An efficient catalytic asymmetric synthesis of a β2-amino acid on multikilogram scale

Article abstract of DOI:10.1021/op4002966

We describe herein a scalable catalytic asymmetric hydrogenation process for the multikilogram-scale production of a β ²-amino acid. A short and efficient synthesis of the starting unsaturated N-Boc-protected β ²-enamide was developed followed by extensive catalysis screening and optimization studies that identified a simple Ru-BINAP catalyst system to directly afford the (S) product in high enantiomeric excess and yield. The final process enabled the multikilogram production in >99% ee, to be used as a key component for one of our clinical candidates.

Full text of DOI:10.1021/op4002966

Article
pubs.acs.org/OPRD
An Efficient Catalytic Asymmetric Synthesis of a β2‑Amino Acid on
Multikilogram Scale
Travis Remarchuk,*,† Srinivasan Babu,† Jeffrey Stults,† Antonio Zanotti-Gerosa,‡ Stephen Roseblade,‡
Shaohui Yang,§ Ping Huang,§ Chunbo Sha,§ and Youchu Wang§
†Small Molecule Process Chemistry, Genentech Inc., A member of the Roche Group, 1 DNA Way, South San Francisco, California
94080
‡Catalysis and Chiral Technologies, Johnson Matthey, Unit 28, Cambridge Science Park, Cambridge CB4 0FP, United Kingdom
§WuXi AppTec Co., Ltd., No. 1 Building, #288 FuTe ZhongLu, WaiGaoQiao Free Trade Zone, Shanghai 200131, P.R. China
S
* Supporting Information
ABSTRACT: We describe herein a scalable catalytic asymmetric hydrogenation process for the multikilogram-scale production
of a β2-amino acid. A short and efficient synthesis of the starting unsaturated N-Boc-protected β2-enamide was developed
followed by extensive catalysis screening and optimization studies that identified a simple Ru-BINAP catalyst system to directly
afford the (S) product in high enantiomeric excess and yield. The final process enabled the multikilogram production in >99% ee,
to be used as a key component for one of our clinical candidates.
accessing β2-amino acid (S)-2 (Figure 2). Installation of the
chirality in the last step of the synthesis is a strategic advantage
INTRODUCTION
■
The efficient synthesis of chiral β2-amino acid derivatives, as key
components of medicinally active compounds, remains a
challenge especially for kilogram-scale production.1 Several
methods are known for the synthesis of these types of amino
acids, and a few examples are illustrated in Figure 1. One of the
Figure 1. Selected asymmetric methods to access β2-amino acid
derivatives.
most widely used methods includes the asymmetric amino-
methylation, or Mannich reaction, that employs chiral
auxiliaries such as the Evans oxazolidinone to efficiently set
the stereochemistry.2a More recent advances include a report
by Davies and co-workers of an enantioselective intermolecular
C−H insertion reaction of amino methyl derivatives with aryl
diazoesters and catalytic chiral dirhodium complexes, that
shows high utility.2b Also noteworthy is the organocatalytic
asymmetric transfer hydrogenation reactions of β-nitroacrylates
reported by List and co-workers.2c New methodologies
continue to be developed for the synthesis of this important
class of compounds. From a process perspective, however, the
catalytic asymmetric hydrogenation (CAH)3 of β2-enamide
derivatives was envisioned to be the most efficient method for
Figure 2. Process chemistry routes for scale-up of β2-amino acid (S)-2.
to minimize the loss of chiral product. In addition, numerous
chiral transition metal catalysts are available for asymmetric
hydrogenation on large scale, and this became the focus of our
development efforts starting with the β2-enamide precursor
(E)-1.
Received: October 17, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/op4002966 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Z-olefin geometries) as determined by HPLC analysis. Boc
protection of the vinylogous carbamate was performed at this
stage of the process since attempted ester hydrolysis with
NaOH resulted in a retro-Claisen/hydrolysis byproduct.7
Reaction of the enamine with Boc anhydride was sluggish
and required up to 8 equiv of this reagent in the presence of
catalytic DMAP and with heating in order to completely
consume the mixture of enamine esters 4a/4b.8 Interestingly,
during the Boc-protection step, both isomers 4a/4b were
consumed, but only one olefin isomer of 5 was isolated.
Presumably, this is a result of isomerization of the less reactive
Z-isomer and concomitant protection of the E-isomer although
this was not confirmed.9 The unsaturated ester (E)-5 was
isolated in 68% yield overall from the starting material ECPA.
The final step in the synthesis of (E)-1 was hydrolysis of the
ethyl ester group with aqueous NaOH to afford the unsaturated
carboxylic acid product in 90% yield and >99 A% purity by
HPLC as a single geometrical isomer. The E-olefin geometry
was confirmed by single-crystal X-ray structural analysis, and
this substrate was used for all catalyst screenings and
development work leading up to the scale-up of (S)-2.
RESULTS AND DISCUSSION
■
Our first-generation process route to (S)-2 involved use of the
Evans’ chiral (R)-4-benzyloxazolidinone auxiliary in compound
(R)-6 (Figure 2) to direct the asymmetric Mannich reaction
with the iminium ion derived from N-(alkoxymethyl)carbamate
7.4 Using this approach, the initial kilogram quantities of
enantiomerically pure (S)-2 was produced in ∼20% overall
yield from 2-(4-chlorophenyl)-acetic acid. Although the key
asymmetric Mannich reaction provided alkylation product
(S,R)-8 in high diastereoselectivity (>20:1 dr) this route
required the use of stoichiometric auxiliary, and undesirable
reagents such as MOMBr, TiCl4, H2O2, etc. In addition,
purification of the crude 2 was laborious and required Boc-
deprotection and formation of the sodium carboxylate salt (S)-
9-Na which was key for purification as well as crystallization to
increase the enantiopurity from ∼98% ee to >99% ee. Finally,
the Boc group had to be put back on in the last step. This was a
fit-for-purpose route to supply our initial kilogram quantities of
(S)-2; however, it was not a practical long-term manufacturing
synthesis.
Retrosynthetically, the second-generation approach we
investigated was the catalytic asymmetric hydrogenation of
Boc-protected enamide (E)-1 illustrated in Figure 2. A very
large number of catalysts are available for the asymmetric
hydrogenation of α-aminoesters(acids) and β-substituted-β-
aminoesters(acids).5a A more limited number of examples are
available for the hydrogenation of α-methylaminoacrylates
leading to β2-type amino acids.5b At the start of our work, there
were no reported examples for the asymmetric hydrogenation
of a hindered Boc-protected alkyl amine substrate possessing an
unsaturated α-aryl carboxylic acid of the type 2.5c It was also
unclear how important the olefin bond geometry would be and
if the hindered amino group would be detrimental to finding a
practical catalyst system to apply on large scale.
Chiral Catalyst Screening. A catalyst screen was designed
to examine common and readily available chiral metal
complexes of Rh, Ru, and Ir that would produce β2-amino
acid 2 in high enantiomeric excess. Over 70 experiments with
>25 different Rh and Ru metal complexes were performed, and
a representative subset of these experiments is shown in Table
1 (illustrated in eq 1). At this preliminary stage the screen was
Synthesis of Boc-Enamide Substrate (E)-1. Preparation
of enamide (E)-1 substrate from commercially available ethyl 4-
chlorophenyl acetate (ECPA) is illustrated in Scheme 1.
Formylation of ECPA with tert-butoxide and ethyl formate gave
the known Claisen product 3 in 83% HPLC assay yield.6
Treatment of crude 3 with isopropylamine and acetic acid
afforded a ∼70:30 mixture of enamine products 4a/4b (E- and
Scheme 1. Preparation of Boc-enamide (E)-1
performed using either CH2Cl2 or MeOH as solvent in the
presence, or absence, of triethylamine as a base additive under
mild reaction conditions (35 °C, 10 bar H2 with substrate to
catalyst (S/C) loading = 50:1 mol/mol). Among the rhodium
10a
catalysts tested, [Phanephos Rh(COD)]BF4 and [Ph-BPE
10b
Rh(COD)]BF4
catalysts gave complete conversion in
MeOH in the presence of 0.7 equiv of triethylamine, with
modest enantioselectivities (entries 1 and 2, Table 1). The
10c
rhodium complex [(R,R)-BDPP-RhCOD]BF4 gave 80% ee
of (R)-2 in complete conversion (entry 3, Table 1). Higher
enantioselectivities were observed using the chiral ruthenium
BINAP complexes with both enantiomers of [BINAP-RuCl2]-
10d
(DMF)n to give (R)- or (S)-2 in >95% ee with complete
conversion (entries 4B, 4C and 5, Table 1). Contrary to the
case of rhodium catalysts, the addition of Et3N to the carboxylic
acid substrate was found to be detrimental to both conversion
and ee. Chiral ruthenium BINAP complexes provided the
highest enantioselectivity among the catalysts tested.11 They
B
dx.doi.org/10.1021/op4002966 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
differences in performance among different ruthenium
complexes with increasing S/C ratio. We concluded that the
ligands on the ruthenium precatalysts would lead to higher
conversion and enantioselectivity in the order of benzene > p-
cymene > DMF.
Having identified [BINAP-RuCl(benzene)]Cl as the most
active and selective catalyst, optimization continued with this
system to determine the effectiveness of additives to improve
both reactivity and enantioselectivity. The addition of various
salts including NaOTf, AgOTf, LiBF4, AgBF4, and KPF6 was
examined to evaluate the potential for catalyst activation by
displacing the chloride anion from the coordination sphere of
the cationic ruthenium complex.14 These additives led to an
increase of reactivity by the catalyst and allowed for lower
catalyst loadings (up to 2500/1) and higher concentrations of
substrate (up to 0.6 M). The results in Table 3 show that, at
Table 1. Representative Rh and Ru catalysts screened for
conversion of (E)-1 to (S)- or (R)-2.
a
b
entry
1
catalyst
solvent
Et3N, equiv conv., %
ee, %
[(R)-Phanephos-
Rh(COD)]BF4
A. CH2Cl2
B. MeOH
MeOH
0.7
72
>99
>99
17
53 (S)
40 (S)
57 (R)
37 (R)
80 (R)
0.7
2
3
[(R,R)-Ph-BPE-
Rh(COD)]BF4
A. 0.7
B. none
0.7
[(R,R)-BDPP-Rh
COD]BF4
MeOH
CH2Cl2
>99
[(S,S)-BDPP-Rh
0.7
12
52 (S)
COD]BF4
4
5
[(S)-BINAP-
RuCl2](DMF)n
A. MeOH
B. MeOH
C. CH2Cl2
MeOH
0.7
35
>99
>99
>99
35 (S)
>95 (S)
>95 (S)
>95 (R)
none
none
none
[(R)-BINAP-
RuCl2](DMF)n
a
b
All reactions were conducted in a Biotage Endeavor. The (S)
configuration was confirmed by comparison to a reference standard
prepared from the first-generation synthesis of (S)-2, illustrated in
Figure 2.
Table 3. Additives screened with [BINAP-
a
RuCl(benzene)]Cl
pressure,
entry
1
additive
LiBF4
S/C
[S]/M
0.375
bar
conv., %
>99
ee, %
also had the advantage of providing lower catalysts’ cost
contribution to the process over other catalysts.
1000/1
10
96 (R)
(4 mol %)
NaOTf
(2 mol %)
AgBF4
(2 mol %)
AgOTf
(2 mol %)
KPF6
(2 mol %)
LiBF4
(5 mol %)
NaOTf
(5 mol %)
LiBF4
(2 mol %)
LiBF4
(2 mol %)
LiBF4
(5 mol %)
After having identified ruthenium-BINAP as the most
promising chiral catalyst in Table 1, the nature of the
ruthenium precatalyst, the catalyst loading, and effect of
substrate concentration were examined using the (R)-
enantiomer of the catalyst. Table 2 shows representative
examples from eq 1, of reactions at 50 °C and 10 bar hydrogen
in MeOH and EtOH, both solvents having emerged from a
solvent screen as leading to the highest reactivity.12 Cationic
complexes of the type [(R)-BINAP-RuCl (L)]Cl (L= benzene
or p-cymene)13 gave a slight improvement in enantioselectivity
relative to the less characterized [BINAP-RuCl2](DMF)n
complex at a S/C = 100/1 and substrate concentration of
0.05 M (entry 1A vs entries 2A and 3A, Table 2). In the case of
the [(R)-BINAP-RuCl (p-cymene)]Cl complex, the use of
MeOH gave lower enantioselectivities and yields than the use
of EtOH (entries 3A and 3E, Table 2). In general, lower
catalyst loadings (S/C >100/1) required the use of higher
pressures (30 bar) and temperatures (up to 80 °C; entry 2D,
Table 2) to achieve decent conversions and to obtain lower
enantioselectivities. Increased substrate concentration up to 0.5
M with lower catalyst loading (entry 2C, Table 2) led to
complete conversion to product at S/C = 1000/1 but with only
92% ee. The results from Table 2 suggest there were significant
2
1000/1
1000/1
1000/1
1000/1
1000/1
1000/1
2500/1
5000/1
2500/1
0.375
0.375
0.375
0.375
0.6
10
10
20
10
5
>99
99
96 (R)
96 (R)
94 (R)
95 (R)
94 (S)
94 (S)
97 (S)
96 (S)
96 (S)
3
4
>99
95
5
6
93
7
0.6
5
64
8
0.6
20
20
25
61
9
0.6
15
10
0.5
>99
a
Reaction temperature = 65 °C and analyzed after 24 h in a Parr
autoclave.
0.375 M substrate concentration and S/C = 1000/1, three
additives gave identical conversion and enantioselectivity
(entries 1−3, Table 3). Of these three salts, LiBF4 was
a
Table 2. Optimization experiments with BINAP-ruthenium catalysts to afford (R)-2
entry
1
catalyst
S/C
[S]/M
solvent
conv.,%
ee, %
[(R)-BINAP-RuCl2](DMF)n
A. 100/1
B. 250/1
A. 100/1
B. 500/1
C. 1000/1
D. 1000/1
A. 100/1
B. 250/1
C. 500/1
D. 1000/1
E. 100/1
0.05
0.13
0.05
0.37
0.5
A. EtOH
B. MeOH
EtOH
EtOH
>99
>99
>99
>99
>99
>99
>99
>99
63
95
90
98
2
[(R)-BINAP-RuCl (benzene)]Cl
[(R)-BINAP-RuCl (p-cymene)]Cl
b
95
b
EtOH
92
c
0.2
EtOH
88
3
0.05
0.13
0.25
0.5
EtOH
EtOH
99
98
d
EtOH
96
b
EtOH
30
84
0.05
MeOH
32
73
a
b
All reactions were conducted in a Biotage Endeavor. Hydrogen reaction pressure = 30 bar and reaction temperature 65 °C using a Parr autoclave.
c
d
Hydrogen reaction pressure = 30 bar and reaction temperature 80 °C using a Parr autoclave. Hydrogen pressure = 30 bar.
C
dx.doi.org/10.1021/op4002966 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 3. Hydrogen update curves for conversion of (E)-1 to (S)-2. S/C = 1000/1 mol/mol, EtOH (0.375 M), H2 (20 bar), 60 °C, 16 h, 2−4 mol %
additive.
shown to be superior to NaOTf at [S]/M = 0.6 M (entries 6
and 7, Table 3) and was favored over the use of the silver salts.
Upon further optimization of S/C loading and substrate
concentration, the limit using LiBF4 additive was found to be
S/C = 2500/1 at 0.5 M, 65 °C, and 25 bar pressure of
hydrogen to afford complete conversion and 96% ee. The
hydrogen uptake was also measured at 60 °C and 20 bar
hydrogen to visualize the effect of the additives on reaction
profiles, and the plots from the addition of 2 mol % AgOTf
(blue) and 4 mol % LiBF4 (red) are shown in Figure 3.15 There
was a significant increase in the rate of hydrogen uptake with
these additives relative to no additives added (green curve).
Scale-up of β2-Amino Acid (S)-2. On initial 20 g scale, the
optimized reaction conditions were used with minor
modifications to suit the available equipment: S/C = 2000/1,
EtOH (0.6 M, 5 vol), H2 (35 bar), and 5 mol % LiBF4, the
hydrogenation of (E)-1 proceeded in 95% conversion after 17 h
and gave (S)-2 in 98.9% ee. After recrystallization from
heptane, the optical purity of (S)-2 increased to >99% and 86%
yield.16 With increasing scale, a similar catalyst loading was
found to achieve similar full conversion. Six batches ranging
from 4−35 kg of (E)-1 (Table 4) were subjected to reaction
produce a total of 126 kg of (S)-2 in >99.9% ee and >99A%
purity by HPLC (Scheme 2).
Process Mass Intensity (PMI).17 The process mass
intensity (PMI) defined as the total mass (kg) of materials
that go into each step divided by the mass of product (kg) was
used to compare our first- and second-generation process
routes to (S)-2. In the first-generation auxiliary route, a total of
8.9 kg of (S)-2 was produced (in seven synthetic steps) and
required on average 575 kg of materials input to produce one
kilogram of product (blue bar in Figure 4). In contrast, the
second-generation catalytic asymmetric hydrogenation (CAH)
route (six synthetic steps) used on average 146 kg of input for
every kilo of product. The lower PMI for the CAH route
represents an overall 4-fold reduction in mass input (and waste
stream) leading to a greener process. A major factor
contributing to the higher PMI in the chiral auxiliary route
was the purification of (S)-2 for ee enhancement. As described
earlier, Boc deprotection, formation of the sodium salt (S)-9-
Na, crystallization, and Boc protection was required that added
significant processing. In contrast, purification of crude (S)-2 in
the CAH route was achieved by an efficient crystallization.
Although the CAH route required up to 8 equiv of Boc2O to
prepare compound (E)-5, further optimization is ongoing to
reduce the number of equivalents which could further reduce
the overall PMI of this route.18 In a direct route comparison of
the organic solvents used (defined as preferred, usable,
undesirable, and excluding water), there is an overall 6-fold
reduction in solvents used in the second generation hydro-
genation process.19
a
Table 4. Scale-up batches of crude (S)-2
(E)-1 (S)-2 in EtOH
purity (HPLC
area) (%)
purity (w/w
assay) (%)
%
ee
(kg)
(kg)
1
2
3
4
5
6
4
20
75
98.8
98.2
98.6
98.2
98.7
98.4
20
20
21
21
20
21
98
97
97
97
97
98
15
35
178
163
167
76
33
33
15.2
CONCLUSION
■
a
Reactions were performed under identical conditions as described in
In summary, we have reported an efficient and robust catalytic
enantioselective hydrogenation process to generate hundred
kilogram quantities of (S)-2 in high optical purity using a
readily available BINAP-Ru catalyst. Production of the
penultimate (E)-1 olefin substrate was achieved in a four-
step, two-pot process from commercially available ethyl 4-
chloro phenylacetate in ∼50% overall yield. Obtaining (E)-1 as
a single geometric isomer and developing hydrogenation
conditions suitable for the use of BINAP-Ru at acceptably
low catalyst loadings were key to the success of this project.
Scheme 2 and gave ∼100% yield of crude product. See Supporting
Information for analytical methods.
conditions similar to those above, with S/C = ∼2000/1. All six
reactions proceeded smoothly with complete conversion to
afford crude 2 with 97−98% ee. The combined batches of
crude product were treated with SiliaMetS thiol to scavenge
residual ruthenium and were crystallized from heptane to
D
dx.doi.org/10.1021/op4002966 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 2. Kilogram scale-up synthesis of (S)-2
4a+4b/5 = 8.7 A%. Additional triethylamine (54 kg, 535 mol),
(Boc)2O (95 kg, 436 mol) and DMF (43 kg) were added over
3 h at 60 °C, resulting in 4a+4b/5 = 5.4 A% by HPLC.
The reaction mixture was cooled to 25 °C and charged with
MTBE (900 kg), water (840 kg), and 10% aqueous citric acid
(800 kg) with agitation for 1.5 h. The aqueous layer was
separated, and the organic layer was washed with 25% aqueous
NaCl (820 kg) with agitation for 1.5 h. The aqueous layer was
removed, and the solution was concentrated via vacuum
distillation to a minimum working volume of ∼250 L (<50 °C).
Heptane (80 kg) was charged to the reactor, and the solution
was cooled to ∼5 °C and stirred for 5 h. The crystallized
product was filtered and washed with heptane (12 kg) and
dried under vacuum to afford (E)-5 (155.95 kg, 96.1 A% purity
by HPLC) in 68% yield (over three steps) as white solid. Loss
1
in the mother liquors was ∼9%. H NMR (500 MHz, DMSO-
Figure 4. Solvent usage and PMI for auxiliary and catalytic asymmetric
d6) 7.76 (s, 1H), 7.42 (d, J = 8.6 Hz, 2H), 7.22 (d, J = 8.6 Hz,
2H), 4.13 (q, J = 7.1 Hz, 2H), 3.74 (h, J = 6.8 Hz, 1H), 1.32 (s,
9H), 1.18 (t, J = 7.1 Hz, 3H), 1.08 (d, J = 6.8 Hz, 6H); 13C
NMR (125 MHz, DMSO-d6) 166.7, 152.0, 138.8, 133.8, 132.0,
131.5 (2C), 128.0 (2C), 116.7, 81.6, 60.4, 49.7, 27.5 (3C), 19.9
(2C), 14.0; HRMS: m/z calculated for C19H26ClNO4 [M + H]+
= 368.1623; found 368.1616; mp = 88.6 °C (onset) by DSC.
Preparation of N-Boc-Enamide Carboxylic Acid (E)-1.
A 20% solution of NaOH (88 kg in 370 kg H2O) was charged
to a solution of (E)-5 (200.0 kg, 543.6 mol) and EtOH (450
kg) with agitation at 25 °C. The mixture was warmed to 45 °C
and maintained until a clear solution was formed (∼3 h). In-
process analysis by HPLC showed 1/5 = not detected. The
mixture was concentrated under vacuum to a minimum
working volume at ≤60 °C. The mixture was cooled to ∼20
°C and charged with EtOAc (1375 kg) and aqueous 2 N HCl
(880 kg) to obtain a pH between 4−5. The aqueous layer was
separated, and the organic layer was washed with 25% aqueous
NaCl (1210 kg). The organic layer was concentrated under
vacuum (<50 °C) and then charged with n-heptane (280 kg);
this process was then repeated. The solids were filtered, washed
with n-heptane (3 × 24 kg), and then dried under vacuum at
∼45 °C for 10 h to afford (E)-1 (133 kg, 87% yield, 99.9A%
purity by HPLC) as white solid. Losses in the mother liquor
were found to be 7.6%. 1H NMR (500 MHz, DMSO-d6) 12.38
(bs, 1H), 7.70 (s, 1H), 7.41 (d, J = 8.6 Hz, 2H), 7.21 (d, J = 8.6
Hz, 2H), 3.78 (h, J = 6.8 Hz, 1H), 1.30 (s, 9H), 1.08 (d, J = 6.8
Hz, 6H); 13C NMR (125 MHz, DMSO-d6) 168.1, 152.0, 138.3,
134.3, 131.8, 131.4 (2C), 127.9 (2C), 118.0, 81.3, 49.5, 27.5
(3C), 20.0 (2C); KF = 0.1%; ROI = 0.02%). HRMS: m/z
calculated for C17H22ClNO4 [M − H]− = 338.1165; found
338.1161; mp = 163.8 °C (onset) by DSC.
hydrogenation routes.
EXPERIMENTAL SECTION
■
Preparation of N-Boc-enamide Ethyl Ester (E)-5. A
mixture of MTBE (766.7 kg) and t-BuOK (114.6 kg, 1023 mol,
2.0 equiv), was cooled to −5 °C. A solution of ethyl formate
(93.6 kg, 1263.2 mol, 2.5 equiv), MTBE (142 kg), and ethyl 4-
chlorophenylacetate (100 kg, 503 mol, 1.0 equiv) was charged
≤5 °C. The mixture was stirred between 0−10 °C for 1.5 h and
deemed complete by HPLC analysis: 3 = 0.8 A% by HPLC.20
The reaction mixture was added to aqueous HCl (35%, 114.6
kg in 454 L H2O) at ≤10 °C. The mixture was stirred for 30
min between 0−10 °C (pH = 1.5). The layers were separated,
and the organic layer was washed with 25% aqueous NaCl
solution (500 kg). HPLC analysis showed starting material and
a mixture of keto−enol tautomers of compound 3 correspond-
ing to 79.7% assay yield.
The reaction mixture was cooled to −5 °C, and isopropyl-
amine (62 kg, 1051 mol, 2.0 equiv) was charged followed by
the slow addition of AcOH (62 kg, 1033 mol, 2.0 equiv) at a
rate to maintain internal temperature <10 °C. The mixture was
stirred for 3 h between 0−10 °C and then monitored for
consumption of β-formyl ester 3. In-process analysis of the
reaction mixture by HPLC showed 3/4a+4b = 0.9 A%. The
organic layer was washed consecutively with H2O (640 kg),
15% aqueous Na2CO3 (350 kg), and then 25% aqueous NaCl
(500 kg). The separated organic layer was analyzed by HPLC:
4a+4b = 92.7A% purity.
The organic layer was concentrated by vacuum distillation to
∼250 L (<50 °C), and then the reactor was charged with DMF
(299 kg), and the organics were concentrated again to ∼250 L
(<50 °C). The reaction mixture was charged with DMAP (12
kg, 98.4 mol, 0.2 equiv) and triethylamine (152.7 kg, 1512 mol,
3 equiv) and heated to 60 °C. A solution of (Boc)2O (755 kg,
3457 mol, 7 equiv) and DMF (165 kg) were slowly added over
24 h at this temperature. In-process analysis by HPLC showed
Hydrogenation of N-Boc-(E)-Enamide Acid 1. Hydro-
genation Screen in Biotage Endeavor. Catalyst and substrate
1 were weighed into glass Endeavor reaction vials. The vials
were purged with nitrogen, and solvent (anhydrous, Fluka) was
added. The reaction was purged with hydrogen and run at the
E
dx.doi.org/10.1021/op4002966 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
required temperature and pressure for 18 h. A drop of the
reaction mixture was diluted in EtOH and analyzed by HPLC.
Column: Diacel Chiralpak AD-H, 0.46 cm × 25 cm, eluent: n-
hexane/i-PrOH 97/3, 1 mL/min, T = 30 °C, wavelength
detection = 223 nm, Rt of (R)-2 = 9.8 min; Rt of (S)-2 = 12.1
min.
Hydrogenation in Parr Autoclave. [(R)-BINAP-RuCl-
(benzene)]Cl (2.5−5 mg) and substrate 2 (1.0 g, 2.9 mmol)
were weighed into a 25 mL Parr autoclave fitted with overhead
stirrer. The autoclave was sealed and purged with nitrogen.
EtOH (6−8 mL) was added via syringe, and the reaction was
purged with hydrogen and heated to 65 °C under 30 bar H2
pressure for 18 h with stirring. The vessel was then cooled to
room temperature, vented, and sampled for HPLC analysis.
The reaction mixture was filtered through a short pad of Celite
and washed with EtOAc (10−5 mL), and the solvents were
removed in vacuo. Drying at 50−60 °C under high vacuum (<5
mbar) gave the product as a viscous oil.
3H); 13C NMR (125 MHz, DMSO-d6) 173.4, 154.1, 136.8,
131.8, 130.1 (2C), 128.3 (2C), 78.7, 51.0, 48.4, 47.8, 27.9 (3C),
20.2 (2C); Metal analysis for Ru: <10 ppm; ROI = 0.04%;
HRMS: m/z calculated for C17H24ClNO4 [M − H]−
=
340.1321; found 340.1316; [α]24 = −97.63 (c 0.2, MeOH).
D
ASSOCIATED CONTENT
* Supporting Information
■
S
General information, analytical methods, and additional
characterization data of compounds (E)-5, (E)-1, (S)-2. This
material is available free of charge via the Internet at http://
pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: travisr@gene.com
■
Notes
The authors declare no competing financial interest.
Large-Scale Hydrogenation Representative Procedure. A
mixture of (E)-1 (15.0 kg, 44.1 mol), EtOH (56 kg), LiBF4
(0.21 kg, 2.2 mol, 5 mol %), and [(S)-BINAP-RuCl(benzene)]
Cl (0.020 kg, 0.023 mol; S/C = 1919:1 mol/mol = 750/1 wt/
wt) were charged to a reactor.21 The agitated mixture was
vacuum degassed with nitrogen (evacuated to ≤−0.08 MPa and
bleeding nitrogen to atmospheric pressure) for five times,
evacuated to ≤−0.08 MPa and bleeding hydrogen to ∼34 bar
once, then evacuated to ≤−0.08 MPa again and pressurized
with hydrogen gas at ∼34 bar and maintained at ∼55 °C for
∼26 h. In-process analysis by HPLC showed complete
consumption of enamide (E)-1. The mixture was cooled to
∼25 °C and transferred to a holding drum and then analyzed
by HPLC to have an assay = 20.3%, purity = 98.2%, and 97.4%
ee.
ACKNOWLEDGMENTS
■
We thank Angela Zhou for outsourcing support, Alan Deese for
NMR spectral data, Christine Gu for HRMS, Michael Dong
and Derrick Yazzie for analytical; Francis Gosselin, Remy
Angelaud and David Askin for comments and suggestions. In
addition, we thank Dr. Hans Nedden (JM CCT) for the
preparation of catalyst [(R)-BINAP-RuCl(benzene)]Cl.
REFERENCES
■
(1) For reviews, see: (a) Juaristi, E., Soloshonok, V. Enantioselective
Synthesis of β-Amino Acids, 2nd ed.; Wiley-VCH: New York, 2005.
(b) Saladino, R.; Botta, G.; Crucianelli, M. Mini-Rev. Med. Chem. 2012,
12, 277−300. (c) Weiner, B.; Szyman
A. J.; Feringa, B. L. Chem. Soc. Rev. 2010, 39, 1656−1691. (d) Seebach,
D.; A. Beck, A. K.; Capone, S.; Deniau, G.; Groselj, U.; Zass, E.
Synthesis 2009, 1−32.
(2) (a) Evans, D. A.; Urpi, F.; Somers, T. C.; Clark, J. S.; Bilodeau,
M. T. J. Am. Chem. Soc. 1990, 112, 8215−8216. (b) Davies, H. M. L.;
Venkataramani, C. Angew. Chem., Int. Ed. 2002, 41, 2197−2199.
(c) Martin, N. J. A.; Cheng, X.; List, B. J. Am. Chem. Soc. 2008, 130,
13862−13863.
́
ski, W.; Janssen, D. B.; Minnaard,
A total of six batches were processed under identical reaction
conditions described above, and the results are summarized in
Table 4. All reactions proceeded to ≥100% assay yield and
totalled 138.7 kg in EtOH solution. The six batches of crude
(S)-2 were combined, scavenged for Ru, and crystallized to
produce one batch as described below.
̌
Metal Scavenging and Purification of Combined Batches
of Crude (S)-2. An EtOH solution of crude (S)-2 (∼20% assay
in EtOH solution = 138.7 kg in 680 kg EtOH) was
concentrated under vacuum via distillation (<50 °C) to ∼250
L to which was then added EtOAc (999 L). The mixture was
washed with 25% aqueous NaCl (3 × 700 L) and then the
organic layer concentrated under vacuum to ∼600 L at ≤40 °C.
To the solution was charged SiliaMetS thiol (8.30 kg), and the
mixture was agitated for 14 h at ∼50 °C. After cooling to ∼30
°C the mixture was filtered and washed with EtOAc (40 L).
The filtrate was concentrated to ∼140 L at ≤42 °C and n-
heptane (2 × 485 L) was charged in portions with continuous
distillation to form a suspension. The suspension was agitated
for 1.5 h at ∼50 °C and then for 16 h at ∼0 °C. The product
was collected by filtration and washed with n-heptane (4 × 229
L). The filter cake was dried under vacuum at ∼45 °C for 10 h
to afford compound β2-amino acid (S)-2 (126.24 kg, 91% yield,
99.6A% purity by HPLC and >99.9% ee) as white solid. The
filtrate contained (S)-2 product in ∼80% ee; attempted
recrystallization from EtOAc afforded (S)-2 with 86% ee and
40% yield recovery. 1H NMR (500 MHz, DMSO-d6) 12.60 (bs,
1H), 7.40 (d, J = 8.46, 2H), 7.31 (d, J = 8.46, 2H), 3.86 (bs,
1H), 3.63 (dd, J = 6.8, 14.2 Hz, 1H), 3.35 (dd, J = 7.2, 14.0 Hz,
1H), 1.36 (s, 9H), 1.01 (d, J = 6.6 Hz, 3H), 0.90 (d, J = 6.8 Hz,
(3) (a) Shimizu, H.; Nagasaki, I.; Matsumura, K.; Sayo, N.; Saito, T.
Acc. Chem. Res. 2007, 40, 1385−1393. (b) Blaser, H.-U.; Spindler, F.;
Studer, M. Top. Organomet. Chem. 2012, 42, 65−102. (c) Ager, D. J.;
de Vries, A. H. M.; de Vries, J. G. Chem. Soc. Rev. 2012, 41, 3340−
3380.
(4) The discovery synthesis of (S)-2 using the Evans chiral auxiliary
has been reported: Blake, J. F.; Xu, R.; Bencsik, J. R.; Xiao, D.; Kallan,
N. C.; Schlachter, S.; Mitchell, I. S.; Spencer, K. L.; Banka, A. L.;
Wallace, E. M.; Gloor, S. L.; Martinson, M.; Woessner, R. D.; Vigers,
G. P. A.; Brandhuber, B. J.; Liang, J.; Safina, B. S.; Li, J.; Zhang, B.;
Chabot, C.; Do, S.; Lee, L.; Oeh, J.; Sampath, D.; Lee, B. B.; Lin, K.;
Liederer, B. M.; Skelton, N. J. Discovery and Preclinical Pharmacology
of a Selective ATP-Competitive Akt Inhibitor (GDC-0068) for the
Treatment of Human Tumors. J. Med. Chem. 2012, 55, 8110−8127.
(5) (a) For a recent review on the synthesis of amines via catalytic
asymmetric hydrogenation see: Xie, J.-H.; Zhu, S.-F.; Zhou, Q.-L.
Chem. Rev. 2011, 111, 1713−1760. (b) For a review on asymmetric
hydrogenation of β-substituted-β-(acylamino)acrylates see: Bruneau,
C.; Renaud, J.-L.; Jerphagnon, T. Coord. Chem. Rev. 2008, 252, 532−
544. (c) Phospholane-based Rh catalysts gave 58−67% ee on α-aryl-β-
amidoacrylates: Elaridi, J.; Thaqi, A.; Prosser, A.; Jackson, W. R.;
Robinson, A. J. Tetrahedron: Asymmetry 2005, 16, 1309−1319. For
alternative examples of synthesis of β2-amino acid/ester via rhodium-
catalyzed asymmetric hydrogenation of α-(acylaminomethylen)-β-
substituted acrylates see: (d) Luhr, S.; Holz, J.; Zayas, O.; Wendisch,
̈
V.; Borner, A. Tetrahedron: Asymmetry 2012, 23, 1301−1319 and the
̈
F
dx.doi.org/10.1021/op4002966 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
references therein. (e) Hoen, R.; Tiemersma-Wegman, T.; Procuranti,
B.; Lefort, L.; de Vries, J. G.; Minnaard, A.; Feringa, B. L. Org. Biomol.
Chem. 2007, 5, 267−275. (f) Qiu, L.; Prashad, M.; Hu, B.; Prasad, K.;
Repic, O.; Blacklock, T. J.; Kwong, F. Y.; kok, S. H. L.; Lee, H. W.;
Chan, A.S. C. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 16787−16792.
(g) Deng, J.; Hu, X. P.; Huang, J. D.; Yu, S. B.; Wang, D. Y.; Duan, Z.
C.; Zheng, Z. J. Org. Chem. 2008, 73, 2015−2017. (h) Huang, H.; Liu,
X.; Deng, J.; Qiu, M.; Zheng, Z. Org. Lett. 2006, 8, 3359−3362.
(6) Preparation of 3 was reported using NaH and ethyl formate:
(a) Xiao, Z.-P.; Fang, R.-Q.; Li, H.-Q.; Xue, J.-Y.; Zheng, Y.; Zhu, H.-L.
Eur. J. Med. Chem. 2008, 43, 1828−1836.
(19) (a) ACS GCI Pharmaceutical Roundtable Solvent Selection Guide,
Version 2.0; American Chemical Society Green Chemistry Institute:
Washington, D.C., Issued April 1, 2011. (b) Henderson, R. K.;
Jimenez-Gonzalez, C.; Constable, D. J. C.; Alston, S. R.; Inglis, G. G.
́ ́
A.; Fisher, G.; Sherwood, J.; Binksa, S. P.; Curzons, A. D. Green Chem.
2011, 13, 854−862.
(20) This reaction is reversible and longer reaction times did not
result in complete consumption of starting material.
(21) The reactor was pretreated with an acetone/NaHCO3 mixture,
followed by acetone/water, and then concentrated HNO3. The reactor
was thoroughly rinsed with water until pH = 7, and then a final EtOH
reflux was performed.
(7) Attempts to hydrolyze the ethyl ester mixture of 4a/4b with 20%
aqueous NaOH in EtOH from 20−60 °C resulted in retro-Claisen/
hydrolysis to afford 4-chlorophenyl acetic acid.
(8) Alternative solvents to DMF for the Boc protection, such as
acetonitrile and THF, did not improve the conversion. Noteworthy is
the Boc-protection of compound (S)-9-Na from Figure 2 that
proceeded in quantitative conversion to product (S)-2 using only
1.7 equiv of (Boc)2O. This observation suggests the unsaturation of
enamine ester 4 possesses a steric constraint for the Boc-protection,
leading to enamide (E)-5.
(9) By HPLC analysis of the crude reaction mixture, there is no
evidence for decomposition of isomer 4b; see Supporting Information
for the in-process control (IPC) of step 3 [4a and 4b/5 = 5.4%].
(10) (a) PhanePhos = 4,12-Bis(diphenylphosphino)-[2.2]-para-
cyclophane; Rh-Phanephos: Pye, P. J.; Rossen, K.; Reamer, R. A.;
Tsou, N. N.; Volante, R. P.; Reider, P. J. J. Am. Chem. Soc. 1997, 119,
6207−6208. (b) Ph-BPE = 1,2-Bis(2,5-diphenylphospholano)ethane:
Pilkington, C. J.; Zanotti-Gerosa, A. Org. Lett. 2003, 5, 1273−1275.
(c) BDPP = (2S,4S)-(-)-2,4-Bis(diphenylphosphino)-pentane: Bakos,
J.; Toth, I.; Marko, L. J. Org. Chem. 1981, 46, 5427−5428. (d) BINAP
= 2,2′-Bis(diphenylphosphino)-1,1′-binaphthyl: Kitamura, M.;
Tokunaga, M. Ohkuma, T.; Noyori, R. Organic Syntheses; Wiley and
Sons: New York; 1998; Collect. Vol. 9, p 589; 1998; Vol. 71, p 1.
(11) Two chiral iridium catalysts, (a) ([(R)-iPr-PHOX Ir(COD)]
BArF and (b) [(R)-PPhos Ir(COD)Cl]), have also been tested under
similar reaction conditions but were unreactive. For reference to (a)
see: Lightfoot, A.; Schnider, P.; Pfaltz, A. Angew. Chem., Int. Ed. 1998,
37, 2897−2899. For reference to (b) the catalyst was prepared mixing
[Ir(COD)Cl2]2 and P-Phos: P-Phos: 2,2′,6,6′-tetramethoxy-4,4′-
bis(diphenylphosphino)-3,3′-bipyridine: Pai, C.-C.; Lin, C.-W.; Lin,
C.-C.; Chen, C.-C.; Chan, A. S. C.; Wong, W. T. J. Am. Chem. Soc.
2000, 122, 11513−11514.
(12) Solvents including THF, EtOAc, and toluene were also screened
with the [BINAP-RuCl2](DMF)n catalyst under the conditions
described in Table 2; however, <5% conversion to product occurred.
Similar to MeOH and EtOH, CH2Cl2 also afforded high optical purity
of the product but was not desirable to use on large scale due to the
environmental impact of the waste stream.
(13) [BINAP-RuCl (L)]Cl: Mashima, K.; Kusano, K.; Sate, N.;
Matsumura, Y.; Nozaki, K.; Kumobayashi, H.; Sayo, N.; Hori, Y.;
Ishizaki, T.; Akutagawa, S.; Takaya, H. J. Org. Chem. 1994, 59, 3064−
3076.
(14) Isolated complexes of the type [BINAP-RuCl (L)]X (X = BF4,
I) are known; see ref 13.
(15) The initial negative values are attributed to equilibration of the
equipment to heating, during which some gas expansion occurs before
hydrogen uptake begins.
(16) The residual Ru in this lot was ∼100 ppm before crystallization
from heptane and ∼30 ppm after crystallization. Metal scavenging was
performed before the crystallization for all subsequent scale-up
batches.
(17) Jimenez-Gonzalez, C.; Ollech, C.; Pyrz, W.; Hughes, D.;
́ ́
Broxterman, Q. B.; Bhathela, N. Org. Process Res. Dev. 2013, 17, 239−
246.
(18) For comparison, use of only 1 equiv (Boc)2O vs 8 equiv used in
the current process, would reduce the overall PMI from 146 to 142
(see Table S-3 of Supporting Information), but has a significant impact
on the total cost of reagents used.
G
dx.doi.org/10.1021/op4002966 | Org. Process Res. Dev. XXXX, XXX, XXX−XXX