Convenient method for synthesis of N-protected α-amino epoxides: Key intermediates for HIV protease inhibitors

Article abstract of DOI:10.1021/op100174j

A convenient method for synthesis of 2R,3S and 2S,3S N-Boc phenylalanine epoxides using readily available allylamine is described. Previous methods employed multistep synthetic routes from l-phenyl alanine that include use of m-chloroperbenzoic acid (m-CPBA) and a chromatographic method for purification of the desired diastereomers. Column purification could be eliminated by bringing in much improvement in the existing process. The process was further enhanced by replacing m-CPBA with oxone, an ecofriendly reagent advantageous for commercial application. The overall green process discussed involves the recovery and recycling of enantiomers of chiral allyl amines and judicial separation of diastereomers of the epoxides using simple economical methods.

Full text of DOI:10.1021/op100174j

ARTICLE
pubs.acs.org/OPRD
Convenient Method for Synthesis of N-Protected r-Amino
Epoxides: Key Intermediates for HIV Protease Inhibitors
A. John Blacker,§,‡ Mita Roy,*,† Sivaramkrishanan Hariharan,† Catherine Headley,§,# Abhay Upare,†
Ashutosh Jagtap,† Karuna Wankhede,† Sushil Kumar Mishra,† Dagadu Dube,† Sanjay Bhise,†
Sandesh Vishwasrao,† and Nitin Kadam†
†Piramal Healthcare Ltd., Nirlon Complex, Goregaon-E, Mumbai 400 063, India
§Piramal Healthcare (UK) Ltd., Leeds Road, Huddersfield, West Yorkshire HD1 9GA, United Kingdom^
S Supporting Information
b
ABSTRACT: A convenient method for synthesis of 2R,3S and 2S,3S N-Boc phenylalanine epoxides using readily available
allylamine is described. Previous methods employed multistep synthetic routes from L-phenyl alanine that include use of m-
chloroperbenzoic acid (m-CPBA) and a chromatographic method for purification of the desired diastereomers. Column purification
could be eliminated by bringing in much improvement in the existing process. The process was further enhanced by replacing m-
CPBA with oxone, an ecofriendly reagent advantageous for commercial application. The overall green process discussed involves the
recovery and recycling of enantiomers of chiral allyl amines and judicial separation of diastereomers of the epoxides using simple
economical methods.
1. INTRODUCTION
Optically active R-amino alkyl epoxides are key building
blocks for the synthesis of active pharmaceutical ingredients. In
particular, the two diastereomers 2R,3S and 2S,3S N-Boc phe-
nylalanine epoxides (1 and 2 respectively) form essential inter-
mediates for HIV protease inhibitors viz. Atazanavir1 (involves
2R,3S) as well as Saquinavir,2b Nelfanavir,3 and Fosamprenavir4
(involves 2S,3S). Several synthetic approaches reported in lite-
rature5a (Figure 1) start from L-phenyl alanine and build up to
synthons such as aldehyde (I),5b,5c 1,2-diols (II),5d,5e halomethyl
derivatives (III),5f-5h or R-substituted allyl amines (IV).5i
Figure 1. Conversion of amino acid to intermediates required for
synthesis of epoxides.
a technique adds not only to the number of steps but also to the
cost. This led us to consider classical resolution, and the scheme
intended was for making racemic R-alkyl allyl amine. A route was
designed (Scheme 2) using a simple ketone such as benzophe-
none for making a Schiff base of allyl amine 3 which on alkylation
with benzyl halide and subsequent hydrolysis would form
methods include the multistep transformations from the amino
Difficulties associated with the industrial application of these
racemic R-benzyl allyl amine 5. A “low waste green process”
approach was followed using diastereomeric crystallization cou-
pled with racemisation. The final step of epoxidation leads to the
desired diastereomers 1 and 2. Literature describes epoxidation
using m-chloroperbenzoic (m-CPBA) acid and use of a chroma-
tographic method for separation of the diastereomers. The haza-
rds associated with m-CPBA and cost ineffectiveness of column
acid and use of expensive reagents. Our approach was to use an
alternative route with a minimum number of steps, starting from
easily available raw material other than an amino acid and then
adopting a green synthetic process along the pathway. Com-
pound IV was chosen as the preferred precursor. Retrosynthetic
analysis of IV (Scheme 1) suggested commercially available allyl
amine, as a possible starting raw material.
R-Alkylation of allyl amine is enabled by its conversion to a
Schiff base. Stereoselective alkylation reaction is possible at this
stage using sterically hindered or chiral allyl imines.6 Use of such
Received: June 23, 2010
Published: September 23, 2010
r
2010 American Chemical Society
331
dx.doi.org/10.1021/op100174j Org. Process Res. Dev. 2011, 15, 331–338
|

Organic Process Research & Development
ARTICLE
Scheme 1. Retrosynthetic analysis for N-protected 1-substituted amine
Scheme 2. Synthetic route for the synthesis of 1 and 2
chromatography for an industrial large scale process incited us to
look for an alternate reagent and an effective separation method.
Thus herein we report a green industrially feasible process for
preparation of pharmaceutically important 2S,3S and 2R,3S N-
Boc phenylalanine epoxides using a single synthetic route.
with the chloride derivative. A number of bases were screened for
the reaction (Figure 2). Bases such as LiNH2, PhLi, and LiHMDS
showed practically no formation of alkylated imine 4 and a dis-
proportionate presence of only ∼60% unreacted imine 3. On the
other hand moving from sec-BuLi, to a mixture of n-BuLi and t-
BuOK, to only n-BuLi, a >40% increase in formation of 4 (44% to
84%) was observed. The studies led to the choice of n-BuLi as the
base for alkylation.
A subsequent solvent selection screening (Figure 3) showed
THF to be the best solvent for the reaction. Thus with benzyl
chloride in the presence of n-BuLi in THF, amine 5 could be iso-
lated in 74% yield with g97% purity (GC, a/a).
To develop a method for resolution of the racemic amine 5,
readily available chiral resolving agents, such as L-tartaric acid
(TA), L-ditolyl tartaric acid (DTTA), and N-acetyl-L-leucine
(NAL) were tried. The tartrate salt of amine (()-5 was found
to be highly soluble in various alcoholic solvents viz. methanol,
isopropanol, and n-propanol, thereby making crystallization
ineffective. A fruitful result was obtained with DTTA in methanol
whereby the DTTA salt of 5a was isolated with 90% ee
(corresponding to free base) in 27% yield (Table 1). Efforts to
2. RESULTS AND DISCUSSION
Reaction of allyl amine with benzophenone is described in
literature in the presence of TiCl4 with a reaction time of 48 h.
Replacing TiCl4 with a cheaper reagent such as Ti(OPri)4
resulted in the reduction of reaction time to 6 h and an increase
in yield from 88-92%7,8 to >98% (purity >96% GC a/a).
Another obvious advantage was easy handling of Ti(OPri)4
which enhanced the scope of application of the reaction on a
commercial scale (Scheme 2). The benzophenone used was
recovered in the subsequent reaction and purified for further
reuse.
As alkylation of allyl imine at the C-3 position is known to be
kinetically controlled, all reactions were carried out at -50 ꢀC.
The nature of the alkylating reagent was also a significant factor,
and the yield improved from 50% using benzyl bromide to 74%
332
dx.doi.org/10.1021/op100174j |Org. Process Res. Dev. 2011, 15, 331–338

Organic Process Research & Development
ARTICLE
Figure 2. Benzylation of 3 with benzyl chloride at -50 ꢀC in THF using different bases. Compound 3 (1 equiv) in THF (5 v/w) was cooled to -50 ꢀC.
A respective base (1 equiv) was added at -50 ꢀC over a period of 1 h. The reaction mixture was maintained at -50 ꢀC for an additional 1 h, followed by
addition of benzyl chloride (1 equiv) in THF (1 v/w). The reaction mass was then stirred for 1 h, quenched in EtOH, and analysed on GC.
Figure 3. Benzylation of 3 with benzyl chloride at -50 ꢀC in different solvents; n-BuLi was used as a base. Compound 3 (1 equiv) in respective solvent
(5 v/w) was cooled to -50 ꢀC. n-BuLi (1.6 M, 1 equiv) was added at -50 ꢀC over a period of 1 h. The reaction mixture was maintained at -50 ꢀC for an
additional 1 h, followed by addition of benzyl chloride (1 equiv) in respective solvent (1 v/w). The reaction mass was then stirred for 1 h, quenched in
EtOH, and analysed on GC.
Table 1. Resolution using chiral reagent (0.55 equiv) in
methanol
Table 3. Resolution using NAL (0.55 equiv) in different
solvents
Sr. No
Reagent
Volume (w/v)
%ee (S)
Yielda
Sr. No
Solvent
MeOH
Volume (w/v) %ee (S)
Yielda
1
2
3
TA
5
25
12
-
0
27
28
1
2
3
4
5
6
7
5
82
80
-
28
35
DTTA
90
n-PrOH
20
NAL
82
IPA
5
Salt highly soluble
a The yields are of the isolated products, based on racemate.
n-BuOH
20
68
80
84
80
33
35
40
36
MeOH/Toluene
n-PrOH/Toluene
n-PrOH/Toluene
3/20
7/20
10/10
Table 2. Resolution using DTTA (0.55 equiv) in different
solvents
a The yields are of the isolated products, based on racemate.
Sr. No
Solvent
MeOH
Volume (w/v)
%ee (S)
Yielda
1
2
3
4
5
25
90
50
0
27
33
solvent for this resolution (Table 3). In two crystallizations NAL
salt of 5a having 98% ee (corresponding to free base) was obtai-
ned in 35% overall yield. Recovery and reuse of binary solvent
was also studied. Binary solvent was recovered in 80% yield; the
desired ratio was readjusted and reused to achieve reproducible
results.
MeOH/water
MeOH/n-BuOH
IPA
15/2
20/6
12
38
0.1
100
90
IPA/water
13/2
17
a The yields are of the isolated products, based on racemate.
Thus N-acetyl-L-leucine (NAL) seems to be the chiral reagent
of choice for the resolution of racemic amine 5. Operating
through a combination of two crystallizations with maximizing
yield in the first and chiral purity in the second, an efficient reso-
lution process with NAL has thus been developed.
improve the yield further by using different solvents (Table 2) or
upgrade ee by a second crystallization were not successful.
N-Acetyl-L-leucine was then revisited and a subsequent sol-
vent selection screen showed n-PrOH/Toluene to be the best
333
dx.doi.org/10.1021/op100174j |Org. Process Res. Dev. 2011, 15, 331–338

Organic Process Research & Development
ARTICLE
Figure 4. Resolution of racemic allyl amine (()-5 using NAL.
Scheme 3. Racemization of 5b
As depicted in Figure 4 additional experimentation with the
filtrate of resolved S isomer 5a resulted in (i) recovery of the
racemic mixture, (ii) isolation of the R isomer 5b (Scheme 3)
having 98% ee, (iii) recovery of R-isomer 5b having 40% ee, (iv) a
97% material balance. The racemic mixture was recycled into the
process.
A scheme was designed for recycling of the R-isomer based on
racemisation of the N-acetyl or N-Boc derivative, followed by
extraction of the chiral proton using an alkoxide base such as t-
BuOK (Scheme 3). N-Acetyl derivative 8 was found to be unsta-
ble in the presence of t-BuOK, whereas N-Boc derivative 7 race-
mised effectively to a 60:40 ratio of R/S.
The scheme was then integrated into a semicontinuous “reso-
lution-recycle” process as depicted in Figure 5.
Epoxidation of allyl amine (S isomer) reported in literature
makes use of m-CPBA9 to form selectively the RS-isomer with
74% de. With long reaction hours and excess m-CPBA, the minor
SS epoxide is preferentially decomposed improving the de of RS-
epoxide to >98%. Disadvantages of the route include (i) a low
purity of the desired product and (ii) use of chromatographic
method for purification. These factors are due to generation of
byproducts oxazolidinones 9 and 10 formed from the decom-
position of epoxides (Figure 6).10
Investigation of the process revealed that other than oxazoli-
dinones excess m-CPBA also led to formation of a byproduct viz.
bis-benzoyl peroxide 11. It was observed that reducing the molar
equivalents from 4 to 1.5 enhanced the purity of the RS-isomer
334
dx.doi.org/10.1021/op100174j |Org. Process Res. Dev. 2011, 15, 331–338

Organic Process Research & Development
ARTICLE
Figure 5. “Resolution-recycle” process of amine (()-5.
Table 4. Epoxidation10 of 6 using m-CPBA in MDC at 40 ꢀC
HPLC (a/a)
m-CPBA
Time
(h)
Purity of
Sr. No
(equivalent)
IV
epoxide (%)
RS/SS
1
2
3
4
8
11
23
-
-
-
67
79
65
99:1
1.5
1
99:1
99.6:0.4
the epoxidation reaction, and the reaction was terminated just as
the allyl amine IV was consumed. Screening various solvents
suitable for isolation led to hexane, and RS-epoxide was isolated
in hexane in 90% purity (HPLC, a/a) and 80% de. Differential
decomposition of the isomers was then tried with benzoic acid,
acetic acid, and formic acid. In benzoic acid and acetic acid the
rate of decomposition was very slow. A comparable result with m-
CPBA (1.5 equiv) was obtained by using formic acid (1.0 equiv)
whereby in 8 h the de of RS-epoxide was >99% with 80% HPLC
purity (Table 5). Optimization with molar equivalents of formic
acid and reaction temperature (Table 6) resulted in a sustained
purity of the RS-epoxide of ∼90% (HPLC, a/a) with de > 99.5%!
The results enabled us to circumvent chromatographic purifica-
tion which was not possible earlier and use crystallization to
obtain RS-epoxide with g97% purity and 99.6% de albeit at 25%
yield (based on the diastereomeric mixture).
In pursuit of a greener process with “no decomposition” and
consequently a commendable yield, we were encouraged to
investigate other oxidizing reagents milder than m-CPBA viz.
H2O211 and oxone.12 The reaction rate was found to be very slow
in H2O2, and epoxidation remained incomplete. Reaction with
oxone was found to be much cleaner with no impurity formation
(even after a lengthy reaction time) and most importantly with
little difference in selectivity achieved with m-CPBA. As a result
epoxide 6 was isolated in quantitative yield with an RS/SS ratio of
73:27 (Figure 7). Triturating the above mixture with hexane at
low temperature allowed precipitation of a solid enriched in the
SS isomer. The RS isomer was subsequently isolated from the
filtrate in 71% yield (91% yield based on diastereomeric purity)
with 97% purity (HPLC, a/a) and 94% de. Thus by utilizing
Figure 6. Impurities formed during m-CBPA epoxidation.
while maintaining its de of >99% (Table 4). Reducing the
quantity of m-CPBA further to 1.0 mol equiv increased the
reaction time required to decompose the SS-isomer to <1%
which again affected the purity of the RS-isomer. Thus keeping
a balance between the reaction time and quantity of m-CPBA,
1.5 mol equiv of m-CPBA was derived as the optimal condition
for epoxidation. But even under these conditions purification
by any method other than column chromatography failed to
succeed.
An approach was needed which would allow formation of the
RS-isomer with the desired de while sustaining its purity, which as
a corollary would facilitate the purification. Due to the fact that
the differential rate of decomposition of the diastereomers was
acid initiated, the search for a mild acid which would generate
minimal byproduct was a viable option. Experiments were plan-
ned where m-CPBA was used only to the point of completion of
335
dx.doi.org/10.1021/op100174j |Org. Process Res. Dev. 2011, 15, 331–338

Organic Process Research & Development
ARTICLE
epoxidation with oxone rather than m-CPBA, significant im-
provements were achieved: (i) decomposition and formation
of byproducts were avoided; (ii) a high yield was achieved; (iii)
column chromatography was circumvented; (iv) the process
was simplified by use of crystallization for purification. A
further added benefit was the isolation of the SS-epoxide 2 in
90% de and 98% purity (HPLC, a/a) in 18% yield. This
amounts to an overall 89% recovery of combined diastereo-
mers. Thus an ultimate green process was developed, which
allowed judicial segregation of the two respective 2R,3S and
2S,3S N-Boc phenylalanine epoxides of equal pharmaceutical
importance.
3. CONCLUSIONS
In conclusion, by employing readily available allylamine and
an ecofriendly reagent, a convenient industrially feasible method
for the synthesis of the amino epoxides in high diastereomeric
purity has been developed. The scheme includes resolution at
two stages: (i) enantiomers of allylamine and (ii) diastereomers
of epoxide. In recognition of the necessity for favorable process
economics, an approach of “minimal loss and maximum exploi-
tation” in resolution was adapted. This led to (1) the recycling of
waste enantiomer and (2) a process for making optically pure
chiral building blocks: (i) S-allyl amine 5a, (ii) R-allyl amine 5b,
(iii) RS-amino epoxide 1, and (iv) SS-amino epoxide 2.
Table 5. Decompositiona of epoxides in different acids (1
mol equiv) in MDC at 40 ꢀC
4. EXPERIMENTAL SECTION
Instrumentation and Materials. 1H NMR. ABruker300MHz
instrument with tetramethylsilane as reference was used. Che-
mical shifts were reported as (δ) values in parts per million. The
following abbreviations were used: s, singlet; d, doublet; m,
multiplet; dd, doublet of doublets; and br s, broad signal.
13C NMR. Spectra were recorded at 75 MHz.
MS. A Shimadzu QP 2010 instrument was used.
Melting Points. A Lab India MR-V/S melting point apparatus
was used, and the measurements are uncorrected.
FT-IR. Perkin-Elmer spectrophotometer.
HPLC (a/a)
Time
(h)
Purity of
Sr. No
Acid
pKa
epoxide (%)
RS/SS
1
2
3
4
m-CBPA
3.82
4.2
23
24
63
8
65
84
88
80
99.6:0.4
94:6
Benzoic acid
Acetic acid
Formic acid
4.76
3.77
97:3
99:1
a Crude epoxide (RS/SS 90:10) purity 90% (HPLC, a/a) was taken in
MDC. To this respective acid, 1 equiv was added and heated to 40 ꢀC.
Optical Rotations. Jasco P-1020 polarimeter at λ 589 nm and
25 ꢀC.
Techniques. Reactions involving air- or moisture-sensitive
reagents or intermediates were performed under nitrogen.
Solvents. THF (anhydrous), MTBE (methyl tertiary butyl
ether), toluene, MeOH, CH2Cl2, EtOAc, and n-propanol were
purchased from RFCL Ltd.
Table 6. Decompositiona study of epoxide in formic acid in
MDC
Temp
Time
(h)
HPLC (a/a)
Sr. No
Mol equiv
(ꢀC)
Purity of 1 (%)
RS
Reagents. Allyl amine, benzophenone, titanium isopropox-
ide, N-acetyl-L-leucine, and DTTA were purchased commercially
from Arrow Biochem. Benzyl bromide was purchased from
Spectrochem. n-BuLi was purchased from ACROS ORGANICS.
GC Analytical Method. For the GC analysis of allyl amine, 3,
4, 5, IV, and 7, a Perkin-Elmer Clarus 500 GC chromatograph
1
2
3
2
3
3
40
40
25
4
2
8
86
86
89
99.9
99.9
99.8
a Crude epoxide (1 equiv) was taken in MDC. To this formic acid, 1
equiv was added and heated to 40 ꢀC.
Figure 7. Schematic presentations for isolation of epoxides.
336
dx.doi.org/10.1021/op100174j |Org. Process Res. Dev. 2011, 15, 331–338

Organic Process Research & Development
ARTICLE
equipped with an FID detector was used. The GC method used
included the following: capillary column, DB-1 30 m ꢀ 0.32 mm,
1.0 μm with USP G1 packing; carrier gas, nitrogen with inlet
pressure of 12 psi; temperature of injector and detector, 250 ꢀC; oven
programme, 50 ꢀC, 5 min hold, ramped to 300 ꢀC @10 ꢀC/min,
10 min hold. Retention times of the compounds observed were
as follows: Allyl amine, Rt 1.8 min; 3, Rt 24.5 min; 4, Rt 29.8 min;
5, Rt 16.2 min; IV and 7, Rt 22.0 min.
HPLC Analytical Methods. HPLC Method for Achiral Analysis
for 1 and 2. Column, Eclipse XDB 4.6 mm ꢀ15 cm, 5 μm, with USP
L1 packing; flow rate, 1.5 mL/min; mobile phase, water-acetonitrile;
programme, 40% acetonitrile kept for 15 min and then increased to
80% in next 15 min, then kept constant for 7 min, and re-equilibrate
for 5 min. Sample was prepared at 1000 ppm concentration in
acetonitrile. Retention times of the compounds observed were as
follows: 1, Rt 10.9 min; 2, Rt 10.1 min; IV and 7, Rt 21.5 min.
HPLC Method for Chiral Analysis. Column, Chiralcel OD-H
4.6 mm ꢀ 25 cm, 5 μm; flow rate, 0.5 mL/min. Mobile phase used
for compound 5a and 5b, n-hexane/methyl tert-butyl ether/trifluor-
oacetic acid (800:200:2), 960 mL of this solvent mixture with 40 mL
of EtOH, and 1.0 mL of diethyl amine; retention times of 5a and 5b,
Rt 15.1 and 16.2 min. Mobile phase used for compounds 1, 2, IV,
and 8, n-hexane/IPA (98:2); flow rate for 1 and 2, 0.9 mL/min and
that for IV and 7, 0.5 mL/min; retention times as follows: 1, Rt 11.0
min; 2, Rt 12.0 min; IV, Rt 13.5 min; 7, Rt 13.0 min. Mobile phase
used for compound 7, n-hexane/EtOH/diethylamine (920:80:0.1);
flow rate, 0.4 mL/min; retention time of 8, Rt 15.3 min.
Preparation of 1,1-Diphenyl-2-aza-1,4-pentadiene (3). To a
solution of benzophenone (300.0 g, 1.64 mol) in toluene (1500 mL)
was added allyl amine (140.7 g, 2.46 mol) followed by titanium
isopropoxide (300.3 g, 1.05 mol). After heating at 80 ꢀC for 6 h, the
mass was cooled to 25-26 ꢀC, and water (35 mL) and toluene
(1200 mL) were added. The reaction mass was stirred for 0.5 h, and
the white suspension formed (of TiO2) was filtered through Celite.
The residue was washed with toluene (2 ꢀ 750 mL). The
combined organic layer was concentrated under vacuum at
60-65 ꢀC to give 356 g (98%) of 3 as a yellow oil: 97% GC (a/
a) purity (Rt 24 min). IR (neat): 3080, 1661, 1622, 1597, 1313,
1287 cm-1. 1HNMR(300MHz, CDCl3):δ7.78-7.76 (m, 2H, Ar
—H), 7.44-7.28 (m, 6H, Ar—H), 7.18-7.15 (m, 2H, Ar—H),
6.20-6.07 (m, 1H, dN—CHdCH2), 5.25 (dd, 1H, J = 1.5, 9.6
Hz, dN—CHdCH2), 5.15 (dd, 1H, J = 1.5, 9.6 Hz, dN—
CHdCH2), 4.10 (d, 2H, J = 5.4 Hz, —CH2—Ph) ppm. 13C NMR
(75 MHz, CDCl3) δ 142.3, 138.7, 129.3, 128.3, 126.2, 113.5, 55.3,
44.1 ppm. MS m/z 221 [M]þ.
Preparation of (()3-Amino-4-phenyl-1-butene (5). A
solution of compound 3 (336 g, 1.52 mol) in THF (1680 mL)
was cooled to -50 ꢀC, and n-BuLi (953 mL, 1.52 mol) was added
dropwise over a period of 1 h. During the addition a wine red colored
suspension was formed. The reaction mass was stirred at -50 ꢀC for
an additional 1 h. Benzyl chloride (192 g, 1.52 mol) in THF (192
mL) was then added to the mixture, over a period of 10 min. The
reaction mass was stirred for an additional 1 h, and the formation of
compound 4 was monitored by GC. The reaction was quenched
with ethanol (60 mL). KHSO4 solution (20%, 1700 mL) was added
to the reaction mass at a rate that maintains the internal temperature
below -30 ꢀC. The mixture was warmed to room temperature and
stirred for 9-10 h (for hydrolysis). During this period two layers
were formed. Jthe organic layer (THF) was separated, and the
aqueous layer containing the product was washed with methyl tert-
butyl ether (350 mL) followed by CH2Cl2 (2 ꢀ 350 mL). The
aqueous layer was cooled to 0 ꢀC and basified with 50% NaOH
solution (412 mL) to pH 9-10. The product was extracted in
CH2Cl2 (2 ꢀ 350 mL). The combined CH2Cl2 layer was washed
with brine (2 ꢀ 350 mL), dried over Na2SO4, and concentrated
in vacuum togive164g(74%) of5as a light brown colored oil: 96%
GC(a/a) purity (Rt 16.2 min.), enantiomeric ratio S/R 55:45. IR
(neat): 3367, 3289, 1642, 1602, 1584 cm-1. 1H NMR (300 MHz,
CDCl3) δ 7.03-7.16 (m, 5H, Ar—H), 5.67-5.79 (m, 1H,
CH2dCH—CH—), 5.01 (dd, 1H, J = 1.2, 10.5 Hz, —N—
CHdCH2), 4.90 (d, 1H, J = 1.2, 10.5 Hz, —N—CHdCH2), 3.4
(q, 1H, J = 6, 13.2, —N—CHdCH2), 2.6 (dd, 1H, J = 5.4, 13.2
Hz, —CH2Ph), 2.46 (dd, 1H, J = 8.1, 13.2 Hz, —CH2Ph) ppm.
13C NMR (75 MHz, CDCl3) δ 142.3, 138.7, 129.3, 128.3, 126.2,
113.5, 55.3, 44.1 ppm. MS m/z 147 [M]þ.
Procedure for Recovery of Benzophenone. The methyl
tert-butyl ether extract was combined with the organic layer
(THF) and concentrated to obtain 345 g of a brown colored oil
containing 80% benzophenone (GC, a/a). The oil was distilled
under reduced pressure (0.5 mm) at 160 ꢀC. The fraction
distilling out at 117-125 ꢀC was collected to give 250 g of
benzophenone with 96% purity (GC, a/a).
General Procedure for Resolution. To a solution of 1 equiv
of amine 5 in toluene (20 volumes) and n-PrOH (7 volumes) was
added N-acetyl-L-leucine (0.55 equiv). The reaction mixture was
heated to 95 ꢀC and then cooled slowly at 5 ꢀC/h to 20 ꢀC. The
reaction mass was stirred for 2-3 h at 20 ꢀC. The filtered solid
(NAL salt) was dried under vacuum and underwent a second
crystallization following the above procedure. The solid thus
obtained was hydrolyzed with 10% NaOH, and the free base was
extracted in CH2Cl2 to afford optically pure amine.
Resolution of (S)-3-Amino-4-phenyl-1-butene (5a). Compound
5 (100 g, 0.680 mol) was resolved with N-acetyl-L-leucine (64 g,
37.4 mmol) in 2000 mL of toluene and 700 mL of n-PrOH as per
procedure mentioned above to give 35g (35%) of 5a as a colorless
oil: 98% GC(a/a) purity (Rt 16.2 min), ee 98% (Chiral HPLC).
[R]25D = þ15.1 (c 1, CHCl3).
Resolution of (R)-3-Amino-4-phenyl-1-butene (5b). The
mother liquour from the first crystallization was concentrated to
obtain 80 g of residue. The residue was stirred in hexane (300
mL) and filtered to obtain the NAL salt of (R)-3-amino-4-
phenyl-1-butene in 82% ee (32 g). The hexane filtrate was
evaporated to obtain the free amine of R-isomer 5b in 50% ee,
which on resolution with NAL afforded the NAL salt of the R-
isomer (27 g) in 85% ee. The combined NAL salt (59 g) was
underwent a second crystallization followed by hydrolysis. (R)-3-
Amino-4-phenyl-1-butene (5b) was obtained as a colorless oil
(23.2 g, yield 23%); purity 98% (GC a/a, Rt 15.1 min), ee 98%
(HPLC); [R]25D = -15.3 (c 1, CHCl3).
General Procedure for Preparation of N-Boc Derivative.
To a stirred solution of amine (1 equiv) in CH2Cl2 (5 volumes)
was added triethylamine (1.25 equiv) followed by the dropwise
addition of di-tert-butyl dicarbonate (1.25 equiv) over a period of
1 h. The reaction was monitored by GC. After 28 h the reaction
mixture was quenched with 5% KHSO4 solution. The organic
layer was washed with water followed by brine, dried over
Na2SO4, and concentrated. Hexane was added to the residue,
and the product was isolated from hexane.
Preparation of (S)-2-(t-Boc-amino)-1-phenylbut-3-ene
(IV). White solid; 55 g (95% yield), 98% GC (a/a, Rt 22 min);
mp = 67-68 ꢀC, [R]25D = þ36.9 (c 1, CHCl3).
Preparation of (R)-2-(t-Boc-amino)-1-phenylbut-3-ene
(7). White solid; yield 54.5 g (94%), purity 98% (GC a/a, Rt
22 min); mp 67-68 ꢀC, [R]25D = -34.8 (c 1, CHCl3).
337
dx.doi.org/10.1021/op100174j |Org. Process Res. Dev. 2011, 15, 331–338

Organic Process Research & Development
ARTICLE
Racemization of 7. To a stirred solution of 7 (20 g, 0.081 mol)
in toluene (200 mL) was added t-BuOK (4.5 g, 0.041 mol). The
reaction mixture was heated to reflux for 48 h. The reaction
was monitored by chiral HPLC. On completion of racemization
after 48 h, the toluene layer was washed with water and dried over
Na2SO4. The toluene evaporated to obtain a crude residue which
was then hydrolysed using 6 N HCl (120 mL) in MeOH
(30 mL). The reaction was monitored by TLC. After 4 h the
aqueous layer was washed with CH2Cl2 (150 mL) and basified
with 20% NaOH (450 mL). The product was extracted with
CH2Cl2 (3 ꢀ 150 mL). All CH2Cl2 layers were combined and
washed with a brine solution (200 mL), dried over Na2SO4, and
evaporated to dryness affording 6 g (67%) of racemic 5 as an oil:
98% GC (a/a) purity.
Development, University of Leeds, Woodhouse Lane, Leeds LS2
9JT, United Kingdom.
#The University of Manchester, Sackville Street Building, Sack-
ville Street, Manchester M13 9PL, United Kingdom.
DISCLOSURE
^This manufacturing unit for Piramal has been closed.
’ ACKNOWLEDGMENT
Authors are thankful to Piramal Life Sciences Ltd., India for
providing NMR facility.
’ REFERENCES
Preparation of N-((S)-1-Benzyl-allyl)-acetamide (8). To a
stirred solution of amine 5b (20 g, 0.136 mol) in dichloro-
methane (100 mL) was added triethylamine (17.1 g, 0.17 mol)
followed by acetic anhydride (15.2 g, 0.145 mol) dropwise over a
period of 1 h. The reaction was monitored by GC. After 28 h the
reaction mixture was quenched with 5% KHSO4 solution. The
organic layer was washed with water and a brine solution, dried
over Na2SO4, and evaporated. The crude solid was stirred with
hexane (60 mL) and filtered to obtain 20.6 g (80%) compound 8
as a white solid: 99% GC (a/a) purity (Rt 15.9 min). Mp =
116-117 ꢀC, [R]25D = -49.5 (c 1, CHCl3).
(1) (a) Charrier, N.; Clarke, B.; Cutler, L.; Demont, E.; Dingwall, C.;
Dunsdon, R.; East, P.; Hawkins, J.; Howes, C.; Hussain, I.; Jeffrey, P.;
Maile, G.; Matico, R.; Mosley, J.; Naylor, A.; O’Brien, A.; Redshaw, S.;
Rowland, P.; Soleil, V.; Smith, K. J.; Sweitzer, S.; Theobald, P.; Vesey, D.;
Walter, D. S.; Wayne, G. J. Med. Chem. 2008, 51, 3313–3317. (b) Ghosh,
A. K.; Kumaragurubaran, N.; Hong, L.; Kulkarni, S.; Xu, X.; Miller, H. B.;
Reddy, D. S.; Weerasena, V.; Turner, R.; Chang, W.; Koelsch, G.; Tang,
J. Bioorg. Med. Chem. Lett. 2008, 18, 1031–1036.
(2) (a) Parkes, K. E. B.; Bushnell, D. J.; Dunsdon, S. J.; Freeman,
A. C.; Funnm, M. P.; Hopkin, R. A.; Lambert, R. W.; Martin, J. A.;
Merrett, J. H.; Redshaw, S.; Spurden, W. C.; Thomas, G. J. J. Org.
Chem. 1994, 59, 3656–3664. (b) Goehring, W.; Gokhale, S.; Hilpert, H.;
Roessler, F.; Schalengeter, M.; Vogt, P. Chimia 1996, 50, 532–537.
(3) (a) Al-Farhan, E.; Deininger, D. D.; McGhie, S. S.; O’Callaghan, J.;
Robertson, M. S.; Rodgers, K.; Rout, S. J.; Singh, H.; Tung, R. D. Glaxo
group Ltd. PCT Int. Appli., WO9948885. (b) The Organic Chemistry of
Drug Synthesis; Lednicer, D., Ed.; Wilely Interscience: New York, 2007; Vol. 7.
(4) Cunico, W.; Gomes, C. R. B.; Moreth, M.; Manhanini, D. P.;
Figueiredo, I. H.; Penido, C.; Henriques, M. G. M. O.; Varotti, F. P.;
Krettli, A. U. Eur. J. Med. Chem. 2009, 44, 1363–1368.
(5) (a) Izawa, K.; Onishi, T. Chem. Rev. 2006, 106, 2811–2827. (b)
Green, B. E.; Chen, X.; Norbeck, D. W.; Kempf, D. J. Synlett 1995, 613.
(c) Nogami, H.; Kanai, M.; Shibasaki, M. Chem. Pharm. Bull. 2003, 51,
702. (d) Xu, Z.; Singh, J.; Schwinden, M. D.; Zheng, B.; Kissick, T. P.;
Patel, B.; Humora, M. J.; Quiroz, F.; Dong, L.; Hsieh, D.-M.; Heikes,
J. E.; Pudipeddi, M.; Lindrud, M. D.; Srivastava, S. K.; Kronenthal, D. R.;
Mueller, R. H. Org. Process Res. Dev. 2002, 6, 323.(e) Koshigoe, T.; Satoh,
H.; Yamamoto, K. European Patent 657446 , 1995.(f) Liu, C.; Ng, J. S.;
Behling, J. R.; Yen, C. H.; Campbell, A. L.; Fuzail, K. S.; Yonan, E. E.;
Mehrotra, D. V. Org. Process Res. Dev. 1997, 1, 45. (g) Proctor, L. D.;
Warr, A. J. Org. Process Res. Dev. 2002, 6, 884.(h) Malik, A. A. European
Patent 1050532A2 , 2000.(i) Branalt, J.; Kvarnstrom, I.; Classon, B.;
Samuelsson, B.; Nillroth, U.; Danielson, U. H.; Karlen, A.; Hallberg, A.
Tetrahedron Lett. 1997, 38, 3483.
Preparation of (2R,3S)-3-(tert-Butoxycarbonyl)amino-1,2-
epoxy-4-phenylbutane (1) and (2S,3S)-3-(tert-Butoxycarbo-
nyl)amino-1,2-epoxy-4-phenylbutane (2). A solution of
compound IV (20 g, 8.1 mmol) in acetone (235 g, 8.1 mol)
was added to sodium bicarbonate (340 g, 4.05 mol) solution in
mixture of water (3400 mL) and ethyl acetate (1700 mL). Oxone
(797 g, 1.296 mol) was added to this biphasic mixture in 8 lots
(each lot 99.6 g) at intervals of 1 h. The reaction was monitored
by HPLC. The reaction mixture was filtered after 16 h, and the
inorganic residue was washed with ethyl acetate (2 ꢀ 200 mL).
The combined organic layer was washed with brine (2 ꢀ 200 mL),
dried over Na2SO4, and concentrated under vacuum. Hexane
(40 mL) was added to the residue and was filtered to give
21.5 g (100%) of 6 as a white solid: 96% HPLC (a/a) purity;
diastereomeric ratio RS/SS 73:27 (HPLC). The solid was taken
in hexane (160 mL) and heated to 45 ꢀC. The solution was then
cooled to 5 ꢀC, stirred for 1 h, and filtered. Both the solid (7.3 g)
and the filtrate were treated separately. The filtrate was concen-
trated to 75% of the volume, cooled to -5 ꢀC, and filtered to give
14.2 g (71%) of 1 as a white solid: 97.2% HPLC (a/a) purity (Rt
12.7 min); diastereomeric ratio ratio RS/SS 97:3 (HPLC).
The solid (7.3 g) was recrystallized thrice from 45 mL of
hexane to give 3.0 g (14%) of 2 as a white solid: 99% HPLC (a/a)
purity (Rt 11.7 min): diastereomeric ratio SS/RS 98:2 (HPLC).
(6) (a) Stakemeier, H.; Wurthwein, E.-U. Liebigs Ann. 1996, 1833–
1843and references cited therein. (b) Barrett, A. G. M.; Seefeld, M. A.;
White, A. J. P.; Williams, D. J. J. Org. Chem. 1996, 61, 2677–2685.
(7) Wolf, G.; Wurthwein, E.-U. TetrahedronLett.1988, 29, 3647–3650.
(8) Moretti, I.; Torre, G. Synthesis 1970, 141.
(9) Luly, J. R.; Dellaria, J. F.; Plattner, D. J.; Soderquist, J. L.; Yi, N.
J. Org. Chem. 1987, 52, 1487–1492.
’ ASSOCIATED CONTENT
(10) Rich, H.; Romeo, S. Tetrahedron Lett. 1994, 35, 4939–3942.
(11) (a) Lianhe Shu, L.; Yian Shi, Y. Tetrahedron 2001, 57, 5213–
521. (b) Lane, B. S.; Vogt, M.; J. DeRose, V. J.; Kevin Burgess, K. J. Org.
Chem. 1983, 48, 890–892.
S
Supporting Information. Experimental details. This ma-
b
terial is available free of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
(12) Hashimoto, N.; Kanda, A. Org. Process Res. Dev. 2002, 6, 405–406.
Corresponding Author
*E-mail: mita.roy@piramal.com.
’ NOTE ADDED AFTER ASAP PUBLICATION
In the version published September 23, 2010, there was an author
omitted, and the affiliation for one of the other authors was not clear.
These have been corrected in the version posted February 10, 2011.
Present Addresses
‡School of Chemistry, School of Process, Environmental
and Materials Engineering, Institute of Process Research and
338
dx.doi.org/10.1021/op100174j |Org. Process Res. Dev. 2011, 15, 331–338