Efforts towards a Noyori reduction of a 3-fluorpiperidin-4-one with dynamic kinetic resolution

Article abstract of DOI:10.1016/j.tetlet.2020.152764

A Noyori reduction of racemic 1-Boc-3-fluoropiperidin-4-one has been achieved under dynamic kinetic resolution conditions that results in a single cis enantiomer being obtained with both diastereo- and enantioselectivity > 90%. This medicinal chemistry building block can be generated on multi-gram scale using the developed conditions to a single enantiomer in a 60% yield after employing a crystallisation procedure.

Full text of DOI:10.1016/j.tetlet.2020.152764

Tetrahedron Letters 66 (2021) 152764
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Efforts towards a Noyori reduction of a 3-fluorpiperidin-4-one
with dynamic kinetic resolution
Jacques LePaih a, Dane A. Goff b, Rajinder Singh b, Simon J. Shaw b,
⇑
a Johnson-Matthey Plc, 28 Cambridge Science Park, Milton Road, Cambridge CB4 0FP, UK
b Rigel Pharmaceuticals, Inc., 1180 Veterans Boulevard, South San Francisco, CA 94607, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
A Noyori reduction of racemic 1-Boc-3-fluoropiperidin-4-one has been achieved under dynamic kinetic
resolution conditions that results in a single cis enantiomer being obtained with both diastereo- and
enantioselectivity > 90%. This medicinal chemistry building block can be generated on multi-gram scale
using the developed conditions to a single enantiomer in a 60% yield after employing a crystallisation
procedure.
Received 11 October 2020
Revised 10 December 2020
Accepted 14 December 2020
Available online 6 January 2021
Ó 2021 Elsevier Ltd. All rights reserved.
Keywords:
Noyori dynamic kinetic resolution
3-Fluoropiperidin-4-ol
Diastereoselective
Enantioselective
Building block
Piperidines are a common building block in medicinal chem-
istry. They are found in both natural product and synthetic drugs,
where medicinal chemists prize their ability to act as pharma-
cophores and the basic amine can be used to improve pharmacoki-
netic properties including the aqueous solubility of the molecule
through lowering the lipophilicity. In addition the conformational
rigidity allows the piperidine to play a role as a scaffold for display-
ing substituents with defined vectors [1]. Marketed agents include
the anti-cancer alkaloid vinblastine (VeblanÒ) isolated from the
Madagascar Periwinkle, the stimulant methylphenidate (RitalinÒ)
and the Janus kinase inhibitor (XeljanzÒ) tofacitinib (Fig. 1).
Indeed, an analysis by Njardarson in 2014 found the piperidine
to be the most commonly found nitrogen-containing heterocycle
in FDA-approved pharmaceuticals [2], while 25% of all kinase inhi-
bitors licensed by the FDA in the United States as of April 2020 con-
tain a piperidine [3].
Although the basicity of the piperidine has benefits, it can also
lead to off-target liabilities. For example, it has been implicated
in the inhibition of the hERG potassium ion channel [1], which
can lead to cardiac arrhythmias that can be fatal. Modulation of
the basicity of the piperidine has been demonstrated to reduce
the hERG liabilities and an increasingly common tactic has been
the introduction of a 3-fluoro-group [4]. The addition of a fluoro
group can improve both the off-target liabilities and on-target
potency as well as improving the pharmacokinetic properties of
the molecule by improving permeability and adsorption [4–7].
The orientation of the fluorine group has a significant influence
on the piperidine basicity with the equatorial fluoro-group reduc-
ing the pKa by approximately 2 units, while a smaller reduction of
about 1 unit is observed with an axial fluoro-group [4–7]. These
properties continue to be exploited in medicinal chemistry pro-
grams [8,9]. In one of our own drug discovery programs we were
keen to probe 3-fluoropiperidines for their effects on hERG and
PK. These compounds are usually synthesised in a racemic fashion
followed by a chiral separation. Thus, MacMillan’s reported chiral
fluorination of 1-Boc-piperidin-4-one 1 using a modified Cinchona
alkaloid catalyst appeared to be very useful [10]. We were able to
use this chemistry to generate either enantiomer of cis-1-Boc-3-
fluoropiperidin-4-ol 2 through use of the catalyst derived from
either quinine or the pseudoenantiomer quinidine followed by
reduction of the ketone 3 with enantiopurities of 95% [11]. A
recrystallization procedure was employed to obtain enantiopure
material in an overall yield of 44%. We further showed that similar
enantioselective fluorination could be obtained by employing sim-
ple benzylamines (for example
a-methylbenzylamine) as catalyst
albeit with lower enantioselectivities.
As part of our evaluation of alternative routes to enantiopure
(3S, 4R)-1-Boc-3-fluoropiperidin-4-ol (+)-2 we were intrigued by
the possibility of carrying out a dynamic kinetic resolution (DKR)
of racemic 1-Boc-3-fluoropiperidinone rac-3 under reducing con-
ditions. The Noyori dynamic kinetic resolution is a powerful
⇑
Corresponding author.
E-mail address: sshaw@rigel.com (S.J. Shaw).
https://doi.org/10.1016/j.tetlet.2020.152764
0040-4039/Ó 2021 Elsevier Ltd. All rights reserved.

J. LePaih, D.A. Goff, R. Singh et al.
Tetrahedron Letters 66 (2021) 152764
substrates for the DKR reduction with Noyori showing that a 3-
phenylpiperidin-4-one could be reduced enantioselectively [17].
In addition workers at Hoffmann-La Roche were able to use the
methodology to effect reduction of 4-benzyl-3-piperidinone [18].
In each of these examples the cis diastereomer was obtained, the
desired diastereomer in our situation.
In our studies we focused on attempting to obtain the (3S, 4R)-
1-Boc-3-fluoropiperidinol (+)-2 and began by developing a chiral
hplc method to separate the two cis enantiomers from the trans
diastereomers (the two trans enantiomers did not separate under
the column conditions) and starting ketone (see Supplementary
Material for details). A series of ruthenium pre-catalysts containing
a chiral diphosphine ligand and a chiral diamine, which are known
to be effective in such reactions, were screened [12,19]. The reac-
tions were carried out with a substoichiometric amount of potas-
sium tert-butoxide as base under a high pressure of hydrogen at
elevated temperature. Under the conditions screened many cata-
lyst combinations resulted in almost complete consumption of
the ketone with the cis diastereomer being favoured in moderate
ee’s (Table 1). The S,S catalyst system gave the desired enantiomer
(+)-2 with encouraging enantioselectivity of up to 71%, while the
enantiomeric catalysts resulted in the undesired 3R, 4S enantiomer
(À)-2 being preferred (Entry 9 vs 10). The mis-matched catalysts
resulted in poor enantioselectivity and often low diastereoselectiv-
ity (Entries 2 vs 12 and 5 vs 13).
Having identified the (S,S) catalyst system as leading to the
desired enantiomer we investigated a smaller set of catalyst com-
binations looking to improve the catalyst loading (Table 2). At a
five-fold lower loading the diastereoselectivity was retained but
the enantioselectivity was improved and we were able to reach
84%ee using the [(S)Xyl-PPhos RuCl2 (S)DAIPEN] catalyst (entry
2). Reducing the amount of potassium tert-butoxide under these
catalyst conditions did not change the product ratio but reducing
the catalyst loading to 1:500 resulted in a 91%ee of the desired
enantiomer (+)-2 in improved diastereoselectivity. There are sev-
eral interesting observations from the screen. The BINAP ligand
does not perform to the same level as the Chan dipyridylphosphine
ligand (entries 6 vs 2, 3 vs 4). Also, using the bulkier Xyl-PPhos
ligand with the DIAPEN diamine improves the enantioselectivity
(84 vs 73%ee, entries 2 vs 4) while with the DPEN diamine the
smaller PPhos ligand is clearly superior (entries 1 vs 5).
Fig. 1. Piperidine-containing pharmaceuticals.
method for setting stereochemistry and relies on establishing an
equilibrium between the two enantiomers that can be differenti-
ated by the catalyst system to effect reduction of the ketone [12–
14]. In this way racemic material can be converted to a single enan-
tiomer. Since our previous experience suggested that if we could
obtain >90% ee of (3S, 4R)-1-Boc-3-fluoropiperidin-4-ol (+)-2, we
would be able to recrystallize to enantiopurity, we proceeded to
investigate the potential for such a DKR (Scheme 1).
To our knowledge the dynamic kinetic resolution of an
ropiperidinone has not been reported although enzymatic
a-fluo-
a
transanimation strategy has recently been published [15]. Given
the small difference in size between hydrogen and fluorine we con-
sidered enantioselective reduction to be challenging but wondered
whether the rigidity of the piperidinone ring may be beneficial.
Notably, cis-diastereoselective reduction of 2-methylcyclohex-
anone has been reported along with dynamic kinetic resolution
of 2-isopropylcyclohexanone [16]. Further, we noted that there
have been successful examples of substituted piperidinones as
Doubling the concentration from [0.17 M] to [0.34 M] while
halving the catalyst ratio to 1000:1 (0.1 mol%) and using only
10 mol% potassium tert-butoxide resulted in good cis diastereose-
lectivity (86%de) but a reduction in enantioselectivity to 85%ee.
Interestingly using the enantiomeric catalyst [(R)Xyl-PPhos RuCl2
(R)DAIPEN] under these same 1000:1 conditions resulted in 80%
de in favour of the cis diastereomer with the (3R, 4S)-enantiomer
Scheme 1. Comparison of previous enantioselective fluorination with the dynamic
kinetic resolution.
Table 1
Initial screen of catalysts for the reduction of rac-3.
Entry
Catalyst
Conversion (%)
trans-2 (%)
cis-2 (%)
ee (cis) (%)
de (%)
Other (%)
1
2
3
4
5
6
7
8
[(S)Xyl-PPhos RuCl2 (S)DAIPEN]
[(S)PPhos RuCl2 (S)DAIPEN]
[(S)Xyl-BINAP RuCl2 (S)DAIPEN]
[(S)dm-SegPhos RuCl2 (S)DAIPEN]
[(S)BINAP RuCl2 (S)DAIPEN]
[(S)Tol-BINAP RuCl2 (S,S)DPEN]
[(S)PPhos RuCl2 (S,S)DPEN]
[(S,S)ChiraPhos RuCl2 (S,S)DPEN]
[(S)Xyl-PPhos RuCl2 (S,S)DPEN]
[(R)Xyl-PPhos RuCl2 (R,R)DPEN]
[(R)PPhos RuCl2 (R,R)DACH]
[(R)PPhos RuCl2 (S)DAIPEN]
[(S)BINAP RuCl2 (R)DAIPEN]
100
100
100
100
100
100
86
100
96
84
8
11
13
15
10
1
4
48
4
91
89
86
83
88
99
81
51
87
76
70
73
89
71
64
30
25
25
58
48
25
6
À42
À14
À18
À24
84
78
74
69
80
98
91
3
91
83
65
78
80
1
0
1
2
2
0
1
1
5
1
0
1
1
9
10
11
12
13
7
15
9
85
83
100
10
Conditions: Substrate/catalyst ratio 50:1, 20 mol% KOtBu in iPrOH [0.17 M], H2 (30 bar), 50 °C, 18 h.
2

J. LePaih, D.A. Goff, R. Singh et al.
Tetrahedron Letters 66 (2021) 152764
Table 2
Reduced catalyst loading for the reduction of rac-3.
Entry
Catalyst
Conversion (%)
trans-2 (%)
cis-2 (%)
ee (cis) (%)
de (%)
Other (%)
1
2
3
4
5
6
7
8
[(S)Xyl-PPhos RuCl2 (S,S)DPEN]
[(S)Xyl-PPhos RuCl2 (S)DAIPEN]
[(S)BINAP RuCl2 (S)DAIPEN]
[(S)PPhos RuCl2 (S)DAIPEN]
100
100
100
100
100
100
100
100
29
9
1
4
5
6
9
4
70
90
92
95
94
92
90
95
54
84
56
73
68
50
82
91
41
82
97
92
90
88
82
92
1
1
7
1
1
2
1
1
[(S)PPhos RuCl2 (S,S)DPEN]
[(S)Xyl-BINAP RuCl2 (S)DAIPEN]
[(S)Xyl-PPhos RuCl2 (S)DAIPEN]*
[(S)Xyl-PPhos RuCl2 (S)DAIPEN]**
Conditions: Substrate/catalyst ratio 250:1, 20 mol% KOtBu in iPrOH [0.17 M], H2 (30 bar), 50 °C, 18 h; * 10 mol% KOtBu; ** Substrate/catalyst ration 500:1.
Scheme 2. Large scale formation of (+)-2 from rac-3 by Noyori reduction followed by crystallisation.
(À)-2 being the major product but with a lower enantioselectivity
of 61%ee. By reducing the concentration back to [0.2 M] allowed us
to match our best conditions to date achieving 94%de and 91%ee
with catalyst ratio of 1000:1 and 10 mol% potassium tert-butoxide.
Reduction could also be achieved under transfer hydrogenation
conditions without the need for high pressure. However, using tri-
ethylamine as base, the reductions showed little diastereoselectiv-
ity with low enantioselectivity of the cis product being observed.
We assume that under these conditions there is little epimeriza-
In summary, we have been able to develop a Noyori reduction
under dynamic kinetic resolution conditions of 1-Boc-3-fluo-
ropiperidin-4-one which is both diastereoselective and enantiose-
lective, a reaction that to our knowledge has not been reported
despite the significance of the resulting fluoropiperidinol to the
medicinal chemistry community. The reaction can be carried out
on a multi gram scale without impact to the selectivity with cata-
lyst loadings of 0.1 mol%. By carrying out a crystallisation protocol
the fluoropiperidinol can be obtained in >99%ee in a yield that
competes well with our previous enantioselective fluorination
procedure.
tion of the
a-fluoro stereocenter occurring. Using potassium tert-
butoxide in these reactions did increase the ratio of the cis
diastereomer; however, there was considerable decomposition in
these reactions and further studies were not carried out.
Declaration of Competing Interest
The reactions discussed above were carried out using a parallel
multi-reactor synthesizer on a 0.5–1 mmol scale [20]. Attempts to
lower the pressure of the reaction at catalyst loading of 0.1 mol%
had resulted in incomplete conversion after 18 h but did go to
completion with similar product ratios at higher catalyst concen-
trations. To make the reaction more useful we turned to a high-
pressure reaction vessel to carry out multi-gram reactions, where
it was hoped that the increase in scale would allow for lower
hydrogen pressures. In a series of studies with the [(S)Xyl-PPhos
RuCl2 (S)DAIPEN] catalyst on an increased scale it was shown that
the hydrogen pressure could be reduced down to 5 bar without
effecting the diastereo- or enantioselectivity. Indeed, we were able
to achieve complete conversion on a 16 g scale with only 5 bar of
hydrogen (Scheme 2). On completion of the reaction, the mixture
was filtered through a pad of silica and recrystallised using the
conditions developed previously [11]. The crystallization removed
the trans impurity and the undesired enantiomer to obtain (+)-2
with >99%ee, albeit in three crops, in an overall yield of 60%, which
compares well with the previously reported enantioselective fluo-
rination method (see Supplementary Material for experimental
details).
The authors are or were shareholders of either Johnson-Matthey
or Rigel Pharmaceuticals
Acknowledgments
We thank Mark Irving (Rigel) for chiral hplc analysis and
Richard Brzozowski (Johnson-Matthey) for assistance.
Appendix A. Supplementary data
Supplementary data to this article can be found online at
https://doi.org/10.1016/j.tetlet.2020.152764.
References
[1] M Baumann, I. R.. Baxendale, Beilstein J. Org. Chem. 9 (2013) 2265–2319.
[2] E. Vitaku, D.T. Smith, J.T. Njardarson, J. Med. Chem. 57 (2014) 10257–
10274.
[3] R. Roskoski, Jr. Blue Ridge Institute for Medical Research April 12, 2020 www.
brimr.org/PKI/PKI.htm.
3

J. LePaih, D.A. Goff, R. Singh et al.
Tetrahedron Letters 66 (2021) 152764
[4] C.D. Cox, P.J. Coleman, M.J. Breslin, D.B. Whitman, R.M. Garbaccio, M.E. Fraley,
C.A. Buser, E.S. Walsh, K. Hamilton, M.D. Schaber, R.B. Lobell, W. Tao, J.P.
Davide, R.E. Diehl, M.T. Abrams, V.J. South, H.E. Huber, M. Torrent, T.
Prueksaritanont, C. Li, D.E. Slaughter, E. Mahan, C. Fernandez-Metzler, Y. Yan,
L.C. Kuo, N.E. Kohl, G.D. Hartman, J. Med. Chem. 51 (2008) 4239–4252.
[5] M.B. van Niel, I. Collins, M.S. Beer, H.B. Broughton, S.K.F. Cheng, S.C. Goodacre,
A. Heald, K.L. Locker, A.M. MacLeod, D. Morrison, C.R. Moyes, D. O’Connor, A.
Pike, M. Rowley, M.G.N. Russel, B. Sohal, J.A. Stanton, S. Thomas, H. Verrier, A.P.
Watt, J.L. Castro, J. Med. Chem. 42 (1999) 2087–2104.
[10] P. Kwiatkowski, T.D. Beeson, J.C. Conrad, D.W.C. MacMillan, J. Am. Chem. Soc.
133 (2011) 1738–1741.
[11] S.J. Shaw, D.A. Goff, L.A. Boralsky, M. Irving, R. Singh, J. Org. Chem. 78 (2013)
8892–8897.
[12] For
a review of Noyori reductions of ketones including dynamic kinetic
resolutions see R. Noyori, T. Ohkuma Angew. Chem. Int. Ed. 40 (2001) 40–73.
[13] R. Noyori, T. Ikeda, T. Ohkuma, M. Widhaim, M. Kitamura, H. Takaya, S.
Akutagawa, N. Sayo, T. Saito, T. Taketomi, H. Kumobayashi, J. Am. Chem. Soc.
111 (1989) 9134–9135.
[6] E.P. Gillis, K.J. Eastman, M.D. Hill, D.J. Donnelley, N.A. Meanwell, J. Med. Chem.
58 (2015) 8315–8359.
[7] N.A. Meanwell, J. Med. Chem. 61 (2018) 5822–5880.
[14] M. Kitamura, M. Tokunaga, R. Noyori, J. Am. Chem. Soc. 115 (1993) 144–152.
[15] C. Molinaro, E.M. Phillips, B. Xiang, E. Milczek, M. Shevlin, J. Balsells, S. Ceglia, J.
Chen, L. Chen, Q. Chen, Z. Fei, S. Hoerrner, J. Qi, M. de Lera Ruiz, L. Tan, B. Wan, J.
Yin, J. Org. Chem. 84 (2019) 8006–8018.
[16] T. Ohkuma, H. Ooka, M. Yamakawa, T. Ikariya, R. Noyori, J. Org. Chem. 61
(1996) 4872–4873.
[17] T. Ohkuma, J. Li, R. Noyori, SynLett (2004) 1383–1386.
[18] M. Scalone, P. Waldmeier, Org. Proc. Res. Dev. 7 (2003) 418–425.
[19] J. Wu, A.S.C. Chan, Acc. Chem. Res. 39 (2006) 711–720.
[20] Reactions were carried out using a BiotageÒ EndevorTM Instrument.
[8] S.K. Bagal, M. Andrews, B.M. Bechle, J. Bian, J. Bisland, D.C. Blakemore, J.F.
Braganza, P.J. Bungay, M.S. Corbett, C.N. Cronin, J.J. Cui, R. Dias, N.J. Flanagan, S.
E. Greasley, R. Grimley, K. James, E. Johnson, L. Kitching, M.L. Kraus, I.
McAlpine, A. Nagata, S. Ninkovic, K. Omoto, S. Scales, S.E. Skerratt, J. Sun, M.
Tran-Dubé, G.J. Waldron, F. Wang, J.S. Warmus, J. Med. Chem. 61 (2018) 6779–
6800.
[9] V. Spahn, G. Del Vecchio, D. Labuz, A. Rodriguez-Gaztelumendi, N. Massaly, J.
Temp, V. Durmaz, P. Sabri, M. Reidelbach, H. Machelska, M. Weber, C. Stein,
Science 355 (2017) 966–969.
4