A practical synthesis of an aminopyridine as a component of a BTK inhibitor

Article abstract of DOI:10.1021/op400046y

Development of a new route to the BTK intermediate, 5-(1-(azetidin-1-yl)-2- methylpropan-2-yloxy)pyridin-2-amine (7), has resulted in significant improvements in terms of yield, purity, and operability over the previously known synthesis. The new route was demonstrated on a multihundred-gram scale, and the overall yield of 7 from 2-chloro-5-hydroxypyridine (1) was improved to 72%. Lithium aluminum hydride, a Swern oxidation, cesium carbonate, azetidine free base, and four chromatographic purifications used in the earlier synthesis were eliminated.

Full text of DOI:10.1021/op400046y

Technical Note
pubs.acs.org/OPRD
A Practical Synthesis of an Aminopyridine as a Component of a BTK
Inhibitor
Lady Mae Alabanza, Yan Dong, Ping Wang, James A. Wright, Yingchao Zhang, and Andrew J. Briggs*
Process Research and Synthesis, Hoffmann-La Roche, Inc., 340 Kingsland Street, Nutley, New Jersey 07110, United States
ABSTRACT: Development of a new route to the BTK intermediate, 5-(1-(azetidin-1-yl)-2-methylpropan-2-yloxy)pyridin-2-
amine (7), has resulted in significant improvements in terms of yield, purity, and operability over the previously known synthesis.
The new route was demonstrated on a multihundred-gram scale, and the overall yield of 7 from 2-chloro-5-hydroxypyridine (1)
was improved to 72%. Lithium aluminum hydride, a Swern oxidation, cesium carbonate, azetidine free base, and four
chromatographic purifications used in the earlier synthesis were eliminated.
INTRODUCTION
(odor, cryogenic temperatures), alkylation with cesium carbo-
nate (cost, hygroscopicity), and reductive amination using
azetidine free base (cost). In addition, chromatographic
purifications of compounds 3, 5, 6, and 7 were required to
control the purity. Consequently, a new and improved route to 7
was developed that addressed the scalability issues associated
with the initial route. The new process was demonstrated on a
multihundred-gram scale to provide 7 in 72% overall yield. The
lithium aluminum hydride reduction, Swern oxidation, cesium
carbonate alkylation, azetidine free base, and all of the
chromatographic purifications were eliminated. Minimization
of a side product in a borane reduction, as well as optimization of
a Buchwald−Hartwig animation, is discussed.
■
Bruton’s tyrosine kinase (BTK) plays a key role in multiple cell-
signaling pathways, and small-molecule BTK inhibitors are
currently being pursued for the treatment of autoimmune
1
diseases. Several BTK inhibitors were under investigation at
Roche for the treatment of rheumatoid arthritis in which
aminopyridine 7 is a common component.
Scheme 1 depicts the route developed by colleagues in
Discovery Chemistry to prepare gram quantities of amino-
2
2
Scheme 1. Discovery route for preparing 7
RESULTS AND DISCUSSION
■
The first approach investigated was based on a nucleophilic
substitution of 5-chloro-2-nitropyridine with 1-(azetidin-1-yl)-2-
methylpropan-2-ol. However, instead of displacement of the
chlorine atom, the reaction occurred with loss of the nitro group.
A second approach using a Weinreb amide was then studied to
circumvent the undesirable reduction−oxidation sequence to
generate aldehyde 5. The corresponding Weinreb amide of ester
2
was prepared, but the nucleophilic substitution with
hydroxypyridine 1 was not clean under several conditions tried.
The successful, large-scale route to 7 (Scheme 2) proceeded
through hydrolysis of ester 2, amide formation, and reduction to
amine 6, replacing the original sequence of lithium aluminum
hydride reduction, Swern oxidation, and reductive amination.
The alkylation of 2-chloro-5-hydroxypyridine (1) was first
optimized. Screening experiments showed that potassium
carbonate in DMF at 85 °C could replace cesium carbonate in
acetonitrile at room temperature. However, switching to the less
3
expensive base required an adjustment in the stoichiometry of
the bromoester. A reaction run with 1.5 equiv of the bromide still
showed over 20% of 1, even after the addition of a further 0.2
equiv of the reagent. Fortunately, with 2 equiv the reaction could
pyridine 7 for initial screening. With the need to rapidly prepare
larger quantities of this key intermediate, the scalability of the
original route was evaluated. Several of the reagents and
conditions are undesirable for large scale, including the reduction
with lithium aluminum hydride, oxidation using a Swern protocol
4
be driven to ∼94% completion (Table 1, entry 2).
Received: February 20, 2013
©
XXXX American Chemical Society
A
dx.doi.org/10.1021/op400046y | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Technical Note
Scheme 2. Borane reduction route to 7
These issues were addressed by minimizing the excess of CDI
while maintaining a high level of conversion. This was
accomplished in two ways. First, the reaction rate was accelerated
by increasing the concentration from 0.063 to 0.17 g/mL.
Second, formation of the acylimidazole of 8 was monitored by
quenching an aliquot with n-butylamine. Initially, CDI was added
as a solid to a solution of 8 in dichloromethane. However, for ease
of operation on a large scale this was modified to adding the
solution of 8 to a slurry of CDI (1.1 equiv). Typically, after 2 h
the acylimidazole formation was complete. Then, azetidine
hydrochloride was added. This reagent has a significant cost
advantage over azetidine free base and was also available in larger
10
pack sizes. The acylimidazole can be added to a slurry of
azetidine hydrochloride in dichloromethane, but this mode of
addition resulted in considerably higher solvent volumes.
Addition of triethylamine and refluxing for 2 h provided amide
9
. While azetidine hydrochloride is hygroscopic, this reaction is
not sensitive towards moisture; material containing 8−13 wt % of
water can be used without negatively affecting the reaction.
Workup with 1 M sulfuric acid and water, solvent exchange to
heptanes, and filtration provided 9 as a white solid in 95% yield
and 98% purity.
11
12
The amide reduction of 9 was investigated using LiAlH , Red-
4
Al, and borane−THF complex, of which the last gave the best
result (Scheme 3). HPLC analysis of the reaction samples
Table 1. Alkyation of hydroxypyridine 1
a
a
entry
bromoester (equiv)
1 (%)
3 (%)
Scheme 3. Amide reduction
1
2
3
1.5 + 0.2
2.0
20.5
5.6
79.5
94.4
94.8
b
2.2
5.2
a
b
Normalized to 1 and 3. Slow addition of bromoester.
As the reaction is exothermic, in order to enable better
temperature control on a larger scale, a slow addition of the
bromide to a preheated mixture of 1 and potassium carbonate in
DMF was used. The reaction was scaled up using 2.2 equiv of
halide and 2.6 equiv of potassium carbonate. Workup with
heptane, MTBE, and water was followed by solvent exchange
into heptanes to obtain 3 as a white solid in 91% yield and 99%
5
purity.
Hydrolysis of ester 3 was carried out with lithium hydroxide in
6
aqueous THF. No displacement of the chlorine atom was
observed under these conditions. The hydrolysis required 1.5
equiv of base to reach completion in a reasonable time frame.
Acidification, extraction with MTBE, and solvent exchange into
heptanes provided 99% pure 8 in 99% yield.
showed an additional peak which was identified as 12 by LC/MS.
Treatment of these samples with a small amount of concentrated
hydrochloric acid conveniently converted 12 to 6. The major
byproduct from the reduction was hydroxypyridine 1. A
considerable amount of 1 (24 area % by HPLC) formed when
2.4 equiv of borane−THF complex was added to a THF solution
of 9 at room temperature and the mixture was allowed to heat to
44 °C (Table 2, entry 1). Although removal of 1 was not an issue,
there was an obvious impact on yield.
7
Although propane phosphonic acid anhydride (T3P) was
initially used for converting the acid to amide 9, carbon-
8
yldiimidazole (CDI) produced fewer side products by HPLC
analysis. Small-scale experiments gave 9 in over 90% isolated
yield, but upon initial scale-up to 18 g the yield fell to 80%.
Samples from these early experiments also showed several small
1
peaks in the H NMR which were attributed to urea 10 and
Table 2. Minimization of the formation of 1
9
imidazole 11 (Figure 1).
a
a
entry
BH −THF (equiv)
T (°C)
1 (%)
6 (%)
3
b
1
2.4
20−44
20−29
45−55
45
24
18
8
75
82
92
96
b
2
2.0 + 1.0
b
3
1.5 + 0.5 + 0.5
1.0 + 0.5 + 0.5 + 0.5
b
4
4
a
Determined by HPLC analysis of a reaction sample treated with
b
Figure 1.
hydrochloric acid. All reactions proceeded to <1% starting material.
B
dx.doi.org/10.1021/op400046y | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Technical Note
a
Adding the borane in two portions (Table 2, entry 2) led to a
reduction in the amount of the side product formed. Although
there is no direct evidence, it seemed likely that the amide 9 and
not the azetidine 6 was the source of 1. We reasoned that, by
using a slow addition and limiting the excess of borane present at
any given time, the level of 1 could be reduced. Such an approach
was expected to have a negative impact on the overall reaction
rate, but this could be compensated for by increasing the reaction
temperature. Making these changes led to a significant reduction
in the level of the production of 1 to 8% (Table 2, entry 3). The
final conditions (Table 2, entry 4), used for the larger batches,
employed four additions each over 3/4 to 1 h at approximately 1-
h intervals. Heating at 45 °C for an additional 6 h was required to
complete the reaction. After cooling to room temperature, excess
borane−THF was quenched by the slow addition of methanol so
as to control the extensive gas evolution. The mixture then
required heating under reflux for at least 3 h in order to break the
boramine complex 12. After concentration, the residue was
partitioned between MTBE and 0.5 M sodium hydroxide. The
organic layer was separated and washed with water. In order to
recover the remaining 6 in the aqueous layers, a back extraction
was required. The solvent was exchanged for toluene, and the
resulting solution of 98% pure 6 was used directly in the next
step.
Table 3. Scale-up of Buchwald−Hartwig reaction
scale
(g)
ligand
(equiv)
Pd (dba)3
LHMDS
(equiv)
2
b
entry
(equiv)
solvent
6 (%)
1
2
0.1
0.4
0.04
0.06
0.02
1.5
THF
THF
<1
0.02
+ 0.01
1.3
+ 0.5
2.9
+
0.03
3
4
5
6
0.4
0.3
0.06
0.02
1.3
THF
4.1
c
+ 0.03
+ 0.01
+ 0.5
d
0.06
0.02
1.5
THF/
PhMe
trace 7
<1
12.0
4.0
0.03
0.01
+ 0.01
1.0
+ 0.5
THF
+
0.03
0.03
0.01
1.2
THF/
PhMe
<1
a
Reactions at 65 °C, using 2-(dicyclohexylphosphino)biphenyl unless
b
c
otherwise noted. HPLC area % at 230 nm. 2-(Di-tert-
d
butylphosphino)biphenyl was used. NaHMDS (at 110 °C) was
used instead of LHMDS.
butylphosphino)biphenyl ligand, which appeared to give a
18
slower reaction (Table 3, entry 3). NaHMDS gave only a
trace amount of the desired product, consistent with the lack of
literature precedent for the use of this base as an ammonia
equivalent. After some adjustment of the ligand/palladium/base
amounts, together with the timing of the second catalyst charge,
complete consumption of starting material was eventually
obtained on a 12-g scale (Table 3, entry 5). A more satisfactory
solution came through switching the solvent from THF to
toluene and by slowly adding the LHMDS/THF solution (Table
3, entry 6). This change resulted in complete conversion and
eliminated the inconvenience of charging additional reagents to
an air-sensitive reaction. These conditions were subsequently
scaled up over 150-fold with no further issues. Workup with
water and concentration of the organic phase were followed by
partioning of 7 into dilute acid. Extraction of nonbasic impurities,
then addition of sodium hydroxide and extraction with Me-THF
provided crude 7. The crystallization system eventually selected
was a mixture of cyclohexane and heptane, due to the high
solubility of 7 in many solvents and the propensity of 7 to oil out
from other mixed-solvent systems containing heptane. Initial
crystallization from cyclohexane, with subsequent slow addition
of heptane avoided this last problem, and provided 7 in 98%
purity and 85% overall yield from 9. A total of 520 g, prepared in
two runs of 80 and 440 g, respectively, provided sufficient 7 to
continue development of the BTK program. No efforts were
made to remove residual palladium since downstream processing
also had several palladium-catalyzed transformations.
The scalability of the palladium-catalyzed amination reaction
1
3
using LHMDS was of concern as these conditions had
previously provided disappointing results (Scheme 4). This has
Scheme 4. Buchwald−Hartwig amination
14
also been observed by others. In another program, an
15
CONCLUSION
amidation reaction using tert-butyl carbamate had proved to
■
16
be very successful for an unactivated heteroaromatic chloride.
The new process to BTK intermediate 7 has provided significant
improvements in terms of yield, purity, and operability. The
overall yield of aminopyridine 7 from hydroxypyridine 1 was
increased to 72% on a multihundred-gram scale. The new route
eliminated lithium aluminum hydride, a Swern oxidation, cesium
carbonate, azetidine free base, and the four chromatographic
purifications used in the previous synthesis.
Unfortunately, this previously effective combination of the ligand
t-BuXPhos, catalytic palladium, and sodium tert-butoxide in
17
toluene failed to produce any 13. Other literature conditions
using XPhos as the ligand and cesium carbonate in dioxane were
also unsuccessful. Perhaps not surprisingly, aromatic substitution
using potassium tert-butoxide in NMP also failed. Consequently,
optimization of the LHMDS-based system was undertaken
EXPERIMENTAL SECTION
(
Table 3).
■
On small scale (Table 3, entry 1), complete conversion was
General Methods. Reactions were monitored by reverse-
phase HPLC (Zorbax Eclipse XDB C8 column, 4.6 mm × 50
mm, 1.8 μm particle size, 0.5 mL/min flow, detection at 220 nm,
5−100% acetonitrile/water containing 0.1% TFA over 10 min)
or UPLC (Zorbax SB-C8 RRHD column, 2.1 mm × 50 mm, 1.8
μm particle size, 0.7 mL/min, detection at 220 nm, 10−90%
observed after heating for 2 h using Pd (dba) and 2-
2
3
(
dicyclohexylphosphino)biphenyl. However, on only 4 times
the scale, the reaction stalled and required additional charges of
reagents to achieve good conversion (Table 3, entry 2). No
advantage could be gained from the bulkier 2-(di-tert-
C
dx.doi.org/10.1021/op400046y | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Technical Note
+
acetonitrile/water containing 0.1% H PO over 3.5 min, hold 0.5
1.46 (s, 6H); HRMS calculated for C H N O Cl (M + H)
3
4
12 16
2
2
min).
255.08948, found 255.08910.
Ethyl 2-(6-Chloropyridin-3-yloxy)-2-methylpropa-
noate (3). To a mixture of 2-chloro-5-hydroxypyridine (400 g,
.09 mol) and DMF (3.2 L) was added potassium carbonate
325 mesh; 1.28 kg, 8.05 mol). After heating to 85 °C, ethyl 2-
5-(1-(Azetidin-1-yl)-2-methylpropan-2-yloxy)-2-chlor-
opyridine (6). To a solution of 9 (665 g, 2.61 mol) in THF (1.0
L) at 45 °C was added borane−THF (2.66 L of a 1 M solution in
THF, 2.66 mol) over 45 min. After 45 min, a second portion of
borane-THF (1.33 L of a 1 M solution in THF, 1.33 mol) was
added over 50 min. After a further 45 min, a third portion of
borane−THF (1.33 L of a 1 M solution in THF, 1.33 mol) was
added over 90 min. Then, after 60 min, a final portion of borane−
THF (1.33 L of a 1 M solution in THF, 1.33 mol) was added over
3
(
bromoisobutyrate (1.34 kg, 1.01 L, 6.87 mol) was added over 100
min. After heating overnight at 85 °C, the mixture was cooled and
diluted with MTBE (4.0 L), n-heptane (4.0 L), and water (8.0 L).
The aqueous layer was separated and back extracted with 50%
MTBE/n-heptane (4.0 L). The organic layers were combined
and washed with water (2.0 L). The organic layer was then
concentrated under vacuum (50 °C water bath); n-heptane (3.0
L) was added; the mixture was reconcentrated, diluted with n-
heptane (3.0 L), and finally concentrated to a thick slurry. The
resulting suspension was cooled to 0 °C and filtered, and the
collected solids were washed with n-heptane (2.0 L) and dried at
7
0 min, and the mixture was heated at 45 °C for 6 h. After the
mixture was allowed to cool to rt overnight, methanol (2.66 L)
was added over 90 min. The addition of the first 0.6 L caused a
vigorous gas evolution and was also exothermic (the temperature
increased to 38 °C). The mixture was heated under reflux for 3 h,
then concentrated under vacuum to an oil, and diluted with
MTBE (4.7 L) and 0.5 M sodium hydroxide (2.2 L). The organic
layer was separated and washed with brine (2.2 L) and water (2 ×
4
0 °C to give 685 g (91% yield) of 3 as a white solid (99.0%
1
HPLC pure): H NMR (300 MHz, DMSO-d ) δ 8.01 (dd, J =
6
2
.2 L). The water washes were back extracted with MTBE (2.0
3
8
(
.01, 0.38 Hz, 1H), 7.43 (dd, J = 8.67, 0.38 Hz, 1H), 7.33 (dd, J =
.67, 3.01 Hz, 1H), 4.18 (q, J = 7.03 Hz, 2H), 1.55 (s, 6H), 1.17
L). The combined organic extracts were concentrated under
vacuum and then coevaporated with toluene (2 × 1.5 L) to give
t, J = 7.06 Hz, 3H); HRMS calculated for C H NO Cl (M +
11 15 3
1
+
615 g (97.9% yield) of 6 as an orange oil (97.8% HPLC pure): H
H) 244.07350, found 244.07291.
-(6-Chloropyridin-3-yloxy)-2-methylpropanoic Acid
8). To a solution of 3 (865 g, 3.55 mol) in THF (5.39 L) was
NMR (300 MHz, DMSO-d ) δ 8.07 (dd, J = 2.83, 0.57 Hz, 1H),
6
2
7
3
.51 (dd, J = 8.69, 2.83 Hz, 1H), 7.41 (dd, J = 8.69, 0.57 Hz, 1H),
(
.22 (t, J = 6.99 Hz, 4H), 2.52 (s, 2H), 1.91−2.03 (m, 2H), 1.19
added aqueous lithium hydroxide (5.39 L of a 1 M solution, 5.39
mol), and the mixture was stirred at room temperature for 2.5 h.
Hydrochloric acid (2.7 L of a 2 M solution, 5.4 mol) was added
with cooling, and the mixture was diluted with MTBE (6.0 L).
The organic layer was separated, washed with water (2.0 L), and
concentrated at 35 °C under vacuum to a low volume while
adding n-heptane (2.0 L), during which time the product
crystallized. Additional n-heptane (4.0 L) was added, and the
+
(
s, 6H); HRMS calculated for C H N OCl (M + H)
12 18 2
2
41.11022, found 241.10979.
-(1-(Azetidin-1-yl)-2-methylpropan-2-yloxy)pyridin-
-amine (7). To a degassed (3 × vac/nitrogen) solution of 6
5
2
(
(
(
607 g, 2.33 mol) in toluene (2.43 L) were added Pd (dba)
21.4 g, 23.3 mmol) and 2-(dicyclohexylphosphino)biphenyl
24.5 g, 70.0 mmol). The solution was heated to 45 °C, then
2 3
LHMDS (2.57 L of a 1 M solution in THF, 2.57 mol) was added.
slurry was concentrated to dryness to give 756 g (98.8% yield) of
The first 0.65 L was added over 5 min; the temperature rose to 48
1
8
as a light-yellow solid (99.0% HPLC pure): H NMR (300
°
C. The mixture was then heated to 62 °C, and the remainder of
MHz, DMSO-d ) δ 13.31 (br, 1H), 8.01 (d, J = 3.11 Hz, 1H),
6
the LHMDS was added over 90 min. The reaction mixture was
heated at 63 °C overnight and then cooled to 30 °C, and ice/
water (1.4 L) was added. The organic layer was separated and
filtered through a thin layer of Celite to remove palladium black,
and the pad was washed with THF. The combined filtrates were
concentrated to a dark oil and diluted with MTBE (1.8 L), and 3
N hydrochloric acid (1.6 L) was slowly added while keeping the
temperature below 34 °C. The aqueous layer was separated, and
the organic layer was extracted with 1 N hydrochloric acid (2.2
L). The combined aqueous layers were diluted with MTBE (3.6
L), and with cooling, the pH was adjusted to 11.5 using 10 N
sodium hydroxide (0.82 L). Sodium chloride (1.5 kg) was added
and then MTBE (1.0 L) and Me-THF (3.8 L). The aqueous layer
was separated and then extracted with Me-THF (3.0 L), and the
combined organic extracts were concentrated under vacuum to
half volume. After coevaporation with Me-THF (2 × 1.0 L) to ∼4
L, the resulting solution was filtered through a Celite pad, and the
pad was washed with Me-THF (1.0 L). After concentration
under vacuum to a low volume, dilution with cyclohexane (2.0 L)
followed by the addition of seed crystals gave a slurry. This slurry
was concentrated until near dryness, diluted with cyclohexane
(0.6 L), and heated to 45 °C. Then, n-heptane (2.4 L) was added
slowly, and the resulting mixture was allowed to cool to rt
overnight. The slurry was filtered, and the collected solids were
washed with 50% cyclohexane/n-heptane (1.0 L) and then n-
heptane (2 × 0.5 L), and the solids were dried under vacuum at
40 °C to give 443 g (85.8% yield) of 7 as a dark-brown solid
7
1
2
.44 (dd, J = 8.85, 0.38 Hz, 1H), 7.34 (dd, J = 8.76, 3.11 Hz, 1H),
+
.54 (s, 6H); HRMS calculated for C H NO Cl (M + H)
9
11
3
16.04220, found 216.04169.
-(Azetidin-1-yl)-2-(6-chloropyridin-3-yloxy)-2-meth-
1
ylpropan-1-one (9). To a slurry of CDI (508 g, 3.13 mol) in
dichloromethane (3.07 L) was added a solution of 8 (614 g, 2.85
mol) in dichloromethane (1.84 L) over ∼40 min, during which
time the temperature rose from 14 to 18 °C. After aging for 2 h,
the reaction mixture was cooled to 8 °C, and azetidine
hydrochloride (346 g, 3.7 mol) was added in three portions
under nitrogen, followed by the slow addition of triethylamine
(
663 g, 6.55 mol). The mixture was then heated under reflux for 2
h. After cooling to rt, 1 M sulfuric acid (4.5 L) was slowly added
exothermic). The aqueous layer was separated and extracted
(
with dichloromethane (2.0 L). The combined organic layers
were washed with water (4.9 L). The organic layer was then
concentrated at 40 °C under vacuum to ∼5 L, n-heptane (2.0 L)
was added, and concentration continued to give a thick slurry.
Additional n-heptane (3.0 L) was added and the mixture
reconcentrated. The slurry was cooled to 4 °C and filtered, and
the collected solids were washed with n-heptane (2 × 1.0 L) and
then dried under vacuum at 40 °C to afford 687 g (94.7% yield)
1
of 9, a white solid (98.1% HPLC pure): H NMR (300 MHz,
DMSO-d ) δ 8.01 (dd, J = 3.21 Hz, 0.38 Hz, 1H), 7.46 (dd J =
6
8
7
.69, 0.38 Hz, 1H), 7.32 (dd, J = 8.69, 3.21 Hz, 1H), 4.18 (t, J =
.74 Hz, 2H), 3.90 (t, J = 7.74 Hz, 2H), 2.14 (t, J = 7.74 Hz, 2H),
D
dx.doi.org/10.1021/op400046y | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Technical Note
1
(
(
(
98.1% HPLC pure): H NMR (300 MHz, DMSO-d ) δ 7.55
6
dd, J = 2.83, 0.57 Hz, 1H), 7.04 (dd, J = 8.69, 2.83 Hz, 1H), 6.35
dd, J = 8.69, 0.57 Hz, 1H), 5.60 (s, 2H), 3.20 (t, J = 6.99 Hz,
4
H), 2.43 (s, 2H), 1.88−2.03 (m, 2H), 1.07 (s, 6H); HRMS
+
calculated for C H N O (M + H) 222.16009, found
1
2
20
3
222.15956.
AUTHOR INFORMATION
Corresponding Author
E-mail: andrew.briggs@roche.com
■
*
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank Fred Roberts for analytical support and Masa Okabe
for helpful discussions.
■
REFERENCES
■
(
1) Lou, Y.; Owens, T. D.; Kuglstatter, A.; Kondru, R. K.; Goldstein, D.
M. J. Med. Chem. 2012, 55, 4539.
(
2) Billedeau, R. J.; Kondru, R. K.; Lopez-Tapia, F. J.; Lou, Y.; Owens,
T. D.; Qian, Y.; So, S.-S.; Thakkar, K. C.; Wanner, J. Inhibitors of
Bruton’s tyrosine kinase. U.S. Patent Application 2012/0295885, Nov
2
(
2, 2012, CA157:734758.
3) Recent pricing for Cs CO was $31/mol (Chem-Impex), and for
2
3
K CO , 325 mesh was $8/mol (SigmaAldrich).
2
3
(
(
(
4) Presumably there is competing elimination to ethyl methacrylate.
5) Purity refers to HPLC area%.
6) Evans, D. A.; Ellman, J. A.; Dorow, R. L. Tetrahedron Lett. 1987, 28,
1
123.
(
7) Wissmann, H.; Kleiner, H.-J. Angew. Chem., Int. Ed. 1980, 19, 133.
8) Staab, H. A. Angew. Chem., Int. Ed. 1962, 1, 351.
(
(
9) A multiplet at δ 3.6 ppm was attributed to 10, the remaining signals
at ∼2 ppm were hidden by those of the product. Signals at δ 8.09 (m),
.50 (m), 7.10 (m), and 2.30 (m) were attributed to 11. In addition
there were signals at ∼4 ppm but no clear assignment could be made.
10) Recent catalog pricing for azetidine was $1710/mol (5 g,
Oakwood Products) and for azetidine hydrochloride was $636/mol
100 g, Combi-Blocks).
11) Providing allowance is made for the low assay of the azetidine
hydrochloride.
12) Work-up with citric acid led to a lower yield, whereas hydrochloric
7
(
(
(
(
acid required additional washes to remove unreacted azetidine from the
organic phase.
(
13) (a) Lee, S.; Jorgensen, M.; Hartwig, J. F. Org. Lett. 2001, 3, 2729.
b) Huang, X.; Buchwald, S. L. Org. Lett. 2001, 3, 3417.
14) Leahy, D. K.; Desai, L. V.; Deshpande, R. P.; Mariadass, A. V.;
(
(
Rangaswamy, S.; Rajagopal, S. K.; Madhavan, L.; Illendula, S. Org.
Process Res. Dev. 2012, 16, 244.
(
15) Bhagwanth, S.; Waterson, A. G.; Adjabeng, G. M.; Hornberger, K.
R. J. Org. Chem. 2009, 74, 4634.
16) Lu, X. Hoffmann-La Roche, Inc., Nutley, NJ. Personal
communication.
17) Qin, L.; Cui, H.; Zou, D.; Li, J.; Wu, Y.; Zhu, Z.; Wu, Y.
Tetrahedron Lett. 2010, 51, 4445.
18) Samples taken after 20 min also showed a lower conversion for the
(
(
(
bulkier di-t-butyl phosphine ligand as compared to the dicyclohexyl
analog.
E
dx.doi.org/10.1021/op400046y | Org. Process Res. Dev. XXXX, XXX, XXX−XXX