One-Step Synthesis of 2-Chloropyrimidin-4-ol Derivatives: An Unusual Reactivity of Thiophosgene

Article abstract of DOI:10.1021/acs.orglett.5b02375

A novel, high-yielding, one-step synthesis of 2-chloroquinazolin-4-ols and analogous bicycles from 2-aminoamides using thiophosgene is described. The scope of the reaction includes aminothioamides, amino acids, and fused heterocycle derivatives, furnishing quinazolines, oxazinones, and substituted fused pyrimidine bicycles, respectively. On the basis of observed results with substituted analogues, a mechanism for this transformation is thought to occur via an isothiocyanate intermediate followed by an unexpected chemoselective reaction of thiophosgene on the thiol intermediate.

Full text of DOI:10.1021/acs.orglett.5b02375

Letter
pubs.acs.org/OrgLett
One-Step Synthesis of 2‑Chloropyrimidin-4-ol Derivatives: An
Unusual Reactivity of Thiophosgene
Michael Callingham, Francesca Blum, and Gregoire Pave*
́
́
CRT Discovery Laboratories, Jonas Webb Building, Babraham Research Campus, Cambridge CB22 3AT, United Kingdom
S
* Supporting Information
ABSTRACT: A novel, high-yielding, one-step synthesis of 2-
chloroquinazolin-4-ols and analogous bicycles from 2-aminoamides
using thiophosgene is described. The scope of the reaction includes
aminothioamides, amino acids, and fused heterocycle derivatives,
furnishing quinazolines, oxazinones, and substituted fused pyrimidine
bicycles, respectively. On the basis of observed results with substituted
analogues, a mechanism for this transformation is thought to occur via
an isothiocyanate intermediate followed by an unexpected chemo-
selective reaction of thiophosgene on the thiol intermediate.
ith the discovery of the biological potency of
pyrimidine-containing compounds over the last 50
Table 1. Optimization of Reaction Conditions
W
years, there have been a number of different methods used in
their syntheses.1−3 In this Letter we report a novel, facile one-
step synthesis of pyrimidine-containing compounds from
aromatic aminoamides and their analogues, using thiophosgene.
Syntheses of 2-chloropyrimidin-4-ols are often carried out in
three steps by treating 2-aminoamides with phosgene, followed
by phosphorus oxychloride to obtain the dichloroquinazoline.4
Careful addition of sodium hydroxide selectively hydrolyzes the
chlorine at the 4-position. The work described here represents a
simple, fast, one-step transformation of readily accessible
starting materials to biologically relevant 2-chloropyrimidin-4-
ol intermediates, which can be regioselectively diversified in one
pot.
The initial reaction conditions were optimized based on the
procedure used by Gochman and Wei starting from
anthranilamide and thiophosgene.5 The authors described the
formation of 2-thiopyrimidin-4-ol 4 and a tetracyclic byproduct
5, exclusively. Interestingly, in our hands these reaction
conditions led to the major compound 2-chloropyrimidin-4-ol
2, which was not described by Gochman and Wei. To the best
of our knowledge, this represents a novel transformation which
prompted our further investigations. The effects of solvent and
stoichiometric equivalents of thiophosgene upon this reaction
were investigated (Table 1).
When performing the reaction, the major side products
obtained identified by LCMS only are thought to be
quinazolin-2,4-diol 3, thiol derivative 4, and the tetracyclic
species 5. Addition of excess thiophosgene (Table 1, entry 1) in
1,4-dioxane gave excellent 95% conversion to 2 (65% isolated
yield), which was improved to 100% conversion (86% isolated
yield) with addition of only 2.1 equiv of thiophosgene (entry
3), with no observed side products. When thiophosgene was
added in two separate portions (entry 2), very similar results
were obtained, but the presence of compound 4 was clearly
seen by LCMS after addition of the first portion. Using THF as
a
conversion
b
solvent
CSCl2
2
3
4
5
yield
(equiv)
(%)
(%) (%) (%)
(%)
1
2
3
4
5
6
7
8
9
1,4-dioxane
1,4-dioxane
1,4-dioxane
THF
2.5
95
97
0
3
0
2
4
6
1
0
4
2
0
3
0
65
57
86
70
84
c
1.1 + 1.1
2.1
100
93
0
0
2.5
4
0
THF
2.1
89
6
0
d
MeCN
2.5
51
4
3
nm
toluene
2.5
63
2
28
9
43
d
CH2Cl2
2.5
33
44
11
nm
d
cyclohexane
2.5
33
15
nm
a
Crude conversion by UV was measured using liquid chromatog-
raphy−mass spectrometry (LCMS) after heating at reflux for 1 h.
Isolated yield of 2. Second portion of 1.1 equiv CSCl2 was added
after 30 min stirring at room temperature. nm (not measured): the
purity of 2 was low, and therefore, the material was not isolated.
b
c
d
solvent also led to the formation of side products (entries 4 and
5), as did acetonitrile (entry 6) which showed significant
presence of unidentifiable side products, as well as 3, 4, and 5,
and therefore, no material was isolated for this reaction. Use of
Received: August 19, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.5b02375
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
toluene or cyclohexane (entries 7 and 9) gave significant
amounts of undesired side products, notably the tetracycle 5
(most likely formed by the reaction of 2 and additional starting
material), which appears to be favored in less polar solvents.
The formation of 4 appears to be favored when using CH2Cl2
(entry 8). The optimal conditions were therefore found to be
2.1 equiv of thiophosgene in 1,4-dioxane, stirring at room
temperature for 1 h and subsequently heating at 105 °C for 1 h.
At that point, the 2-chloro-4-hydroxyquinazoline 2 can easily be
isolated as its hydrochloride salt by addition of Et2O on the
cold reaction mixture and collection by filtration (see
Supporting Information).
group was present (compound 6d), most likely due to the poor
solubility, which made purification challenging.
The solubility of the products obtained with heterocyclic
aminoamides (Figure 1, compounds 6e−l) was in general very
poor, which made purification difficult and contributed to the
lower yields.
Thiophenes (compounds 6e−g and 6i) were tolerated in this
reaction with good yields obtained. However, no conversion to
the 2-chloro-4-yl-3H-thieno[3,2-d]yrimidin-4-ol was seen de-
spite an additional 18 h of heating at 105 °C (compound 6h).
No conversion to the desired bicycle from the corresponding
2-pyridyl amino amide was observed (compound 6j). However,
for the 3-pyridyl isomers (compounds 6k and 6l), the desired
azaquinazoline was isolated in reasonable yields of 29% and
61%. An NH on the heterocyclic ring of the substrate
(compounds 6m and 6n) was also not tolerated; in the case
of compound 6n, this may be because the proposed reactive
isothiocyanate intermediate cannot form (see mechanism in
Scheme 1).
To examine the scope of the reaction, various aminoamides
were treated with thiophosgene using the optimized conditions
(Figure 1). We found that the 2-chloro position of several of
Scheme 1. Postulated Mechanism for First Step of
Thiophosgene Addition
Interestingly, thiophosgene addition to the methyl-substi-
tuted amide gave the desired bicycle 6o in good yield; however,
the secondary amine (compound 6p) did not react at all,
despite additional heating at 105 °C for 24 h and addition of a
further equivalent of thiophosgene.
From the experimental results, we postulate a mechanism
(Scheme 1). In the initial optimization reaction (Table 1, entry
2), the addition of the first portion of 1.1 equiv of thiophosgene
gave principally thiol intermediate 4, with full conversion to the
desired 2-chloropyrimidin-4-ol 2 only upon addition of the
second portion. The thiol intermediate 4 was also observed in
all the other optimization reactions after stirring at room
temperature for 30 min, suggesting that the reaction proceeds
via this intermediate.
The first step (step a) consists of the attack of the amine on
one molecule of thiophosgene followed by the rearrangement
to the corresponding isothiocyanate in step b. The existence of
the intermediate isothiocyanate is postulated because when the
amine is methyl-substituted, no reaction is observed (Figure 1,
compound 6p). In this case, it is not possible to form the
isothiocyanate intermediate (Scheme 1, step b, R = CH3).
However, methylated amides can still perform the intra-
molecular ring-closure (Figure 1, compound 6o). In the next
step (step c), intramolecular ring closure gives the 2-
thioquinazolinol 4. The thiol function is proposed to attack a
second equivalent of thiophosgene (step d), forming a leaving
group which can be readily displaced by the released chlorine
anion, giving the observed 2-chloropyrimidin-4-ol 2. It is
Figure 1. Scope of the reaction. Conditions: (a) No reaction despite
heating at 105 °C for 24 h and addition of a further 1 equiv of
thiophosgene; (b) additional heating at 105 °C for 18 h; (c) purity
85%, mixture with dihydroxylated byproduct; (d) Morpholino product
not isolated but the corresponding 2-chloro derivative was clearly seen
by LCMS.
these analogues was prone to hydrolysis, meaning that it was
difficult to assess accurately the conversion to the desired
product. Therefore, for ease of isolation, the crude reaction
mixtures were concentrated under reduced pressure and the
crude products obtained were subsequently reacted with an
excess of morpholine in CH2Cl2 in order to obtain an accurate
measure of conversion.
The morpholine addition was assumed to be quantitative
relative to the conversion of the starting aminoamide to the 2-
chloropyrimidin-4-ol species in the first step, and therefore,
isolated yields of the 2-morpholine derivatives were expected to
reflect conversion to the 2-chloropyrimidine derivatives.
Compounds 6a, 6b, and 6c (Figure 1) suggest that electron-
withdrawing and electron-donating substituents are tolerated in
this reaction, with excellent isolated yields of 82−94%.
However, the yield was significantly lower when a nitro
B
DOI: 10.1021/acs.orglett.5b02375
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
interesting to note that, during the reaction, only the thiol
moiety will attack the second equivalent of thiophosgene, not
the hydroxyl group. Moreover, the desired 2-chloropyrimidin-4-
ol 2 can also be formed by addition of one equivalent of
thiophosgene to the commercial 2-thiopyrimidine-4-ol 4, with
observed UV (LCMS) conversion of 100% (Scheme 2).
Scheme 3. Reaction with Thioamide
significant amounts of the 2-chloropyrimidin-4-ol 11 (30%
conversion by UV) and dihydroxy 12 (5% conversion by UV)
byproducts after heating at 105 °C for 1 h. We believe that
during the reaction, the major product formed is the 3-chloro
derivative 10. But due to its instability and under LCMS
condition, 10 is readily hydrolyzed to the corresponding mono
and bis-hydroxy analogues 11 and 12 in 30% and 5% yield,
respectively. Further hydrolysis at the 2-position was observed
during an attempt to purify the compound using flash column
chromatography. Hydrolysis at the 2- and 4-chloro positions
was facile, and therefore, any derivatization at these positions
would require careful optimization of conditions to avoid this
rapid hydrolysis.
Scheme 2. Conversion of 2-Thio-4-hydroxyquinazoline 4 to
2-Chloro-4-hydroxyquinazoline 2
Amino acid substrates are also tolerated in this reaction
(Table 2). The unsubstituted benzo[d]oxazinone 8a (entry 1)
Table 2. Reaction with Amino Acids
We have reported a one-pot thiophosgene-promoted
cyclization of readily available 2-aminoamides to furnish 2-
chloro-4-hydroxypyrimidine derivatives. These highly versatile
building blocks contain key pharmacologically active motifs and
can be readily and regioselectively functionalized at both the 2-
and 4-positions. We have demonstrated the highly selective
reactivity of thiophosgene toward 2-thiopyrimidines in the
presence of hydroxyl functionality. This allows us to consider
the thiol group as a masking group for a chlorine moiety, which
can then be readily displaced as shown above or cross-coupled
(data not shown). Further investigation is underway to expand
the scope of this new transformation.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.or-
glett.5b02375.
Experimental procedures and characterization for all new
compounds (PDF)
a
Desired product not observed.
AUTHOR INFORMATION
Corresponding Author
■
was obtained in excellent yield and the bromo-substitution 8b
(entry 2) also allowed good conversion of 59%. Two molecules
of morpholine are added in compound 8c when there is an
electron-donating methoxy group present (entry 3); this was
not observed for entries 1 and 2. Although the desired product
5-methoxybenzo[d][1,3]oxazin-2,4-dione was observed by
LCMS, the stability of this ring system was poor after the
addition of morpholine, and addition of a second molecule of
morpholine was observed.
The thiopheno-amino acid substrate (7d, entry 4) did not
show any conversion to the chloro intermediate 8d after the
addition of thiophosgene. An interesting point to note is that
upon addition of thiophosgene to these amino acid derivatives,
formation of the 2-chloro intermediate was limited. Under
LCMS conditions, the major products seen were the 2-hydroxy-
substituted products in each case (Table 2, entries 1−3).
The reaction of 2-amino-4-chlorothiobenzamide was also
successful, furnishing the 2,4-dichloroquinazoline species 10
directly (Scheme 3). Although the desired trichloro product 10
was present (30% conversion by UV), there were also
*E-mail: gpave01@gmail.com
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
The authors would like to thank Drs Tom Sheppard
(University College London), Katherine Fowler and Jon Roffey
(Cancer Research Technology) and Gregory Ott (Teva
Pharmaceutical) for useful discussions.
REFERENCES
■
(1) Connolly, D. J.; Cusack, D.; O’Sullivan, T. P.; Guiry, P. J.
Tetrahedron 2005, 61, 10153−10202.
(2) Bergman, J.; Brynolf, A. Tetrahedron 1990, 46, 1295−1310.
(3) Larksarp, C.; Alper, H. J. Org. Chem. 2000, 65, 2773−2777.
(4) Lee, S. J.; Konishi, Y.; Yu, D. T. J. Med. Chem. 1995, 38, 3547−
3557.
(5) Bell, S.; Gochman, C.; Wei, P. J. Heterocycl. Chem. 1975, 12,
1207−1210.
C
DOI: 10.1021/acs.orglett.5b02375
Org. Lett. XXXX, XXX, XXX−XXX