2 provides entry to another heterocycle,
1
0
11
4
1
3
with 5 was found to stop at heterocycle 2 (Scheme
).3a,c A second isolable product from the reaction was
,5-diethoxy-1,2,4-thiadiazole (3, Scheme 1), and the
9
X
Abstract published in Advance ACS Abstracts, August 1, 1996.
(
1) (a) University of Minnesota. (b) Louisiana State University. (c)
Permanent Address: Department of Natural Sciences, Edgewood
College, Madison, WI 53711.
(5) Previous preparative and mechanistic studies from our labora-
tory related to N-R-1,2,4-dithiazolidine-3,5-diones [R ) H, alkyl, and
H N-R′ corresponding to amino acids and peptides] are described in
(2) (a) Xu, Q.; Musier-Forsyth, K.; Hammer, R. P.; Barany, G. In
2
Peptides: Chemistry, Structure & Biology. Proceedings of the 14th
American Peptide Symposium; Kaumaya, P. T. P.; Hodges, R. S. Eds.;
Mayflower Scientific: Kingswinford, England, 1996; pp 123-124. (b)
Xu, Q.; Musier-Forsyth, K.; Hammer, R. P.; Barany, G. Nucleic Acids
Res. 1996, 24, 1602-1607.
the following: (a) Barany, G.; Merrifield, R. B. J . Am. Chem. Soc. 1977,
99, 7363-7365. (b) Słomczy n´ ska, U.; Barany, G. J . Heterocycl. Chem.
1984, 21, 241-246. (c) Barany, G.; Słomczy n´ ska, U.; Mott, A. W.
Abstracts of Papers, 187th National Meeting of the American Chemical
Society, St. Louis, MO, April 8-13, 1984; American Chemical Society:
Washington, DC, 1984; ORG 32. (d) Zalipsky; S.; Albericio, F.;
Słomczy n´ ska, U.; Barany, G. Int. J . Pept. Protein Res. 1987, 30, 740-
783. (e) Hammer, R. P.; Barany, G. Abstracts of Papers, 195th National
Meeting of the American Chemical Society, Toronto, Canada, J une
5-10, 1988; American Chemical Society: Washington, DC, 1988; ORG
137.
(6) The dithiasuccinoyl (Dts) amino protecting group, removable by
thiolysis and other reductive methods, was described first in ref 5a
and has been applied in numerous other papers from our laboratories,
e.g.: (a) Barany, G.; Merrifield, R. B. J . Am. Chem. Soc. 1980, 102,
3084-3095. (b) Barany, G.; Albericio, F. J . Am. Chem. Soc. 1985, 107,
4936-4942. (c) Albericio, F.; Barany, G. Int. J . Pept. Protein Res. 1987,
30, 177-205. (d) Hammer, R. P.; Albericio, F.; Gera, L.; Barany, G.
Int. J . Pept. Protein Res. 1991, 36, 31-45. (e) J ensen, K. J .; Hansen,
P. R.; Venugopal, D.; Barany, G. J . Am. Chem. Soc. 1996, 118, 3148-
3155 and references cited in all of these papers.
(3) (a) Review: Zumach, G.; K u¨ hle, E. Angew Chem., Int. Ed. Engl.
1
1
970, 9, 54-63. (b) Dahms, G.; Haas, A.; Klug, W. Chem. Ber. 1971,
04, 2732-2742. (c) The method sketched in ref 3a was described more
exactly in a letter dated Dec 19, 1973, from G. Zumach to G. Barany:
3
Et
2
1% yield for reaction in diethyl ether of 4 and 5 in the presence of
3
N (1 equiv) to give 2, mp 56-57 °C (iPrOH); 30% yield for conversion
to 1, mp 142-144 °C (toluene). (d) The present work describes
substantially improved yields of 1 or 2 over those reported in ref 3c,
defines an alternative product 3, and indicates that the presence of
base is not critical to achieve good yields and purities of 2. The initial
reaction of 4 and 5 does not give any C NMR-detectable 1, regardless
of the presence or absence of base. The ether needs to be reasonably
dry to minimize the formation of the hydrolysis product O-ethyl
1
3
1
13
carbamate [ H NMR (CDCl
CDCl ) δ 157.3, 60.9, 14.4].
4) Compound 1 is also claimed in two patents, but as stated in a
3
) δ 4.12 (q, 2 H), 1.24 (t, 3 H); C NMR
(
3
(
review (ref 3a), the methods that are general for N-R-1,2,4-dithiazo-
lidine-3,5-diones (R ) alkyl, aryl) fail for the special case R ) H. The
patents referred to describe the following. (a) R′O(CdS)NHR +
Cl(CdO)SCl: Zumach, G.; Weiss, W.; K u¨ hle, E., Belgian Patent 682,-
(7) For a review of Gabriel synthesis of amines (alkylation of
phthalimide salts, followed by dephthaloylation), see: Gibson, M. S.;
Bradshaw, R. W. Angew. Chem., Int. Ed. Engl. 1968, 7, 919-930.
(8) Compound 2 also has attractive properties for efficient sulfur-
ization (see ref 2b and Scheme 4 of the present paper) and in Gabriel-
like syntheses: Chen, L.; Hammer, R. P.; Barany, G., work in progress.
(9) Compound 3 has been described previously: (a) Holmberg, B.
Chem. Zentralbl. 1930, 1925-1926. (b) Kresze, G.; Horn, A.; Philipp-
son, R.; Trede, A. Chem. Ber. 1965, 98, 3401-3409. The paper by
Kresze and co-workers includes some interesting mechanistic experi-
ments, including crossover studies.
9
91, J une 23, 1966; British Patent 1,136,737, J une 21, 1966; Chem.
Abstr. 1969, 70, 77951p. (b) H(CdO)NHR + 2 Cl(CdO)SCl: Zumach,
G.; Weiss, W.; K u¨ hle, E., Farbenfabriken Bayer AG. Belgian Patent
6
82,820, J une 20, 1966; Chem. Abstr. 1968, 68, 105203a. For a
comprehensive description of this and the related patent literature,
see ref 3a and the following: (d) K u¨ hle, E. The Chemistry of the Sulfenic
Acids; Georg Thieme: Stuttgart, Germany, 1973.
S0022-3263(96)00723-2 CCC: $12.00 © 1996 American Chemical Society

6
640 J . Org. Chem., Vol. 61, No. 19, 1996
Chen et al.
Sch em e 1
Ta ble 1. P r od u ct Distr ibu tion fr om Rea ction of O-Eth yl
Th ioca r ba m a te (4) a n d (Ch lor oca r bon yl)su lfen yl
a
Ch lor id e (5) u n d er Va r iou s Con d ition s
concentration (M)
conv of 4 (%)
entry solvent
4
Et
3
N
5
order of addition
2
3
b
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
Et
Et
Et
Et
Et
Et
Et
Et
Et
Et
CHCl
CHCl
CHCl
CHCl
CHCl
CHCl
2
2
2
2
2
2
2
2
2
2
O
O
O
O
O
O
O
O
O
O
2
2
1
1
0
2
1
1
0
0
0
0.4
0.4
0.8
1
0.2 4 to 5
0.2 (4 + Et
78
75
46
35
67
80
68
84
75
75
3
1
5
b
3
3
N) to 5
N) to 5
9
12
24
18
6
10
5
9
13
52
32
41
63
c
1
1
1
(4 + Et
5 to (4 + Et
3
N)
1
4 to 5
0.4
0.4
0.4
0.4
0.4
1
0.4
0.4
0.4
0.4
0.2
0.4 4 to 5
0.4 5 to 4
0.4 (4 + Et
3
N) to 5
0.4 5 to (4 + Et
3
N)
N) to 5
N)
1
1
1
1
1
1
1
0.4 (4 + 2 × Et
3
3
3
3
3
3
3
1
5 to (4 + Et
3
0
0.4 5 to 4
0.4
0.4
0.8
0.2
0.4 (4 + Et
0.4 5 to (4 + Et
3
N) to 5
N)
14
3
5
N) <1
7
0.4 5 to (4 + 2 × Et
3
b
0.2 5 to (4 + Et N)
3
69
a
Unless indicated otherwise, experiments were carried out in
the indicated solvent on 2 mmol scales [each of 4 and 5 in a 1:1
molar ratio, and (as needed) Et3N]. Two stock solutions at the
indicated concentrations were combined in the indicated order of
addition, at a rate to maintain the temperature at 10 ( 5 °C.
Results were qualitatively the same, except for slightly lower
absolute conversions, in those cases where experiments were
carried out at ambient temperature without cooling. Each com-
pleted reaction mixture was diluted as necessary with additional
solvent (to ∼0.1 M product), washed with 1 N aqueous HCl (3×),
dried (MgSO4), and concentrated. The residue was then taken up
in CH2Cl2, filtered to remove elemental sulfur, and evaporated.
Product yields and distributions reflect conversion of starting 4,
e.g., when 10 mmol of 4 gives 2 mmol of 2 and 3 mmol of 3, this
is reported as 20% of 2 and 60% of 3. Compound 1 was not
detected (e3%) in any of these reactions. b Larger scale, see
relative amounts of 2 and 3 formed depended on solvent
in a striking fashion (Table 1). In addition, the absolute
amounts of 2 and/or 3 formed were influenced markedly
by the presence or absence of a tertiary amine base
(Scheme 1, Table 1) and somewhat by the concentrations
and order of addition of reactants.
Reactions of equimolar amounts of 4 with 5 in chloro-
form (CHCl or CDCl ) at 0-25 °C were rapid, with
3 3
immediate transformation of 4 to final products and the
absence of intermediates that might have been detectable
1
13
by in situ H and C NMR monitoring. When no base
was present, the principal products were the volatile
Experimental Section for full procedure directed toward pure 2
1
2
ethyl chloride (EtCl), cyanic acid (HNdCdO), hydrogen
or 3. c Based on starting 4, 85% mass recovery; complicated
chloride (HCl), and carbonyl sulfide (COS), as well as
spectrum with multiple ethyl-containing products, the major one
1
insoluble elemental sulfur ( /
8
S
8
); no 1 and negligible
of which was 4, formed by decomposition of the actual products
1
13
upon workup. H and C NMR spectra of the reaction product
amounts of 2 or 3 were formed. The same reaction
3
d
mixture, taken before workup, showed unreacted 4, intermediate
carried out in anhydrous diethyl ether (Et
2
O) at 10 °C
1
1, and EtCl as the three major products, accounting respectively
(external cooling to control the spontaneous exotherm)
for 21, 56, and 6% of the ethyl groups. The expected 2 and 3 were
not formed to any appreciable extent, as revealed by spectroscopic
examination both before and after workup.
provided a crude product mixture that comprised prin-
cipally dithiazoline 2 in 78% absolute yield (52% after
recrystallization), as well as 3, at 5% of the molar amount
of 2 (Table 1, entry 1).
3
addition, i.e., 5 to a mixture of 4 and Et N), the major
product was 3, with only a minor amount of 2. In this
latter case, the purified yield of 5 was 54%, taking into
account that 2 molecules of 4 are required to form the
product (Table 1, entry 16). Results upon addition of a
second equivalent of base13 were solvent-dependent: in
Addition of an ethereal mixture of 4 (1 equiv) and
triethylamine (Et
in Et O (Table 1, entry 2) again gave 2 as the predomi-
N, 1 equiv)13 to a solution of 5 (1 equiv)
3
2
nant product in initial purity and overall yield (75%
crude, 63% after recrystallization) comparable to those
Et
2
O, yields of 2 were only slightly worse than optimal
, the extra base
of the base-free reaction.3
c,d
In contrast, when the
(opposite order of
(entries 10 vs 8 vs 6), while in CHCl
3
reaction was carried out in CHCl
3
contributed to formation of a complicated mixture of
acyclic products and recovered starting material (Table
(
10) (a) Blankenhorn, E. J . Prakt. Chem. 1877, 16, 358-384. (b)
1
, entry 15 and footnote c).
Wheeler, H. L.; Barnes, B. Am. Chem. J . 1899, 22, 141-151. (c) Davies,
W.; MacLauren, J . A. J . Chem. Soc. 1951, 1434-1438. (d) Goerdeler,
J .; Schulze, A. Chem. Ber. 1982, 115, 1252-1255.
As could be predicted, the relative amount of “dimer-
like” 3 with respect to “monomer-like” 2 increased at
higher concentration of starting reactants (Table 1,
(
11) (a) Weiss, W. German Patent 1,224,720, Nov 11, 1966; Chem.
Abstr. 1966, 65, 12112h. (b) Barany, G.; Schroll, A. L.; Mott, A. W.;
Halsrud, D. A. J . Org. Chem. 1983, 48, 4750-4761.
entries 3 and 4 vs 8 and 9 in Et
CHCl ) and was somewhat higher when the order of
addition was sulfenyl chloride 5 to the mixture of 4 and
Et N (entries 4 vs 3 and 9 vs 8 in Et O; entries 14 vs 13
in CHCl ).
The differences observed in product distribution be-
tween Et O and CHCl do not appear to be a simple
2
O; entries 11 vs 14 in
(
12) Cyanic acid (bp 23.5 °C) readily trimerizes to cyanuric acid
3
(
2,4,6-trihydroxy-1,3,5-triazine), a material with limited or negligible
solubility in organic solvents that decomposes prior to melting.
Review: Smolin, E. M.; Rapoport, L. In The Chemistry of Heterocyclic
Compounds; Weissberger, A. Ed.; Wiley Interscience: New York, 1959;
Vol. 13, pp 17-48. Some of the reactions described in this work gave
a byproduct, which, on the basis of solubility properties and mass
spectrometry, is assigned as cyanuric acid.
3
2
3
2
3
(
13) Balanced equations for reactions of 4 and 5 to give either 2 or
nominally show formation of 2 equiv of HCl, which may need to be
neutralized by Et
N. As shown in the text, a full 2 equiv of Et N has
little effect on the outcome when the reaction is conducted in Et O,
whereas when CHCl (or CDCl ) is the solvent, the major product
solvent polarity effect, as the respective dielectric con-
3
stants14 and solvent parameters are not widely different
14
3
3
N
N
2
(
0
Et
O, ꢀ
) 4.20, ET ) 0.117; CHCl
, ꢀ
) 4.80, ET
)
2
r
3
r
3
3
.259). As a measurement of the sensitivity of this
appears to be 11 (Scheme 2), which decomposes upon workup back to
starting material 4.
reaction to solvent polarity, several other solvents were

Studies Related to 1,2,4-Dithiazolidine-3,5-dione
J . Org. Chem., Vol. 61, No. 19, 1996 6641
N
tested: 1,2-dimethoxyethane (ꢀ
methyl tert-butyl ether (ꢀ
results comparable to those with Et
r
) 7.20, ET ) 0.231) and
Sch em e 2
N
r
) 4.5, ET ) 0.148) gave
O, toluene (ꢀ ) 2.38,
ET ) 0.099) gave results comparable to those with
2
r
N
N
CHCl
3
, and results with dioxane (ꢀ ) 2.21, ET ) 0.164)
r
N
and tetrahydrofuran (ꢀ
r
) 7.58, ET ) 0.231) were inter-
mediate. Thus, the observed results are likely to reflect
other factors, such as the solubility of intermediates, e.g.,
8
, in the particular solvent (see discussion of mechanism
that follows, as well as Scheme 2).
Mech a n ism s of Dith ia zolin on e a n d Th ia d ia zole
F or m a tion . Insights into the mechanisms of the afore-
mentioned transformations came, in part, from experi-
ments generating the same core heterocyclic systems, e.g.,
1
5
3
-methyl-1,2,4-dithiazolin-5-one (6) by the reaction of
thioacetamide with (chlorocarbonyl)sulfenyl chloride (5),
and 3,5-dimethyl-1,2,4-thiadiazole (7)1 by the oxidation
of thioacetamide with appropriate reagents (see Scheme
6,17
1
, inset). A relevant precedent is the formation of N,N′-
dialkyl-1,2,4-thiadiazolidine-3,5-diones as side products
during the preparation of N-alkyl-1,2,4-dithiazolidine-3,5-
diones from 5 and the appropriate O-alkyl′, N-alkyl
thiocarbamate.18 The bis-ethoxy-substituted thiadiazole
sulfenyl chloride 5 generate respectively the formimidoyl-
20
(chlorocarbonyl)disulfane 8 or the formimidoylsulfenyl
chloride 9. In our proposed mechanism, route a predomi-
nates in Et O, and route b is preferred in CHCl
2
3
.
3
encountered in the present work (optimized yield 69%;
Alternatively, disulfane 8 may be formed initially in
either of these solvents, but we would postulate that, in
see Table 1, entry 16) was prepared previously by
9
a
oxidation of thiocarbamate 4 with hydrogen peroxide
25% yield) or N-sulfinyl-p-toluenesulfonamide9b (61%
CHCl
, route c) before it can collapse and cyclize to dithiaz-
olinone 2 (Scheme 2, route d, the preferred pathway in
Et O). The proposal continues with the bimolecular
reaction of 9 with a second equivalent of 4, providing
3
, 8 rearranges readily to sulfenyl chloride 9 (Scheme
(
2
yield). These precedents raise the relatively straightfor-
ward explanation that formation of 3 is due to the action
2
of 5 as an oxidizing agent, a role it is known to take in
other reactions [Cl(CdO)SCl ≡ “Cl
2
” + COS].1
9
21
symmetrical bis(formimidoyl)disulfane 10.
Subse-
While it is difficult to predict which product should be
preferred depending on whether Et O or CHCl is the
quently, 10 rearranges, with loss of elemental sulfur, via
a six-membered transition state to give acyclic intermedi-
ate 11, in analogy to the rearrangement of N,N′-dialkyl-
2
3
reaction solvent, we are able to propose a mechanism for
formation of 2 as well as 3 that involves as the common
initial step an electrophilic attack on the sulfur of the
thiocarbamate 4 (Scheme 2). Thus, reactions of the
thiocarbonyl moiety of 4 with either the sulfur (Scheme
2
1b
bis(formimidoyl)disulfanes studied by Barrett.
In the
final step of our proposed mechanism, the thiocarbonyl
of 11 is activated22 with a second equivalent of 5 to give
the sulfenyl chloride 12, which then closes rapidly to
provide thiadiazole 3.
2
, route a) or chlorine (Scheme 2, route b) atoms of
We have been able to isolate several of the proposed
intermediates, i.e., 8, 10, and 11, and have shown that
their chemistry is consistent with Scheme 2. Authentic
disulfide 10 was prepared by reaction of 4 (2 equiv) and
(
14) Data taken from the following: Reichardt, C. Solvents and
Solvent Effects in Organic Chemistry; VCH Verlagsgesellschaft: Wein-
heim, 1988; pp 408-410.
(15) This general class of compounds has been mentioned in a review
article (ref 3a) and
documentation (e.g., yield, mp). See the present paper for our procedure
to prepare 6 (due to the low solubility of thioacetamide in both Et
and CHCl , the results are not strictly comparable to our observations
a
book (ref 4d), without any experimental
iodine (1 equiv) in the presence of Et
reaction was cleanest with Et O as solvent (salts removed
by filtration).2 Compound 10 in CDCl
rearranged to
3
N (2 equiv); the
2
2
O
1b
3
3
on the reactions of 4 and 5). Physical and spectral properties of 6 are
also reported herein.
11 upon addition of trifluoroacetic acid (partial decom-
position to 4 occurred under the same conditions), again
in analogy to reactions observed by Barrett.2 Interest-
ingly, the rather unstable 11 is the major component
detected directly in the complicated product mixture from
(
16) Literature routes to 7, and some mechanistic proposals, are
1b
described in ref 9b and the following: (a) Walter, W. J ustus Liebigs
Ann. Chem. 1960, 633, 49-55. (b) El-Wassimy, M. T. M.; J orgensen,
K. A.; Lawesson, S.-O. Tetrahedron 1983, 39, 1729-1734. (c) Hu, N.
K.; Aso, Y.; Otsubo, T.; Ogura, F. Chem. Lett. 1985, 603-606.
reaction in CDCl
presence of Et
N (2 equiv),13 and 10 is formed (present
along with unreacted starting material) from reaction in
3
of 4 and 5 (1 equiv each) in the
(
17) In ref 16c, it was shown that oxidation of thioamides R(CdS)-
NH
2
with polystyrene-bound selenoxide in acetic acid produced exclu-
3
sively dialkythiadiazoles (e.g., 7 for R ) Me), while reactions in other
solvents such as EtOH, benzene, and CH CN produced only the
3
corresponding nitriles, RC≡N.
(
18) I.e., for R * H, the reaction of R’O(C)S)NHR + Cl(C)O)SCl
(
20) The reactions of thiocarbamates with sulfenyl chlorides to
(
compare to refs 3a and 4a) gives rise to low levels of carbamoylsulfenyl
chlorides ClS(CdO)NHR, which cyclodimerize with loss of HCl + / S
2
provide formimidoyldisulfanes were described first by Harris: Harris,
J . F., J r. J . Am. Chem. Soc. 1960, 82, 155-158. They are supported
by ample examples and mechanistic studies from K u¨ hle (refs 3 and 4)
and us (ref 5 and herein). See text for experimental evidence supporting
the structure of intermediate 8 and studies of its further transforma-
tions.
1
to provide the heterocycle drawn below; see ref 5b for detailed
mechanistic studies including crossover experiments.
(
21) Symmetrical disulfanes of this type are well known and are
generally prepared by slow addition of 0.5 equiv of an oxidant, e.g., I
2
,
to thiocarbamates or thioamides. For examples, see: (a) McGowan,
G. J . Chem. Soc. 1886, 190-196. (b) Barrett, A. G. M.; Barton, D. H.
R.; Colle, R. J . Chem. Soc., Perkin Trans 1 1980, 665-671.
(22) An analogue of 11 with ethoxy groups replaced by phenyl groups
has been isolated and subsequently converted to a thiadiazole by
oxidation, see ref 9b.
(
19) As an example where 5 acts merely as an oxidant, in ref 3a,
Zumach and K u¨ hle report that Cl(CdO)SCl (5) reacts with thiophenol
PhSH) to give diphenyl disulfide (PhSSPh) plus COS and HCl. A more
complicated further example is given in ref 5c.
(

6
642 J . Org. Chem., Vol. 61, No. 19, 1996
Chen et al.
Ta ble 2. Electr on Im p a ct Ma ss Sp ectr a of 1,2,4-Dith ia zolid in e-3,5-d ion e (1) P r ep a r ed u n d er Va r iou s Con d ition sa
HNCOS2+
b
M•+ of 1b
(M + 2) / [(M) + (M + 2)]c
S2•+ b
64
entry
reaction conditions
107
109
135
137
HNCOS2+
M•+
1
2
3
4
1,2,4-dithiazolidine-3,5-dione (1)
100
100
100
100
14
14
12
12
1
3
4
5
52
41
27
29
5
9
10
15
0.07
0.17
0.25
0.29
0.09
0.18
0.27
0.34
1
8
1 + HCl/ OH2, 1 h, 100 °C
1
8
2 + HCl/ OH2 f 1, 15 min, 100 °C
1
8
2 + HCl/ OH2 f 1, 1 h, 100 °C
a
See text for procedure, specifically with regard to entry 3; 83% yield for 1. For the control experiment on entry 2, on same scale but
with a 4-fold longer reaction time, the recovery of 1 was 64%, indicating that some hydrolysis had taken place (the recovered 1 had the
same mp as the pure starting 1). The experiment on entry 4, with the longer reaction time, was worked up less efficiently, i.e., crystals
were collected directly rather than evaporating to dryness. Therefore, the isolated yield of 30% cannot be compared to the other, more
3
4
b
quantitative, experiments. Entry 1 is a standard of 1, and the (M + 2) peaks reported are due to the natural abundance of S. Intensities
of the mass of this ion are indicated as percent of the most intense ion. c Ratios of the intensities of indicated species.
Sch em e 3
CDCl
Et N (2 equiv).
As already described, the reaction of 4 and 5 in Et
in the absence of tertiary amine eventually gives hetero-
cycle 2, which is soluble in Et O. However, a white
3
of 4 (2 equiv) and 5 (1 equiv) in the presence of
3
2
O
F igu r e 1. ORTEP representations of 1 and 2 (40% ellipsoids,
H atoms in idealized positions). Selected bond distances, bond
angles, and torsion angles are listed in Table 3.
2
precipitate was observed to form shortly after both
reactants had been combined, and this highly unstable
solid was isolated by rapid filtration under suction. The
conditions (Table 2, entry 2), we conclude that route a is
favored over route b by a factor of approximately 4:1.
1
observed mass (CIMS) and H NMR (CDCl
3
) spectra were
consistent with the structure of the intermediate formim-
Route a is directly analogous to a process which otherwise
idoyl(chlorocarbonyl)disulfane 8 (or its hydrochloride salt,
a,4,5a
(
R * H) occurs essentially spontaneously.3
In the
8
‚HCl).23 After 30 min in CDCl
3
solution, 8 had com-
special case of interest here, reaction at 25 °C in CDCl
3
pletely disappeared and showed EtCl (85% of ethyl
groups) as the main product (presumed coproducts COS
and elemental S; only small amounts of 2 and 11 formed),
solution of 2 (0.6 M) saturated with anhydrous HCl was
shown to give EtCl and 1, albeit very slowly (5% conver-
sion after 12 h), and hence not of any preparative value.
in accord with results in CHCl
base (Scheme 1, Table 1, entry 12). Rapid addition of
freshly collected solid 8 to a CDCl solution containing
Et
3
in the absence of external
X-r a y Cr ysta llogr a p h ic Str u ctu r es of 1,2,4-Dith ia -
zolid in e-3,5-d ion e (1) a n d 3-Eth oxy-1,2,4-d ith ia zo-
lin -5-on e (2), a n d Mech a n istic Im p lica tion s. Struc-
3
3
N (∼1 equiv) showed the major product to be again
2
4
tures of 1 and 2 were solved (Figure 1, Table 3), showing
EtCl (50% of ethyl groups), but 2 (15% of ethyl groups)
and 11 (27% of ethyl groups) also formed. Again, these
results with isolated 8 mirror the ultimate solvent-
dependent product distribution from the reactions of 4
and 5 in the absence and in the presence of tertiary
amine.
2
5
the planar, aromatic structure of both heterocycles.
Bond lengths and angles of 1 and 2 were typical for five-
2
5
membered cyclic disulfides with aromatic character.
Whereas the preferred bond angle (C-S-S) and torsion
angle (C-S-S-C) in acyclic disulfides are around 105°
and (90°, respectively, the heterocycles 1 and 2 have
corresponding bond angles of 90-95° and torsion angles
of 0 ( 5°. These values help to account for the facile
desulfurization reactions (Scheme 4 and ref 2) of 1 and
Con ver sion of 3-Eth oxy-1,2,4-d ith ia zolin -5-on e (2)
to 1,2,4-Dith ia zolid in e-3,5-Dion e (1). The desired title
heterocycle 1 was obtained by heating 2 with concen-
trated aqueous hydrochloric acid at reflux. Two plausible
pathways for this transformation involve protonation of
heterocycle 2, followed by attack of chloride on the ethyl
group, to give 1 and EtCl (Scheme 3, route a) or,
alternatively, nucleophilic addition of water to the imino
carbon with ultimate loss of ethanol (Scheme 3, route b).
2
. Thus, addition of a soft nucleophile, e.g., a phospho-
rus(III) species, to one of the sulfur atoms in the rings is
an enthalpically favorable processes (loss of strain en-
ergy), accentuated by favorable entropic components
(three molecules formed from two molecules). However,
1
8
under acidic or neutral conditions and over long-term
ambient storage, both 1 and 2 are kinetically stable for
weeks or even years. Finally, the fact that the basic
imino nitrogen of 2 is coplanar with the ethoxy group is
Conversion of 2 to 1 was carried out in HCl-H
2
O (Table
2
1
, entries 3 and 4) and found to give product of both m/ z
35 and 137 (EIMS), indicating significant incorporation
of O in 1. Taking into account the S contribution to
m/ z 137 in dithiazolidine 1 (Table 2, entry 1), as well as
background oxygen exchange of 1 under the reaction
1
8
34
(24) The authors have deposited atomic coordinates for the struc-
tures of compounds 1 and 2 with the Cambridge Crystallographic Data
Centre. The coordinates can be obtained, on request, from the Director,
Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge,
CB2 1EZ, UK.
(25) Barany, G. Cryst. Struct. Commun. 1982, 11, 913-928 and
references cited therein. This work provides a relevant discussion of
planarities, bond lengths, bond angles, and dihedral angles in a number
of five-membered heterocyclic disulfides.
(
23) The fact that the compound precipitates from Et
strongly that it is 8‚HCl (ammonium chlorides have very low solubility
in Et
2
O suggests
2
O). The δ ) 4.92 ppm for the methylene protons of the Et group
indicates a strongly electropositive center and is also consistent with
an ammonium form [in contrast, bis(formimidoyl)disulfane 10 has δ
)
4.29 ppm for the methylene protons of the Et group].

Studies Related to 1,2,4-Dithiazolidine-3,5-dione
J . Org. Chem., Vol. 61, No. 19, 1996 6643
Ta ble 3. X-r a y Cr ysta llogr a p h ic P a r a m eter s a n d Selected Bon d Len gth s, Bon d An gles, a n d Tor sion An gles for
,2,4-Dith ia zolid in e-3,5-d ion e (1) a n d 3-Eth oxy-1,2,4-d ith ia zolin -5-on e (2)
1
compound
1
2a
Mw
space group
135.16
I41/a
19.556(6)
19.556(6)
4.996(2)
90
90
90
163.21
Pc (No. 7)
3.939(2)
35.84(1)
18.789(4)
90
90.81(3)
90
2652(3)
16
a, Å
b, Å
c, Å
R
â
γ
3
V, Å
Z
1910.6(11)
16
temp (K)
Fcalcd, g/cm
radiation
crystal shape
size (mm)
total reflections
unique reflections
obsd reflections
R1 (on F)
297
1.880
Mo KR ) 0.710 73 Å
prism
172
1.635
Mo KR ) 0.710 73 Å
needle
0.60 × 0.20 × 0.15
6067
5968
5753
0.087
3
0.60 × 0.45 × 0.35
3297
1088
1088
0.032
0.080
2
ωR2 (on F )
0.181
Selected Bond Lengths (Å)
N1-C1/N1-C11
C1-O1/C11-O11
C1-S1/C11-S11
S1-S2/S11-S12
S2-C2/S12-C12
C2-N1/C12-N1
C2-O2/C12-O12
1.367(2)
1.208(2)
1.764(2)
2.0584(8)
1.761(2)
1.369(2)
1.208(2)
1.36(2)
1.23(2)
1.81(2)
2.053(6)
1.76(2)
1.26(2)
1.31(2)
Selected Bond Angles (deg)
113.5(1)
N1-C1-S1/N1-C11-S11
C1-S1-S2/C11-S11-S12
S1-S2-C2/S11-S12-C12
S2-C2-N1/S12-C12-N1
116(1)
93.3(5)
91.9(5)
123(1)
95.42(5)
95.45(6)
113.6(1)
Selected Torsion Angles (deg)
S1-S2-C2-O2/S11-S12-C12-O12
S1-S2-C2-N1/S11-S12-C12-N1
S1-C1-N1-C2/S11-C11-N1-C12
S2-S1-C1-O1/S12-S11-C11-O11
S2-S1-Cl-N1/S12-S11-C1l-N1
S2-C2-N1-C1/S12-C12-N1-C11
N1-C12-O12-C13
-179.6(1)
175(1)
-2(2)
6(2)
180(2)
-4(2)
-5(3)
-7(3)
1.5(1)
-1.7(2)
-177.8(2)
2.3(1)
-0.1(2)
b
a
The asymmetric unit contains eight chemically identical molecules that are not crystallographically equivalent. Pseudotranslations
between the eight molecules render much of the data weak, and the noncentrosymmetric nature of the structure also limits the data:
b
parameter ratio. Not applicable.
Sch em e 4
Exp er im en ta l Section
Gen er a l. 1H, 13C, and 31P NMR spectra were observed in
the indicated deuterated solvents with IBM NR 200 AF or NR
3
00 AF or Varian VXR 300 instruments. Resonances are
expressed as ppm downfield from TMS; coupling constants of
ethyl groups were approximately 7.1 Hz and are not specified.
Exchangable protons are not reported. IR spectra were
obtained on a Perkin Elmer 1600 Series FTIR spectrophotom-
eter. Electron ionization mass spectra were obtained on a
Finnigan MAT 95 instrument at the indicated eV and a source
temperature of 200 °C. Positive and negative methane chemi-
(
26) (a) The reference spectral positions for isocyanurate 14 were
consistent with the major proposed pathway for conver-
sion of 2 to 1 (Scheme 3, route a).
established on a reasonably pure crystalline sample, mp 75-88 °C,
which was obtained from the residue of a redistillation of the contents
of an old bottle of ethyl isocyanate (EtNdCdO). (b) The corresponding
cyanurate isomer, 2,4,6-triethoxy-s-triazine, was made by reaction of
cyanuric chloride with NaOH (3 equiv in EtOH), as outlined in the
following: Dudley, J . R.; Thurston, J . T.; Schaefer, F. C.; Holm-Hansen,
D.; Hull, C. J .; Adams, P. J . Am. Chem. Soc. 1951, 73, 2986-2990.
The cyanurate was readily distinguishable [ H NMR (CDCl
3
q, 2 H), 1.28 (t, 3 H); C NMR (CDCl ) δ 172.9, 64.1, 14.1]. (c)
Oth er Tr a n sfor m a tion s of 3-Eth oxy-1,2,4-d ith ia -
zolin -5-on e (2) (Sch em e 4). When heated past the
melting point, 2 undergoes (most likely bimolecular)
rearrangement to give N-ethyl-1,2,4-dithiazolidine-3,5-
dione (13).5b A minor byproduct of this reaction is 1,3,5-
1
3
) δ 4.31
1
3
(
Literature mp’s: 95 °C for 14; 30 °C for the cyanurate isomer; see:
Dictionary of Organic Compounds, 5th ed.; Chapman & Hall: New
York, 1982; C-10334 and T-02371, pp 131 and 5388.
triethyl-2,4,6-trioxohexahydro-s-triazine (triethyl iso-
cyanurate) (14),2
6a,c
the trimer of ethyl isocyanate
(
27) For relatively complicated routes to EtOC≡N, see: (a) J ensen,
(EtNdCdO). The rapid desulfurization of 2 by a variety
K. A.; Holm, A. Acta Chem. Scand. 1964, 18, 826-828. (b) Martin, D.;
Mucke, W. Chem. Ber. 1965, 98, 2059-2062. For alkyl cyanate
chemistry, see: (c) J ensen, K. A.; Due, M.; Holm, A.; Wentrup, C. Acta
Chem. Scand. 1966, 20, 2091-2106.
of phosphines in chloroform or acetonitrile provides a
novel entry to O-ethyl cyanate (15).2
,8,27

6
644 J . Org. Chem., Vol. 61, No. 19, 1996
Chen et al.
cal ionization mass spectra were recorded under the specified
conditions on the Finnigan MAT 95 instrument. Low-resolu-
tion fast atom bombardment mass spectroscopy (FABMS) was
carried out on a VG Analytical 707E-HF low-resolution double-
focusing mass spectrometer equipped with a VG 11/250 data
system, operated at a resolution of 2000. Elemental analyses
were performed by M-H-W Laboratories (Phoenix, AZ).
X-r ay Data Collection , Solu tion , an d Refin em en t (Table
residue was redissolved in CH
2
Cl
2
(100 mL), filtered to remove
an insoluble white solid (0.4 g) believed to be cyanuric acid
(2,4,6-trihydroxy-1,3,5-triazine),12 and reconcentrated to pro-
1
vide a light-yellow solid (7.22 g, 89% purity by H NMR with
the remainder being 3). Recrystallization from Et O (15 mL)
2
at -20 °C gave off-white needles (3.6 g). The mother liquor
provided an additional portion of light yellow needles (0.6 g):
1
total yield, 4.2 g (52%); H NMR (CDCl
3
) δ 4.68 (q, 2 H), 1.47
(t, 3 H); 13C NMR (CDCl
2990 (w), 1705 (s), 1541 (vs), 1470 (m), 1393 (w), 1365 (w),
3
). Crystals were mounted in an Enraf-Nonius CAD4 diffrac-
tometer. Intensity data for independent reflections (0° < 2θ
60°) were collected at the indicated temperature (Table 3)
3
3
) δ 187.3, 179.6, 73.5, 14.1; IR (CDCl )
-
1
<
1297 (w), 1259 (s), 1239 (m), 1153 (w), 1004 (w) cm ; positive
methane CIMS (source 160 °C, solid probe 20 °C, 0.1 mm) m/ z
by the ω scan technique. Data reduction was performed on a
PDP 11/34 computer using an Enraf-Nonius SDP program
library. The structures were solved by direct methods (MUL-
TAN) and refined by full-matrix least-squares with SHELXTL
+
+
164 [(M + H) , 100], 136 [(M + H) - CO, 67]; negative
-
methane CIMS m/ z 162 [(M - H) , 59], 133 (14), 102 [(M -
-
H) - COS, 100]; EIMS (source 200 °C, solid probe 30 °C) m/ z
•
+
•+
•+
•+
(version 5) program on a Pentium PC using all measured data.
163 (M , 24), 135 (M - CO, 13), 131 (M - S, 10), 107 (M
•
+
The final R values based on data having I > 2σ|I| are given in
- CO - C
(21), 29 (C
2
H
H
4
, 18), 103 (M - COS, 8), 70 (79), 64 (29), 60
+
Table 3.
2
5
, 100).
4 5 2 2
Anal. Calcd for C H NO S (MW 163.22): C, 29.44; H, 3.09;
1
,2,4-Dith ia zolid in e-3,5-d ion e (1). A stirred suspension
2
8
N, 8.58; S, 39.29. Found: C, 29.23; H, 3.14; N, 8.68; S, 39.41.
Samples of crystalline 2 were maintained under ambient
conditions, exposed to open atmosphere, for up to 2 years
without any signs of decomposition as judged by H NMR,
NMR, and mp redetermination; routine storage is carried out
at -20 °C.
Th er m a l Decom p osit ion a n d R ea r r a n gem en t of
3-Eth oxy-1,2,4-d ith ia zolin -5-on e (2). Title substrate 2 (89
mg, 0.53 mmol) was melted and heated at 100 °C in a closed
7-mL screw-cap tube with a Teflon-lined cap. After 5 days,
of recrystallized 3-ethoxy-1,2,4-dithiazolin-5-one (2) (13.9 g,
5 mmol) in concentrated aqueous HCl (80 mL) was brought
8
over 45 min to 110 °C and 10 min later passed while hot
through a glass fritted funnel to remove elemental sulfur. The
filtrate was concentrated to dryness to provide a solid, which
was crystallized from toluene: yield, 8.5 g (74%), mp 141-
1
13
C
3
b
3c
13
1
43 °C (lit. mp 141 °C; lit. mp 142-144 °C); C NMR
1
3
(CDCl
3
) δ 167.7; C NMR (CD
w), 1735 (s), 1700 (vs), 1683 (s, sh), 1287 (m) cm ; methane
3
3
CN) δ 168.7; IR (CDCl ) 3360
-
1
(
CIMS (source 160 °C, solid probe 60 °C, 0.1 mm) m/ z 136 [(M
+
+
+
+
H) , 100), 108 [(M + H) - CO, 29], 93 (14), 64 (S
2
, 19).
3
the tube was cooled, vented, taken up in CDCl , and filtered
to provide an insoluble residue (6 mg, 0.2 mmol assuming
elemental sulfur) and a soluble portion (66 mg) that, according
to 1H and C NMR, was devoid of starting material and
Anal. Calcd for C HNO (MW 135.16): C, 17.77; H, 0.74;
2
2 2
S
N, 10.36; S, 47.44. Found: C, 17.91; H, 0.84; N, 10.44; S, 47.31.
Samples of crystalline 1 were maintained under ambient
conditions for 2 years or more without any signs of decomposi-
13
5
b
comprised N-ethyl-1,2,4-dithiazolidine-3,5-dione (13) (0.28
1
tion as judged by 13C NMR, repeat elemental analysis, and
mmol, 53%) [ H NMR (CDCl
3
) δ 3.81 (q, 2 H), 1.23 (t, 3 H);
1
3
C NMR (CDCl
trioxohexahydro-s-triazine (triethyl isocyanurate, 14)
3
) δ 167.3, 41.5, 12.7], 1,3,5-triethyl-2,4,6-
mp redetermination; routine storage is carried out at -20 °C.
Con ver sion of 3-Eth oxy-1,2,4-d ith ia zolin -5-on e (2) to
2
6a,c
(0.03
) δ 3.88 (q, 2 H),
) δ 148.6, 38.1, 13.1], and an
1
mmol, 16% of ethyl groups) [ H NMR (CDCl
1.18 (t, 3 H); 13C NMR (CDCl
unknown N-ethyl derivative (0.09 mmol, 17%) [partial H NMR
3
1
,2,4-Dith ia zolid in e-3,5-d ion e (1) by Hyd r ogen Ch lor id e
1
8
18
3
in [ O]Wa ter (Ta ble 2). H
open 7-mL screw-cap tube and saturated for 1 h with dry HCl
2
O (∼0.5 mL) was placed in an
1
1
3
2
9
(CDCl
NMR (CDCl
In a separate experiment, substrate 2 (89 mg, 0.55 mmol)
3
) δ 4.09 (q) (corresponding triplet not resolved);
C
gas (produced by slowly dropping 50 mL of concentrated
sulfuric acid onto 100 g of NaCl). Compound 2 (100 mg) was
added, and after bubbling had subsided (15 min), the tube was
sealed and brought to 100 °C for 15 min reaction. Next, the
still-hot reaction mixture was filtered through glass wool, and
evaporation provided white crystals (76 mg, 83%): mp 136-
3
) δ 160.4, 40.5, 30.0].
1
3
in CDCl (0.5 mL) was refluxed for 7 days and shown by H
and 13C NMR to be a 19:1 mixture of unchanged 2 (major) plus
rearrangement product 13 (minor).
3
,5-Dieth oxy-1,2,4-th ia d ia zole (3). A solution of (chlo-
rocarbonyl)sulfenyl chloride (5) (6.25 mL, 75 mmol) in CHCl
375 mL) was added over 30 min into a stirred and externally
chilled mixture of O-ethyl thiocarbamate (4) (7.88 g, 75 mol)
and Et N (10.7 mL, 75 mmol) in CHCl (375 mL), at a rate to
maintain the reaction temperature at 5-10 °C. After a further
5 min, the reaction mixture was washed with 1 N aqueous
HCl (2 × 1 L) and water (2 × 1 L), dried (MgSO ), concen-
trated, taken up in CHCl (50 mL), filtered to remove a
substantial mass of elemental sulfur, and concentrated again.
The crude product (5.4 g) comprised title product (52 mmol,
9% of ethyl groups in starting 4) and 2 (5.5 mmol, 7%);
extrapolating from results of a small-scale reaction in CDCl
EtCl (17%) and O-ethyl carbamate (3%) also were formed but
1
39 °C, mass spectral data in Table 2, entry 3.
-E t h oxy-1,2,4-d it h ia zolin -5-on e (2). Met h od A.
mixture of O-ethyl thiocarbamate (4) (21.0 g, 0.2 mol) and Et
28 mL, 0.2 mol) in Et
3
3
A
N
(
3
(
2
O (100 mL) was added dropwise over 1
3
3
h into a stirred and externally chilled solution of (chlorocar-
bonyl)sulfenyl chloride (5) (16.8 mL, 0.2 mol) in ethyl ether (1
L), at a rate to maintain the reaction temperature below 10
1
4
°
C. After an additional 4.5 h of stirring, the precipitated
3
Et
3
N‚HCl (30.2 g, quantitative) was removed by filtration. The
filtrate was concentrated to provide a light-yellow solid (26.8
g, 91% purity by 1H NMR with the remainder being 3).
6
Recrystallization from ether (100 mL) at -20 °C gave off-white
3
c
3
,
needles (16.4 g), mp 51-53 °C (lit. mp 56-57 °C). The
mother liquor provided an additional portion of light yellow
needles (4.1 g): mp 42-46 °C; total yield, 20.5 g (63%).
Meth od B. A solution of thiocarbamate 4 (5.25 g, 50 mmol)
were removed in the workup [the NMR study also showed that
decomposed at 25 °C on a time scale of days to EtCl, H NMR
1
3
1
3
13
δ 3.56 (q), 1.48 (t), C NMR δ 40.2, 18.8, and COS, C NMR
δ 153.1; see Scheme 1]. A portion of the crude material (3.5
g) was subjected to careful short-path distillation, bp 68-71
in Et
2
O (25 mL) was added dropwise over 40 min into a stirred
and externally chilled solution of 5 (4.2 mL, 50 mmol) in Et
2
O
(
1
250 mL), at a rate to maintain the reaction temperature below
°
C (0.2 mm) or bp 50-53 °C (0.1 mm), leaving 2 and some
0 °C. A white precipitate formed20 but disappeared as the
sulfur in the residue and providing analytically pure title
product (2.3 g, extrapolated yield 54% based on ethyl groups
in starting 4), which solidified in the receiver, mp 41-43 °C
reaction progressed with an additional 1.5 h of stirring at 25
C. Solvent was evaporated at reduced pressure, and the
°
(
lit.9 mp 48-49 °C); H NMR (CDCl
a,b
1
3
) δ 4.52 (q, 2 H), 4.38
(q, 2 H), 1.45 (t, 3 H), 1.41 (t, 3 H); C NMR (CDCl ) δ 189.7,
) 2985 (w), 1542 (vs),
512 (s), 1476 (w), 1377 (m), 1328 (vs), 1252 (m), 1228 (w),
1
3
(28) On a 20 mmol scale, crude 2 from the reaction of 4 and 5 was
3
treated as described in the text with concentrated aqueous HCl.
Surprisingly, relatively little 1 was isolated (∼15% based on 4), and
the main product (∼40% of the crude product) was cyanuric acid
1
1
1
65.5, 69.8, 64.8, 14.2, 14.1; IR (CDCl
3
-
1
078 (w), 1013 (w) cm ; positive methane CIMS (source 160
C, solid probe 50 °C, 0.1 mm) m/ z 175 [(M + H) , 100], 147
2 4
[(M + H) - C H , 90], 72 (27); EIMS (source 200 °C, solid
(
compare to ref 12).
+
°
(29) The Merck Index, 11th ed.; Budavari, S., Ed.; Merck & Co.:
+
Rahway, NJ 1989; No. 4721, p 759.

Studies Related to 1,2,4-Dithiazolidine-3,5-dione
J . Org. Chem., Vol. 61, No. 19, 1996 6645
probe 50 °C, 30 eV) m/ z 174 (M•+, 28%), 146 (M•+ - C
H
,
portions. One portion was added to a solution of Et
(4 mL). After 30 min of stirring, the
2
4
3
N (0.14
•+
3
4
0%), 118 (M - 2C
4 (19), 43 (20), 29 (C , 71).
10 2 2
Anal. Calcd for C H N O S (MW 174.22): C, 41.36; H, 5.79;
2
H
4
, 100), 75 (47), 72 (25), 70 (29), 57 (14),
mL, ∼1 equiv) in CDCl
3
+
2
H
5
distribution of major compounds was 2 (∼16%), 11 (27%),
6
O-ethyl carbamate (5%), and EtCl (50%). The other portion
N, 16.08; S, 18.40. Found: C, 41.11; H, 5.86; N, 15.97; S, 18.23.
3
was treated with twice the amount of Et N, and the distribu-
O-Eth yl Th ioca r ba m a te (4). Following the method of
Davies and MacLauren, solutions of potassium ethyl xan-
thate (192 g, 1.2 mol) in water (500 mL) and chloroacetic acid
113 g, 1.2 mol) plus sodium hydroxide (48 g, 1.2 mol) in water
450 mL) were combined and stirred at 25 °C throughout a
full working day, following which concentrated ammonium
tion of major compounds was 2 (∼16%), 11 (17%), 4 (6%),
1
0c
O-ethyl carbamate (5%), and EtCl (50%).
Bis(eth oxyfor m id oyl) Disu lfa n e (10) a n d Its Ch em is-
(
(
21b
tr y. Modeled after the procedure of Barrett et al., a solution
of iodine (635 mg, 2.5 mmol) in Et
dropwise over 5 min to an ice-bath-chilled solution of O-ethyl
thiocarbamate (4) (535 mg, 5 mmol) and Et N (0.7 mL, 5 mmol)
in Et
2
O (15 mL) was added
hydroxide (100 mL, ∼1.5 mol) was added, and stirring con-
3
2
tinued overnight. The product was then extracted into Et O
2
O (5 mL). Slightly more iodine (∼20 mg) was added to
(
2 × 500 mL), and the organic phase was washed with equal
get a brown endpoint. After 10 min, the reaction mixture was
filtered to remove salts and concentrated at reduced pressure
volumes of H O and 1 N aqueous HCl, dried (MgSO ), and
2
4
concentrated. The resultant oil was placed under petroleum
ether at -20 °C, providing a white solid. The purification
process removed the S-ethyl isomer, which was present as a
1
to provide semisolid 10 (>90% pure by H NMR): yield, 0.47
1
13
g (90%); H NMR (CDCl
NMR (CDCl
3
) δ 4.29 (q, 2 H), 1.26 (t, 3 H);
C
3
) δ 161.8, 66.8, 14.2; FABMS [3-nitrobenzyl
7
3
(
% contaminant in the crude product: yield, 78.5 g (62%); mp
+
alcohol (MNBA) matrix] m/ z 209 (M + H ), and some peaks
with one or two additional sulfurs.
1
0c
10d
1
6-38 °C (lit. mp 40-41 °C; lit. mp 38-39 °C); H NMR
CDCl
S-ethyl isomer]; C NMR (CDCl
-Meth yl-1,2,4-d ith ia zolin -5-on e (6). A solution of thio-
acetamide (0.73 g, 10 mmol) and Et N (2.8 mL, 20 mmol) in
,2-dimethoxyethane (5 mL) was added dropwise over 20 min
3
) δ 4.48 (q, 2 H), 1.34 (t, 3 H) [no δ 2.90, diagnostic of
Neat compound 10 decomposed upon overnight standing at
1
3
3
) δ 192.0, 67.3, 13.8.
4
°C to give a complicated mixture, of which the main
3
identifiable components were thiocarbamate 4 and (to a lesser
extent) isocyanurate 14, as well as a significant yellow
insoluble fraction, which was presumably elemental sulfur. In
3
1
to a stirred and externally chilled solution of (chlorocarbonyl)-
sulfenyl chloride (5) (0.84 mL, 10 mmol) in 1,2-dimethoxy-
ethane (50 mL), at a rate to maintain the reaction temperature
below 10 °C. After an additional 1 h of stirring, the precipi-
3
an NMR tube experiment, freshly prepared 10 in CDCl was
treated with a drop of trifluoroacetic acid. An immediate
exotherm was observed, along with intense coloration and
precipitation of sulfur. Spectral examination indicated forma-
tion in a 2:1 molar ratio of 4 and a new species assigned to
tated Et
concentrated at reduced pressure to provide a dark-brown solid
0.9 g), which was dissolved in ethyl acetate (6 mL). Two-
thirds of this solution was applied to flash chromatography
hexane-ethyl acetate ) 4:1) to produce an off-white solid:
yield, 0.29 g (29%, extrapolated to purification of all of the
3
N‚HCl was removed by filtration. The filtrate was
ethoxythiocarbonyl O-ethyl imidate (11): 1H NMR (CDCl
3
) δ
(
13
4
.46 (q, 2 H), 4.32 (q, 2 H), 1.42 (t, 3 H), 1.35 (t, 3 H); C NMR
(
3
CDCl ) δ 190.1, 165.6, 70.5, 65.6, 14.2, 14.0; FABMS [3-ni-
(
+
trobenzyl alcohol (MNBA) matrix] m/ z 177 (M + H ).
O-Eth yl Cya n a te (15) by Desu lfu r iza tion of 3-Eth oxy-
1
crude material); mp 64-67 °C; H NMR (CDCl ) δ 2.73 (s, 3
3
1
3
,2,4-d ith ia zolin -5-on e (2). A solution of 2 in CDCl (0.2 M)
1
3
H); C NMR δ 192.7, 188.0, 23.1.
Anal. Calcd for C NOS (MW 103.18): C, 27.06; H, 2.27;
N, 10.52; S, 48.14. Found: C, 26.88; H, 2.20; N, 10.66; S, 48.12.
was treated with ∼1 equiv of triphenyl phosphite, and spectra
3
H
3
2
1
were recorded shortly thereafter: H NMR (CDCl
3
) δ 4.49 (q,
31
2
H), 1.40 (t, 3 H); 13C NMR (CDCl
NMR (CDCl ) δ 54.9 (triphenyl thiophosphate). Similar results
3
) δ 188.1, 70.8, 13.5;
P
2
0
F or m im id oyl(ch lor oca r b on yl)d isu lfa n e (8) a n d It s
3
Ch em istr y. A solution of O-ethyl thiocarbamate (4) (0.21 g,
were obtained with other trivalent phosphorus compounds,
e.g., triphenylphosphine and trimethyl phosphite, and with the
solvent CD CN.
3
2
mmol) in Et
a stirred and externally chilled solution of (chlorocarbonyl)-
sulfenyl chloride (5) (168 mL, 2 mmol) in Et O (5 mL), at a
rate to maintain the reaction temperature below 10 °C.
white precipitate formed instantaneously, and after another
min stirring while ice-bath-chilled, the 8 was collected by
2
O (5 mL) was added dropwise over 5 min into
2
A
Ack n ow led gm en t. We are indebted to Professor
Doyle Britton (University of Minnesota) for his expertise
in obtaining and solving the crystallographic data and
to Yen-Hsiang Liu (LSU) and Dr. Frank Fronczek (LSU)
for assistance with the preparation of Figure 1 and
Table 3 and submission of X-ray data to the Cambridge
Crystallographic Data Centre. We further thank Steven
J . Eastep, David A. Halsrud, Lydia Ong, and Robert L.
Walsky for expert technical assistance. Special grati-
tude is expressed to Qinghong Xu for her initiative with
regard to studies reported in ref 2, which reminded us
of the importance of title compound 1 and established
a new application for compound 2. Financial support
was from NIH (GM 28934 and 43552) and the Univer-
sity of Minnesota Graduate School.
5
rapid filtration and drying under aspirator suction (essentially
quantitative formation of solid; the filtrate was concentrated
and shown to comprise <5% of the total mass). Spectral data
1
were recorded immediately: H NMR (<5 min in CDCl
3
) δ 4.92
1
3
(
q, 2 H), 1.54 (t, 3 H); C NMR (CDCl
3
) unable to detect 8
because of decomposition during acquisition; positive isobutane
CIMS (source 160 °C, solid probe 25 °C, 0.1 mm) m/ z 200 (M
+
+
H ).
Solid compound 8 decomposed within 10 min upon standing
at 25 °C to give heterocycle 2, hydrolysis product O-ethyl
carbamate, gaseous byproducts (COS, EtCl), and a significant
insoluble fraction, which was presumably elemental sulfur. A
1
solution of 8 in CDCl
3
was re-examined by H NMR after 30
min at 25 °C, showing 2 (∼5%), 11 (10%), O-ethyl carbamate
3%), and EtCl (81%). In a separate experiment, freshly
prepared 8 from the same scale reaction was divided into equal
(
J O960723U