Development of a scaleable synthesis of NDT 9533750, a key intermediate to a series of novel subtype preferring GABAA partial agonists

Article abstract of DOI:10.1021/op700084h

A scaleable route to 6-chloro-4-[2-(3-fluoropyridin-2-yl)-imidazol-1- ylmethyl]-5-propyl-pyrimidine (NDT 9S33750), a key intermediate to a series of novel subtype preferring GABA_A partial agonists, is described in which various scaleup issues were addressed to provide an efficient and robust route for the preparation of kilogram quantities of the compound.

Full text of DOI:10.1021/op700084h

Organic Process Research & Development 2007, 11, 716−720
Development of a Scaleable Synthesis of NDT 9533750, a Key Intermediate to a
Series of Novel Subtype Preferring GABA Partial Agonists
A
†
†
†
‡
†
†
Yuelian Xu, Bingsong Han, Zhe-Qing Wang, Kenneth R. Shaw, Bertrand L. Chenard, George Maynard, and
Linghong Xie*,†
Department of Chemistry, Neurogen Corporation, Branford, Connecticut 06405, U.S.A.
Abstract:
Scheme 1
A scaleable route to 6-chloro-4-[2-(3-fluoropyridin-2-yl)-imi-
dazol-1-ylmethyl]-5-propyl-pyrimidine (NDT 9533750), a key
intermediate to a series of novel subtype preferring GABA
A
partial agonists, is described in which various scaleup issues
were addressed to provide an efficient and robust route for the
preparation of kilogram quantities of the compound.
Introduction
GABA (γ-aminobutyric acid) is the major inhibitory
neurotransmitter in the mammalian central nervous system,
binding to three different receptor types, GABA
A B
, GABA ,
Scheme 2
and GABA . In addition to binding GABA, each receptor
C
possesses a number of allosteric sites that bind a variety of
ligands, leading to a modulation of the action of GABA.
The Bz (benzodiazepine) binding site is found in GABA
A
receptors containing â and γ2 subunits in conjuction with
an R1, R2, R3, or R5 (but not R4 or R6) subunit. Bz site full
agonists which nonselectively interact at R1-, R2-, R3-, and
R5-containing subtypes enhance the inhibitory effects of
GABA and display sedative hypnotic, anaesthetic, muscle-
relaxant, anticonvulsant, and anxiolytic activities, but they
are also associated with side effects such as amnesia,
tolerance, dependence, and alcohol potentiation.1
,2
In an effort to discover and develop novel subtype
preferring GABA partial agonists for the treatment of sleep
A
disorders with less side effects, an efficient large-scale
synthesis of a key intermediate 6-chloro-4-[2-(3-fluoropy-
ridin-2-yl)-imidazol-1-ylmethyl]-5-propyl-pyrimidine (NDT
3
Results and Discussion
9533750) (Scheme 1) was required. NDT 9533750 was
constructed from the two key intermediates 5-propyl-6-
halomethyl-4-chloropyrimidine (1) and 3-fluoro-2-(1H-imi-
dazol-2-yl)-pyridine (2). The preparation of these two key
intermediates and their coupling posed several synthetic
challenges for a large-scale synthesis. This paper describes
how these challenges were overcome to provide a robust and
efficient large-scale synthesis of NDT 9533750.
Preparation of 5-Propyl-6-halomethyl-4-chloropyri-
midine (1). The initial small scale synthesis of NDT 9533750
involved 5-propyl-6-bromomethyl-4-chloropyrimidine (1a)
(Scheme 2) which was prepared from 5-propyl-6-methyl-4-
hydroxypyrimidine (3) in ca. 40% yield by chlorination and
subsequent regioselective bromination of the intermediate
5
-propyl-6-methyl-4-chloropyrimidine (4) with 1 equiv of
bromine in HOAc at 85 °C. While conversion of 4-hydroxy-
pyrimidine 3 to 4-chloropyrimidine 4 with POCl was a
*
To whom correspondence should be addressed. Telephone: 203 488 8201.
3
Fax: 203 481 5290. E-mail: lxie@nrgn.com.
†
straightforward procedure, reaction of 4 with bromine in
HOAc was not amenable to scaleup. Careful chromatographic
purification was required to separate 1a from three major
impurities: dibromomethylpyrimidine 5 (∼10%), chloride
to bromide exchanged byproduct 6 (∼40%), and starting
material 4 (∼10%). Attempts to drive the reaction to
Neurogen Corporation.
Marinus Pharmaceuticals.
‡
(1) Tallman, J. F.; Cassella, J. V.; White, G.; Gallager, D. W. Neuroscientist
1
999, 5 (6), 351-361.
(
(
2) Mehta, A. K.; Ticku, M. K. Brain Res. ReV. 1999, 29, 196-217.
3) (a) Xie, L.; Han, B.; Xu, Y.; Maynard, G.; Chenard, B.; Shaw, K.; Gao, Y.
Int. Patent. Appl. WO /2005012306, 2005. (b) Xie, L.; Han, B.; Xu, Y.;
Maynard, G. Int. Patent. Appl. WO /2004041808, 2004.
7
16
•
Vol. 11, No. 4, 2007 / Organic Process Research & Development
10.1021/op700084h CCC: $37.00 © 2007 American Chemical Society
Published on Web 06/27/2007

Figure 1. Composition of the reaction mixture of methylpyrimidine 4 with TCC (1.5 mol equiv) in MTBE at 40 °C as a function
of time based on LC-MS.
Table 1. Solvent, temperature, and time screening reactions
completion by using an excess amount of bromine or higher
of methylpyrimidine 4 with TCC
temperatures led to the predominant formation of dibro-
momethylpyrimidine 5.
entry
solvent
temperature time
1b
7
4
To avoid the halogen-halogen exchange side reaction,
optimization work was focused on the preparation of the
corresponding chloride, 5-propyl-6-chloromethyl-4-chloro-
pyrimidine (1b). The inexpensive chlorinating agent tri-
chloroisocyanuric acid (TCC), a component of some indus-
trial deoderants and household cleaning products under the
1
2
3
4
5
6
7
8
9
heptane
heptane
heptane
t-BuOMe
t-BuOMe
AcOH
toluene
CHCl3
DCM
60 °C
80 °C
reflux
40 °C
60 °C
60 °C
60 °C
60 °C
rt
10 h 30%
18 h 90%
0% 70%
7%
3%
0%
2%
0%
0%
0%
0%
2%
2%
2%
5 h 90% 10%
8 h 82% 16%
5 h 70% 30%
5 h 60% 40%
10 h 88% 10%
10 h 90% 10%
6 d 96%
24 h 96%
15 h 96%
4
trade name Chloreal, was reported to chlorinate N-hetero-
cyclic compounds such as 2-methylpyridine, 2-methylquino-
line, and 2-methylquinoxaline to the corresponding chlo-
2%
2%
2%
1
1
0
1
DCM
DCM
30 °C
reflux
5
romethyl derivatives in good yields. In our system, initial
exploration indicated that TCC was clearly superior to other
chlorination reagents such as Cl , 1,3-dichloro-5,5-dimeth-
2
ylhydantoin, and NCS in terms of reactivity and selectivity.
Thus, TCC was chosen as the chlorination reagent for further
investigation.
We found that 1.5 mol equiv of TCC (4.5 equiv of
chlorine) were required for good conversion of methylpy-
rimidine 4 to chloromethylpyrimidine 1b. It seemed that only
the first chlorine of TCC was consumed and the resulting
dichloroisocyanuric acid was not as reactive as TCC. At the
early stages of the reaction only chloromethylpyrimidine 1b
was formed. However, overchlorination to dichlorometh-
ylpyrimidine 7 started before methylpyrimidine 4 was
completely consumed as indicated by the initial evaluation
in methyl tert-butyl ether (MTBE) at 40 °C (Figure 1). Thus,
efforts were focused on finding optimal reaction conditions
to enhance the yield and purity of 1b.
(
∼ 96%). The reaction time course of methylpyrimidine 4
with TCC in DCM at 30 °C was illustrated in Figure 2.
Our kiloscale reaction was carried out in DCM at 30 °C
for 24 h. The reaction was stopped by cooling to 5 °C and
subsequent filtration to remove excess TCC and the resultant
6
dichloroisocyanuric acid. The filtrate was washed with
aqueous sodium bisulfite solution to remove residual chlo-
rinating agents. Chloromethylpyrimidine 1b was obtained
upon removal of solvent in high yield (>96%) and purity
(
96-98%).
Preparation of 3-Fluoro-2-(1H-imidazol-2-yl)-pyridine
(2). The original discovery research preparation of 2 (Scheme
7
3
3
) involved regioselective C-2 lithiation and formylation of
-fluoropyridine 8 to generate pyridyl-2-carboxaldehyde 9,
which was condensed with glyoxal and ammonium hydrox-
ide to afford 2 in ca. 28% overall yield. The need for low
temperatures (-76 °C to -50 °C) and column chromatog-
raphy to purify the aldehyde 9 in the first step presented
challenges for large-scale operations. A more robust and
practical method to 3-fluoro-2-(1H-imidazol-2-yl)-pyridine
An array of parallel reactions between methylpyrimidine
4
and TCC (with a molar ratio of 1 to 1.5) was undertaken
to evaluate various solvents, temperatures, and reaction times.
Reactions were monitored by HPLC (Table 1). The best
results were observed with dichloromethane (DCM) as the
solvent (Table 1, entries 8-10) providing the highest ratio
of monochloride 1b/dichloride 7 (∼ 96:2) and highest yield
(2) was needed.
Since we could access bulk quantities of 3-fluoropyridine-
-carbonitrile (10) from commercial sources, we turned our
2
(
6) TCC dust is extremely destructive to eyes, skin, and lungs. Operations should
be carried out in a fume hood. Filtered residual chlorinating agent was
destroyed with aq. bisulphate solution.
(4) Stecher P. G. The Merck Index, 8th ed.; Merck & Co.: 1968.
(
5) Jeromin, G. E.; Orth, W.; Rapp, B.; Weiss, W. Chem. Ber. 1987, 120, 649-
6
51.
(7) Marsaia, F.; Gueguiner, G. Tetrahedron 1983, 39, 2009.
Vol. 11, No. 4, 2007 / Organic Process Research & Development
•
717

Figure 2. Composition of the reaction mixture of methylpyrimidine 4 with TCC (1.5 mol equiv) in DCM at 30 °C as a function
of time based on LC-MS.
Scheme 3
Scheme 4
attention to the classic imidazole formation method reported
by Lawson from imidates and R-amino acetals. It is worth
noting that conversion of nitriles directly to imidines with a
Our attempts to convert 3-fluoropyridine-2-carbonitrile (10)
to the corresponding imidate with gaseous HCl in methanol
failed due to lack of reactivity, possibly as a result of the
basicity of the pyridine nucleus.
8
9
9
10
promoting agent such as AlCl
3
, ZnCl
2
, aluminium amides,
11
12
lanthanide(III) triflates, or Cu(I) has been reported in the
literature. More recently, Frutos et al. have reported an
expedient imidazole synthesis by the direct CuCl-mediated
reaction of nitriles with R-amino acetals followed by
cyclization, avoiding the formation of imidates. However,
attempts to utilize these methods in our system were
fruitless: nucleophlic displacement of fluorine competed
significantly with the nucleophilic addition to the nitrile when
The most commonly used alkoxide in the literature for
imidate formation is sodium methoxide. Treatment of 10 with
1 equiv of methanolic sodium methoxide at ambient tem-
perature followed by refluxing with 2,2-dimethoxyethylamine
and acetic acid gave imidine 12, which, upon acidification
with 5 M HCl and heating, was converted to imidazole 2
(Scheme 4). This three-step one-pot procedure afforded
3-fluoro-2-(1H-imidazol-2-yl)-pyridine (2) in 40% overall
yield along with ca. 30% of 3-methoxypyridine-2-carbonitrile
(13) as byproduct, formed by methoxide displacement of
fluorine.
In order to minimize the formation of 13 and improve
the yield of imidazole 2, various metallic methoxides (sodium
methoxide, lithium methoxide, and magnesium methoxide)
and different reaction temperatures were evaluated. Two
optimal conditions were obtained: magnesium methoxide
at ambient temperature and sodium methoxide at -20 °C.
Both conditions provided the highest ratio of imidazole 2 to
13
1
0 was treated with 2,2-dimethoxyethylamine under these
conditions.
Nitriles were reported to be converted to imidates by either
8
14
gaseous HCl in alcoholic solvents or alcoholic alkoxide.
(8) Lawson, A. J. Chem. Soc. 1957, 4225.
(9) Oxley, P.; Partridge, M. W.; Short, W. F. J. Chem. Soc. 1947, 1110.
(10) Garigipati, R. S. Tetrahedron Lett. 1990, 31, 1969.
(
11) Forsberg, J. H.; Spaziano, V. T.; Balasubramanian, T. M.; Liu, G. K.;
Kinsley, S. A.; Duckworth, C. A.; Poteruca, J. J.; Browm, P. S.; Miller, J.
L. J. Org. Chem. 1987, 52, 1017.
(
(
12) Rousselet, G.; Capdevielle, P.; Maumy, M. Tetrahedron Lett. 1993, 34, 6395.
13) Frutos, R. P.; Gallou, I.; Reeves, D.; Xu, Y.; Krishnamurthy, D.; Senanayake,
C. H. Tetrahedron Lett. 2005, 46, 8369.
3
-methoxypyridine-2-carbonitrile (13) (70:15). Due to a
(14) Clews, J.; Morgan, N. G.; Ramsden, C. A. Synthesis 2001, 10, 1546.
supply issue with obtaining a large quantity of magnesium
718
•
Vol. 11, No. 4, 2007 / Organic Process Research & Development

Scheme 5
to aqueous basic conditions and decomposed slowly itself.
Thus, a set of experiments were carried out to determine
the optimal ratio of chloromethylpyrimidine 1b to imidazole
2
. The best yield was obtained when 1.2 equiv of 1b were
used. Water was added to precipitate the product, which was
isolated in quantitative yield. Crude NDT 9533750 was
recrystallized from i-PrOAc and TBME and isolated as a
white solid in 86% yield (>99% purity).
methoxide at the time, the kilogram scale reaction was
conducted with sodium methoxide.
Conclusions
3
-Fluoropyridine-2-carbonitrile (10) was dissolved in
methanol and treated with 1 equiv of sodium methoxide at
20 °C, resulting in the formation of imidate 11 and
-methoxypyridine-2-carbonitrile (13) (Scheme 4). Fortu-
This paper describes the preparation of 6-chloro-4-[2-(3-
fluoropyridin-2-yl)-imidazol-1-ylmethyl]-5-propyl-pyrimi-
dine (NDT 9533750), a key intermediate to a series of novel
-
3
A
subtype preferring GABA partial agonists, in a robust and
nately, only the fluorine of the nitrile 10 was prone to
replacement, but not the fluorine of the imidate 11. Further-
more, 13 did not react further with methoxide to form its
corresponding imidate under the reaction conditions. After
efficient manner suitable for kilogram scale. The original
medicinal chemistry synthesis was modified to eliminate
column chromatography and low-temperature reactions.
Productivity and yields were greatly improved. This, together
with the optimization of reaction conditions allowed NDT
4
h at -20 °C equilibrium was reached consisting of a
mixture of imidate 11 (70%), methoxypyridine 13 (15%),
and starting material 10 (15%). Thus, it was important to
wait a full 4 h before adding 2,2-dimethoxyethylamine and
acetic acid. Upon heating this mixture at 60 °C for 30 min
the intermediate imidine 12 was formed.
9
533750 to be prepared in an overall yield of ca. 75% from
hydroxypyrimidine 3.
Experimental Section
Imidazole ring closure was realized by acidification with
M HCl followed by heating. Besides reducing the overall
yield of product, formation of 13 created a secondary
problem, by liberating fluoride. When HCl was added,
corrosive HF formed and severely etched the glass reaction
All reagents and solvents were commercially available.
1H and C NMR spectra were recorded with a Varian-300
(300 MHz) spectrometer using tetramethylsilane as an
internal standard. LC-MS analysis was performed using a
Micromass Time-of Flight LCT, equipped with a Waters 600
pump, Waters 996 photodiode array detector, Gilson 215
autosampler, and Gilson 841 microinjector. Elemental analy-
ses were carried out by Robertson Microlit Laboratories, Inc.
5-Propyl-4-chloro-6-methyl-pyrimidine (4). 5-Propyl-
13
5
vessel. This problem was resolved by adding CaCl
2
(10 mol
%
based on 10), a commonly used reagent for fluoride
15,16
removal,
to sequester free fluoride before acidification
with HCl.
1
7
After the cyclization was complete, methanol was re-
moved by distillation. After cooling, 30% NaOH was added
until pH > 12 was achieved. Adjusting the pH to greater
than 12 was critical to dissolving all the inorganics, especially
6-methyl-4-hydroxypyrimidine (3, 1000 g, 5.58 mol) and
toluene (5.5 L) were charged to a nitrogen-purged reactor
and heated to 100 °C with mechanical stirring. POCl (899
3
g, 5.86 mol) was added at an internal temperature of 100 °C
CaF
subsequent organic solvent extraction. The product 3-fluoro-
-(1H-imidazol-2-yl)-pyridine (2) is not very soluble in many
organic solvents and has significant aqueous solubility.
Consequently, many solvents were screened, and it was found
that n-butanol extraction gave very good separation from
aqueous solution and afforded the best recovery. A single
recrystallization of the crude 2 from acetonitrile gave the
product as a yellow solid in 52% yield over the three-step,
one-pot reaction, with >99% purity.
Preparation of 6-Chloro-4-[2-(3-fluoropyridin-2-yl)-
imidazol-1-ylmethyl]-5-propyl-pyrimidine (NDT 9533750).
Final Coupling Reaction. The final nucleophilic displace-
ment reaction between 3-fluoro-2-(1H-imidazol-2-yl)-pyri-
dine (2) and chloromethylpyrimidine 1b (Scheme 5) was
carried out in N,N-dimethylacetamide (DMA). Sodium
hydroxide was chosen as the base due to the much shorter
reaction time required for complete conversion compared to
potassium carbonate (30 min vs 24 h) although both bases
provided similar yields. It was observed that 1b was sensitive
2
. This step allowed a clean phase separation in the
within 30 min. The reaction mixture was stirred at 100 °C
for 4 h. The solution was cooled to 10 °C. Water (1.5 L)
was added to the solution over a period of 30 min. The
mixture was treated with 30% of NaOH (2.5 L) over a period
of 75 min while maintaining the temperature below 50 °C.
The pH of the solution was around 10. The layers were
separated, and the aqueous layer was extracted with toluene
(2 L). The combined extracts were washed with water (1
L). The organic layer was filtered over a Buchner funnel,
which was charged with silica gel (500 g). The silica plug
was washed with toluene (2 L). The combined organic
solution was concentrated under reduced pressure to dryness
2
1
to afford 4 as a light yellow oil (834 g, 87.5%). H NMR
(CDCl ): δ 1.04 (3H, t, J ) 7.5 Hz), 1.60 (2H, m), 2.57
3
(3H, s), 2.73 (2H, t, J ) 7.5 Hz), 8.68 (1H, s). 13C NMR
(CDCl ): δ 14.23, 21.49, 22.42, 31.08, 132.42, 155.39,
3
160.81, 167.22. Mass (e/z): 171.14 (M + 1). Anal. Calcd
for C H ClN : C, 56.31; H, 6.50; N, 16.42. Found: C,
8
11
2
56.57; H, 6.51; N, 16.48.
(
17) 5-Propyl-6-methyl-4-hydroxypyrimidine (3) was purchased from Fontarome
Chemical. The material contained ca. 17% inorganic salt which had no
influence on the outcome of the chlorination reaction.
(15) Huang, G.; Zou, L.; Lu, D. Huagong Shikan 1999, 13 (7), 20.
(16) Singh, M.; Kanvinde, V. Y. AdV. Chem. Eng. Nucl. Process Ind. 1994, 289.
Vol. 11, No. 4, 2007 / Organic Process Research & Development
•
719

5
-Propyl-6-chloromethyl-4-chloropyrimidine (1b). To
washed with t-BuOMe (1 L × 2), and dried under vacuum.
a mixture of 5-propyl-4-chloro-6-methyl-pyrimidine (4, 987
g, 5.784 mol) and DCM (5.9 L) was added TCC (2016 g,
The product (2, 696 g) was obtained as a yellow solid (52%).
1
3
H NMR (CDCl ): δ 7.21-7.26 (3H, m), 7.51 (1H, td, J )
8
.67 mol) in portions over a period of 30 min. The reaction
8.4, 1.2 Hz), 8.33 (1H, dt, J ) 4.5, 1.5 Hz), 11.00 (1H, br).
C NMR (CDCl ): δ 124.30 (d, J ) 4.6 Hz), 125.15 (d, J
3
) 19.8 Hz), 137.12 (d, J ) 9.9 Hz), 143.16 (d, J ) 9.9 Hz),
144,90, 144.95, 155.30, 157.95. Mass (e/z): 164.17 (M +
1
3
mixture was heated at 30 °C for 23 h. The reaction mixture
was cooled to 5 °C and filtered. The solid cake was washed
with DCM (1.5 L). Saturated Na SO (aq.) solution was
2 3
added until the pH reached ∼8 and the mixture was then
stirred for 1 h. The solution was filtered through Celite, and
the layers were separated. The organic layer was washed
with water (1.5 L) and concentrated to dryness to afford
product 1b (1189 g, 96-98% purity with 4-chloro-6-
dichloromethyl-5-propylpyrimidine as the major impurity),
8 6 3
1). Anal. Calcd for C H FN : C, 58.89; H, 3.71; N, 25.76.
Found: C, 58.82; H, 3.66; N, 25.80.
4-Chloro-6-[2-(3-fluoropyridin-2-yl)-imidazol-1-yl-
methyl]-5-propyl-pyrimidine (NDT 9533750). 3-Fluoro-
2-(1H-imidazol-2-yl)-pyridine (2, 742 g, 4.548 mol) was
charged in a flask under nitrogen, followed by the addition
of DMA (2.8 L) and 3 N NaOH solution (1.744 L, 5.23 mol).
Upon cooling to 10-15 °C, 5-propyl-6-chloromethyl-4-
chloropyrimidine (1b, 1120 g, 5.458 mol) was added
streamwise to the mixture in ca. 30 min to give a brownish
slurry, which was stirred at room temperature for an
additional 30 min. Ice water (2.5 kg) was added, and the
resulting mixture was stirred for another 30 min and filtered.
The collected solid was rinsed with water (300 mL × 3)
and dried in a hood overnight to give the crude product (NDT
9533750, 1.6 kg containing ∼6% water) in 99% yield,
>95.9% HPLC purity.
which was used for the next step reaction without further
1
purification. H NMR (CDCl
3
): δ 1.00 (3H, t, J ) 7.5 Hz),
1
(
1
.61 (2H, m), 2.75 (2H, t, J ) 7.5 Hz), 4.57 (2H, s), 8.74
1H, s). 13C NMR (CDCl
): δ 14.47, 22.47, 30.81, 43.48,
33.41, 156.18, 162.57, 163.99.
-Fluoro-2-(1H-imidazol-2-yl)-pyridine (2). 2-Cyano-
-fluoropyridine (9, 1000 g, 8.196 mol) was charged into a
3
3
3
three-neck flask under nitrogen at room temperature, fol-
lowed by the addition of MeOH (5.0 L). The solution was
cooled to ca. -20 °C. A solution of NaOMe (25%, 1772 g,
8.196 mol) in MeOH was added Via an additional funnel
within 30 min, while maintaining the internal temperature
at ca. -20 °C. The mixture was stirred at -20 °C for an
additional 4 h.
Aminoacetaldehyde dimethylacetal (775 g, 7.376 mol)
was added to the suspension at -20 °C within 5 min,
followed by the addition of acetic acid (978 g, 16.3 mol)
over a period of 5 min. The resulting mixture was heated at
Recrystallization. Crude NDT 9533750 (2.35 kg, con-
taining ∼6% water, from 2 lots) was mixed with i-PrOAc
(4.0 L) and t-BuOMe (4.0 L) under nitrogen. The mixture
was slowly heated to a gentle reflux. After stirring at the
same temperature (about 55-58 °C) for 1 h, the mixture
was cooled to 30 °C and filtered. The collected solid was
rinsed with i-PrOAc (150 mL × 3) and dried in air overnight
6
5 °C for 30 min and was then cooled to room temperature
and stirred overnight.
CaCl (90 g, 0. 818 mol) was added to the solution,
to give NDT 9533750 (1.89 kg) as a white solid in 86%
1
yield (>99% purity). H NMR (CDCl
3
): δ 0.99 (3H, t, J )
2
7.5 Hz), 1.56 (2H, m), 2.75 (2H, t, J ) 7.5 Hz), 5.86 (2H,
s), 7.11 (1H, d, J ) 1.2 Hz), 7.21-7.26 (1H, m), 7.31 (1H,
d, J ) 1.2 Hz), 7.52 (1H, td, J ) 8.4, 1.2 Hz), 8.23 (1H, dt,
followed by the addition of 5 N HCl (6 L). The suspension
was heated at 65 °C for 4 h. MeOH (∼4 L) was removed at
reduced pressure. The residue was cooled to room temper-
ature. Water (4 L) was added, followed by the addition of
1
3
J ) 4.5, 1.5 Hz), 8.64 (1H, s). C NMR (CDCl ): δ 14.43,
3
21.79, 30.28, 50.29, 123.96, 124.37 (d, J ) 4.6 Hz), 125.11
(d, J ) 19.8 Hz), 129.44, 131.72, 138.72 (d, J ) 9.2 Hz),
140.95 (d, J ) 8.4 Hz), 143.94 (d, J ) 5.3 Hz), 156.12,
157.66 (d, J ) 267.8 Hz), 161.65, 164.37. Mass (e/z): 332.07
3
0% NaOH solution to pH ca. 12 upon cooling with an ice-
water bath.
To the suspension was added n-BuOH (5 L). The layers
were separated. The aqueous layer was extracted with
n-BuOH (2 L). The organic layers were combined and
concentrated to dryness at reduced pressure. To the residue
was added MeCN (2.5 L). The suspension was heated to 80
5
(M + 1). Anal. Calcd for C16H15ClFN : C, 57.92; H, 4.56;
N, 21.11. Found: C, 57.81; H, 4.64; N, 21.27.
Received for review April 13, 2007.
OP700084H
°
C and filtered while hot. The filtrate was cooled to 0 °C to
allow crystallization. The product was collected by filtration,
720
•
Vol. 11, No. 4, 2007 / Organic Process Research & Development