Asymmetric Synthesis of cis -(S, R)-3-Amino-4-fluoro-1-methylpyrrolidine

Article abstract of DOI:10.1055/s-0037-1611553

The development of the stereoselective synthesis of cis -(S, R)-3-amino-4-fluoro-1-methylpyrrolidine is described starting from chiral, non-racemic 1-[(3 S,4 S)-3-azido-4-hydroxypyrrolidin-1-yl]-2,2,2-trifluoroethan-1-one. Two sets of deoxyfluorination co

Full text of DOI:10.1055/s-0037-1611553

SYNLETT
0
9
3
6
-
5
2
1
4
1
4
3
7
-
2
0
9
6
© Georg Thieme Verlag Stuttgart · New York
2019, 30, A–C
letter
A
en
Z. Fei et al.
Letter
Synlett
Asymmetric Synthesis of cis-(S,R)-3-Amino-4-fluoro-1-methylpyr-
rolidine
Zhongbo Feia
Xin Xionga
F
DAST, DCM
O
N
–60 °C to rt, 16 h
66%
CF3
N3
Chiming Cheunga
Weipeng Liua
Qirong Shenb
Jinzhu Zhangb
Hongjun Gaob
Jianwei Bian*a
HO
N3
O
complete retention
N
CF3
SO2F2 (1 atm)
tetramethylammonium
2,6-dimethylphenoxide
F
F
O
N
N
Me
CF3
DMF, rt, 24 h
72%
N3
H2N
2HCl
complete inversion
a Chemical and Analytical Development, Suzhou Novartis
Pharma Technology Co., Ltd, Changshu, Jiangsu
215537, P. R. of China
jianwei.bian@novartis.com
b Zhejiang Jiuzhou Pharmaceutical Technology Co., Ltd.,
Jiangling Road 88, Binjiang, Hangzhou, Zhejiang
310000, P. R. of China
Received: 14.03.2019
Accepted after revision: 30.04.2019
Published online: 20.05.2019
group participation could be prevented by installing an
electron-withdrawing group on the amino group prior to
deoxyfluorination, or by employing substrate 3 possessing
the azido group as an amino surrogate for the deoxyfluori-
nation, given that azides have been proven to be non-par-
ticipating neighboring groups in sugar chemistry.5
DOI: 10.1055/s-0037-1611553; Art ID: st-2019-v0148-l
Abstract The development of the stereoselective synthesis of cis-
(S,R)-3-amino-4-fluoro-1-methylpyrrolidine is described starting from
chiral, non-racemic 1-[(3S,4S)-3-azido-4-hydroxypyrrolidin-1-yl]-2,2,2-
trifluoroethan-1-one. Two sets of deoxyfluorination conditions are de-
veloped for achieving inversion of the chiral center with high or com-
plete stereoselectivity.
HO
NBoc
BnHN
F
2
Key words stereoselective, azides, deoxyfluorination, sulfuryl fluo-
ride, pyrrolidines
N
Me
Deoxyfluorinations
H2N
2HCl
1
HO
N3
O
N
cis-(S,R)-3-Amino-4-fluoropyrrolidines are common
structural moieties often found in biologically important
compounds.1 However, there are limited reports of asym-
metric syntheses of these compounds.2 In support of our
drug development programs, we were interested in produc-
ing cis-(S,R)-3-amino-4-fluoro-1-methylpyrrolidine (1) in
an efficient manner. Herein, we report an asymmetric syn-
thesis of compound 1, which was developed as part of these
efforts. Our strategy for the asymmetric synthesis of 1 was
to utilize an appropriate chiral and easily accessible sub-
strate reported in the literature for stereospecific deoxyfluo-
rination. After carefully reviewing literature reports, we di-
rected our efforts to examine deoxyfluorination on sub-
strates of type 2 and 3, as shown in Scheme 1. Substrate 2,
in its chiral non-racemic form, was reported by Tsuzuki et
al. and generated via chemical resolution.3 Substrate 3 has
been reported by Jacobsen et al. and prepared via de-
symmetrization of the corresponding epoxide.4 It was ex-
pected that the nucleophilic amino group on 2 would par-
ticipate in substitution to form the aziridine in the course of
deoxyfluorination. We envisioned that this neighboring
CF3
3
Scheme 1 Synthetic plan
We prepared substrate 2 according to Tsuzuki’s proce-
dure,3 and then converted it into the acetylated substrate 4
in 98% yield (Scheme 2). When subjecting 4 to deoxyfluori-
nation using DAST (diethylaminosulfur trifluoride), we ob-
tained aziridine 5 in 75% yield. This suggested that either
the electron-withdrawing acetyl group did not prevent the
electron pair of the amino group from participating in the
substitution, or that anchimeric assistance via the amide
oxygen occurred. In theory, there would be an opportunity
to screen other electron-withdrawing groups, but this
would be less attractive in comparison to using substrate 3
for the deoxyfluorination. Nevertheless, aziridine 5 could
serve as a reliable intermediate to access the trans-3-ami-
no-4-fluoropyrrolidine, a reference sample for confirming
the relative configuration of the desired cis product. We
therefore proceeded with the synthesis by ring-opening of
© Georg Thieme Verlag Stuttgart · New York — Synlett 2019, 30, A–C

B
Z. Fei et al.
Letter
Synlett
Table 1 Deoxyfluorination of 3
HO
N3
F
F
O
O
O
Conditions
N
N
N
+
CF3
CF3
CF3
N3
N3
3
9
10
Entry
Conditions
Results
1
2
3
DAST, DCM
DAST, pyridine, DCM
SO2F2, tetramethylammonium 2,6-dimethylphenoxide, DMF
only trans 9, 66% yield
trans 9/cis 10 = 1:11, 70% yield
only cis 10, 72% yield
aziridine 5 with gentle warming in Et3N·3HF at 80 °C to give
trans-substituted 6 in 65% yield. Using pyridine·HF instead
led to a messy reaction. Removal of the Boc protecting
group followed by methylation provided 7. Debenzylation
of 7 under hydrogenation conditions afforded the trans-3-
amino-4-fluoropyrrolidine 8 in racemic form.
strates.2 As Paulsen et al. had shown that azides serve as
non-participating neighboring groups for the synthesis of
1,2-cis glycosides,5 we suspected that this reaction proceed-
ed through an SNi pathway, and not an SN2 process.6 In this
SNi pathway, the substrate 3 was activated to form 3a, but
the desired rear-face attack of fluoride was hindered by the
neighboring azido group, which allowed the formation of
intimate ion pair 3b. The concerted release of the sulfura-
midous fluoride and fluoride-bonding to the carbon from
the same face (the front face) gave trans product 9 (Scheme
3). The SNi pathway is reported to be inhibited by using pyr-
idine as an additive.7 Indeed, by adding one equivalent of
pyridine to this deoxyfluorination, the desired cis product
10 was favored over trans product 9 with an 11:1 ratio (en-
try 2). With pyridine present, the transient intermediate
was assumed to be 3c, the productive pathway of which for
trans product 9 was inhibited due to preferential release of
pyridine instead of fluoride from the sulfite if the intimate
ion pair of 3c had indeed formed (Scheme 3). Alternatively,
we sought to perform the deoxyfluorination with cheap
and readily available sulfuryl fluoride, as the reported ob-
servation of direct deoxyfluorination of -hydroxy acetates
Bn
N
BnHN
Ac2O
DAST
Ac
NBoc
BnN
NBoc
NBoc
DCM
75%
K2CO3, DCM, rt
98%
HO
HO
5
2
4
BnHN
H2N
BnHN
1. TFA
Et3N⋅3HF
Pd/C, H2
NBoc
N Me
N
Me
2. MeI
44%
EtOH
92%
80 °C
65%
F
F
F
(
)
( )
(
)
8
6
7
Scheme 2 Synthesis of trans-3-amino-4-fluoropyrrolidne 8
Our efforts were then focused on the deoxyfluorination
of 3. Enantiomerically enriched 3 (93% ee) was smoothly
prepared by following the reported protocols.4 However,
deoxyfluorination of 3 using DAST still provided the trans
product 9 (Table 1, entry 1), as observed for analogous sub-
F
F
H
–
O
H
F
N
S
F
O
N
S
F
O
O
O
SNi
N
N
N
CF3
CF3
CF3
N3
N3
N3
HF
HF
9
3a
3b
DAST
F–
F
S
H
HO
N3
F
N
O
SN2
O
O
O
DAST, pyridine
N
N
N
N
CF3
CF3
CF3
N3
N3
3
10
3c
SN
2
O
O
Me4NF
O
O
S
O
S
O
F
F
F
N
Me
O
CF3
O
N3
S
F
O
3d
Me
Me
O–
NMe4
O
O
+
S
F
F
Me
Scheme 3 Plausible pathways for the deoxyfluorination of 3
© Georg Thieme Verlag Stuttgart · New York — Synlett 2019, 30, A–C

C
Z. Fei et al.
Letter
Synlett
drew our attention.8 Simple exposure of the substrate, as a
solution in DMF, to gaseous sulfuryl fluoride as reported did
not provide any conversion, but subsequent addition of te-
tramethylammonium 2,6-dimethylphenoxide to this solu-
tion led to formation of the desired cis product 10 in 72%
yield (entry 3). Sanford’s group reported the in situ genera-
tion of anhydrous tetramethylammonium fluoride from the
combination of sulfuryl fluoride with tetramethylammoni-
um 2,6-dimethylphenoxide.9 We therefore proposed that
the formed intermediate 3d underwent SN2 reaction with
in situ generated anhydrous tetramethylammonium fluo-
ride to give the product (Scheme 3). To the best of our
knowledge, this is a new application of Sanford’s conditions
for the deoxyfluorination of an aliphatic alcohol.
Funding Information
All work described in this paper was funded by Novartis, Inc.()
Acknowledgment
We acknowledge Fabrice Gallou, Wei Li and Ning Ye for helpful dis-
cussions. We thank Yanjie Li for HRMS analyses.
Supporting Information
Supporting information for this article is available online at
https://doi.org/10.1055/s-0037-1611553.
S
u
p
p
orit
n
gInformati
o
n
S
u
p
p
orit
n
gInformati
o
n
From 10, we completed the synthesis of target 1 as
shown in Scheme 4. Reduction of the azido group under hy-
drogenation in the presence of Boc2O directly converted 10
into 11 in 88% yield. Methanolysis of the latter provided 12,
and subsequent reductive methylation afforded 13. Finally,
acid treatment of 13 to remove the Boc protecting group
and purification by recrystallization provided 1 as a single
diastereomer in 94% yield and with 99.9% ee.10
References and Notes
(1) (a) Coleman, P. J.; Cox, C. D. WO 2005017190, 2005. (b) Kawato,
H.; Miyazaki, M.; Sugimoto, Y.; Naito, H.; Okayama, T.; Soga, T.;
Uoto, K. WO 2008072655, 2008. (c) Corminboeuf, O.; Pozzi, D.
WO 2013171694, 2013. (d) Guerin, D. J.; Bair, K. W.; Caravella, J.
A.; Ioannidis, S.; Lancia, D. R. Jr.; Li, H.; Mischke, S.; Ng, P. Y.;
Richard, D.; Schiller, S. E. R.; Shelekhin, T.; Wang, Z. WO
2017139778, 2017.
(2) (a) Li, Q.; Wang, W.; Berst, K. B.; Claiborne, A.; Hasvold, L.; Raye,
K.; Tufano, M.; Nilius, A.; Shen, L. L.; Flamm, R.; Alder, J.; Marsh,
K.; Crowell, D.; Chu, D. T. W.; Plattner, J. J. Bioorg. Med. Chem.
Lett. 1998, 8, 1953. (b) Bouzard, D.; Di Cesare, P.; Essiz, M.;
Jacquet, J. P.; Kiechel, J. R.; Remuzon, P.; Weber, A.; Oki, T.;
Masuyoshi, M.; Kessler, R. E.; Fung-Tomc, J.; Desiderio, J. J. Med.
Chem. 1990, 33, 1344. (c) Wang, C.-S.; Li, T.-Z.; Cheng, Y.-C.;
Zhou, J.; Mei, G.-J.; Shi, F. J. Org. Chem. 2019, 84, 3214.
(3) Tsuzuki, Y.; Chiba, K.; Mizuno, K.; Tomita, K.; Suzuki, K. Tetrahe-
dron: Asymmetry 2001, 12, 2989.
F
F
F
O
cat. PtO2, H2
O
K2CO3
MeOH
N
N
NH
CF3 Boc2O, MeOH
88%
CF3
N3
BocHN
BocHN
12
11
10
F
F
CH2O, NaBH4
HCl
N
Me
N
Me
MeOH
94%
MeOH
73% over
two steps
BocHN
2HCl
H2N
1
13
Scheme 4 Completion of the synthesis of 1
(4) (a) Martinez, L. E.; Leighton, J. L.; Carsten, D. H.; Jacobsen, E. N.
J. Am. Chem. Soc. 1995, 117, 5897. (b) Jacobsen, E. N.; Tokunaga,
M.; Larrow, J. F. US Patent 6262278, 2001.
(5) Beckmann, H. S. G.; Wittmann, V. In Organic Azides: Syntheses
and Applications; Bräse, S.; Banert, K., Ed.; John Wiley & Sons:
Chichester, 2010, 469–490.
(6) Lewis, E. S.; Boozer, C. E. J. Am. Chem. Soc. 1952, 74, 308.
(7) Cram, D. J. J. Am. Chem. Soc. 1953, 75, 332.
(8) Ishii, A.; Yasumoto, M. WO 2010047266, 2010.
In summary, we have developed a stereoselective syn-
thesis of cis-(S,R)-3-amino-4-fluoro-1-methylpyrrolidine
(1) starting from known chiral non-racemic 3, which is eas-
ily accessible by applying Jacobsen’s desymmetrization of
an epoxide. The deoxyfluorination of 3 was studied toward
the cis selectivity. While DAST alone led to the trans isomer
exclusively, the conditions were tuned by adding pyridine
to improve the cis selectivity. Eventually, complete cis selec-
tivity was achieved by applying the combination of SO2F2
(9) Cismesia, M. A.; Ryan, S. J.; Bland, D. C.; Sanford, M. S. J. Org.
Chem. 2017, 82, 5020.
(10) Characterization of 1. Yield: 99 mg (94%); white solid; mp 264–
266 °C. 1H NMR (400 MHz, D2O):  = 5.55 (dt, J = 51.7, 3.5 Hz, 1
H), 4.36 (dtd, J = 23.4, 8.9, 3.5 Hz, 1 H), 4.01 (m, 1 H), 3.88 (m, 1
H), 3.82–3.52 (m, 2 H), 3.00 (s, 3 H). 13C NMR (101 MHz, D2O):
 = 90.4 (d, J = 183.82 Hz), 59.6 (d, J = 21.21 Hz), 54.4, 50.3 (d, J =
17.17 Hz), 42.6. 19F NMR (376 MHz, D2O):  = –196.7. 19F NMR
(376 MHz, CD3OD):  = –198.1. HRMS (ESI): m/z [M + H – 2
HCl]+ calcd for C5H12FN2: 119.0979; found: 119.0978.
with
tetramethylammonium
2,6-dimethylphenoxide
(Sanford conditions). Due to its unique properties, we be-
lieve that the synthesis of this highly valuable chiral build-
ing block will be of significant interest to chemists and
pharmacologists.
© Georg Thieme Verlag Stuttgart · New York — Synlett 2019, 30, A–C