The ring opening of 3,4-dichloro-1,2,5-thiadiazole with metal amides. A new synthesis of 3,4-disubstituted-1,2,5-thiadiazoles

Article abstract of DOI:10.1016/j.tetlet.2006.09.102

We have developed a new synthesis of 3,4-disubstituted-1,2,5-thiadiazoles. The methodology is based on the ring opening of readily available 3,4-dichloro-1,2,5-thiadiazole with metal amides to afford a stable synthon, which is then transformed into the 3,

Full text of DOI:10.1016/j.tetlet.2006.09.102

Tetrahedron Letters 47 (2006) 8285–8288
The ring opening of 3,4-dichloro-1,2,5-thiadiazole with metal
amides. A new synthesis of 3,4-disubstituted-1,2,5-thiadiazoles
Alain Merschaert,* Pascal Boquel, Hugo Gorissen, Jean-Pierre Van Hoeck, Alfio Borghese,
Luc Antoine, Vincent Mancuso, Anne Mockel and Michel Vanmarsenille
Lilly Research Laboratories, Chemical Product Research and Development, Rue Granbonpre´ 11, 1348 Mont-Saint-Guibert, Belgium
Received 27 July 2006; revised 18 September 2006; accepted 19 September 2006
Available online 9 October 2006
Abstract—We have developed a new synthesis of 3,4-disubstituted-1,2,5-thiadiazoles. The methodology is based on the ring opening
of readily available 3,4-dichloro-1,2,5-thiadiazole with metal amides to afford a stable synthon, which is then transformed into the
3,4-disubstituted-1,2,5-thiadiazole derivatives via two consecutive reactions with O-, S-, N- or C-nucleophiles.
Ó 2006 Elsevier Ltd. All rights reserved.
Substituted 1,2,5-thiadiazole derivatives have found use-
ful applications in many different technological areas.1
In the pharmaceutical industry, the exceptional pharma-
cological properties of the 1,2,5-thiadiazole nucleus has
been exploited in the development of drugs or drug
candidates.2 Therefore, the development of scalable
processes for the preparation of this heterocycle has
been the subject of an extensive research for many
years.3 Despite this substantial effort, most of the meth-
ods used for the synthesis of 1,2,5-thiadiazole derivatives
remain difficult to scale-up because they involve the use
of highly toxic (e.g., cyanogen), corrosive (e.g., S2Cl2) or
hazardous (S4N4) reagents and the production of large
amounts of by-products (e.g., sulfur) and wastes. On
the other hand, 3,4-dichloro-1,2,5-thiadiazole 1, a start-
ing material in the synthesis of the Merck’s b-adrenergic
blocking agent Timolol, is readily available at an indus-
trial scale but there are few examples of successful trans-
formations of 1 into its unsymmetrical derivatives.4 In
particular, the reaction of 1 with Grignard,5 alkyl-
lithium6 or metal amides7 reagents leads mostly to
ring-opened decomposition by-products through an
attack on the sulfur atom. In this sense, we reported
recently the synthesis of a series of alkyl-, alkenyl-,
alkynyl-, aryl- and heteroaryl-1,2,5-thiadiazoles via the
Pd-catalyzed cross-coupling reaction of 1 with organo-
stannanes or -boranes.8
We wish to report herein a new method for the synthesis
of 3,4-disubstituted-1,2,5-thiadiazoles via an unprece-
dented ring opening/substitution/ring closure process
(Fig. 1).9
We found that hindered metal amides at low tempera-
ture allow to control the ring opening of 1 and to avoid
the subsequent decomposition of the ring-opened prod-
ucts. Therefore we sought to use these latter compounds
as useful unsymmetrical synthetic equivalents of 1 for
the synthesis of 3,4-disubstituted-1,2,5-thiadiazoles. We
initially investigated the reaction of 1 with 1 equiv of
LDA in THF at À78 °C. Under these conditions, the
ring-opened product 2a was obtained in low yield
(<20%) after preparative TLC separation. It appeared
that LDA attacks the sulfur atom of 2a (Fig. 2) similarly
to what has been reported in the reaction of 1 with lith-
ium (trimethylsilyl)acetylide.7
To avoid this side reaction, we sought to utilize more
bulky and/or less nucleophilic metal amides for the ring
N
N
1
1
Nu1
Cl
2
2
Cl
Cl
Nu2
Nu1
4
3
4
3
N
N
Nu1
Nu2
R1R2NLi
S
N
S
2
5
2
5
N
N
N
N
S
N
S
1
R2
1
R1
R2
R1
R1, R2 = Alk or (Alk)3Si
Nu1, Nu2 = RO-, RS-, R- or R2N-
1
*
Corresponding author. Tel.: +32 10477021; fax: +32 10476315;
e-mail: merschaert_alain@lilly.com
Figure 1.
0040-4039/$ - see front matter Ó 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2006.09.102

8286
Cl
A. Merschaert et al. / Tetrahedron Letters 47 (2006) 8285–8288
a
use them directly or after purification. It is worth
Cl
b
N
mentioning that no detectable degradation of the crys-
talline compound 2b has been observed after storage
for several years at À20 °C.10
Cl
-Cl-
-Cl-
N
N
N
NC-CN
+
S
N
S
N
S
N
LDA
a
b
1
CN
N
After having prepared this collection of ring-opened
products, we evaluated their reactivity with O-, N-,
S- and C-nucleophiles. Complete decomposition was
observed with Grignard reagents, stabilized carbanions
(potassium diethylmalonate) or potassium cyanide, but
good yields were obtained with O-, N- and S-nucleo-
philes. From our initial screening study, we concluded
that compounds 2b–d were the more useful for
synthetic purposes and we exemplified their reactivity
with a series of classical nucleophiles (Table 2). As
shown in Table 2, compounds 2a–d reacted with a wide
range of O-, N- and S-nucleophiles to afford the sub-
stitution products 3–6. The tetramethylpiperidine deriv-
ative 2d gave a high yield with all types of nucleophiles
(entries 16–19). High yields were also obtained for
sulfur and nitrogen nucleophiles with the bis-trimethyl-
silyl compound 2b (entries 4, 5, 8–10). With alkoxides
(entries 6 and 7, PhCH2OÀ ꢀ Ph2CHOÀ), we observed
a competitive attack on silicon leading to the formation
of cyclized products (Fig. 3). The relative reactivity
order: 2c > 2b > 2d was established for both O- and
S-nucleophiles by the competitive substitution of these
compounds with excess nucleophiles.
N
2a
Figure 2.
opening reaction. The ring opening products 2a–k were
obtained by reaction of 1 with 1 equiv of lithium metal
amide in solvents like methyl-tert-butyl ether, diethyl
ether, THF or n-heptane (Table 1). In general, the ring
opening reactions were carried out at À78 °C, but
higher T ° (À40/À10 °C) were required for the more
bulky metal amides (entries 4–6 and 10). Moderate
yields were obtained with weakly hindered metal
amides (entries 3, 9 and 11) due to competitive S-attack
on 2c,2i and 2k.
Better results were obtained with the bis-trialkylsilyl
amide derivatives (entries 2, 5 and 6) and particularly
with readily available LiHMDS (entry 2). Depending
on the purity of the crude reaction mixture and on the
stability of the ring-opened products, we could either
Table 1. Reaction of 3,4-dichloro-1,2,5-thiadiazole with metal amides
Li
N
N
1
Cl
2
Cl
Cl
R2
R1
N
S
N
N
N
S
R2
R1
2a-k
1
Entry –R1
–R2
Experimental conditionsa
Product
(Yield %)b (ppm)
13C NMR-C1 13C NMR-C2 IR–CN UV–kmax
(ppm)
(cmÀ1
)
(nm)
1
2
3
4
–CH(CH3)2
–Si(CH3)3
–CH(CH3)2
–Si(CH3)3
THF or Et2O/ À78 °C, 30 min
2a (20)c
100.3
99.4
111.1
111.8
111.5
111.7
104.5
100.0
2229
2227
2236
2225
nd
n-Heptane or Et2O/À78 °C, 30 min 2b (90)d
n-Heptane or Et2O/À78 °C, 30 min 2c (47)e
320.8
294.0
300.8
–(CH2)2–O–(CH2)2–
–C(CH3)2–(CH2)3–
C(CH3)2–
t-BuOMe/À40 °C, 20 min
2d (55)e
5
6
7
–SiCH3(Ph)2 –SiCH3(Ph)2 t-BuOMe–THF/À10 °C, 30 min
2e (88)e
2f (75)e
2g (60)f
100.0
99.9
112.4
111.5
111.6
100.0
2227
2227
2222
nd
nd
322.8
–Si(CH3)2Ph –Si(CH3)2Ph t-BuOMe–THF/À10 °C, 30 min
–Si(CH3)2–(CH2)2–
Si(CH3)2–
t-BuOMe–THF/À10 °C, 30 min
8
9
10
–Si(CH3)3
–CH2–Ph
–(CH2)2–S–(CH2)2–
–C(CH3)2–CH2–O–CH2–
C(CH3)2–
t-BuOMe–THF/À10 °C, 30 min
t-BuOMe/À40 °C, 20 min
t-BuOMe/À40 °C, 20 min
2h (50)e
2i (36)e
2j (61)e
101.5
111.3
111.6
111.5
104.1
101.4
nd
2229
2231
nd
294.1
295.5
11
–(CH2)2–NCH3–(CH2)2–
t-BuOMe/À40 °C, 20 min
2k (30)e
117.1
109.4
2215
nd
nd: not determined.
a The metal amide (1 equiv), prepared from the corresponding amine and 1 equiv of n-BuLi (commercial solution of LDA and LiHMDS were used),
was added to 1 in the indicated solvent at the indicated T °C. Reactions were monitored by TLC (n-hexane or n-hexane/EtOAc 98:2).
b Isolated (unoptimized).
c Unstable.
d Bp 69–72 °C/0.3 mbar (mp 29–30 °C).
e Chromatography.
f Bp 72–73 °C/0.3 mbar (mp 38–39 °C).

A. Merschaert et al. / Tetrahedron Letters 47 (2006) 8285–8288
Table 2. Reaction of 2-chloro-2-[(dialkylaminothio)]imino acetonitriles with nucleophiles
8287
N
N
1
1
Nu1
Cl
2
2
Nu1
N
N
S
S
N
N
R2
R2
R1
R1
3-6
2a-d
Entry
Substrate
Nucleophilea Nu1
Experimental
conditionsb
Product
Yieldc (%)
13C NMR-C1
(ppm)
13C NMR-C2
(ppm)
IR-CN
(cmÀ1
)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
2a
2a
2a
2b
2b
2b
2b
2b
2b
2b
2c
2c
2c
2c
2c
2d
2d
2d
2d
n-PrSNa
À20 °C, 1 h
À20 °C, 1 h
À20 °C, 1 h
À20 °C, 1 h
À20 °C, 1 h
À70 °C, 10 min
À70 °C, 10 min
20 °C, 2 h
20 °C, 2 h
20 °C, 2 h
À40 °C, 15 min
À40 °C, 15 min
0 °C, 2 h
3a
3b
3c
4a
4b
4c
4d
4e
4f
20
25
15
91
84
40
70
95
90
100
50
60
35
56
30
100
85
85
100
110.5
110.7
108.2
110.9
111.2
109.6
109.1
107.9
110.5
107.9
110.3
110.7
110.4
108.3
107.2
111.1
111.3
108.6
107.9
122.9
122.0
123.6
121.6
120.6
123.2
121.9
125.4
123.3
124.6
126.2
125.4
126.3
125.5
127.6
122.7
121.6
123.6
125.4
2218
2216
2227
2218
2215
2236
2231
2220
2236
2221
2220
2218
nd
2226
2226
2216
2215
2225
2217
(4-Me)PhCH2SNa
PhCH2OK
n-PrSNa
(4-Me)PhCH2SNa
PhCH2OK
Ph2CHOK
Me2NH
PhCH2NH2
Morpholine
n-PrSNa
(4-Me)PhCH2SNa
PhSNa
PhCH2OK
Morpholine
n-PrSNa
(4-Me)PhCH2SNa
PhCH2OK
Piperidine
4g
5a
5b
5c
5d
5e
6a
6b
6c
6d
À40 °C, 10 min
20 °C, 1 h
À40 °C, 15 min
À40 °C, 15 min
À40 °C, 15 min
20 °C, 2 h
a All reactions were performed in t-BuOMe under nitrogen. Thiolates were prepared from the corresponding thiols and 1 equiv of NaH (n-PrSNa is
commercially available); alkoxides were prepared from the corresponding alcohols and 1 equiv of t-BuOK. The nucleophiles (ca. 1.1 equiv) were
added to a solution (ca. 0.5 M) of the substrate.
b The reactions were monitored by TLC.
c Isolated (chromatography).
with carbanions (entries 2–10) including protected ami-
N
R
no acids (entries 8–10). For both O- and S-nucleophiles,
we carried out the cyclization reactions with a catalytic
amount of base as alcohols or thiols are progressively
deprotonated by the metal amide released during the
reaction. This method allowed us to minimize the side
reactions.
R
NH2
N
S
N
BnO-
Me3Si
N
N
S
SiMe3
R = Cl, OBn
Figure 3.
In conclusion, we have developed an original approach
for the synthesis of 3,4-disubstituted-1,2,5-thiadiazoles
from the cheap and readily available 3,4-dichloro-1,2,
5-thiadiazole. Although unoptimized, our methodology
is already more convenient than most known proce-
dures. The key step is a controlled ring opening reaction
with hindered metal amides leading to stable original
compounds for the atom connectivity of which was
unknown. The reactivity of these intermediates towards
nucleophiles is modulated by the bulkiness of the metal
amide used in the first step. The most versatile interme-
diate 2b is also the most readily obtained. In addition,
our methodology offers perspectives for a high through-
put solid-phase combinatorial chemistry. Indeed, if a
polymer-supported metal amide were used in the first
step, the ring-opened product would be grafted to a
solid support. After substitution, the unsymmetrical
3,4-disubstituted-1,2,5-thiadiazoles would be released
from the solid support by reaction with the second
nucleophile.
We performed the ring-closure of compounds 4–6 by
two different pathways: (a) cyclization induced by
O-, N-, S- and C-nucleophiles through an attack on
nitrile and (b) cyclization induced by attack on silicon
(Table 3). As observed before for substitution, the rate
of the ring-closure reaction is highly sensitive to steric
hindrance around sulfur (cyclization rate: 5a–e > 4a–
g > 6a–d). The cyclization of the 2-substituted-deriva-
tives 4–6 with O-, N-, S- or C-nucleophiles afforded
the 3,4-disubstituted-1,2,5-thiadiazoles in moderate to
high yields. The N,N-bis-trimethylsilyl compounds 4a–
f are particularly attractive as they can be cyclized by
both pathways. Indeed, we have shown that the desilyl-
ation-induced synthesis of 3-amino-4-substituted-1,2,5-
thiadiazoles was particularly effective (entries 18–21).
On the other hand, the N,N-bis-trimethylsilyl derivatives
(e.g., 4a) are also the most appropriate for reactions

8288
A. Merschaert et al. / Tetrahedron Letters 47 (2006) 8285–8288
Table 3. Cyclization of 2-substituted-2-[(dialkylaminothio)]imino acetonitriles with nucleophiles
a)
b)
N
1
H2N
Nu1
Nu1
Nu2
Nu1
2
Nu2
HO- or RO-
N
N
N
N
N
S
S
S
R1, R2 = Si(Alk)3
N
7a-n
8a-d
R2
R1
4-6
Entry
Substrate
Nucleophile (Nu2)/solvent
HMDS + LiHMDS/THF
T (°C)
Product
Yield (%)
Exact mass calcd/found
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
4a
4a
4a
4a
4a
4a
4a
4a
4a
4a
4d
4f
65
7a
7b
7c
7d
7e
7f
70
73
40
88
94
84
68
86
89
30
85
86
90
80
50
40
77
94
95
62
95
319.102848/319.103500
260.044192/260.043909
256.052421/256.053273
300.096622/300.096620
260.065322/260.065019
303.107521/303.108504
329.123171/329.123757
289.091871/289.091397
nd
Phenylacetylene + LDA/THF
TMS-acetylene + LDA/THF
c-Hex–CO2Me + LDA/THF
Me2–CH–CO2Me + LDA/THF
Me2N–CH2–CHMe–CO2Me + LDA/THF
N-Me-ethylpipecolinate + LDA/THF
Me2N–CH2–CO2Et + NaDA/THF
Bn2N–CH2–CO2Et + NaDA/THF
Ph2C@N–CH2–CO2Et + LDA/THF
n-PrSH + n-PrSNa/THF
n-PrSH + n-PrSNa/THF
BnOH + t-BuOK/THF
BnOH + t-BuOK/THF
n-PrSH + n-PrSNa/THF
n-PrSH + n-PrSNa/THF
BnOH + t-BuOK/THF
NaOH/THF–H2O or K2CO3/EtOH
NaOH/THF–H2O or K2CO3/EtOH
NaOH/THF–H2O or K2CO3/EtOH
NaOH/THF–H2O or K2CO3/EtOH
À50 to 20
À25 to 0
À25
À25
À25
À25
7g
7h
7i
À25
À30 to À10
À30 to 0
7j
nd
20
7k
7l
342.086057/342.088500
265.070741/265.071800
297.093584/294.092400
266.054757/266.053900
nd
nd
nd
nd
20
20
20
20
60
20
20
20
4f
7m
7n
7n
7n
7n
8a
8b
8c
8d
5a
5d
6c
6a
4a
4c
4d
4f
207.046634/207.046800
283.077934/283.078900
206.059246/206.060100
20
20
All compounds were fully characterized by 1H and 13C NMR spectroscopy. Compounds 7i and 7j were not analyzed for exact mass (converted to the
corresponding amino-acids). Compounds from entries 15–18 were not analyzed for exact mass (entries 15–17 same as entry 14, entry 18 known
compd).
1967, 32, 2823–2829; (c) Begland, R. W.; Hartter, D. R. J.
Org. Chem. 1972, 37, 4136–4145; (d) Yoon, S. C.; Cho, J.;
Kim, K. J. Chem. Soc., Perkin Trans. 1 1998, 109–116.
4. (a) Weinstock, L. M.; Mulvey, D. M.; Tull, R. J. Org.
Chem. 1976, 41, 3121–3214; (b) Hanasaki, Y.; Watanabe,
H.; Tsuzuki, K. JP 03193771, 1991; Chem. Abstr. 1992,
116, 59363.
5. Unpublished results. The reactivity of alkyl- and aryl-
thiadiazoles with Grignard reagents has been reported.
See: De Munno, A.; Bertini, V.; Picci, N. Heterocycles
1986, 24, 1131–1136, and references cited therein.
6. Kouvetakis, J.; Grotjahn, D.; Becker, P.; Moore, S.;
Dupon, R. Chem. Mater. 1994, 6, 636–639.
7. Komin, A. P.; Carmack, M. J. Heterocycl. Chem. 1976,
13, 13–22.
8. Merschaert, A.; Gorissen, H. Heterocycles 2003, 60, 29–
45.
9. Full preparative and characterization details in Borghese,
A.; Mancuso, V.; Merschaert, A. PCT Int. Appl. WO
2006068821, 2006.
10. A kg batch of 2b was produced. Stable for several years at
À20 °C as indicated by NMR spectroscopy.
References and notes
1. (a) Weinstock, L. M.; Shinkai, I. 1,2,5-Thiadiazoles and
their Benzo Derivatives. In Comprehensive Heterocyclic
Chemistry; Katritzky, A. R., Rees, C. W., Potts, K. T.,
Eds.; Pergamon Press: Oxford, 1984; Vol. 6, pp 513–543;
(b) Weinstock, L. M.; Pollak, P. I. Adv. Heterocycl. Chem.
1968, 9, 107–163.
2. (a) Wasson, B. K.; Gibson, W. K.; Stuart, R. S.; Williams,
H. W. R.; Yates, C. H. J. Med. Chem. 1972, 15, 651–655;
(b) Hanasaki, Y.; Watanabe, H.; Katsuura, K.; Taka-
yama, H.; Shirakawa, S.; Yamaguchi, K.; Sakai, S.; Ijichi,
K.; Fujiwara, M.; Konno, K.; Yokota, T.; Shigeta, S.;
Baba, M. J. Med. Chem. 1995, 38, 2038–2040; (c) Ward, J.
S.; Merritt, L.; Calligaro, D. O.; Bymaster, F. P.;
Shannon, H. E.; Mitch, C. H.; Whitesitt, C.; Brunsting,
D.; Sheardown, M. J.; Olesen, P. H.; Swedberg, M. D. B.;
Jeppesen, L.; Sauerberg, P. J. Med. Chem. 1998, 41, 379–
392, and references cited therein.
3. (a) Weinstock, L. M.; Davis, P.; Handelsman, B.; Tull, R.
Tetrahedron Lett. 1966, 7, 1263–1268; (b) Weinstock, L.
M.; Davis, P.; Handelsman, B.; Tull, R. J. Org. Chem.