Convenient and practical one-pot synthesis of 4-chloropyrimidines via a novel chloroimidate annulation

Article abstract of DOI:10.1021/op1002352

Reaction of aromatic or heteroaromatic 2-acyl(amino)nitriles with phosphorus pentachloride triggers a novel chloroimidate cyclization, leading directly to the corresponding annullated 4-chloropyrimidines in good to excellent yields. The reaction lends itself to the telescoped one-pot construction of 4-functionalized pyrimidines from the corresponding (hetero)aromatic 2-aminonitriles. For a pyrazolopyrimidine development intermediate, this reaction was scaled up to multikilogram scale with excellent results. A total of 10 examples with different substrates are provided. This one-pot reaction provides an attractive and sustainable alternative to the commonly used multistep methodology for this transformation.

Full text of DOI:10.1021/op1002352

ARTICLE
pubs.acs.org/OPRD
Convenient and Practical One-Pot Synthesis of 4-Chloropyrimidines
via a Novel Chloroimidate Annulation
Thomas Storz,*,† Richard Heid,‡ Joseph Zeldis,‡ Steven M. Hoagland,§ Vito Rapisardi,§ Susan Hollywood,^
and George Morton#
†Pfizer Worldwide R&D, Chemical Research and Development, Eastern Point Road, Groton, Connecticut 06340, United States
‡Pfizer Chemical Development, §Kilolab Facilities, ^Safety Group, and #Analytical Services, 401 North Middletown Road, Pearl River,
New York 10965, United States
ABSTRACT: Reaction of aromatic or heteroaromatic 2-acyl(amino)nitriles with phosphorus pentachloride triggers a novel
chloroimidate cyclization, leading directly to the corresponding annullated 4-chloropyrimidines in good to excellent yields.
The reaction lends itself to the telescoped one-pot construction of 4-functionalized pyrimidines from the corresponding
(hetero)aromatic 2-aminonitriles. For a pyrazolopyrimidine development intermediate, this reaction was scaled up to multikilogram
scale with excellent results. A total of 10 examples with different substrates are provided. This one-pot reaction provides an attractive
and sustainable alternative to the commonly used multistep methodology for this transformation.
’ INTRODUCTION
chloroimidate (path A) in a nucleophilic (6-endotrig) fashion,
resulting in cyclization to the desired 4-chloropyrimidine
(Scheme 4):12
Duetotheir excellent reactivityinaromatic nucleophilic (SNAr)
substitutions and transition metal-catalyzed cross-coupling reac-
tions, 4-halogenated pyrimidines1-4 are widely used as activated
building blocks for the construction of various pharmacophores.
Traditional synthesis of 4-chloropyrimidines (B) involves reac-
tion of the corresponding pyrimidin-4-one precursor (A) with an
inorganic acid halide (e.g., phosphorus pentachloride, phosphorus
oxychloride, or thionyl chloride)5 (Scheme 1):
For annullated systems, the pyrimidin-4-ones A are usually
obtained by cyclization of a (hetero)aromatic 2-acylamino
carboxamide C, which in turn may be obtained via a variety of
methods, including amidation of the corresponding carboxylic
acid D or partial hydrolysis of a precursor nitrile E (Scheme 2).5f
In the context of one of our development programs, we
assessed the preparation of the azolo-chloropyrimidine intermedi-
ate 4 via a three-step literature6 sequence shown in Scheme 3.
From a process point of view, this stepwise hydrolysis/
cyclization/chlorination protocol, although widely used for the
assembly of highly substituted, annullated 4-chloropyrimidines
(for examples, see refs 6-11), appeared tedious and fraught with
both safety concerns (heating alcoholic mixtures containing hydro-
gen peroxide) and robustness issues (basic hydroperoxide hydro-
lysis of the nitrile in presence of amide). These led us to consider the
direct transformation of 1 to 4.
This was indeed found to be the case. A variety of solvents
were screened, with sulfolane providing the cleanest reactions
and highest yields. Phosphorus pentachloride proved superior to
phosphorus oxychloride (lower yields), whereas phosphorus
trichloride and thionyl chloride led to decomposition. The
desired azolo-(chloro)pyrimidine 4 (vide supra) was thus ob-
tained in 90% yield on 200-g scale directly from the amide 1
(Scheme 5).
Investigation of the thermal safety data obtained by Thermal
Screening Unit (TSu) testing (5 mL suspension of all starting
materials mixed cold, with 2 equiv of PCl5) showed a very
shallow, small exotherm ranging from 70 to 170 ꢀC and a second
smaller exotherm from 170 to 250 ꢀC. A residual pressure of 3 bar
was observed (Figures 1 and 2). Additional ARSST (Advanced
Reactive System Screening Tool, Fauske and Associates, LLC)13
testing on a more concentrated sample, with 4 equiv of PCl5,
showed a small exotherm at 115 to 150 ꢀC and a stronger
exotherm ranging from 152 to 320 ꢀC, whereas the isolated,
crude compound 4 proved benign with regard to thermal safety
(data not shown). On the basis of this initial safety testing result
and the lack of internal and external scale-up experience with this
new reaction type, the amount of PCl5 for initial kilolab scale-up
was limited to 2 equiv, the concentration was adjusted slightly
towards higher dilution, and the batch was divided into three runs
of approximately 2.8 kg each. Another concern was that on lab
scale, under stronger nitrogen sweep flow rates, PCl5 was some-
times observed to crystallize/accumulate in the condenser during
the course of the reaction. This was addressed in the kilolab
’ RESULTS AND DISCUSSION
We felt it was reasonable to assume intermediate imidoyl
chlorophosphate formation from the secondary amide substruc-
ture and a phosphorus halide, in analogy to the first step in the
Vilsmeier reaction of hydroxyheteroaromatics.5 Hydrogen chlo-
ride addition to the nitrile group by one of the pathways shown
below (A vs B) could follow. In this manner, a primary
chloroimidate would be generated which might then attack either
the imidoyl chlorophosphate (path B), or the secondary
Special Issue: Sustainable Process Chemistry
Received: August 27, 2010
Published: November 24, 2010
r
2010 American Chemical Society
918
dx.doi.org/10.1021/op1002352 Org. Process Res. Dev. 2011, 15, 918–924
|

Organic Process Research & Development
ARTICLE
Scheme 1
Scheme 4
Scheme 2
Scheme 5a
Scheme 3a
a Reagents and conditions: a) 4 equiv of PCl5, sulfolane, 120 ꢀC, 14 h,
90% yield (assay-corrected).
proceeds to ∼90-95% completion in the absence of added base;
therefore, we speculate that the neighboring cyano group may act
as an internal base capturing the liberated hydrogen chloride and
forming the chloroimidate to some extent already during the
acylation step (this intermediate can be observed by HPLC in
some cases).
This nucleophilic chloroimidate cyclization may be rather
general, as demonstrated by synthesis of the chloro[benzo]-
diazines 10, 12, and 14 shown below (Scheme 7); not only
chloroquinazolines (10, 14), but also pyrimido-chloropyrimi-
dines (12) are readily accessible.
a Reagents and conditions: a) 30% H2O2, NaOH-EtOH. b) EtOH,
reflux. c) POCl3, cat. DMF.
scale-up through lower agitation rates and a minimized head-
space nitrogen flow.
Each of the three ∼2.8 kg kilolab runs gave a reproducible
assay-corrected yield of ∼85% and 99% HPLC-purity. On this
scale, the two equivalents of PCl5 used in the reaction gave 94%
conversion after only 5 h reaction time, which was attributed to
the minimal overhead nitrogen sweep applied and the reduced
agitation rate (HCl flushout minimized). No exotherms or
accumulation of PCl5 in the condenser were observed under
these conditions. To obtain a homogeneous batch, a crystal-
lization step combining the three sublots in a DMF-water
mixture was added, which also served to completely eliminate
any residual sulfolane trapped in the product and resulted in an
81% assay corrected yield of ∼7 kg of recrystallized 4 (99.9 A%
HPLC purity).
A telescoped process is also achievable in the case of the
six-membered (hetero)aromatic aminonitriles. The aromatic
β-amino nitriles were heated in sulfolane with an acid chloride
in the absence of base and after completed acylation, phosphorus
pentachloride was added, and heating was continued until
the cyclization was complete (typically 8-12 h) (Scheme 8,
compounds 16 and 18). Additionally, acylated aminonitriles
can be heated with phosphorus pentachloride and the reaction
quenched directly with a nucleophile. For instance, 4-aminoqui-
nazolines are readily accessible (Scheme 8, compound 20).
With a secondary amine as nucleophile, the sequence could
even be telescoped over three steps with satisfactory overall yield
(Scheme 8, compound 21).
Substrate scope was explored (Scheme 6), with good tolera-
tion of both a triazole (5) and a benzofuran (7) substrate on
small scale using these conditions (not optimized). It was also
found that both the acylation step from the amino-azole as well as
the subsequent PCl5-cyclization in sulfolane can be carried out in
one pot, without workup or isolation of the intermediate amide
(Scheme 6). At elevated temperatures, the acylation typically
Mechanistically, pathway (A) in Scheme 4 appears more
likely,12 as an exploratory 13C NMR experiment on 2-benzoyla-
mino-benzonitrile showed the initial formation of a chloropho-
sphate imidate species (2JC,P ≈ 160 Hz) upon PCl5 addition
which does not collapse directly to the product but instead goes
through a mixture of non-phosphorus-containing intermediates
before finally converging to the product spectrum. However,
919
dx.doi.org/10.1021/op1002352 |Org. Process Res. Dev. 2011, 15, 918–924

Organic Process Research & Development
ARTICLE
Figure 1. TSu safety data for reaction 1 f 4 (starting materials mixed cold).
Figure 2. TSu safety data for reaction 1 f 4 (starting materials mixed cold).
since this experiment was carried out in CD3CN (deuterated
sulfolane was not available to us at the time), solvent participa-
tion in the mechanism cannot be excluded.
developed for the same transformation, reducing waste and
increasing sustainability.
In conclusion, we have found and demonstrated scale-up
viability for a novel and practical chloroimidate cyclization proto-
col, allowing rapid construction of annullated 4-chloropyrimidines
from the corresponding (hetero)aromatic amino- or acylaminoni-
trile precursor in a straightforward fashion. The reaction lends
itself to telescoped process variants. In most cases, the crude
product isolated after aqueous workup is of good purity (typically
∼95% HPLC). This one-pot reaction sequence represents
significant progress over existing multistep methodology6-11
’ EXPERIMENTAL SECTION
Starting materials, reagents, and solvents were obtained from
commercial suppliers and were used without further purification.
All the melting points are uncorrected and determined on a
1
B€uchi apparatus. H NMR spectra were recorded at 500 MHz,
and 13C NMR spectra were recorded at 125 MHz on a Bruker
DRX-500 NMR instrument. Assignments were confirmed by the
appropriate 2D-NMR experiments. IR spectra were measured on
920
dx.doi.org/10.1021/op1002352 |Org. Process Res. Dev. 2011, 15, 918–924

Organic Process Research & Development
ARTICLE
Scheme 6a
Scheme 8a
a Reagents and conditions: a) p-nitrobenzoyl chloride, sulfolane, 80 ꢀC,
2 h. b) Add 2 equiv of PCl5, 95 ꢀC, 3 h. c) p-Nitrobenzoyl chloride,
sulfolane, 30 ꢀC, 2 h. d) Add 2 equiv of PCl5, 95 ꢀC, 2 h.
Scheme 7a
a Reagents and conditions: a) Nicotinoyl chloride, 100 ꢀC, 12 h.
b) PCl5, sulfolane, 100 ꢀC, 10th. c) Benzoyl chloride, 95 ꢀC, 75 h. d)
PCl5, 100 ꢀC, 12th. e) PCl5, 100 ꢀC, 19th. f) NH3, H2O-CH3CN, rt f
50 ꢀC, 1 h. g) Pivaloyl chloride, 75 ꢀC, 48 h. h) PCl5, 110 ꢀC, 19 h. i)
Morpholine-CH3CH, rt, 1 h.
cell, pressure transducer, and piping—is approximately 10 mL.
Advanced Reactive System Screening Tool (ARSST, Fauske and
Associates, LLC.): The sample is charged into a 10-mL glass cell,
which is then wrapped in a heating mantle and aluminum foil and
fiberglass to minimize thermal losses. The cell is placed in a larger
containment vessel (total containment is 310 mL). The contain-
ment vessel is pressurized with nitrogen gas to prevent solvent
boiling. A thermocouple is immersed in the sample to measure
sample temperature, and pressure is monitored throughout the
run with a transducer. A small magnetic bar is used to induce
sample stirring. The ARSST has an injection port, allowing for the
charging of one of the reagents in situ during the run.
a Reagents and conditions: a) 1.75 equiv of PCl5, sulfolane 110 ꢀC, 16 h.
b) 1.75 equiv of PCl5, 100 ꢀC, 3 h.) 1.75 equiv of PCl5, 100 C, 16 h.
Experimental Procedures for 4-Chloropyrimidines (step-
wise). First Proof-of-Concept Experiment: 4-Chloro-2-phenyl-
quinazoline (10). To a solution of N-(2-cyanophenyl)-benzamide
9 (598 mg, 2.69 mmol) in sulfolane (12 mL) was added
phosphorus pentachloride (1.04 g, 4.98 mmol) in one portion
at 45 ꢀC (yellow, clear solution). Solution was stirred at 110 ꢀC in
a 20-mL septum vial for 16 h, then cooled to 45 ꢀC. HPLC
showed 98 A% product (254 nm). Upon further cooling to to rt,
water (1 mL) was slowly added, followed by H€unig’s base
(3.5 mL, 20 mmol), such that the temperature stayed below
30 ꢀC. The pH was adjusted to 7 with AcOH, and the suspension
was stirred in an ice bath for ∼30 min and then filtered over
a medium porosity glass frit. The filter cake was washed with
water (2 ꢀ 20 mL) and dried to constant weight at 40 ꢀC (high
vacuum). Obtained 588 mg (91%) assay-corrected {using com-
mercial AM-ex-OL (Aldrich) as ref standard} yield of 10 as a white
solid. HPLC (Method A): Rt 4.08 min, 98.8 A% (210 nm).
4-Chloro-2-phenylpyrimido[4,5-d]pyrimidine (12). To a so-
lution of N-(2-cyanophenyl)-4-amino-5-cyanopyrimidine 1115
(800 mg, 3.56 mmol) in sulfolane (12 mL) was added phos-
phorus pentachloride (1.3 g, 6.24 mmol) in one portion at 45 ꢀC
(yellow, clear solution). Stirred at 100 ꢀC in a 20-mL septum vial
for 3 h, then cooled to 45 ꢀC. HPLC showed 98 A% product
(254 nm). Upon aq workup (as for (10)) and filtration, 1.3 g of a
yellow solid was obtained (HPLC: 97 A% (254 nm)). The crude
wetcake purity decreased to 85 A% after overnight drying at
45 ꢀC in the vacuum oven (hydrolysis product detected). FC of
a Bruker IFS660 spectrometer. Exact mass determinations were
performed on a ABI QSTARXL Q-ToF instrument (calibrated in
positive ion mode with CsI and sex pheromone inhibitor iPD1
(Bachem)). Typical experimental procedures for the isolation of
the chloropyrimidine products generally involved neutralization/
precipitation with water, which usually gave material of sufficient
(g95% HPLC) purity. On small scale, some samples were
purified by flash-chromatography (FC)14 onSiO2. HPLC Method
(A): stationary phase: Sunfire C18, 3.5 μm, 4.6 mm ꢀ 150 mm,
column oven temp: 35 ꢀC; flow: 2.5 mL/min; Solvent A: 1900 mL
of H2O, 100 mL of CH3CN, 1 mL of TFA; Solvent B: 1900 mL of
CH3CN, 100 mL of H2O, 1 mL of TFA. Gradient: linear gradient
0 f 100% B over 4 min, then 100% B for 1 min. Method (B):
stationary phase: SunFire C18, 3.5 μm, 4.6 mm ꢀ 150 mm,
column oven temp: 35 ꢀC; flow: 1.0 mL/min. Solvent A: 950 mL
of H2O, 50 mL of CH3CN, 0.5 mL of H3PO4. Solvent B: 1000 mL
of CH3CN, 0.5 mL of H3PO4. Gradient: 70% A, 30% B for 5 min,
then linear gradient 30% f 100% B over 5 min, then 100% B for
5 min. Detection: UV (λ = 220 nm, unless mentioned otherwise).
Safety testing: (A) Thermal Screening Unit (HEL Ltd.): The
sample is charged into a 9-mL (Hastelloy C unless otherwise
indicated) test cell, which is then sealed into an oven. The oven
temperature is ramped upto 250 ꢀC at 2 ꢀC/min, unless otherwise
indicated. Sample temperature is monitored with a thermocouple
immersed into the sample. Pressure and oven temperature are also
monitored throughout the run. The containment volume—test
921
dx.doi.org/10.1021/op1002352 |Org. Process Res. Dev. 2011, 15, 918–924

Organic Process Research & Development
ARTICLE
this material (heptanes/EtOAc 3:1 f 2:1 gradient) provided 550
mg (64%) 12 as a slightly yellowish solid. HPLC (Method A): Rt
2.18 min, 98.2 A% (254 nm). 1H NMR (CDCl3, 400 MHz): δ
7.55 (m, 2H), 7.60 (m, 1H), 8.67 (d, 2H, J = 8.4 Hz), 9.65 (s, 1H),
9.74 (s, 1H). 13C NMR (CDCl3, 100 MHz): δ 114.9, 128.9,
130.1, 133.3, 135.0, 159.9, 162.5, 163.1, 163.3, 167.5. HR-MS:
ionizes as [M þ H2O] (hydrolysis under LC-MS conditions):
calculated for C12H7ClN4 þ H2O: 260.0464; found: 260.0457.
4-Chloro-2-trifluoromethyl-quinazoline (14). To a solution of
N-(2-cyanophenyl)-2,2,2-trifluoroacetamide 13 (1.15 g, 5.37
mmol) (HPLC) (Method A): Rt 2.36 min) in sulfolane
(12 mL) was added phosphorus pentachloride (2.0 g, 9.4 mmol)
in one portion at 45 ꢀC (thin suspension). Stirred at 100 ꢀC in a
20-mL septum vial overnight, then cooled to 40 ꢀC. Reaction is
slower than for 10 and 12: HPLC showed 93% starting material
conversion and 88.3A% (210 nm) product 14 after 15 h at100 ꢀC.
Upon aq workup (as for (10)), filtration, and drying at 40 ꢀC
under high vacuum, 1.0 g (80%) of 14 was obtained as pale-yellow
solid. HPLC (Method A): Rt 3.25 min, 99.4 A% (210 nm). 1H
NMR (CDCl3, 400 MHz): δ 7.91 (dd, 1H, J = 6.8, 8.0 Hz), 8.11
(dd, 1H, J = 6.8, 8.4 Hz), 8.22 (d, 1H, J = 8.4 Hz), 8.37 (d, 1H, J =
8.0 Hz). 13C NMR (CDCl3, 100 MHz): δ 119.3 (q, CF3, J = 275.7
Hz), 124.2, 126.0, 129.6, 131.1, 136.2, 150.4, 151.4, 164.2. 19F
NMR (CDCl3, 376.5 MHz): δ 69.95. HR-MS: calculated for
C9H4F3ClN2: 232.0015; found: 232.0027.
One-Pot Telescoping Examples. 4-Chloro-2-(3-pyridyl)-
quinazoline (16). To a solution of 2-amino-benzonitrile 15 (240
mg, 2 mmol) in sulfolane (2 mL) was added nicotinoyl chloride
hydrochloride (550 mg, 3 mmol). After stirring in a 4-mL septum
vial at 100 ꢀC overnight, HPLC showed full conversion of the
aminonitrile (Rt 2.03 min) to an intermediate (Rt 1.66 min).
Phosphorus pentachloride (730 mg, 3.5 mmol) was added in
one portion, and stirring was continued at 100 ꢀC until HPLC
showed full conversion of the intermediate (∼10 h). After the
usual workup (vide supra, 10), and FC (heptane/EtOAc), 400 mg
(83%) yield of 16 was obtained as a slightly tan solid. HPLC
(Method A): Rt 2.45 min, 99.5A% (254 nm). 1H NMR (CDCl3,
400 MHz): δ 7.44 (dd, 1H, J = 4.8, 8.0 Hz), 7.69 (td, 1H, J = 1.2,
6.8, 8.4 Hz), 7.95 (td, 1H, J = 1.2, 6.8, 8.4 Hz), 8.09 (d, 1H, J = 8.4
Hz), 8.25 (d, 1H, J = 8.4 Hz), 8.74 (d, 1H, J = 4.8 Hz), 8.81 (dt, J =
1.2, 1.2, 8.0 Hz), 9.76 (s, 1H). 13C NMR (CDCl3, 100 MHz): δ
122.7, 123.4, 125.9, 128.8, 129.0, 132.3, 135.1, 135.9, 150.3,
151.65, 151.7, 158.2, 162.8. HR-MS: calculated for C13H8ClN3:
241.0406; found: 241.0372.
4-Chloro-2-phenylpyrido[2,3-d]pyrimidine (18). To a solution
of 2-amino-nicotine-nitrile 17 (500 mg, 4.19 mmol) in sulfolane
(5 mL) was added benzoyl chloride (1.0 mL, 8.8 mmol). After
stirring at 90-95 ꢀC in a 20-mL septum vial for 72 h, LC-MS
showed full conversion of the aminonitrile to an intermediate.
The mixture was cooled to 45 ꢀC before the volume was adjusted
to 10 mL by addition of sulfolane. Phosphorus pentachloride
(1.53 g, 7.53 mmol) was added in one portion, and stirring was
continued at 100 ꢀC until HPLC showed full conversion of the
intermediate (∼12 h). After the usual workup (videsupra, 10), the
yellow wetcake obtained was dried overnight (vacuum oven,
45 ꢀC), the crude product (1.2 g, >100%, HPLC: 80 A%
[210 nm]) was purified by FC (heptane/EtOAc) to give 0.75 g
(74%) of 18 as an off-white solid. HPLC (Method A): Rt 3.20
min, 98.8 A% (254 nm). 1H NMR (CDCl3, 400 MHz): δ 7.50
(m, 2H), 7.52 (m, 1H), 7.60 (dd, 1H, J = 4.4, 8.4 Hz), 8.57 (dd,
1H, J = 8.4, 2.0 Hz), 8.69 (m, 2H), 9.26 (dd, 1H, J = 2.0, 4.4 Hz).
13C NMR (CDCl3, 100 MHz): δ 117.7, 123.6, 128.7, 129.4,
132.1, 135.3, 135.8, 158.7, 159.6, 163.1, 163.4. HR-MS: calculated
for C13H8ClN3: 241.0407; found: 241.0409.
2-Phenylpyrido[2,3-d]pyrimidin-4-amine (20). To a suspen-
sion of 2-benzoylamino-nicotine-nitrile 19 (1 g, 4.47 mmol,
HPLC: Rt 2.12 min) in sulfolane (10 mL) was added phosphorus
pentachloride (0.94 g, 4.47 mmol) in one portion, and the
mixture was stirred overnight in a 20-mL septum vial at 100 ꢀC.
After 19 h, HPLC showed full conversion, clear-yellow solution,
96.7 A%, Rt 3.18 min (chloropyrimidine). This solution was
cooled to 40 ꢀC and added slowly under vigorous stirring into
cold, aq conc. ammonia (20 equiv, open Erlenmeyer flask), such
that the temperature stayed below 20 ꢀC (ice cooling). After
completed addition, the ice bath was removed, and the creamy
suspension was warmed up to rt. Acetonitrile (10 mL) was added,
and the suspension was stirred at 50 ꢀC, until a reddish, clear
solution formed (∼1 h). HPLC showed complete and clean
conversion of the intermediate chloropyrimidine, 96.9 A%, Rt
1.79 min, LC-MS: [MHþ] = 223. The solution was cooled in an
icebath, and the precipitated solid was filtered off and dried under
vacuum to give 20 (0.9 g, 91%) as a slightly tan solid, HPLC
(Method A): 98.3 A%, Rt 1.82 min [210 nm]. 1H NMR (DMSO-
d6, 400 MHz): δ 7.52 (dd, 1H, J = 4.4, 8.4 Hz), 7.54 (m, 3H), 8.19
(br s, 2H), 8.52 (m, 2H), 8.74 (dd, 1H, J = 8.4, 2.0 Hz), 9.03 (dd,
1H, J = 2.0, 4.4 Hz). 13C NMR (DMSO-d6, 100 MHz): δ 107.9,
120.7, 128.1, 128.2, 130.5, 133.2, 138.2, 156.1, 159.4, 162.9, 163.4.
HR-MS: calculated for C13H9N4: 222.0906; found: 222.0902.
2-tert-Butyl-4-morpholinoquinazoline (21). To a suspension
of 2-aminobenzonitrile 15 (1.54 g, 13.03 mmol, HPLC (Method
A): Rt 2.0 min) in sulfolane (5 mL) was added pivaloyl chloride
(1.92 mL, 15.64 mmol), and the mixture was stirred for 48 h at
75 ꢀC in a 20-mL septum vial; HPLC showed 100% conversion at
this point, mainly to the piv-amide (Rt 2.45 min), with some
chloropyrimidine already present (Rt 4.0 min) also. The amber
solution was cooled to room temperature and diluted with
sulfolane (7 mL). Phosphorus pentachloride (4.1 g, 19.5 mmol)
was added, and the suspension was stirred at 110 ꢀC for 12 h,
before more phosphorus pentachloride (0.54 g, 2.6 mmol) was
added, and stirring was continued at 110 ꢀC for 7 h. After cooling
to room temp., HPLC showed 93% conversion of the intermedi-
ate amide; 78.6 A% chloropyrimidine (Rt 4.0 min). The yellow,
clear solution was slowly added to an ice-cold solution of
morpholine (20 equiv) in acetonitrile (25 mL), such that the
temperature stayed below 10 ꢀC (exothermic). After completed
addition, the amber, creamy suspension was stirred at room temp.
for 1 h. HPLC showed complete conversion of the chloropyr-
imidine, 67.3 A%, Rt 2.28 min (product 21). After dilution with
ice water and extraction with EtOAc, the organic phases were
washed with pH 7 phosphate buffer and water. Drying of the
combined organic phases over sodium sulfate, evaporation, and
FC of the crude material (63.8 A%) on SiO2 (heptane/EtOAc,
10:1) gave 21 as a slightly yellowish, crystalline solid (1.7 g, 48%
over three steps). HPLC (Method A) 95.1 A% (210 nm) Rt 2.25
min. 1H NMR (DMSO-d6, 400 MHz): δ 1.39 (s, 9H), 3.70 (t,
4H), 3.80 (t, 4H), 7.49 (m, 1H), 7.78 (m, 1H), 7.80 (d, 1H), 7.98
(d, 1H). 13C NMR (DMSO-d6, 100 MHz): δ 29.4, 38.9, 49.8,
65.9, 114.0, 124.8, 124.9, 128.2, 132.4, 151.6, 163.7, 170.8. HR-
MS: calculated for C16H21N3O: 271.1685; found: 271.1682.
Experimental Procedures for Azolo-4-chloropyrimidines.
4-Chloro-6-(4-nitrophenyl)-1-(2,2,2-trifluoroethyl)-1,2,3-triazolo-
[4,5-d]-pyrimidine (4)
(a) Lab Scale. To a stirred slurry of N-[4-cyano-1-(2,2,2-
trifluoroethyl)-1H-pyrazol-5-yl]-4-nitro-benzamide 16 (200 g,
922
dx.doi.org/10.1021/op1002352 |Org. Process Res. Dev. 2011, 15, 918–924

Organic Process Research & Development
ARTICLE
589.57 mmol) in sulfolane (1200 mL) was added phosphorus
pentachloride (491.09 g, 2.36 mol, 4 equiv) at 30 ꢀC (addition
slightly exothermic) under nitrogen. The mixture was vigorously
stirred at 120 ꢀC under a nitrogen atmosphere for 14 h (HPLC
indicated complete conversion). The mixture was cooled to 55 ꢀC
and slowly quenched into cold water while maintaining the
temperature between 20-30 ꢀC. The solution was basified by
addition of triethylamine under ice cooling, such that the tem-
perature stayed below 25 ꢀC (at temperatures above 45ꢀC,
hydrolysis to the hydroxypyrimidine accelerates). After further cool-
ing to 18-20 ꢀC, the precipitated solid was filtered off, washed
with water, and dried under vacuum at 50 ꢀC. Product 4 was thus
obtained as an off-white solid (193 g, 90% assay-corrected yield),
4-Chloro-2-(4-nitrophenyl)benzo[4,5]furo[3,2-d]pyrimidine
(8). A solution of 3-aminobenzofuran-2-carbonitrile 717 (Rt 5.2
min) (0.64 g, 4.0 mmol) in sulfolane (15 mL) was treated with
p-nitrobenzoyl chloride (0.82 g, 4.4 mmol) at ambient tempera-
ture (25-30 ꢀC) resulting in formation of thick suspension. After
disappearance of the starting material as judged by HPLC (amide,
Rt 7.2 min), phosphorus pentachloride (1.7 g, 8.1 mmol) was
added, and temperature raised to 95 ꢀC. After stirring for 2 h, the
reaction mixture was cooled and worked up as for 4 to afford crude
product (0.64g, assay-correctedyield: 49%over two steps). HPLC
(Method B) 92.5 A% (220 nm) Rt 12.1 min. 1H NMR (CDCl3,
400 MHz): δ 7.57 (m, 1H, J = 1.2, 6.8, 7.8 Hz), 7.75 (dd, 1H, J =
1.2, 7.8 Hz), 7.78 (m, 1H, J = 1.2, 6.8, 7.8 Hz), 8.31 (dd, 1H, J = 1.2,
7.8 Hz), 8.35 (d, 2H, J = 9.0 Hz), 8.74 (d, 2H, J = 9.0 Hz). 13C
NMR (CDCl3, 100 MHz): δ 113.2, 121.6, 122.9, 123.8, 125.0,
129.3, 132.8, 142.3, 142.8, 143.5, 149.2, 152.1, 157.8, 158.6. HR-
MS: calculated for C16H8ClN3O3: 325.0254; found: 325.0252.
1
HPLC (Method A) 96.3 A% (220 nm) Rt 3.85 min. H NMR
(DMSO-d6, 400 MHz): δ 5.59 (q, 2H, 3JH-F = 8.8 Hz), 8.36 (d,
2H, J = 8.8 Hz), 8.65 (d, 2H, J = 8.8 Hz), 8.67 (s, 1H). 13C NMR
(DMSO-d6, 100 MHz): δ 47.5 (CH2, JC-F = 35.2 Hz), 112.3,
123.5 (CF3, JC-F = 280.7 Hz), 123.9, 129.6, 134.7, 141.0, 149.2,
154.5, 154.9, 159.0. HR-MS: calculated for C13H8F3ClN5O2:
358.0313; found: 358.0346.
’ AUTHOR INFORMATION
Corresponding Author
*E-mail: thomas.storz@pfizer.com.
(b) Kilolab Scale-Up. A 50-L Hastelloy reactor was charged with
sulfolane (23.8 kg). Amide 1 (2.71 kg, 7.99 mol) and phosphorus
pentachloride (3.512 kg, 16.0 mol, 2.0 equiv) were charged to the
solvent by means of a split butterfly valve. The mixture was heated
to 122 ꢀC and held under agitation for a period of 5 h. The batch
was cooled to 55 ꢀC and sampled for reaction completion by
HPLC. The batch was cooled to 50 ꢀC and quenched by addition
of N-methyl-2-pyrrolidone (16.7 kg) over a period of 30 min while
maintaining the batch temperature in the range 46-54 ꢀC. The
reaction solution was then cooled to 13-14 ꢀC, and USP water
(21.6 kg) was added over a 1-h period while keeping the batch
temperature between 14-20 ꢀC. The reaction suspension was
then neutralized to pH 7.6 with triethylamine (4.4 kg, 43.5 mol)
over a period of 2.3 h. The mixture was stirred at 23 ꢀC overnight
(11 h). The product was isolated by filtration on a 16-in diameter
filter nutsche. The reactor and cake were rinsed with water (13.9
kg). The solid was dried on the filter under a nitrogen stream. The
water content in the filter cake was followed by KF titration to
<0.5%. Yield: 2.727 kgof4, 86.6 wt % assay, (83.1% assay-corrected
yield), 99.7 A% HPLC purity [220 nm]. Two more similar sized
batches were run that gave 2.85 kg (80.1% assay-corrected yield,
99.4 A% HPLC), and 3.27 kg (85.1% assay-corrected yield, 99.6 A
% HPLC) crude product, respectively. The three crude lots were
combined and recrystallized from DMF/water (3:1) to give an
overall yield of 6.981 kg of 4 (81.4% overall assay corrected yield,
100.4 wt % by assay and 99.8 A% HPLC purity).
3-Benzyl-7-chloro-5-(4-nitro-phenyl)-3H-[1,2,3]triazolo[4,5-d]
pyrimidine (6). A solution of 5-amino-1-benzyl-1H-[1,2,3]
triazole-4-carbonitrile 516 (0.78 g, 3.9 mmol) and p-nitrobenzoyl
chloride (0.76 g, 4.1 mmol) in sulfolane (10 mL) was stirred at
80 ꢀC in a 10-mL septum vial until HPLC showed complete
acylation (∼2 h). Phosphorus pentachloride (1.6 g, 7.8 mmol) was
added, and temperature raised to 95 ꢀC. After HPLC showed
completion of the ring closure (∼2-3 h), the reaction mixture was
cooled and worked up as for 4 to afford crude product (1.8 g, assay
corrected yield: 76% over two steps). HPLC (Method B) 97.9 A%
(220 nm) Rt 10.64 min. 1H NMR (CDCl3, 400 MHz): δ 5.95 (s,
2H), 7.36 (dd, 2H), 7.36 (m, 1H), 7.52 (d, 2H), 8.37 (d, 2H, J =
8.8 Hz), 8.74 (d, 2H, J = 8.8 Hz). 13C NMR (CDCl3, 100 MHz): δ
51.6, 123.9, 128.6, 129.1, 129.5, 130.2, 133.3, 133.8, 141.1, 150.0,
150.7, 154.3, 160.4. HR-MS: calculated for C17H11ClN6O2:
366.063; found: 366.071.
’ ACKNOWLEDGMENT
We are grateful to Dr. Mahendra Suryan for LC-MS assis-
tance and exact mass determinations. Dr. Nathan Ide is thanked
for valuable discussions and Drs. John Ragan and Rajappa
Vaidyanathan for proof reading.
’ DEDICATION
Dedicated to Dr. Michael Kolb, former Vice President Wyeth
Research, Pearl River, NY, on the occasion of his 65th birthday.
’ REFERENCES
(1) Ceide, S. C.; Montalban, A. G. Tetrahedron Lett. 2006, 47, 4415.
(2) Evans, R. F.; Savage, G. P.; Gough, D. A. Aust. J. Chem. 1990,
43, 733.
(3) Li, C.; Rosenau, A. Tetrahedron Lett. 2009, 50, 5888.
(4) Matsumoto, K.; Minatogawa, H.; Toda, M.; Munakata, M.
Heterocycles 1990, 31, 1217.
(5) For reviews, see: (a) Brown, D. J. Rev. Pure Appl. Chem. 1953,
3, 115. (b) Jones, G.; Stanforth, S. P. Org. React. 1997, 49. (c) Meier,
M. S.; Ruder, S. M. e-EROS Encyclopedia of Reagents for Organic Synthesis;
Wiley InterScience: New York, 2001 (phosphorus oxychloride). (d)
Microwave-assisted chlorination of hydroxyheteroaromatics using phos-
phonium salts has also been reported:Tanji, K.-I.; Koshio, J.; Sugimoto,
O. Synthesis 2005, 35, 1983. (e) Takahashi, T.; Sugimoto, O.; Koshio, J.;
Tanji, K.-I. Heterocycles 2006, 68, 1973. (f) For a review, see:Undheim,
K.; Benneche, T. Comprehensive Heterocyclic Chemistry II: A review of the
literature 1982-1995: The structure, reactions, synthesis, and uses of
heterocyclic compounds; Katritzky, A. R., Rees, C. W., Scriven, E. F. V.,
Eds.; Pergamon: Oxford, New York, 1996; Vol. 6, pp 93 and 1177.
(6) Richard, D. J.; Verheijen, J. C.; Curran, K.; Kaplan, J.; Toral-
Barza, L.; Hollander, I.; Lucas, J.; Yu, K.; Zask, A. Bioorg. Med. Chem. Lett.
2009, 19, 6830.
(7) (a) Cheng, C. C.; Robins, R. K. J. Org. Chem. 1958, 23, 191. (b)
Schmidt, P.; Eichenberger, K.; Wilhelm, M. Helv. Chim. Acta 1962,
45, 1620. (c) Burch, H. A. J. Med. Chem. 1967, 11, 79.
(8) (a) Breuer, H.; Schulze, E.; Treuner, U. Ger. Pat. Appl. DE
2136950A1, 1972. (b) Breuer, H.; Schulze, E. Ger. Pat. Appl. DE
2140986A1, 1972. (c) Breuer, H.; Treuner, U. D. Ger. Pat. Appl. DE
2408906A1, 1974.
923
dx.doi.org/10.1021/op1002352 |Org. Process Res. Dev. 2011, 15, 918–924

Organic Process Research & Development
ARTICLE
(9) (a) Higashino, T.; Iwai, Y.; Hayashi, E. Chem. Pharm. Bull. 1976,
24, 3120. (b) Nguyen-Ba, N.; Quimpere, M.; Kong, L. C. C.; Brown,
W. L.; Dionne, G. U.S. Pat. 5,789,394, 1998. (c) Dugar, S.; Chakravarty, S.;
Conte, A.; Axon, J.; Mcenroe, G. Int. Pat. Appl. WO 04087056A2, 2004.
(10) (a) Jones, R. M.; Semple, G.; Xiong, Y.; Shin, Y.-J.; Ren, A. S.;
Calderon, I.; Fioravanti, B.; Choi, J. S. K.; Sage, C. R. Int. Pat. Appl. WO
007658A2, 2005. (b) Simmen, K. A.; Lin, T.-I.; Lenz, O.; Surleraux, D. L.
N. G.; Raboisson, P. J.-M. B. Int. Pat. Appl. WO 06035061A1, 2006.
(11) Zask, A.; Nowak, P. W.; Verheijen, J.; Curran, K. J.; Kaplan, J.;
Malwitz, D.; Bursavich, M. G.; Cole, D. C.; Ayral-Kaloustian, S.; Yu, K.;
Richard, D. J. Int. Pat. Appl. WO 08115974A2, 2008.
(12) To the best of our knowledge, this “inversed reactivity” chlo-
roimidate cyclization reaction has not been reported before (SciFinder,
Beilstein Searches, June 2010); all literature references for the overall
transformation consist of multistep nitrile hydrolysis/cyclization/chlo-
rination sequences (see refs 6-11]). Chloroimidates are highly reactive
C-electrophiles and typically serve as acid chloride surrogates (cf.
Vilsmeyer reagents/Vilsmeyer reaction); for reviews, see:(a) Bonnett,
R. Chemistry of the Carbon-Nitrogen Double Bond; Patai, S., Ed.; Inter-
science Publishers: London, New York, 1970; p 597. (b) Nakajima, N.;
Ubukata, M. Sci. Synth. 2005, 22, 331. Following an extensive literature
search, to our surprise, only a single example of a related ring closure
could be found, albeit for a phosphaheteroaromaticSchmidpeter, A.;
Schindler, N. Chem. Ber. 1969, 102, 2201. (Cyclization of a p-chloro-
N-(2-cyanophenyl)-phosphinimine hydrochloride salt to a 2-phospha-
quinazoline system). At this point, we have no evidence to exclude the
possibility of a 6π-electrocyclic ring closure. An earlier attempt to achieve
this valuable transformation via intramolecular transfer of HCl from an
aromatic amidinium hydrochloride salt to a neighboring nitrile group and
subsequent cyclization to the 4-chloroquinazoline met with only limited
success (25% yield after chromatogr. purification of the product
mixture): Nagarathnam, D.; Asgari, D.; Shao, J.; Liu, X.-G.; Khire, U.;
Wang, C.; Hart, B.; Boyer, S.; Weber, O.; Lynch, M.; Bankston, D. Int.
Pat. Appl. WO 02076976A2, 2002.
(13) See http://www.prosense.net/files/ARSST_4pages.pdf and
http://www.fauske.com/Download%5Cchemical%5Carsst%5CJPBArticle.
pdf.
(14) Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43, 2923.
(15) Made via reaction of 4-amino-5-cyanopyrimidine with benzoyl
chloride in pyridine and flash chromoatography of the crude product;
HPLC (Method A): Rt 2.14 min (96.7 A% [210 nm]), HR-MS:
calculated for C12H8N4O: 224.0698; found: 224.0683.
(16) Hoover, J. R. E.; Day, A. R. J. Am. Chem. Soc. 1956, 78, 5832.
(17) Sangapure, S. S.; Agasimundin, Y. S. Indian J. Chem. 1976,
14B, 688.
924
dx.doi.org/10.1021/op1002352 |Org. Process Res. Dev. 2011, 15, 918–924