Chiral synthesis of an α-tetrasubstituted proline derivative

Article abstract of DOI:10.1080/00397911.2010.501477

Triple reuptake inhibitors (TRIs) have been identified as possible alternative treatments for central nervous system disorders. A practical synthesis was investigated to access a potent TRI, (4-amino-3-chloro-phenyl)- ((S)-2-isobutyl-pyrrlidin-2-yl)- methanone (1), starting from alkylation of D-proline. Improved routes to key intermediates are discussed and demonstrated by an expedient multigram production of 1. Copyright

Full text of DOI:10.1080/00397911.2010.501477

This article was downloaded by: [University Of Pittsburgh]
On: 12 October 2014, At: 13:47
Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Synthetic Communications: An
International Journal for Rapid
Communication of Synthetic Organic
Chemistry
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/lsyc20
Chiral Synthesis of an α-Tetrasubstituted
Proline Derivative
Eric R. Humphreys a , Jun-Bae Hong a & Keena L. Green a
a Roche Palo Alto LLC , Palo Alto , California , USA
Published online: 25 May 2011.
To cite this article: Eric R. Humphreys , Jun-Bae Hong & Keena L. Green (2011) Chiral Synthesis
of an α-Tetrasubstituted Proline Derivative, Synthetic Communications: An International
Journal for Rapid Communication of Synthetic Organic Chemistry, 41:15, 2256-2264, DOI:
10.1080/00397911.2010.501477
To link to this article: http://dx.doi.org/10.1080/00397911.2010.501477
PLEASE SCROLL DOWN FOR ARTICLE
Taylor & Francis makes every effort to ensure the accuracy of all the information (the
“Content”) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.
This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions

Synthetic Communications1, 41: 2256–2264, 2011
Copyright # Roche Palo Alto LLC
ISSN: 0039-7911 print=1532-2432 online
DOI: 10.1080/00397911.2010.501477
CHIRAL SYNTHESIS OF AN a-TETRASUBSTITUTED
PROLINE DERIVATIVE
Eric R. Humphreys, Jun-Bae Hong, and Keena L. Green
Roche Palo Alto LLC, Palo Alto, California, USA
GRAPHICAL ABSTRACT
Abstract Triple reuptake inhibitors (TRIs) have been identified as possible alternative
treatments for central nervous system disorders. A practical synthesis was investigated
to access
a potent TRI, (4-amino-3-chloro-phenyl)-((S)-2-isobutyl-pyrrlidin-2-yl)-
methanone (1), starting from alkylation of D-proline. Improved routes to key intermediates
are discussed and demonstrated by an expedient multigram production of 1.
Keywords Asymmetric alkylation; chiral quaternary carbon; proline
INTRODUCTION
The biogenic amine transmitters serotonin, norepinephrine, and dopamine
have been recognized as links to depressive and anxiolytic disorders. Triple reuptake
inhibitors (TRIs), which selectively block the reuptake of serotonin, norepinephrine,
and dopamine, have emerged as potentially effective treatments for depression
and other central nervous system indications.[1] As a part of a drug discovery
program, a small molecule, (4-amino-3-chloro-phenyl)-((S)-2-isobutyl-pyrrlidin-2-
yl)-methanone (1, Fig. 1), was identified as a potent TRI by our medicinal chemistry
team and investigated as a possible treatment for depression.[2] Herein, we describe
process development toward a chiral a-tetrasubstituted proline derivative 1 that has
demonstrated an efficient and scalable synthesis of this drug candidate.
The use of a bulky alkyl aldehyde to synthesize chirally pure a-tetrasubstitued
prolines is well known in the literature.[3] The tert-butyl group in bicyclic proline
derivative 3 has an S configuration induced by (R)-proline, which in turn regenerates
chirality in a-carbon during the reaction with an electrophile to give a quaternary
carbon center. The cleavage of this alkylated bicyclic intermediate required another
Received April 20, 2010.
Address correspondence to Jun-Bae Hong, Roche Palo Alto LLC, 3431 Hillview Avenue, Palo
Alto, California, USA. E-mail: jun-bae.hong@roche.com
2256

TETRASUBSTITUTED PROLINE
2257
Figure 1. Structure of desired isobutyl a-proline.
strong nucleophile, aryl lithium species in our synthesis, which allowed us to access
to the target molecule. The scale-up of this synthesis was complicated by problems
encountered during the synthesis of intermediates 3, 4, 5, and 6. Improvements that
allowed the synthesis of these intermediates to be scaled are discussed. Initial inves-
tigation of the synthesis showed steps 1 and 2 and the purity of bissilated intermedi-
ate 6 (Scheme 1) limited overall yields of the title compound 1. Additionally, the
ability to scale step 1 was called into question because of initial observation of the
physical properties of the reaction mixture when overhead agitation was employed.
Condensation of pivalaldehyde and D-proline in the presence of catalytic tri-
fluoroacetic acid (TFA) appeared problematic. Specifically, the unfavorable physical
characteristics of the reaction mixture limited its potential scalability. When over-
head agitation was used, the insoluble D-proline formed a large, golf-ball-sized mass
that was difficult to stir on a 10-g scale. As a result, magnetic stirring was required.
In addition, bicyclic aminal 3 was unstable on contact with air because of the spon-
taneous forming of proline.[4] For this reason, compound 3 was immediately
converted to 4 without purification (68% yield, 10-g scale).
Once a solution was demonstrated, improvements in this procedure were
incorporated to enhance the ability to scale this reaction and isolate 4. Standard
Scheme 1. Synthetic approach to 1.

2258
E. R. HUMPHREYS, J.-B. HONG, AND K. L. GREEN
hydrogenation conditions in the presence of Pd=C were employed to convert 4 to 5.
A switch from Pd to Pt was required to limit internalization of the double bond of
compound 4. The coupling of 5 with 6 will not be discussed in any detail because
since once a selective bis-silation of 4-bromo-2-chloroaniline was identified, this
coupling proceeded in good yield to desired isobutyl a-proline 1 (Scheme 1).
Initial attempts at bis-silation of 4-bromo-2-chloroaniline followed accepted
literature methods and used about 2 equiv. of n-butyl lithium at low temperature fol-
lowed by the addition of 2 equivalents of TMSCl. The reaction was then allowed to
warm to room temperature.[5] This method of bis-silation proved to not be selective
enough for our requirements, giving some oversilated by-products. Sodium hydride
was tried, but even when 3 equiv. of NaH followed by TMSCl was used, only the
monosilated material was obtained. A more selective silation method was observed
when about 2 equiv. of LiHMDS were used, followed by 2 equiv. of TMSCl at about
ꢀ70 ꢁC.
RESULTS AND DISCUSSION
Initial batches of 3 were prepared using a procedure described in organic synth-
eses.[6] Crude 4 was then carried into the hydrogenation to yield 5. After hydrogen-
ation of 4, impurities 4a and 4b (Fig. 2) were isolated from the crude hydrogenation
product 5.
The overall conversion of D-proline to 4 was further investigated to develop a
procedure that minimized the amount of impurities formed and that was more suit-
able for scale-up. The combination of D-proline and catalytic trifluoroacetic acid
(0.09 equiv.) in pentane resulted in large clumps that could not be effectively stirred
using overhead agitation. In an attempt to minimize clumping, celite and D-proline
were slurried together in pentane, and catalytic trifluoroacetic acid was added. When
celite was present, the solids remained finely divided. This facilitated the use of mech-
anical overhead agitation, and the reaction time was reduced from 4 days to 24 h in
pentane under reflux. It was found that the amount of pivaladehyde could be
reduced from 6 equiv. to 4.5 equiv. To minimize the formation of impurity 4a, excess
pivaladehyde was removed from crude 3 by concentration under vacuum. The
amount of pivaladehyde present was determined by 1H NMR. In an attempt to mini-
mize the condensation to product 4b, the order of addition was investigated. In pre-
vious reactions where significant amounts 4b was observed, the lithium
diisopropylamide (LDA) solution was added to 3 at ꢀ60 ꢁC, followed by the addition
Figure 2. Impurities from the conversion of 3 to 4.

TETRASUBSTITUTED PROLINE
2259
Scheme 2. Conversion of 2 to 3 in the presence of trimethyl orthoformate.
of 3-bromo-2-methyl propene. When 3 is added to LDA at ꢀ60 ꢁC followed by the
addition of 3-bromo-2-methyl propene, the amount of condensation product 4b is
minimized. This improved procedure for the conversion of D-proline to 4 was
demonstrated on a 25-g scale to give good quality 4.
Then, further evaluation of solvent was performed to replace pentane. The
reaction was run using the optimized conditions, replacing pentane with either cyclo-
hexane or dichloromethane. When cyclohexane was used as the solvent (81 ꢁC, 24 h),
1
3 was obtained as a dark brown oil (4.8 g, 66% yield). The H NMR showed small
amounts of impurities that were not present when pentane was used. When dichlor-
omethane was used as the solvent (41 ꢁC, 24 h, modified Dean–Stark trap), 3 was
1
obtained as an orange oil (5.7 g, 78% yield). The H NMR was comparable to
material generated using pentane.
Interestingly, when the conversion of 2 to 3 was investigated, using 1.2 equiv.
of trimethyl orthoformate as an internal drying agent in tetrahydrofuran (THF) at
20 ꢁC overnight resulted in the reaction proceeding in near quantitative yield giving
1
material 3 (Scheme 2) of comparable quality (by H NMR) to material generated
using pentane as the solvent.
Compound 4 was then reduced to 5 with 5% Pt-C (10 wt%) and hydrogen in
isopropanol at ambient temperature. After 12 h, the NMR spectrum of the crude
reaction mixture showed no remaining starting material. Under the same conditions,
the Pd-catalyzed hydrogenation gave approximately 10% of an internal olefin. This
olefin isomerization product was not observed in the Pt-catalyzed reaction.
The bis-silylation of 4-bromo-2-chloroaniline was carried out to supply aniline
6 for the synthesis of 1. Starting aniline was treated with n-butyl lithium (2.07 equiv.)
Scheme 3. Impurities formed during the bis-silylation of 4-bromo-2-chloroaniline.

2260
E. R. HUMPHREYS, J.-B. HONG, AND K. L. GREEN
Scheme 4. Bis-silylation of 4-bromo-2-chloroaniline with LiHMDS.
and TMSCl at ꢀ78 ꢁC. The resulting reaction mixture was warmed slowly to ambi-
ent temperature. Following an aqueous drowning and separation of the phases, the
organic layer was distilled and replaced with toluene. Removal of toluene gave a
crude oil, which was purified by Kugelrohr distillation to give aniline 6 (80 g, 54%
corrected yield). The product was contaminated with by-products 7, 8, 9, and 10
(Scheme 3). These impurities co-distilled with 6 and were carried into subsequent
transformations, complicating future purifications.
Investigation showed that the aqueous workup was causing decomposition of
6. Thus, the water drowning was eliminated, and THF was removed by displace-
ment with heptane or hexanes. The inorganic salts were then removed via filtration.
These anhydrous workup conditions eliminated the formation of by-products 7 and
8. This reaction was run with slightly less than 2.0 equiv. of n-butyllithium and 2.0
equiv. of TMSCl. This gave desired 6 along with by-products 9, 11, and the recov-
ered starting material. Addition of n-BuLi to a solution of TMSCl (2 equiv.) at
ꢀ78 ꢁC gave incomplete conversion to 6. Attempts to silyl protect the 4-bromo-2-
chloroaniline in the presence of NaH (up to 3 equiv.) resulted in the sole formation
of monosilane of 4-bromo-2-chloroaniline. Following isolation of this monosilane, it
was treated sequentially with LiHMDS (1 equiv.) and TMSCl (1 equiv.) to yield
pure 6 (76%).
The use of lithium bis(trimethylsilyl)amide was found to be superior to n-butyl
lithium for the bis lithiation of the bromo-chloro aniline (Scheme 4). This silylation
reaction was run with 2.1 equiv. of LiHMDS and 2.1 equiv. of TMSCl. This gave
product 6 in good yield and purity [93.4%, 99.3% pure by analytical (AN) high-
performance liquid chromatography (HPLC)].
(4-Amino-3-chloro-phenyl)-((S)-2-isobutyl-pyrrlidin-2-yl)-methanone (1) was
obtained by coupling of 5 with lithiate of 6 by addition of n-BuLi to a cooled
(ꢀ78 ꢁC) solution of the aryl bromide 6. Recrystallization from methyl-tert-butyl
ether (MTBE)=hexane gave the desired final product with excellent chemical and
optical purity (99.8% by AN HPLC purity, 99.5% ee by Chiral AN HPLC).
From 30 g of 5, 17 g of 1 was prepared with high purity and high ee (65%
isolated yield, including 4 g of the second recrystallization product). Use of excess
6 seemed to cause the formation of the side product 2-chloroaniline.
CONCLUSION
Scalable and practical routes to intermediates 3, 4, 5, and 6 are presented here
that made possible multigram synthesis of the title compound 1. This allowed for the
expedient delivery of 30 g of isobutyl a-proline 1.

TETRASUBSTITUTED PROLINE
2261
EXPERIMENTAL
All NMR analysis were carried out on a Bruker 300 spectrometer using
tetramethylsilane (TMS) as an internal standard. All HPLC analysis were carried
out on a Waters 2690 liquid chromatograph equipped with a Zorbax SB-C8:
3.0 ꢂ 100 mm column, component A 0.1% TFA in water; component B 0.1%
TFA in acetonitrile, gradient 10% to 90% B from 0 to 3 min; hold 90% B for
5 min, flow rate: 1.2 mL=min. All gas chromatographic (GC) analysis were carried
out on an Agilent 6890 equipped with a DB-5MS column, 15 m ꢂ 0.25 mm,
0.25 mm packing (serial no. US5326631A), He @ 7 PSI, 60 ꢁC (2 min.), 15 ꢁC=min
min to 30 ꢁC, hold 2 min. FID, 300 ꢁC. All SFC analyses were carried out on a
Berger Analytical supercritical fluid chromatography (SFC) equipped with a
ChiralPak OD-H, 4.6 ꢂ 250 mm, 98:2 CO2:0.4% iPrNH2 in isopropanol,
2.5 ml=min, 125 bar, 40 ꢁC, UV @ 265 nm and analyzed as the bis-trifluoroacetyl
derivative. LC=MS analyses were carried out on an Agilent 1100MSD fitted with
an electrospray ionization source (nitrogen drying gas 13 l=min @ 350 ꢁC, nebu-
lizer @ 60 psi, capillary voltage 4 kV, fragmentor 70 V). GC=MS analyses were
carried out on an Agilent 6890GC w=5973MSD in chemical-ionization mode
using methane as the reagent gas.
(3S,7aR)-3-tert-Butyl-tetrahydro-pyrrolo[1,2-c]oxazol-1-one (3)
In a 2-l round-bottomed flask equipped with an overhead agitator and Dean–
Stark trap, D-proline (25 g, 0.22 mol), celite (17.5 g), and pentane (875 mL) were
combined. Trifluoroacetic acid (1.5 mL, 0.02 mol) was added to the suspension drop-
wise followed by pivaladehyde (106 mL, 0.98 mol). The mixture was heated to reflux
and aged for 24 h until a total of 4 mL of water was collected. The slurry was cooled
to approximately 20 ꢁC and filtered under nitrogen (the product is extremely sensitive
to hydrolysis). The resulting solution was concentrated to an oil under vacuum. The
1
product 2 can be stored under an inert atmosphere in a refrigerator. H NMR
(300 MHz, CDCl3) d ¼ 4.32 (s, 1 H), 3.57 (dd, J ¼ 4.5, 8.9 Hz, 1 H), 3.07–2.91 (m,
1 H), 2.70–2.55 (m, J ¼ 10.5 Hz, 1 H), 2.03–1.71 (m, 2 H), 1.68–1.40 (m, 2 H), 0.71
(s, 9 H).
(3S,7aR)-3-tert-Butyl-tetrahydro-pyrrolo[1,2-c]oxazol-1-one (3) in the
Presence of Trimethylorthoformate
In a 250-mL, round-bottomed flask equipped with an overhead agitator,
D-proline (10 g, 0.087 mol), celite (5 g), and THF (100 mL) were combined.
Trifluoroacetic acid (0.58 mL, 0.0078 mol) was added to the suspension dropwise,
followed by pivaladehyde (42.5 mL, 0.0.39 mol) and trimethylorthoformate
(11.4 mL, 0.10 mol). The suspension was aged at approximately 20 ꢁC for at least
12 h. The reaction mixture was filtered under nitrogen and the resulting solution
was concentrated under vacuum to an orange oil (14.57 g, 91% yield). The
product was analyzed by 1H NMR and found to be approximately 98% pure
[other impurities include residual THF, trimethyl orthoformate (1%), and pivala-
dehyde (1%)]. 1H NMR consistent with 1H NMR of 3 was made without
trimethylorthoformate.

2262
E. R. HUMPHREYS, J.-B. HONG, AND K. L. GREEN
(3S,7aS)3-tert-Butyl-7a-(2-methyl-allyl)-tetrahydro-
pyrrolo[1,2-]oxazole-1-one (4)
Diisopropyl amine (33 mL, 0.24 mol) was dissolved in THF (145 mL), and the
solution was cooled to approximately 90 ꢁC in an isopropanol=dry ice bath. n-Butyl
lithium solution (135 mL, 0.22 mol) was added at a rate to maintain an internal pot
temperature below ꢀ58 ꢁC (addition was complete in 25 min). The solution was
allowed to warm to 0 ꢁC for 30 min, then cooled back to ꢀ78 ꢁC. Aminal 3 was dis-
solved in THF (360 mL) and cooled to ꢀ65 ꢁC in an isopropanol= dry ice bath. The
THF solution of 3 was added to the LDA solution at a rate to maintain an internal
temperature below ꢀ60 ꢁC (addition was complete in 35 min). Reaction mixture was
aged at ꢀ70 ꢁC for 15 min, and then 3-bromo-2-methyl propene (25.7 mL, 0.26 mol)
was added to the mixture over 10 min. The reaction mixture was then aged at
approximately ꢀ70 ꢁC for at least 1 h, and then it was warmed to 0 ꢁC over 1 h.
An aqueous NH4Cl solution (125 mL of saturated ammonium chloride and
125 mL of water) was added to the reaction mixture at a rate to maintain an internal
pot temperature between 0 ꢁC and 5 ꢁC. The layers were separated, and the organic
layer was concentrated to an oily residue under vacuum. The aqueous layer was
extracted with 250 mL of ethyl acetate. The layers were separated, and the organic
layer was used to redissolve the oily residue from the original organic layer. The
combined organic layers were washed with water (1 ꢂ 100 mL and 1 ꢂ 75 mL) and
dried over sodium sulfate. The resulting filtrate was concentrated under vacuum
1
to yield product 4 (37.6 g, 79% yield). H NMR (300 MHz, CDCl3) d ¼ 4.89–4.80
(m, 1 H), 4.76–4.71 (m, J ¼ 0.8 Hz, 1 H), 4.19 (s, 1 H), 2.97–2.82 (m, 1 H), 2.81–
2.70 (m, 1 H), 2.33 (d, J ¼ 12.0 Hz, 1 H), 2.29 (d, J ¼ 12.0 Hz, 1 H), 2.07–1.96
(m, 1 H), 1.95–1.83 (m, 1 H), 1.80 (s, 3 H), 1.75–1.45 (m, 2 H), 0.88 (s, 9 H).
2-(4-Bromo-2-chloro-phenyl)-1,1,1,3,3,3-hexamethyl-disilazane (6)
In a 2-L, round-bottomed flask equipped with an overhead agitator, 50.38 g of
4-bromo-2-chloroaniline and 250 mL of THF were combined. This slurry was cooled
in a dry ice acetone bath to ꢀ70.6 ꢁC. LiHMDS (1 M) solution in THF was added
dropwise over about an hour. After addition of the first equiv. of LiHMDS, the het-
erogeneous reaction mixture was purple in color. After addition of the second equiv.
of LiHMDS, the heterogeneous reaction mixture was pinkish white in color. The
reaction was left to warm to about 0 ꢁC over a 55-min time period by removing
the cooling bath (Tmax ¼ 3.7 ꢁC). The reaction was then cooled back towards
ꢀ78 ꢁC. At ꢀ72 ꢁC, the TMSCl was added over about 12 min, Tmax ¼ ꢀ62.9 ꢁC.
The dark red reaction mixture was then allowed to warm slowly back to room tem-
perature. As the reaction warmed, the color of the mixture lightened to tan. This
crude mixture was left overnight at room temperature without agitation. The follow-
ing morning on a rotary evaporator the THF was stripped off under house vacuum
(ꢃ29 inch of Hg) to a bath temperature of about 45 ꢁC and then displaced with
200 mL of hexanes. The hexanes were then stripped off at a bath temperature of
about 45 ꢁC. The resulting semisolid residue was dissolved in 300 mL of hexanes,
and the insoluble material was removed by filtration through a medium glass
sintered funnel. The resulting filter cake was washed with 200 mL of hexanes.

TETRASUBSTITUTED PROLINE
2263
The combined liquors were then stripped of the hexanes to yield a crude oil (89.18 g).
This oil was Kugelrohor distilled (vacuum at the source less than 100 millitorr). This
yielded 80 g of a water white oil (heart cut) that was 99.33% pure 5 by GC (yield
93.4%). 1H NMR (300 MHz, CDCl3) d ¼ 7.51 (d, J ¼ 1.0 Hz, 1 H), 7.23 (t, J ¼ 1.0 Hz,
1 H), 6.85 (d, J ¼ 1.0 Hz, 1 H), 0.08 (s, 18 H).
3S,7aS)-3-tert-Butyl-7a-Isobutyl-tetrahydro-pyrrolo[1,2-c]oxazol-
1-one (5)
Meta-allyl pyrrolo-oxazolone 4 was reduced to the corresponding isobutyl
pyrrolo-oxazolone 5 with 5% Pt=C (10 wt%) and hydrogen atmosphere in IPA
(5 vol.) at room temperature. After 12 h, the NMR spectrum of the crude reaction
mixture showed no remaining starting material. Upon completion of hydrogenation,
10 wt% Solka-Floc and 10 wt% charcoal were added to the reaction mixture, and the
resulting mixture was stirred at room temperature for 2 h, filtered through a pad of
Solka-Floc, and concentrated under vacuum. This treatment was effective enough to
reduce the level of residual Pt to <1 ppm. Dimer of pyrrolo-oxazolone 4, which was
generated in the allylation step with LDA, was precipitated out after concentration.
This dimer was removed by filtration through a pad of silica gel with 10% EtOAc=
hexane solvent system. From 35 g of meta-allyl pyrrolo-oxazolone 4, 30 g of isobutyl
pyrrolo-oxazolone 5 was obtained with 92% GC purity (yield: 85%). 1H NMR
(300 MHz, chloroform-d) d ¼ 4.22 (s, 1 H), 3.09–2.95 (m, 1 H), 2.92–2.80 (m, 1 H),
2.13–1.99 (m, 1 H), 1.96–1.64 (m, 4 H), 1.61 (d, J ¼ 1.0 Hz, 2 H), 0.98 (t, J ¼ 7.0 Hz,
6 H), 0.93 (s, 9 H).
(4-Amino-3-chloro-phenyl)-((S)-2-isobutyl-pyrrolidin-
2-yl)-methanone (1)
Aryl addition reaction of substrate 5 was performed with TMS-protected
bromo-chloro aniline 6 and n-BuLi. To a solution of 6 (1.4 equivalent based on 5)
in THF (180 mL), 1.6 M n-BuLi (1.4 equivalent based on 5) in hexanes was added
dropwise at ꢀ78 ꢁC for 2 h, then the reaction mixture was stirred for 30 min.
Lithiation was a very exothermic reaction, and after the addition of n-BuLi was fin-
ished, an off-white precipitate was observed in the reaction mixture. A solution of 5
in THF 60 mL was added to the reaction mixture dropwise at ꢀ78 ꢁC for 1 h. The
resulting mixture turned dark yellow. The resulting mixture was stirred for 4 h, then
allowed to warm to 0 ꢁC for 1 h and quenched with saturated NH4Cl solution. All
steps were followed by GC, which showed over 99% GC conversion. The organic
layer was extracted with MTBE and dichloromethane (3 times). The combined
organic layers were dried over MgSO4 and concentrated in vacuo to afford the crude
product, which was deprotected with 1.0 N HCl (150 mL). After stirring for 2 h at
room temperature, the reaction mixture was neutralized with 2.0 N NaOH solution,
and then the organic layer was extracted with ethyl acetate (three times). The com-
bined organic layers were then dried over MgSO4 and concentrated under reduced
pressure to afford the crude product 25 g. This crude material was purified by recrys-
tallization with MTBE (100 mL)=hexane (50 mL) solvent system (cooling down to
0 ꢁC) to afford the desired final product 1 with 99.8% HPLC purity and 99.5%

2264
E. R. HUMPHREYS, J.-B. HONG, AND K. L. GREEN
enantiomeric excess. From 30 g of 5, 17 g 1 was prepared with high purity and high ee
(yield: 65%, includes 4 g from a second recrystallization product). 1H NMR
(300 MHz, CDCl3) d ¼ 8.05 (d, J ¼ 2.3 Hz, 1 H), 7.82 (dd, J ¼ 2.1, 8.5 Hz, 1 H),
6.74 (d, J ¼ 8.3 Hz, 1 H), 4.52 (s, 2 H), 2.99 (ddd, J ¼ 4.3, 6.9, 9.9 Hz, 1 H), 2.81
(ddd, J ¼ 7.0, 8.0, 9.9 Hz, 1 H), 2.32 (ddd, J ¼ 7.0, 8.1, 12.4 Hz, 1 H), 2.07–1.90
(m, 2 H), 1.89–1.75 (m, 2 H), 1.73–1.57 (m, 2 H), 0.86 (d, J ¼ 6.8 Hz, 3 H), 0.64
(d, J ¼ 6.4 Hz, 3 H).
REFERENCES
1. (a) Millan, M. J. Dual- and triple-acting agents for treating core and co-morbid symptoms
of major depression: Novel concepts, new drugs. Neurotherapeutics 2009, 6, 53–77; (b)
Beer, B.; Stark, J.; Krieter, P.; Czobor, P.; Beer, G.; Lippa, A.; Skolnick, P. DOV
216,303, a ‘‘triple’’ reuptake inhibitor: Safety, tolerability, and pharmacokinetics profile.
J. Clin. Pharmacol. 2004, 44, 1360–1367.
2. Iyer, P.; Lin, C. J. J.; Lynch, S. M.; Lucas, M. C.; Madera, A. M.; Ozboya, K. E.; Weikert,
R. J.; Schoenfeld, R. C. Heteroaryl pyrrolidinyl and piperidinyl ketone derivatives and uses
thereof. US Patent 0146607 A1, 2008.
3. (a) Sayago, J. S.; Calaza, M. I.; Jimenez, I. A.; Cativela, C. Versatile methodology for the
synthesis and a-functionalization of (1R,3aS,7aS)-octahydroindole-2-carboxylic acid.
Tetrahedron 2008, 64, 84–91; (b) Seebach, D.; Sting, A. R.; Hoffman, M. Self-regeneration
of stereocenters (SRS): Applications, limitations, and abandonment of a synthetic prin-
ciple. Angew. Chem. Int. Ed. 1996, 35, 2708–2748; (c) Seebach, D.; Beck, A. K.; Badine,
D. M.; Limbach, M.; Eschenmoser, A.; Treasurywala, A. M.; Hobi, R.; Prikoszovich,
W.; Linder, B. Are oxazolidinones really unproductive, parasitic species in proline cataly-
sis? Thoughts and experiments pointing to an alternative view. Helv. Chim. Acta 2007, 90,
425–471.
4. Seebach, D.; Boes, M.; Naef, R.; Schweizer, W. B. Alkylation of amino acids without loss
of the optical activity: Preparation of a-substituted proline derivatives: A case of
self-reproduction of chirality. J. Am. Chem. Soc. 1983, 105, 5390–5398.
5. Chen, G. H.; Hoechestetter, C.; Knochel, P. Mixed copper, zinc 2-amino benzylic organo-
metallics as efficient reagents for the synthesis of heterocycles. Tetrahedron Lett. 1989, 30,
4795–4798.
6. Beck, A. K.; Blank, S.; Job, K.; Seebach, D.; Sommerfeld, T. Synthesis of (S)-2-
methylproline: A general method for the preparation of a-branched amino acids. Org.
Synth. 1995, 72, 62–73.