An Enantioselective Hydrogenation of an Alkenoic Acid as a Key Step in the Synthesis of AZD2716

Article abstract of DOI:10.1021/acs.oprd.5b00382

A classical resolution of a racemic carboxylic acid through salt formation and an asymmetric hydrogenation of an α,β-unsaturated carboxylic acid were investigated in parallel to prepare an enantiomerically pure alkanoic acid used as a key intermediate in the synthesis of an antiplaque candidate drug. After an extensive screening of rhodium- and ruthenium-based catalysts, we developed a rhodium-catalyzed hydrogenation that gave the alkanoic acid with 90% ee, and after a subsequent crystallization with (R)-1-phenylethanamine, the ee was enriched to 97%. The chiral acid was then used in sequential Negishi and Suzuki couplings followed by basic hydrolysis of a nitrile to an amide to give the active pharmaceutical ingredient in 22% overall yield.

Full text of DOI:10.1021/acs.oprd.5b00382

Article
pubs.acs.org/OPRD
An Enantioselective Hydrogenation of an Alkenoic Acid as a Key Step
in the Synthesis of AZD2716
Staffan Karlsson,*,† Henrik Sorensen,† Soren M. Andersen,† Angele Cruz,† and Per Ryberg‡
̈
̈
†CVMD iMed, Medicinal Chemistry, AstraZeneca R&D Molndal, SE-431 83 Molndal, Sweden
̈
̈
‡SP Process Development, Box 36, SE-151 21 Sodertalje, Sweden
̈
̈
S
* Supporting Information
ABSTRACT: A classical resolution of a racemic carboxylic acid through salt formation and an asymmetric hydrogenation of an
α,β-unsaturated carboxylic acid were investigated in parallel to prepare an enantiomerically pure alkanoic acid used as a key
intermediate in the synthesis of an antiplaque candidate drug. After an extensive screening of rhodium- and ruthenium-based
catalysts, we developed a rhodium-catalyzed hydrogenation that gave the alkanoic acid with 90% ee, and after a subsequent
crystallization with (R)-1-phenylethanamine, the ee was enriched to 97%. The chiral acid was then used in sequential Negishi and
Suzuki couplings followed by basic hydrolysis of a nitrile to an amide to give the active pharmaceutical ingredient in 22% overall
yield.
genation of an alkenoic acid, and resolution of a racemic α-
methyl alkanoic acid building block. Furthermore, we wanted to
avoid the use of chromatography and instead identify crystalline
intermediates that could be purified through crystallization.
Alternatively, noncrystalline intermediates could be used as
such and/or telescoped. We believed that the rest of the first-
generation synthesis could be used as such in the scale-up with
only minor modifications such as in the workup.
INTRODUCTION
■
In one of our drug development programs, we planned to run a
full toxicological study on AZD2716, a candidate drug for the
treatment of coronary artery disease (Figure 1).1 Several
hundred grams of the compound was needed for this purpose.
RESULTS AND DISCUSSION
■
In order to obtain the enantiomerically pure carboxylic acid
(R)-2, two different approaches were investigated in parallel:
resolution of racemic acid ( )-2 through salt formation with
chiral amines and asymmetric hydrogenation of alkenoic acid 8
(Scheme 2). On a large scale, the racemic carboxylic acid ( )-2
was prepared using the procedure from the first-generation
synthesis with small modifications (Scheme 2). The alkenoic
acid required for the asymmetric hydrogenation was prepared
via a Baylis−Hillman reaction5 between 3-bromobenzaldehyde
and methyl acrylate to give 7 followed by reduction6 to give 8.
For the resolution of carboxylic acid ( )-2, a selection of
chiral amines were screened (Table 1). From this screen we
found (R)-1-phenylethylamine to be the best choice for
resolution of the racemate ( )-2. Thus, starting from 1.6 kg
of the racemate ( )-2 and using EtOAc as the solvent, we
obtained the salt with 64% ee, which is a slightly better result
than that obtained in the primary small screen (Table 1, entry
7). We found it necessary to perform two recrystallizations of
this salt from a mixture of EtOH and EtOAc to obtain an
acceptable ee (>95% ee). After acidification using HCl and
extractions, (R)-2 was isolated in 27% overall yield with 97% ee.
As mentioned above, in parallel with the work on the
resolution of acid ( )-2 we searched for a more efficient
method for long-term supply of the enantiomerically pure acid
Figure 1. Structure of AZD2716.
The first-generation synthesis of AZD2716 commenced with
an alkylation of diethyl methylpropanedioate with 1-bromo-3-
(bromomethyl)benzene followed by saponification and decar-
boxylation of the intermediate 1 to give the racemic carboxylic
acid ( )-2 (Scheme 1). Compound ( )-2 was next trans-
formed to the corresponding pinacolborate2 ( )-3, which then
was ready for a Suzuki coupling3 with chloride 4 to give
compound ( )-5. Compound 4 was obtained from commer-
cially available 4-bromo-2-chlorobenzonitrile via a Negishi
coupling.4 The synthesis was finalized with a hydrolysis of
nitrile ( )-5 using NaOH followed by separation of the
enantiomers using chiral HPLC to give AZD2716.
Given limited time for further optimization of the first-
generation route, in the scale-up we decided to keep the major
part of the first-generation synthesis, but a few issues had to be
addressed. We realized that separation of the racemic mixture
of the active pharmaceutical ingredient (API) as the last step,
which resulted in more than 50% loss of material, was not
suitable for further scale-up. Thus, we decided to direct our
efforts toward the development of an asymmetric synthesis of
this chiral α-methyl carboxylic acid moiety. This could in theory
be achieved through asymmetric alkylation, asymmetric hydro-
Received: November 19, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.oprd.5b00382
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 1. First-Generation Synthesis of AZD2716
Scheme 2. Two Different Approaches for the Synthesis of Carboxylic Acid (R)-2
(R)-2. Asymmetric hydrogenation of alkenoic acids is a
powerful methodology for the preparation of enantiomerically
pure carboxylic acids.7 The substrate scope is broad, and a wide
variety of carboxylic acids with different substitution patterns
have been prepared on a multikilogram scale by asymmetric
hydrogenation using both ruthenium- and rhodium-based
catalysts.8 As a part of our efforts to develop a more sustainable
approach for the synthesis of enantiomerically pure carboxylic
acid (R)-2, our plan was to identify an efficient catalyst for the
asymmetric hydrogenation of α,β-unsaturated acid 8 (Scheme
2). A literature survey showed that α-methylcinnamates are
difficult substrates in asymmetric hydrogenations and constitute
an exception to the otherwise broad substrate scope. In fact,
only a few examples were found that gave useful enantiose-
lectivities for a range of different α-methylcinnamates.9 Those
reports used chiral ligands that had to be synthesized through a
multistep sequence, and at the time when this work was
conducted, those ligands were not commercially available. For
this reason and to avoid the dependence on patented protocols
to obtain catalysts, we wanted to identify a commercially
available catalyst that also could give a satisfactory enantiose-
lectivity (>95% ee). A broad screen of both Ru- and Rh-based
catalysts in combination with a wide range of different ligands
representing all of the important ligand families was set up
(Figures 2−4).10 The hydrogenation screen of 8 was performed
on a Chemspeed high-throughput experimentation robot with a
high-pressure parallel autoclave.11
and enantioselectivity, and the majority of the Rh catalysts gave
complete conversion to product, in contrast to the Ru catalysts.
Consistently good performance was observed for almost the
entire series of Walphos ligands (W001−W022) with
enantioselectivities in the range of 80−85% ee. Also, some of
the Mandyphos ligands (M001−M012) showed good
selectivity. The overall best hit was for the Josiphos ligand
J505-1, which afforded the product with 94% ee. Interestingly,
out of the 11 Josiphos ligands that were included in the screen
(J001−J505), only the J-505 ligand gave >30% ee.
On the basis of these results, the Rh−J505 catalyst was
selected for further optimization, and the influence of various
solvents and additives on the performance at gradually
decreasing catalyst loadings was investigated. The effect of
solvents, including methanol, isopropanol, n-butanol, and
tetrahydrofuran (THF) was investigated, as was the effect of
additives such as triethylamine, tert-butylamine, and trifluoro-
acetic acid (TFA).12 It was found that methanol was the best
solvent for the reaction with regard to both enantioselectivity
and catalyst activity. At catalyst loadings below 1 mol %,
addition of TFA had a clear positive effect on the catalyst
activity, and at a loading of 0.5 mol % the reaction stopped at
<30% conversion in absence of added TFA but went to full
conversion when 1 mol % TFA was added. A further decrease
in the catalyst loading to 0.2 mol % or less led to incomplete
conversion even in the presence of added TFA (Figure 5).
Furthermore, upon scaling up to 70 g and using the Rh−J505
catalyst, we found that a slightly lower enantioselectivity was
obtained (90% ee) and a higher catalyst loading (1 mol %) was
required to reach full conversion. However, with the efforts
As illustrated in Figures 2 and 4, both the conversion and
enantioselectivity varied considerably among the catalysts.
However, the Rh catalysts showed overall higher reactivity
B
DOI: 10.1021/acs.oprd.5b00382
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
After filtration of the salt obtained, followed by acidification and
extraction, carboxylic acid (R)-2 was isolated in 75% isolated
yield with 98% ee (Scheme 3).
Table 1. Salt Screen for the Resolution of ( )-2
In order to obtain AZD2716, sequential Negishi and Suzuki
couplings followed by hydrolysis of the nitrile were required
(Scheme 3). The Negishi coupling13 of commercially available
4-bromo-2-chlorobenzonitrile and benzylzinc bromide was
found to furnish the homocoupled dimer 3,3′-dichlorobiphen-
yl-4,4′-dicarbonitrile (20% w/w) as a major byproduct along
with the desired product 4. We tried to suppress this side
reaction through variation of the temperature, the loadings of
the catalyst and BnZnBr, and the rate of addition of the latter.
However, no significant positive effect could be observed.
Instead, we searched for conditions to remove this byproduct
from the crude mixture through crystallizations, but because of
the highly crystalline properties of this dimer, that search was
not successful. Although we aimed for a chromatography-free
process, we decided to purify the mixture through preparative
HPLC followed by one crystallization from MeOH/water,
which gave 321 g (>99% by HPLC) of the desired compound
4. In the subsequent steps to the API, we could not identify a
suitable intermediate for crystallization. Instead, we developed a
three-step sequence to AZD2716 in which none of the
intermediates were isolated or purified other than by
extractions. The sequence included the formation of the
boronate ester (R)-3 from (R)-2 followed by a telescoping
sequence in which a Suzuki coupling3 with 4 resulted in
compound (R)-5. After extractions of the crude mixture of (R)-
5, this was concentrated and then directly hydrolyzed to
furnish, after crystallization from EtOAc, the desired API
AZD2716 in 57% isolated yield over three steps with >98%
ee.14
a
b
A 70 mol % loading of the resolving amine was used. Determined
using chiral HPLC of the free acid.
CONCLUSION
■
In comparison with the first-generation synthesis of AZD2716
in 8% overall yield, we developed an enantioselective large-scale
route with an overall yield of 22%. The main improvement was
the replacement of a late-stage enantiomeric separation of a
racemic mixture with an enantioselective hydrogenation of an
already invested in trying to identify a chiral amine for the
resolution of racemic acid ( )-2, we knew that the ee could be
enhanced through salt formation with (R)-1-phenylethanamine.
Thus, after the hydrogenation was complete, this amine was
added to the crude mixture, which resulted in precipitation.
Figure 2. Results from the hydrogenation screen of 8 using various rhodium catalysts.
C
DOI: 10.1021/acs.oprd.5b00382
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 3. Ligands screened for the asymmetric hydrogenation.
Figure 4. Results from the hydrogenation screen of 8 using various ruthenium catalysts.
early-stage alkenoic acid 8. Using the Josiphos ligand (R)-1-
[(S)-2-(di-tert-butylphosphino)ferrocen-1-yl]ethylbis(2-
methylphenyl)phosphine in combination with rhodium as the
catalyst in the hydrogenation, we obtained (R)-2 with 90% ee.
Salt formation with (R)-1-phenylethylamine followed by
crystallization improved the enantioselectivity to 97% ee.
Another key step in the synthesis was a sequential Negishi/
Suzuki coupling. By using crude intermediates directly in the
next steps without purifications, we avoided chromatography in
all but one step. A final crystallization of the API gave 347 g of
AZD2716 with >99% purity and >98% ee.
EXPERIMENTAL SECTION
■
Materials and Instrumentation. All materials were
purchased from commercial suppliers and used as such without
further purification. All reactions were performed under an
atmosphere of nitrogen. Large-scale reactions were performed
D
DOI: 10.1021/acs.oprd.5b00382
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 5. Conversion of 8 as a function of the loading of the Rh−J505 catalyst.
Scheme 3. Second-Generation Synthesis of AZD2716
using glass reactors equipped with an overhead stirrer. IPCs
analyses were recorded on a Waters ZMD, LC column x Terra
MS C8 (Waters) with detection using an HP 1100 MS-detector
diode array.
1
were recorded either by HPLC or H NMR analysis of the
1
crude reaction mixtures. Assays were determined by H NMR
integration using benzyl benzoate as an internal standard. The
enantiomeric purity of (R)-2 was determined using HPLC [t-
Bu-CQN column (4.6 mm × 250 mm) with CH3CN/2-
propanol/AcOH 90/10/0.3 as the eluent]. The enantiomeric
purity of AZD2716 was determined using HPLC [Chiralcel OJ
column (4.6 mm × 250 mm) with heptane/EtOH/formic acid
10/90/0.1 as the eluent]. High-resolution mass spectrometry
was performed on a QTOF 6530 instrument (Agilent) with a
mass precision of 5 ppm. NMR measurements were
performed using a Bruker Avance III spectrometer. LC/MS
Catalyst Screening. Catalyst screening was performed on a
Chemspeed high-throughput experimentation robot equipped
with a high-pressure parallel autoclave. To each of the wells in a
96-well microtiter plate was added 50 μL of methanol followed
by 50 μL of either a 0.00700 M MeOH solution of
Rh(COD)2BF4 or a 0.00700 M MeOH solution of Ru(COD)-
(TFA)2. Finally, 50 μL aliquots of 0.0077 M solutions of the
ligands in toluene were added to the wells to give a total of 43
different Rh catalysts and 43 different Ru catalysts. The
mixtures were agitated at 40 °C for 2 h, after which 200 μL of a
E
DOI: 10.1021/acs.oprd.5b00382
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
0.28 M solution of 8 in MeOH was added to each well and the
reaction plate was inserted into the autoclave. The autoclave
was pressurized to 15 bar H2 and agitated for 16 h. The reaction
mixtures were diluted with isopropanol and then analyzed by
chiral HPLC for determination of the conversion and
enantiomeric excess.
performed from hot EtOH/EtOAc (3 L/5 L) using a
temperature ramp from 80 to 20 °C over 120 min followed
by filtration of the salt and washing with EtOAc (3 × 500 mL).
After drying under reduced pressure, the title salt 6 was
obtained as a white solid (0.61 kg, 1.7 mol, 27%) with 97% ee
1
(free acid) as determined by chiral HPLC analysis. H NMR
3-(3-Bromophenyl)-2-methylpropanoic Acid [( )-2]. A
25 L reactor was charged with diethyl 2-methylmalonate (1.8
kg, 10.4 mol) and ethanol (99.5%, 12 L). The mantle
temperature was set at 0 °C. When the reaction temperature
had reached 8 °C, sodium 2-methylbutan-2-olate (1.2 kg, 10.4
mol) was added portionwise over 20 min. The reaction
temperature reached 22 °C during this addition. The mantle
temperature was then set at 20 °C. To the homogeneous
solution was then added within 10 min a suspension of 1-
bromo-3-(bromomethyl)benzene (2.0 kg, 8.0 mol) in ethanol
(99.5%, 2 L). This addition raised the reaction temperature to
39 °C, and a white suspension was obtained. The mantle
temperature was set at 60 °C. When the reaction temperature
reached 57 °C, a 25% aqueous solution of sodium hydroxide
(4.5 kg, 28.2 mol) was added, and the white suspension was
allowed to stir for 30 min. EtOH (∼5 L) was then allowed to
distill off (400 mbar/60 °C). With the mantle temperature set
at 20 °C, methyl tert-butyl ether (MTBE) (7.5 L) and water (5
L) were added. The layers were separated, and water (5 L) and
MTBE (2.5 L) were added to the organic layer. The pooled
aqueous layer was washed with MTBE (5 L + 2.5 L). The
mantle temperature was set at 0 °C, and to the aqueous layer
was added MTBE (2.5 L) followed by the addition of 12 M
HCl(aq) (3.5 L) within 10 min until a pH of ∼1 was obtained.
The reaction temperature increased to 39 °C during this
addition. The organic layer (6 L) was concentrated to give the
intermediate diacid as a pale-yellow solid. This was charged into
a 25 L reactor, and acetic acid (3.2 L) was added. The yellow
suspension was then heated at reflux for 17 h, during which gas
evolution was observed. The mantle temperature was set at 20
°C, and water (4 L) and MTBE (4 L) were added. The organic
layer was washed with water (3 × 1 L) and then concentrated.
Residual acetic acid and propionic acid were azeotropically
removed using toluene (3 × 500 mL). This furnished the title
acid ( )-2 (1.64 kg, 6.75 mol, 84%) as a viscous oil which was
used as such in the next step (resolution). Analytical data were
in accordance with those given for (R)-2 in the literature.9c
(2R)-3-(3-Bromophenyl)-2-methylpropanoic Acid (R)-
1-Phenylethylammonium Salt (6). A 25 L reactor was
charged with the crude racemic 3-(3-bromophenyl)-2-methyl-
propanoic acid [( )-2] from the previous step (1.5 kg, 6.3
mol) followed by the addition of ethyl acetate (15 L). The
mixture was heated to 60 °C. (R)-1-Phenylethanamine (0.5 kg,
4.1 mol) was added, and a rapid precipitation occurred. The
mixture was heated to reflux, which furnished a homogeneous
solution. The mantle temperature was then allowed to reach 20
°C over 120 min. Seeding crystals were added when the
reaction temperature had reached 65 °C, which initiated a slow
crystallization. After 1 h at 20 °C, the thick suspension was
filtered, and the white salt collected was washed with EtOAc (3
× 1 L). Chiral HPLC analysis of the salt showed an
enantiomeric excess of 64% for the acid. This salt was
recrystallized from hot ethyl acetate (7 L) and 99.5% EtOH
(3 L) using a temperature ramp from 80 to 20 °C over 10 h, a
constant temperature of 20 °C for 3 h, and then filtration and
washing with EtOAc (3 × 800 mL). This gave the acid with an
enantiomeric excess of 92%. A second recrystallization was
(400 MHz, CD3OD): δ 1.08 (d, J = 6.6 Hz, 3H), 1.59 (d, J =
6.8 Hz, 3H), 2.49−2.58 (m, 2H), 2.92−3.00 (m, 1H), 4.38 (q, J
= 6.8 Hz, 1H), 7.12−7.46 (m, 9H).
(2R)-3-(3-Bromophenyl)-2-methylpropanoic Acid [(R)-
2] via the Resolved Salt 6. (2R)-3-(3-Bromophenyl)-2-
methylpropanoic acid (R)-1-phenylethylammonium salt (6)
(0.61 kg, 1.67 mol) was partitioned between MTBE (1 L) and
4 M HCl (0.5 L). The aqueous layer was extracted with MTBE
(0.4 L), and the pooled organic layer was then washed with
water (100 mL) followed by concentration. Residual water was
azeotropically removed using MTBE (2 × 200 mL). This gave
the title compound (R)-2 as a colorless oil (407 g, 1.67 mol,
100%) with 97% ee and assay 100% (w/w). Analytical data
were in accordance with those in the literature.9c
Methyl 2-((3-Bromophenyl)(hydroxy)methyl)acrylate
(7). With the mantle temperature set at 25 °C, the reactor
was charged with 3-bromobenzaldehyde (100 g, 0.54 mol) and
methyl acrylate (140 g, 1.63 mol) followed by the addition of
methanol (0.5 L). A solution of trimethylamine (25% in H2O,
128 g, 0.54 mol) was added, and the mixture was allowed to stir
at 25 °C for 5 days, after which almost full conversion had been
obtained (checked by LC/MS). Most of the MeOH was
removed under reduced pressure. MTBE (0.5 L) was added,
and the pH was then adjusted to ∼3 using HCl solution. The
layers were separated, and the organic layer was washed with
water (0.2 L) and then concentrated under reduced pressure.
Water was azeotropically removed with toluene (2 × 0.4 L)
until the water content was <0.15% (w/w) as determined by
Karl Fischer titration. The title compound 7 was obtained as a
clear oil (144 g) and was used as such directly in the next step.
1H NMR (400 MHz, CDCl3): δ 3.74 (s, 3H), 5.52 (s, 1H), 5.84
(s, 1H), 6.36 (s, 1H), 7.18−7.24 (m, 1H), 7.30−7.32 (m, 1H),
7.40−7.43 (m, 1H), 7.53−7.55 (m, 1H).
(E)-3-(3-Bromophenyl)-2-methylacrylic Acid (8). To the
crude methyl 2-((3-bromophenyl)(hydroxy)methyl)acrylate
(7) from the previous step (144 g) were added THF (1 L)
and I2 (304 g, 1.2 mol). The resulting brown solution was
cooled to 0 °C. NaBH4 (40 g, 1.06 mol) was added in small
portions over 1 h to maintain the reaction temperature below
20 °C. Caution: exothermic reaction! The mantle temperature
was then set at 20 °C, and the white suspension was allowed to
stir for 17 h. MeOH (180 mL) was slowly added over 1.5 h.
The mixture was then concentrated. MeOH (360 mL) was
added, and the mantle temperature was set at 0 °C. To the clear
solution was added a 50% aqueous solution of sodium
hydroxide (143 mL, 2.7 mol). When all had been added, the
mixture was allowed to reach a temperature of 25 °C and was
stirred for 1 h. The mixture was then concentrated to almost
dryness. Toluene (500 mL) and water (500 mL) were added.
The aqueous layer was acidified to pH 2 using 37% aqueous
HCl (162 mL, 1.97 mol), and toluene (1 L) was added. The
organic layer was concentrated, followed by the addition of
ethanol (500 mL) and water (450 mL). The mantle
temperature was set at 20 °C. Crystallization initiated, and
more ethanol (450 mL) was added. The suspension was stirred
at 20 °C for 16 h. The solids were filtered off and dried to
afford the title compound 8 as a white solid (67.6 g, 0.28 mol,
F
DOI: 10.1021/acs.oprd.5b00382
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
52% over two steps). Analytical data were in accordance with
those in the literature.9c
acetate (494 g, 5.04 mol), diboron pinacol ester (543 g, 2.14
mol), (R)-3-(3-bromophenyl)-2-methylpropanoic acid [(R)-2]
(400 g, 1.64 mol), and 1,4-dioxane (4 L). To the white
suspension was then added 1,1′-bis(di-tert-butylphosphino)-
ferrocene palladium dichloride (10.5 g, 0.016 mol). The
suspension immediately turned brown/orange. The mantle
temperature was set at 70 °C, and stirring was performed for 17
h. LC/MS analysis of the crude mixture showed full conversion
of the starting material. The mantle temperature was set at 20
°C, and the mixture of (R)-3 (477 g) was then used as such in
the next step. MS: m/z 289 [M − H]+.
Synthesis of 6 via Asymmetric Hydrogenation of 8.
Under an atmosphere of nitrogen, to a solution of bis(1,5-
cyclooctadiene)rhodium(I) tetrafluoroborate (1.16 g, 2.9
mmol) in methanol (350 mL) was added a solution of (R)-1-
[(S)-2-(di-tert-butylphosphino)ferrocen-1-yl]ethylbis(2-
methylphenyl)phosphine (1.78 g, 3.2 mmol) in toluene (350
mL). To the resulting active catalyst mixture was added a
solution of (E)-3-(3-bromophenyl)-2-methylacrylic acid (8)
(71.1 g, 295 mmol) in methanol (700 mL). The dark-red
mixture was hydrogenated under hydrogen (19 bar) for 48 h,
after which full conversion had been obtained as determined by
LC/MS with 90% ee as determined by chiral HPLC analysis of
the crude mixture. The mixture was concentrated, and to the
residual brown oil were added ethyl acetate (700 mL) and (R)-
1-phenylethanamine (9.4 g, 78 mmol), after which crystal-
lization initiated. The mantle temperature was set at 80 °C and
more (R)-1-phenylethanamine (26 g, 217 mmol) was added.
Ethanol (99.5%, 350 mL) was added at a process temperature
of 72 °C, which resulted in a homogeneous solution. The
mixture was then allowed to reach 20 °C over a period of 2 h.
The suspension was filtered, and the salt collected was washed
with EtOAc (2 × 200 mL). After drying under reduced
pressure, the title salt 6 was obtained as a white solid (80.7 g,
220 mmol, 75%) with 97% ee (free acid) as determined by
chiral HPLC analysis. The analytical data agreed with those
given above for 6.
(R)-3-(5′-Benzyl-2′-cyanobiphenyl-3-yl)-2-methylpro-
panoic Acid [(R)-5]. To the crude mixture of (R)-2-methyl-3-
(3-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)-
propanoic acid [(R)-3] (477 g) in dioxane (4 L) was added a
degassed solution of potassium carbonate (682 g, 4.93 mol) in
water (680 mL) followed by the addition of a degassed
suspension of 4-benzyl-2-chlorobenzonitrile (4) (376 g, 1.64
mol) in 1,4-dioxane (700 mL) and 1,1′-bis(di-tert-
butylphosphino)ferrocene palladium dichloride (10.8 g, 0.016
mol). The reaction mixture was degassed and then heated to 90
°C for 48 h, after which approximately 50% conversion had
been obtained. More 1,1′-bis(di-tert-butylphosphino)ferrocene
palladium dichloride [10.8 g, 0.016 mol + 2 × (7 g, 0.0109
mol)] had to be added sequentially over the next 48 h to reach
almost full conversion of the starting material. The reaction
mixture was concentrated. With the mantle temperature set at
20 °C, toluene (5 L) was added, and to the resulting dark-
brown solution was added a 3.8 M aqueous solution of HCl (3
L, 11.4 mol) over 10 min followed by the addition of water (1
L). The organic layer was washed with 4 L of 15% NaCl
solution. The mantle temperature was then set at 10 °C, and
the mixture was stirred for an additional 17 h. Undissolved
precipitates were removed through filtration (25 μm filter), and
water (2 L) was added to the filtrate. The organic layer was
then concentrated to give the crude title compound (R)-5 (584
g), which was directly used as such in the next step. MS: m/z
354 [M − H]+. 1H NMR (400 MHz, CDCl3): δ 1.26 (d, J = 6.8
Hz, 3H), 2.74−2.91 (m, 2H), 3.20 (dd, J = 6.0, 13.2 Hz, 1H),
4.11 (s, 2H), 7.22−7.45 (m, 11H), 7.70 (d, J = 8.0 Hz, 1H).
(2R)-3-(5′-Benzyl-2′-carbamoylbiphenyl-3-yl)-2-meth-
ylpropanoic Acid (AZD2716). To the crude (R)-3-(5′-
benzyl-2′-cyanobiphenyl-3-yl)-2-methylpropanoic acid [(R)-5]
from the previous step (584 g) was added 2-propanol (7.5 L)
followed by 50% aqueous NaOH (657 g, 8.2 mol). The mantle
temperature was set at 80 °C, and the mixture was stirred for 14
h, after which full conversion of the nitrile had been obtained.
The mixture was concentrated to almost dryness (7.5 L was
distilled off). To the residue were added water (6 L) and
toluene (6 L). The organic layer was extracted with water (1 L).
To the pooled aqueous layer was added 37% aqueous HCl (805
g, 8.17 mol) until a pH of 2.6 was obtained. Toluene (5 L) was
added. The layers were separated, and the organic one was
stirred at 20 °C for 1 h, after which crystallization initiated. The
solid was filtered off through a sintered glass funnel (P3),
washed with toluene (1 L), and then suspended in ethyl acetate
(9.5 L). The mixture was heated to reflux, which resulted in a
clear solution that was filtered through a sintered glass filter
(P3). EtOAc (4 L) was distilled off under atmospheric pressure,
after which crystallization initiated. The mixture was allowed to
reach 20 °C over 120 min and was then stirred for an additional
16 h. The suspension was filtered, and the solid collected was
dried under reduced pressure at 20 °C for 20 h to give 347 g
4-Benzyl-2-chlorobenzonitrile (4). The reactor was
charged with Pd(Ph3P)4 (38 g, 0.032 mol) and 4-bromo-2-
chlorobenzonitrile (0.48 kg, 2.2 mol) followed by the addition
of THF (1.2 L). At 20 °C, a solution of benzylzinc(II) bromide
(0.5 M in THF, 4.8 L, 2.4 mol) was added within 15 min. A
slightly exothermic reaction initated, and the reaction temper-
ature was kept below 57 °C by adjusting the mantle
temperature in the interval of 35 to 45 °C. After 2 h of
stirring, full conversion of the starting material had been
obtained, and the mixture was cooled to 15 °C. A 0.4 M
aqueous solution of hydrochloric acid (1.5 L) was added,
followed by the addition of brine (1.5 L). The organic layer was
washed with brine (3 × 1.5 L), followed by concentration of
the mixture. EtOAc (2 L) was added in order to precipitate the
byproduct 3,3′-dichlorobiphenyl-4,4′-dicarbonitrile. The sus-
pension was stirred over the weekend and then filtered. The
filtrate was concentrated to give a crude impure mixture (>700
g) of 4. This was further purified by chromatography (column
kromasil-SIL, 250 mm × 100 mm) using 20−100% EtOAc in
heptane as the eluent. Fractions were pooled and concentrated.
To the residue was added methanol (3.5 L) followed by heating
to 40 °C. The homogeneous solution was allowed to reach 20
°C, after which a suspension had been formed. Water (0.9 L)
was added. The mixture was cooled at 0 °C for 2 h and filtered,
and the solid was washed with a 50% aqueous solution of
methanol (0.5 L). After drying under reduced pressure at 40
°C, the title compound 4 was obtained as a white solid (321 g,
1
1.4 mol, 64%) wth 99.4% purity as determined by HPLC. H
NMR (400 MHz, CDCl3): δ 4.02 (s, 2H), 7.15−7.37 (m, 7H),
7.58 (d, J = 7.9 Hz, 1H). 13C NMR (100 MHz, CDCl3): δ 41.6,
110.8, 116.0, 126.8, 127.7, 128.8, 128.9, 130.2, 133.8, 136.6,
138.5, 148.5.
(R)-2-Methyl-3-(3-(4,4,5,5-tetramethyl-1,3,2-dioxabor-
olan-2-yl)phenyl)propanoic Acid [(R)-3]. Under an atmos-
phere of nitrogen, a 10 L reactor was charged with potassium
G
DOI: 10.1021/acs.oprd.5b00382
Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
42, 65−102. (e) Blaser, H. U.; Malan, C.; Pugin, B.; Spindler, F.;
Steiner, H.; Studer, M. Adv. Synth. Catal. 2003, 345, 103−151.
(9) (a) Zhou, Q.; Li, S.; Zhu, S.; Zhang, C.; Wang, L. (Zhejiang
Jiuzhou Pharm Co. Ltd.). Application of an Iridium Complex in
Asymmetric Catalytic Hydrogenation of an Unsaturated Carboxylic
Acid. EP2275398, Jan 19, 2011. (b) Li, S.; Zhu, S.-F.; Zhang, C.-M.;
Song, S.; Zhou, Q.-L. J. Am. Chem. Soc. 2008, 130, 8584−8585.
(c) Cheng, X.; Zhang, Q.; Xie, J.-H.; Wang, L.-X.; Zhou, Q.-L. Angew.
Chem., Int. Ed. 2005, 44, 1118−1121. (d) Cheng, X.; Xie, J.-H.; Li, S.;
Zhou, Q.-L. Adv. Synth. Catal. 2006, 348, 1271−1276.
(0.93 mol, 57% over three steps) of the title compound
AZD2716 with >98% ee (chiral HPLC), assay 100% (w/w),
and 99.6% purity (HPLC). 1H NMR (400 MHz, DMSO-d6): δ
1.04 (d, J = 6.6 Hz, 3H), 2.55−2.68 (m, 2H), 2.95 (dd, J = 6.1,
12.8 Hz, 1H), 4.00 (s, 2H), 7.13−7.37 (m, 13H), 7.49−7.54
(m, 1H), 12.2 (s, br, 1H). 13C NMR (151 MHz, DMSO): δ
16.7, 39.1, 40.7, 41.0, 126.3, 126.4, 127.3, 127.8, 128.0, 128.2,
128.7, 128.9, 129.2, 130.3, 135.3, 139.2, 139.5, 140.5, 141.2,
142.7, 171.3, 177.1. HRMS (ESI): [M + H]+ m/z calcd for
C24H24NO3 374.1751, found 374.1748.
(10) The following ligands were screened: SL-J001 [CAS 155806-35-
2]; SL-J002 [CAS 155830-69-6]; SL-J003 [CAS 167416-28-6]; SL-
J004 [CAS 158923-09-2]; SL-J005 [CAS 184095-69-0]; SL-J006
[CAS 849923-15-5]; SL-J008 [CAS 166172-63-0]; SL-J009 [CAS
158923-11-6]; SL-J212 [CAS 849924-41-0]; SL-J216 [CAS 849924-
43-2]; SL-J505 [CAS 849924-76-1]; SL-W001 [CAS 387868-06-6];
SL-W002 [CAS 388079-58-1]; SL-W003 [CAS 388079-60-5]; SL-
W005 [CAS 494227-30-4]; SL-W006 [CAS 494227-31-5]; SL-W008
[CAS 494227-32-6]; SL-W009 [CAS 494227-33-7]; SL-W022 [CAS
849925-29-7]; SL-M001 [CAS 210842-74-3]; SL-M002 [CAS
494227-35-9]; SL-M003 [CAS 494227-36-0]; SL-M004 [CAS
494227-37-1]; SL-M009 [CAS 793718-16-8]; SL-M012 [CAS
831226-37-0]; R-Binaphane [CAS 253311-88-5]; R-C3-Tunephos
[301847-89-2]; Rax-S,S-xylC3*-Tunephos [CAS 1613225-29-8];
S,S,R,R-Tangphos [CAS 470480-32-1]; S-Binapine [CAS 528854-
26-4]; R,R,S,S-Duanphos [CAS 528814-26-8]; catASium D-R [CAS
99135-95-2]; catASium T1 [CAS 868851-47-2]; catASium T3 [CAS
868851-50-7]; R-SDP [CAS 917377-74-3]; R-Xyl-SDP [CAS 917377-
75-4]; R,R-Me-Duphos [CAS 147253-67-6]; R,R-iPr-Duphos [CAS
136705-65-2]; R-Xyl-Segphos [CAS 850253-53-1]; R-Xyl-BINAP
[CAS 137219-86-4]; R-Segphos [CAS 244261-66-3]; S-Xyl-Pane-
phos [CAS 325168-88-5]; R-Xyl-Pphos [CAS 442905-33-1].
(11) The detailed experimental setup is described in the
Experimental Section
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.oprd.5b00382.
■
S
1H NMR, 13C NMR, HRMS, and HPLC analyses and
extended screen of solvents and additives in the
hydrogenation of 8 (PDF)
AUTHOR INFORMATION
Corresponding Author
*E-mail: staffan.po.karlsson@astrazeneca.com.
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
■
Notes
The authors declare no competing financial interest.
(12) Details are given in the Supporting Information
ACKNOWLEDGMENTS
■
(13) (a) Perez-Balado, C.; Willemsens, A.; Ormerod, D.; Aelterman,
W.; Mertens, N. Org. Process Res. Dev. 2007, 11, 237−240.
We thank our Separation Science Laboratory group for
determination of enantiomeric purities and our NMR special-
ists for structure elucidation. We also thank Dr. Anna Pettersen
for providing us with X-ray data for AZD2716.
(b) Palmgren, A.; Thorarensen, A.; Backvall, J.-E. J. Org. Chem.
̈
1998, 63, 3764−3768.
(14) CCDC 1437667 contains the supplementary crystallographic
data for AZD2716. These data can be obtained free of charge from
The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.
uk/data_request/cif.
REFERENCES
■
(1) Giordanetto, F. et al. Discovery of (2R)-3-[3-(5-benzyl-2-
carbamoylphenyl)phenyl]-2-methylpropanoic acid (AZD2716): a
novel, potent secreted phospholipase A2 inhibitor for the treatment
of coronary artery cardiovascular disease. Manuscript in preparation.
(2) Prieto, M.; Mayor, S.; Lloyd-Williams, P.; Giralt, E. J. Org. Chem.
2009, 74, 9202−9205.
(3) Suzuki, A. J. Organomet. Chem. 2002, 653, 83−90.
(4) Negishi, E. I.; Hu, Q.; Huang, Z.; Qian, M.; Wang, G.
Aldrichimica Acta 2005, 38, 71−88.
(5) (a) Baylis, A. B.; Hillman, M. E. D. Chem. Abstr. 1971, 77, 34174.
(b) Cai, J.; Zhou, Z.; Zhao, G.; Tang, C. Org. Lett. 2002, 4, 4723−4725
and references cited therein.
(6) Das, B.; Banerjee, J.; Chowdhury, N.; Majhi, A. Chem. Pharm.
Bull. 2006, 54, 1725−1727.
̀ ́
(7) (a) Verendel, J. J.; Pamies, O.; Dieguez, M.; Andersson, P. G.
Chem. Rev. 2014, 114, 2130−2169. (b) Tang, W.; Zhang, X. Chem.
Rev. 2003, 103, 3029−3069.
(8) (a) Appell, R. B.; Boulton, L. T.; Daugs, E. D.; Hansen, M.;
Hanson, C.; Heinrich, J.; Kronig, C.; Lloyd, R. C.; Louks, D.; Nitz, M.;
Praquin, C.; Ramsden, J. A.; Samuel, H.; Smit, M.; Willets, M. Org.
Process Res. Dev. 2013, 17, 69−76. (b) Sun, Y.; Krska, S.; Shultz, C. S.;
Tellers, D. M. In Asymmetric Catalysis on Industrial Scale: Challenges,
Approaches, and Solutions, 2nd ed.; Blaser, H.-U., Federsel, H.-J., Eds.;
Wiley-VCH: Weinheim, Germany, 2010; pp 333−374. (c) Boogers, J.
A. F.; Felfer, U.; Kotthaus, M.; Lefort, L.; Steinbauer, G.; de Vries, A.
H. M.; de Vries, J. G. Org. Process Res. Dev. 2007, 11, 585−591.
(d) Blaser, H.-U.; Pugin, B.; Spindler, F. Top. Organomet. Chem. 2012,
H
DOI: 10.1021/acs.oprd.5b00382
Org. Process Res. Dev. XXXX, XXX, XXX−XXX