Enantioselective synthesis of cis -3-fluoropiperidin-4-ol, a Building block for medicinal chemistry

Article abstract of DOI:10.1021/jo401352z

The first enantioselective route to both enantiomers of cis-1-Boc-3-fluoropiperidin-4-ol, a highly prized building block for medicinal chemistry, is reported. An enantioselective fluorination is employed, taking advantage of the methodology reported by MacMillan, which uses a modified cinchona alkaloid catalyst. In studying the fluorination reaction, we have shown that the catalyst can be replaced by commercially available primary amines, including α-methylbenzylamine, with similar levels of enantioselectivity. The piperidinols are readily crystallized to obtain enantiopure material.

Full text of DOI:10.1021/jo401352z

Note
pubs.acs.org/joc
Enantioselective Synthesis of cis-3-Fluoropiperidin-4-ol, a Building
Block for Medicinal Chemistry
Simon J. Shaw,* Dane A. Goff, Luke A. Boralsky, Mark Irving, and Rajinder Singh
Rigel, Inc., 1180 Veterans Boulevard, South San Francisco, California, 94080, United States
S
* Supporting Information
ABSTRACT: The first enantioselective route to both enantiomers of cis-1-Boc-3-fluoropiperidin-4-ol, a highly prized building
block for medicinal chemistry, is reported. An enantioselective fluorination is employed, taking advantage of the methodology
reported by MacMillan, which uses a modified cinchona alkaloid catalyst. In studying the fluorination reaction, we have shown
that the catalyst can be replaced by commercially available primary amines, including α-methylbenzylamine, with similar levels of
enantioselectivity. The piperidinols are readily crystallized to obtain enantiopure material.
iperidines are common heterocycles in medicinal chem-
istry; the group is found in many pharmaceuticals both
naturally occurring and synthetic, from the alkaloid coniine
found in poison hemlock, to the opioid pethidine (commonly
known as Demerol) and antidepressants such as paroxetine
(Chart 1). More recently the introduction of a fluorine atom
number of groups.9−13 The recent results from the MacMillan
group14 in the enantioselective fluorination were particularly
striking for the discussion included the piperidinone example,
suggesting for the first time that an enantioselective route to a
3-fluoropiperidin-4-ol might be possible.
P
The MacMillan procedure14 makes use of the modified
cinchona alkaloid, 9-deoxy-9-epi-aminohydroquinidine 1, syn-
thesized using a one-pot Mitsunobu−Staudinger reduction
procedure from dihydroquinidine reported by Connon,15,16 as a
catalyst in 20 mol % with N-fluorobenzenesulfonimide at −10
°C to obtain (3R)-1-Boc-3-fluoropiperidin-4-one 2 from 2
equiv of 1-Boc-piperidin-4-one 3 in 24 h.
Chart 1. Piperidine-Containing Pharmaceuticals
We were able to repeat this fluorination procedure,
maintaining the temperature between −20 and −10 °C;
however, it was not possible to determine the enantiomeric
excess (ee) of the ketone 2; therefore, we reduced the
piperidinone to the corresponding cis-(3R,4S)-1-Boc-3-fluoro-
piperidin-4-ol 4 prior to determination of the ee,17 which was
determined to be 97% ee. As a confirmation, the corresponding
tosylate 5 (96% ee) was also synthesized (Scheme 1). Our
fluorination reaction, in terms of both yields and ee, is in line
with that reported by MacMillan.
In order to obtain the enantiomeric (3S,4R)-fluoropiper-
idinol we generated the pseudoenantiomer of catalyst 1 derived
from dihydroquinine, 9-deoxy-9-epi-aminodihydroquinine 6. In
this instance the (3S,4R)-1-Boc-3-fluoropiperidin-4-ol 7 and the
corresponding tosylate 8 were obtained in 70% ee under similar
conditions, significantly lower than that from the quinidine
series (Scheme 1).
onto the piperidine ring has been described as an effective
method to modulate the pKa of the piperidine. Indeed it has
been reported that an axial fluorine atom on a piperidine ring
lowers the pKa of the nitrogen by one log unit, while an
equatorial piperidine can reduce the pKa by almost two
units.1−3 The unique features of β-fluoroamines have recently
been discussed.4 The introduction of the fluorine atom onto
the piperidine ring has also been associated with improvements
to pharmacokinetic properties, with reports of improved
bioavailability1,5 and apparent permeability.6 In all of the
syntheses reported to date, including those described in the
patent literature, the fluorine atom is installed directly on the
piperidine ring through a racemic approach from the silyl enol
ether using Selectfluor.1,2,7 Indeed the only means of obtaining
single enantiomers has been through the use of preparative
chiral HPLC,3,7 an undesirable route for large scale
manufacture of these building blocks.8
Given the lower enantioselectivity with 6, several cinchona
alkaloid derivatives were screened to investigate enantioselec-
tivity (Table 1; a more detailed version of this table appears in
The introduction of a fluoro or trifluoromethyl group into
organic molecules is a topic of current investigation in a
Received: June 24, 2013
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jo401352z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Note
Scheme 1. Fluorination of Piperidinone
NMR and HPLC does retain a yellow coloration. This yellow
color is not present in any other of the cinchona alkaloid-
derived catalysts.18 In the quinidine-derived series, both
catalysts give similar levels of ee (entries 4 and 5). Both the
cinchonidine-derived (entry 3) and cinchonine-derived (entry
6) catalyst give lower levels of ee.19 The 9-aminoquinine
catalyst (entry 7) generated by initial inversion of the alcohol
through a Mitsunobu-hydrolysis reaction as described by
Skarewski20 resulted in very poor fluorination (5%, for an
overall yield to 4 of 3%) and low ee, and it may be the result of
the uncatalyzed reaction pathway. It was noted that the amine
appears to be much more hindered in this configuration, for
more forcing conditions were required to effect the Staudinger
reduction of the intermediate azide to the amine. The catalyst
has been noted to result in lower conversions in other catalytic
systems.15
Table 1. Fluorination Using Cinchona Alkaloid-Derived
Catalysts
The results led us to question if there were alternative
commercially available amines that may give similar enantio-
selectivities in the fluorination reaction that did not require
preparation, a practical advantage especially for large scale
reactions. Mindful of the original study conducted by the
MacMillan group, we chose to focus primarily on benzylic
amines. Primary-amine catalysis is a field that is under active
study.21−23 Our own screening paradigm began with inves-
tigation of the reaction with the readily available and
inexpensive (R)-α-methylbenzylamine, which has been used
to induce chiral selectivity widely,24 going back to asymmetric
Strecker synthesis.25
Using this catalyst under the same conditions as those used
above, we were able to get catalytic turnover to obtain the
fluoropiperidinone 9 in similar yield to that obtained with the
dihydroquinine-derived catalyst 6. Upon reduction to 7 the
enantioselectivity was determined to be 64% ee, similar to that
obtained with 6. Conversion to the tosylate 8 confirmed this
level of enantioselectivity. Encouraged that we were maintain-
ing catalyst turnover and obtaining appreciable levels of
enantioselectivity with the α-methylbenzylamine, a collection
of commercially available benzylic amines was screened (Table
2).
the Supporting Information). Although the fluoropiperidinone
2 or 9 was isolated by column chromatography, it was not pure,
containing a small quantity of the 3,5-difluoropiperidinone.
However, after subsequent reduction of the ketone, high purity
cis-fluoropiperidinol 4 or 7 was obtained following purification,
making the two-step overall yield a more accurate representa-
tion of material throughput. Use of tetramethylammonium
borohydride as reductant was found to give an improved ratio
of the cis diastereomer over the trans (approximately 4:1),
compared with the more commonly used sodium borohydride,1
which gave diastereomeric ratios of 2:1 to 3:1 in our hands. It
was found that the trans isomer could be obtained as the major
product by using borane as reductant; however, for this study
we focused on the cis diastereomer.
As can be seen in Table 1, there is a striking difference in ee
between the dihydroquinine- and quinine-derived catalysts
(entry 1 vs 2), with the catalyst obtained from quinine giving
both excellent enantioselectivity and yield. To date we do not
have an adequate explanation for this observation; however, it is
noted that in synthesizing the catalyst 6, isolation of the triple
HCl salt according to the Connon procedure by crystallization
results in a gel-like solid, which while appearing pure by both
The results show that enantioselectivities up to 80% ee can
be obtained with benzylic amines (entries 12, 14). Increasing
the size of the aromatic group in the case of the naphthalenes
(entries 11 and 12) and the (R)-α-methyl-o-methoxybenzyl-
amine (entry 13) improves the selectivity. A similar improve-
B
dx.doi.org/10.1021/jo401352z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Note
Table 2. Fluorinations Using Benzylic Amines
Table 3. Enantiomeric Fluorination
ment in enantioselectivity is obtained by the introduction of a
methoxy group onto the α-methyl (entry 14), a compound
readily available from the reduction of phenylglycine.
Furthermore, the results clearly show that the primary amines
are superior to secondary amines (entry 8 vs 10). With these
catalysts, similar yields are obtained to those for the quinine
derivative 6. Although enantiopure material is not obtained
from the reaction, the alcohol 7 can be recrystallized from 70 to
>98% ee in a single crystallization from ether−hexane allowing
practical access to the enantiopure (3S,4R)-3-fluoropiperidin-4-
ol 7 for the first time. The crystals were suitable for X-ray
crystallography (vide infra).
Figure 1. X-ray crystal structures of the fluoropiperidinols.
Given the commercial availability of the benzylic amines,
access to the enantiomeric series could be readily achieved. As
expected, similar levels of enantioselectivity are observed in
each case where both enantiomers of the catalyst were used
(Table 3). Given the importance of the sterics of the aromatic
group in enantioselectivity, it is not surprising that there is a
large drop in enantioselectivity when moving from a benzylic
amine to the phenethylamine (compare entries 18 and 19). The
triphenylethanolamine catalyst (entry 21) shows the highest
level of ee of all the commercial catalysts, which is in line with
the observation that increasing the steric environment around
the amine improves the ee; however, the conversions are low,
which may also be a result of the sterically encumbered
environment of the catalyst.
Again it was possible to obtain a single enantiomer by
crystallization from ether−hexane. X-ray crystallography
confirmed both the relative stereochemistry and the absolute
configuration (Figure 1). In both of the X-ray structures, the
fluorine occupies the axial position, with the vicinal hydroxyl
occupying the equatorial position. The solid state structures are
in line with previous molecular dynamic and NMR studies26 on
related piperidines as well as a crystal structure of the related 3-
fluoro-4,4-diphenylpiperidine,27 which are thought to be
influenced by a charge dipole interaction between the fluorine
and the piperidine nitrogen. In these instances, the 4-hydroxyl
being equatorial increases the preference for the axial fluorine.
However, it should be noted that the solution structure appears
to be more dynamic as evidenced by the broad proton NMR at
room temperature and two resonances in the fluorine NMR;
both phenomena are resolved at 50 °C (see Supporting
Information), an observation that was previously noted with 3-
fluoro-4,4-diphenylpiperidine.27
Finally, we chose to investigate the effect of raising the
temperature above −10 °C used in the original procedure.
Taking both the best catalyst we had identified, the quinine-
derived catalyst, and the most readily available, the inexpensive
(R)-α-methylbenzylamine, under the same conditions at 0 °C
and room temperature resulted in fluoropiperidinone 9
formation, which in each instance was reduced to the
corresponding piperidinol 7. The results show that at 0 °C
the enantioselectivity is not impacted with either catalyst, while
at 20 °C (room temperature) there appears to be a lowering of
enantioselectivity with both catalysts (Table 4). The result is
more pronounced in the case of α-methylbenzylamine. The
results were confirmed by the tosylate 8. Further, when carrying
out the reaction at 0 °C, the amount of catalyst could be
reduced from the original 20 to 10% without seeing a drop in
either yield or ee with both catalysts investigated. It is
interesting to note that using the optimized reaction conditions
with Selectfluor as fluorinating agent and acetonitrile as solvent
failed to yield any fluorinated product. The increased ease of
synthesis provided by a 0 °C reaction temperature and reduced
catalyst loading has practical advantages, particularly in larger
C
dx.doi.org/10.1021/jo401352z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Note
equiv) dropwise. The reaction was stirred at 0 °C for 10 min, and
diphenylphosphoryl azide (40.7 g, 32.0 mL, 148.1 mmol, 1.2 equiv)
was added dropwise. The reaction was stirred at room temperature for
4 h and heated to 45 °C for 2 h. Triphenylphosphine (38.8 g, 148.1
mmol, 1.2 equiv) was added, and the reaction was stirred at 45 °C for
a further 2 h (Note: some evolution of nitrogen was observed on
addition of the triphenylphosphine). Water (40 mL) was added, and
the reaction was stirred at room temperature for 16 h. The reaction
was concentrated, and the residue was partitioned between CH2Cl2
(400 mL) and HCl (2M, 350 mL). The aqueous phase was washed
with CH2Cl2 (2 × 240 mL) and concentrated to dryness to obtain a
yellow solid. The solid was recrystallized from EtOAc−MeOH (1:1,
320 mL) to yield the title compound (44.9 g, 84%) as a pale yellow
Table 4. Effect of Temperature on the Fluorination Reaction
1
solid: H NMR (DMSO-d6) δ 9.75 (2H, br s, NH2), 9.14 (1H, d, J =
5.0 Hz, H-2′ or H-3′), 8.46 (1H, d, J = 5.0 Hz, H-2′ or H-3′), 8.35
(1H, d, J = 9.0 Hz, H-8′), 8.00 (1H, d, J = 2.0 Hz, H-5′), 7.75 (1H, dd,
J = 9.0, 2.0 Hz, H-7′), 5.99 (1H, d, J = 10.0 Hz, H-9), 5.86 (1H, ddd, J
= 17.0, 10.5, 6.5 Hz, CH2CH), 5.27 (1H, d, J = 17.0 Hz,
transCH2CH), 5.13 (1H, d, J = 10.5 Hz, cisCH2CH), 4.80 (1H,
m, H-8), 4.10 (1H, m, 1 × H-6), 4.06 (3H, s, OCH3), 3.72 (1H, dd, J
= 12.5, 10.5 Hz, 1 × H-2), 3.37 (2H, m, 1 × H-2, 1 × H-6), 2.75 (1H,
m, H-3), 1.87 (3H, m, 1 × H-2, H-5), 1.50 (1H, dd, J = 12.5, 9.5 Hz, 1
× H-7), 0.86 (1H, m, 1 × H-7); 13C NMR (DMSO-d6) δ 160.0, 144.2
(2C), 138.3, 129.2, 126.8, 126.0, 122.1, 117.1, 103.7, 59.1, 57.2, 52.4,
48.3, 42.2, 36.3, 26.0, 23.9 (2C); m/z 324 [M + H]+; HRMS m/z [M
+ H]+ calcd for C20H26N3O+ 324.2070, found 324.2072.
Formation of (R)-α-Methylbenzylamine Monotrichloroacetic
Acid Salt Monohydrate. To a solution of (R)-α-methylbenzylamine
(0.236 g, 1.95 mmol, 1.0 equiv) in tetrahydrofuran (10 mL) was added
trichloroacetic acid (0.334 g, 2.05 mmol, 1.05 equiv) and water (0.035
mL, 1.95 mmol, 1.0 equiv). The solution was used directly in the
following reaction.
a
Using 10 mol % of catalyst.
scale synthesis. The refinements in the reaction allowed the
procedure to be scaled-up to produce 23 g of enantiopure 1-
Boc-3S-fluoro-4R-piperidinol 7 in one batch (see the
Experimental Section).
In summary, we have developed the first enantioselective
route to either enantiomer of the important medicinal
chemistry building block 3-flouropiperidin-4-ol using an
enantioselective fluorination of the corresponding piperidinone.
Further, during our investigation we have carried out a screen
of catalysts and shown that the modified cinchona alkaloid
originally reported to catalyze the fluorination reaction may be
replaced by a commercially available primary benzylic amine
and carried out at 0 °C with 10 mol % catalyst without
detriment. The resulting cis-alcohols can be crystallized from
65% ee to enantiomerically pure material in two rounds of
crystallization (from 70% ee and above, only one crystallization
is required). These refinements will make the reaction more
amenable to scale-up for this key building-block.
Formation of 1-Boc-3S-fluoropiperidin-4-one 9 Using (R)-α-
Methylbenzylamine. A similar procedure was used to that reported
by MacMillan.14 To a suspension of freshly ground sodium carbonate
(1.55 g, 14.63 mmol, 1.5 equiv) and N-fluorobenzenesulfonimide
(3.07 g, 9.75 mmol, 1.0 equiv) in tetrahydrofuran (20 mL) at −20 °C
was added (R)-α-methylbenzylamine trichloroacetic acid salt mono-
hydrate (1.95 mmol, 0.2 equiv) as a solution in tetrahydrofuran (10
mL). The reaction was stirred at −20 °C for 10 min before adding 1-
Boc-piperidin-4-one (3.88 g, 19.50 mmol, 2.0 equiv) in three portions.
The reaction was stirred between −20 and −10 °C for 24 h, before
Et2O (30 mL) was added. The reaction was filtered through silica to
remove insolubles, eluting with Et2O (200 mL). The filtrate was
concentrated under reduced pressure. Column chromatography (20 →
60% EtOAc−hexane) yielded the title compound (1.52 g, 72%) as a
white solid: 1H NMR δ 4.87 (1H, ddd, J = 48.0, 9.5, 6.5 Hz, pipH-3),
4.43 (1H, m, 1H of pipH-2, H-6), 4.16 (1H, m, 1H of pipH-2, H-6),
3.23 (2H, m, 2H of pipH-2, H-6), 2.64−2.44 (2H, m, pipH-5), 1.48
(9H, s, C(CH3)3); 19F NMR δ −197.4.
EXPERIMENTAL SECTION
■
All commercial reagents were used without further purification. All
solvents were reagent or HPLC grade. Anhydrous tetrahydrofuran and
dichloromethane were purchased and used directly. All reactions were
carried out under a nitrogen atmosphere in oven-dried glassware.
Analytical TLC was performed on silica gel 60 F254 plates and
visualized by UV if possible and p-anisaldehyde or ceric ammonium
molybdate staining. Flash chromatography was carried out using an
automated system with prepacked silica columns. Yields refer to
1
chromatographically and spectroscpically pure compounds. H NMR,
13C NMR, and 19F NMR spectra were recorded on a 300 MHz
spectrometer at ambient temperature unless otherwise stated.
Chemical shifts are reported in parts per million relative to residual
solvent CDCl3 (1H 7.26 ppm; 13C 77.0 ppm) and unreferenced for
19F. Multiplicities are reported as follows: s = singlet; d = doublet; t =
triplet; m = multiplet; br = broad. Proton assignments were made
based on COSY spectra. High resolution mass spectra were recorded
using a time-of-flight mass spectrometer. Chiral analysis was
performed using either HPLC or supercritical fluid chromatography
(SFC). The methods chosen were based upon ability to obtain
baseline separation of the racemic mixtures for the compounds of
interest. The mobile phase for the HPLC experiments was an isocratic
hexane/alcohol mixture, whereas the SFC mobile phase was an
isocratic mixture of CO2 and alcohol. Stationary phases for both
HPLC and SFC were purchased from commercial suppliers; see the
Supporting Information for more details.
1-Boc-3R-fluoropiperidin-4-one 2. A similar procedure was used
to that for 9 using (S)-α-methylbenzylamine on a 9.75 mmol scale to
1
yield the title compound (1.50 g, 71%) as a white solid: H NMR
(CDCl3) δ 4.87 (1H, ddd, J = 48.0, 9.5, 6.5 Hz, pipH-3), 4.43 (1H, m,
1H of pipH-2, H-6), 4.16 (1H, m, 1H of pipH-2, H-6), 3.23 (2H, m,
2H of pipH-2, H-6), 2.64−2.44 (2H, m, pipH-5), 1.48 (9H, s,
C(CH3)3); 19F NMR δ −197.3.
Formation of 1-Boc-3S-fluoro-4R-piperidinol 7. To a solution
of the piperidinone (1.52 g, 7.00 mmol, 1.0 equiv) in methanol (70
mL) at 0 °C was added tetramethylammonium borohydride (0.75 g,
8.41 mmol, 1.2 equiv) over 20 min. The reaction was stirred at 0 °C
for 4 h before NH4Cl (20 mL) was added, and then mixture was
stirred at room temperature for 45 min. The reaction was concentrated
to remove methanol and partitioned between EtOAc (100 mL) and
NH4Cl (100 mL). The aqueous phase was extracted with EtOAc (2 ×
100 mL). The combined organics were dried (Na2SO4) and
concentrated under reduced pressure. Column chromatography (silica,
20 → 70% EtOAc−hexane) yielded the title compound (0.86 g, 56%)
as a white solid. Chiral HPLC determined the material to be 64% ee:
IR (film) 3423 (br), 2973, 2934, 1709, 1687, 1427, 1366, 1246, 1166,
Formation of 9-Deoxy-9-epi-aminoquinine 3HCl. A similar
procedure is used to that reported in by Connon.15 To a solution of
quinine (40.0 g, 123.5 mmol, 1.0 equiv) and triphenylphosphine (38.8
g, 148.1 mmol, 1.0 equiv) in tetrahydrofuran (400 mL) at 0 °C was
added diisopropyl azodicarboxylate (29.9 g, 29.1 mL, 148.1 mmol, 1.2
D
dx.doi.org/10.1021/jo401352z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Note
1128, 1089, 999, 870 cm−1; 1H NMR δ 4.58 (1H, br ddt, J = 48.5, 6.5,
3.5 Hz, pipH-3), 3.95−3.80 (2H, m, 1H of pipH-2, pipH-4), 3.68 (1H,
m, 1H of pipH-6), 3.39 (1H, m, 1H of pipH-2), 3.14 (1H, m, 1H of
pipH-6), 2.39−2.30 (1H, m, OH), 1.84−1.68 (2H, m, pipH-5), 1.43
(9H, s, C(CH3)3); 13C NMR δ 154.9, 88.5 (d, J = 178.0 Hz), 80.1,
67.9 (d, J = 18.0 Hz), 44.3 (br), 40.0 (br), 29.3, 28.3; 19F NMR δ
−201.8, −202.9; m/z 164 [M + H − C4H8]+, 120 [M + H − CO2 −
C4H8]+; HRMS m/z [M + H − C4H8 − CO2]+ calcd for C5H11FNO+
120.0819, found 120.0820. Recrystallization from 1:1 Et2O−hexane
resulted in white crystals, which were obtained by filtration and dried
under a vacuum. The crystals were determined to be >98% ee by chiral
HPLC: [α]2D0 +16.0 (c 1.40, CHCl3).
calcd for C12H17FNO3S+ 274.0908, found 274.0922. The reaction was
carried out on enantiopure piperidinol 4 to obtain the tosylate 5,
which was determined to be 97% ee by chiral HPLC: [α]2D0 −10.6 (c
1.08, CHCl3);
Scale-up Procedure for the Formation of 9. To a suspension of
freshly ground sodium carbonate (59.9 g, 563.3 mmol, 1.5 equiv) and
N-fluorobenzenesulfonimide (99.5 g, 379.9 mmol, 1.0 equiv) in
tetrahydrofuran (800 mL) at 0 °C was added 9-deoxy-9-epi-
aminoquinine trichloroacetic acid salt monohydrate (44.6 mmol,
0.12 equiv) as a solution in tetrahydrofuran (200 mL). The reaction
was stirred at 0 °C for 10 min before 1-Boc-piperidin-4-one (150.0 g,
753.8 mmol, 2.0 equiv) was added in three portions. The reaction was
stirred at 0 °C for 24 h, before Et2O (400 mL) was added. The
reaction was filtered through silica to remove insolubles, eluting with
Et2O (500 mL). The filtrate was concentrated under reduced pressure.
Column chromatography (silica, 20 → 60% EtOAc−hexane) yielded
the title compound (53.8 g, 66%) as a white solid: data agrees with
that stated.
Scale-up Procedure for the Formation of 7. To a solution of
the 1-Boc-3S-fluoropiperidin-4-one (53.8 g, 247.9 mmol, 1.0 equiv) in
methanol (1000 mL) at 0 °C was added tetramethylammonium
borohydride (22.1 g, 247.9 mmol, 1.0 equiv) over 1.25 h. The reaction
was stirred at 0 °C for 3 h before NH4Cl (300 mL) was added, and the
mixture was stirred at room temperature for 1 h. The reaction was
concentrated to remove methanol and partitioned between EtOAc
(500 mL) and NH4Cl (300 mL). The aqueous phase was extracted
with EtOAc (3 × 300 mL). The combined organics were dried
(Na2SO4) and concentrated under reduced pressure. Column
chromatography (silica, 20 → 70% EtOAc−hexane) yielded the title
compound (30.6 g, 56%) as a white solid, which was determined to be
96% ee by chiral HPLC. Recrystallization from Et2O−hexane (1:1, 500
mL) yielded white crystals (23.0 g, 75%) that were isolated by
filtration and dried under a vacuum. The crystals were shown to be
>97% ee by chiral HPLC: data agrees with that stated.
1-Boc-3R-fluoro-4S-piperidinol 4. A similar procedure was used
to that for 7 using 1-Boc-3S-fluoropiperidinone 2 on a 6.91 mmol scale
to yield the title compound (0.74 g, 49%) as a white solid. Chiral
HPLC determined the material to be 66% ee: IR (film) 3419 (br),
2977, 2934, 1683, 1446, 1426, 1366, 1244, 1167, 1126, 1087, 1001,
871 cm−1; 1H NMR δ 4.58 (1H, br d, J = 48.0 Hz, pipH-3), 3.96−3.82
(2H, m, 1H of pipH-2, pipH-4), 3.69 (1H, m, 1H of pipH-6), 3.38
(1H, m, 1H of pipH-2), 3.15 (1H, m, 1H of pipH-6), 2.29 (1H, d, J =
6.0 Hz, OH), 1.88−1.66 (2H, m, pipH-5), 1.44 (9H, s, C(CH3)3); 13
C
NMR δ 154.9, 88.5 (d, J = 177.0 Hz), 80.1, 67.9 (d, J = 18.0 Hz), 44.6
(br), 40.0 (br), 29.2, 28.3; 19F NMR δ −201.8, −202.9; m/z 164 [M +
H − C4H8]+, 120 [M + H − C4H8 − CO2]+; HRMS m/z [M + H −
C4H8 − CO2]+ calcd for C5H11FNO+ 120.0819, found 120.0821.
Recrystallization from 1:1 Et2O−hexane (two rounds) resulted in
crystals, which were determined to be >98% ee by chiral HPLC: [α]D20
−16.4 (c 1.73, CHCl3).
Formation of cis-1-Boc-3S-fluoro-4R-toluenesulfonylpiper-
idine 8. To a solution of 1-Boc-3S-fluoro-4R-piperidinol (0.037 g,
0.169 mmol, 1.0 equiv) and p-toluenesulfonyl chloride (0.039 g, 0.203
mmol, 1.2 equiv) in dichloromethane (1.5 mL) was added
triethylamine (0.036 mL, 0.253 mmol, 1.5 equiv) and dimethylami-
nopyridine (0.002 g, 0.017 mmol, 0.1 equiv). The reaction was stirred
at room temperature for 14 h before pouring into NaHCO3 (20 mL).
The organics were extracted with CH2Cl2 (3 × 20 mL), combined,
dried (Na2SO4) and concentrated under reduced pressure. Column
chromatography (20 → 60% EtOAc−hexane) yielded the title
compound (0.051 g, 81%) as a white solid. Chiral HPLC determined
the material to be 61% ee: IR (film) 2977, 1699, 1423, 1367, 1244,
On purifying the piperidinol, two further compounds were
obtained, 1-Boc-3,5-difluoro-4-piperidinol (4.4 g, 8%) and the 1-Boc-
3S-fluoro-4S-piperidinol (5.1 g, 9%).
ASSOCIATED CONTENT
■
1
1191, 1177, 1010, 967, 879, 842, 673 cm−1; H NMR (50 °C) δ 7.80
S
* Supporting Information
(2H, d, J = 8.5 Hz, 2H of C6H4Me), 7.33 (2H, d, J = 8.0 Hz, 2H of
C6H4Me), 4.73 (1H, dddd, J = 19.0, 9.0, 4.0, 2.5 Hz, pipH-4), 4.58
(1H, dddd, J = 47.5, 6.0, 3.0, 2.5 Hz, pipH-3), 3.93 (1H, dt, J = 14.5,
6.5 Hz, 1H of pipH-2), 3.76 (1H, m, 1H of pipH-6), 3.34 (1H, ddd, J
= 24.0, 14.0, 2.0 Hz, 1H of pipH-2), 3.15 (1H, br dd, J = 10.5, 9.5 Hz,
1H of pipH-6), 2.44 (3H, s, C6H4CH3), 2.16−2.02 (1H, m, 1H of
pipH-5), 1.76−1.69 (1H, m, 1H of pipH-5), 1.44 (9H, s, C(CH3)3);
13C NMR (50 °C) δ 154.6, 144.9, 134.2, 129.8, 127.7, 118.3, 85.5 (d, J
= 187.0 Hz), 80.4, 77.1, 40.1, 28.2, 27.4, 21.5; 19F NMR (50 °C) δ
−201.2; m/z 274 [M + H − C4H8 − CO2]+; HRMS m/z [M + H −
C4H8 − CO2]+ calcd for C12H17FNO3S+ 274.0908, found 274.0915.
The reaction was carried out on enantiopure piperidinol 7 to obtain
the tosylate 8, which was determined to be >98% ee by chiral HPLC:
[α]2D0 +11.4 (c 1.03, CHCl3).
An expanded version of Table 1 is given. NMR spectra of the 9-
deoxy-9-epi-aminoquinine 3HCl, compounds 4, 5, 7, and 8, and
a comparison of the 1H and 19F NMR of the 3-fluoropiperidin-
4-ol at both room temperature and 50 °C are included. COSY
spectra of 4, 8, and 9-deoxy-9-epi-aminoquinine 3HCl are
provided. Detailed chiral chromatography conditions, chroma-
tograms for the racemic mixtures of both 4/7 and 5/8 are given
along with examples of the chromatograms obtained in the
reactions and after crystallization. Crystallographic data for
compounds 4 and 7 is provided, along with ellipsoid plots and
CIF files. This material is available free of charge via the
Internet at http://pubs.acs.org.
1-Boc-3R-fluoro-4S-toluenesulfonylpiperidine 5. A similar
procedure was used to that for 1-Boc-3S-fluoro-4R-toluenesulfonylpi-
peridine 8 on a 0.169 mmol scale using 1-Boc-3R-fluoro-4S-piperidinol
4 to yield the title compound (0.050 g, 79%) as a white solid. Chiral
HPLC determined the material to be 67% ee: IR (film) 2977, 2934,
1698, 1424, 1366, 1244, 1176, 1010, 967, 879, 842 cm−1; 1H NMR (50
°C) δ 7.81 (2H, d, J = 8.5 Hz, 2H of C6H4Me), 7.34 (2H, d, J = 8.0
Hz, 2H of C6H4Me), 4.73 (1H, dddd, J = 20.5, 9.5, 4.0, 2.5 Hz, pipH-
4), 4.58 (1H, dddd, J = 47.5, 6.0, 3.5,2.0 Hz, pipH-3), 3.92 (1H, br dt,
J = 14.0, 7.0 Hz, 1H of pipH-2), 3.76 (1H, m, 1H of pipH-6), 3.33
(1H, ddd, J = 23.5, 14.0, 2.0 Hz, 1H of pipH-2), 3.15 (1H, br dd, J =
9.5, 8.5 Hz, 1H of pipH-6), 2.45 (3H, s, C6H4CH3), 2.14−2.02 (1H,
m, 1H of pipH-5), 1.77−1.65 (1H, m, 1H of pipH-5), 1.44 (9H, s,
C(CH3)3); 13C NMR (50 °C) δ ; 19F NMR (50 °C) δ −200.9; m/z
274 [M + H − C4H8 − CO2]+; HRMS m/z [M + H − C4H8 − CO2]+
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: sshaw@rigel.com.
Notes
The authors declare the following competing financial
interest(s):The authors are all employees of Rigel, Inc.
ACKNOWLEDGMENTS
■
We thank Van Ybarra and Duayne Tokushige for high
resolution mass spectral determinations. X-ray crystallography
was performed by Dr. Peter S. White at the University of North
Carolina, Chapel Hill.
E
dx.doi.org/10.1021/jo401352z | J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Note
(23) For a recent example, see: Lalonde, M. P.; McGowan, M. A.;
Rajapaksa, N. S.; Jacobsen, E. N. J. Am. Chem. Soc. 2013, 135, 1891−
1894.
(24) For reviews, see: Juaristi, E.; Leon-Romo, J. L.; Reyes, A.;
Escalante, J. Tetrahedron: Asymmetry 1999, 10, 2441−2495. Bandala,
Y.; Juaristi, E. Aldrichimica Acta 2010, 43, 65−78.
(25) Harada, H. Nature 1963, 200, 1201.
(26) Lankin, D. C.; Chandrkumar, N. S.; Rao, S. N.; Spangler, D. P.;
Snyder, J. P. J. Am. Chem. Soc. 1993, 115, 3356−3357.
(27) Sun, A.; Lankin, D. C.; Hardcastle, K.; Snyder, J. P. Chem.Eur.
J. 2005, 11, 1579−1591.
REFERENCES
■
(1) van Niel, M. B.; Collins, I.; Beer, M. S.; Broughton, H. B.; Cheng,
S. K. F.; Goodacre, S. C.; Heald, A.; Locker, K. L.; MacLeod, A. M.;
Morrison, D.; Moyes, C. R.; O’Connor, D.; Pike, A.; Rowley, M.;
Russell, M. G. N.; Sohal, B.; Stanton, J. A.; Thomas, S.; Verrier, H.;
Watt, A. P.; Castro, J. L. J. Med. Chem. 1999, 42, 2087−2104.
(2) Cox, C. D.; Coleman, P. J.; Breslin, M. J.; Whitman, D. B.;
Garbaccio, R. M.; Fraley, M. E.; Buser, C. A.; Walsh, E. S.; Hamilton,
K.; Schaber, M. D.; Lobell, R. B.; Tao, W.; Davide, J. P.; Diehl, R. E.;
Abrams, M. T.; South, V. J.; Huber, H. E.; Torrent, M.;
Prueksaritanont, T.; Li, C.; Slaughter, D. E.; Mahan, E.; Fernandez-
Metzler, C.; Yan, Y.; Kuo, L. C.; Kohl, N. E.; Hartman, G. D. J. Med.
Chem. 2008, 51, 4239−4252.
́
(3) Koudih, R.; Gilbert, G.; Dhilly, M.; Abbas, A.; Barre, L.;
Debruyne, D.; Sobrio, F. Eur. J. Med. Chem. 2012, 53, 408−415.
(4) Zimmer, L. E.; Sparr, C.; Gilmour, R. Angew. Chem., Int. Ed. 2011,
50, 11860−118671.
(5) Walpole, C.; Liu, Z.; Lee, E. E.; Yang, H.; Zhou, F.; Mackintosh,
N.; Sjogren, M.; Talyor, D.; Shen, J.; Batey, R. A. Tetrahedron Lett.
2012, 53, 2942−2947.
(6) Zak, M.; Mendonca, R.; Balazs, M.; Barrett, K.; Bergeron, P.;
Blair, W. S.; Chang, C.; Deshmukh, G.; DeVoss, J.; Dragovich, P. S.;
Eigenbrot, C.; Ghilardi, N.; Gibbons, P.; Gradl, S.; Hamman, C.;
Hanan, E. J.; Harsted, E.; Hewitt, P. R.; Hurley, C. A.; Jin, T.; Johnson,
A.; Johnson, T.; Kenny, J. R.; Koehler, M. F. T.; Kohli, P. B.;
Kulagowski, J. J.; Labadie, S.; Liao, J.; Liimatta, M.; Lin, Z.; Lupardus,
P. J.; Maxey, R. J.; Murray, J. M.; Pulk, R.; Rodriguez, M.; Savage, S.;
Shia, S.; Steffek, M.; Ubhayakar, S.; Ultsch, M.; van Abbma, A.; Ward,
S. I.; Xiao, L.; Xiao, Y. J. Med. Chem. 2012, 55, 6176−6193.
(7) Darout, E.; Deninno, M. P.; Futatsugi, K.; Guimaraes, C. R. W.;
Lefker, B. A. WO2010/128425A1.
(8) We note that an enantioselective synthesis of a 1,2-amino-
fluorocyclohexane has been reported from an acyclic intermediate.
Schulte, M. L.; Lindsley, C. W. Org. Lett. 2011, 13, 5684−5687.
(9) Enders, D.; Faure, S.; Potthoff, M.; Runsink, J. Synthesis 2001,
2307−2319.
(10) Phipps, R. J.; Hiramatsu, K.; Toste, F. D. J. Am. Chem. Soc. 2012,
134, 8376−8379.
(11) Pham, P. V.; Nagib, D. A.; MacMillan, D. W. C. Angew. Chem.,
Int. Ed. 2011, 50, 6119−6122.
(12) Fujiwara, Y.; Dixon, J. A.; O’Hara, F.; Daa Funder, E.; Dixon, D.
́
D.; Rodriguez, R. A.; Baxter, R. D.; Herle, B.; Sach, N.; Collins, M. R.;
Ishihara, Y.; Baran, P. S. Nature 2012, 492, 95−99.
(13) Morimoto, H.; Tsubogo, T.; Litvinas, N. D.; Hartwig, J. F.
Angew. Chem., Int. Ed. 2011, 50, 3793−3798.
(14) Kwiatkowski, P.; Beeson, T. D.; Conrad, J. C.; MacMillan, D. W.
C. J. Am. Chem. Soc. 2011, 133, 1738−1741.
(15) McCooey, S. H.; Connon, S. J. Org. Lett. 2007, 9, 599−602.
(16) Cassani, C.; Martin-Rapun, R.; Arceo, E.; Bravo, F.; Melchiorre,
P. Nat. Protoc. 2013, 8, 325−344.
(17) MacMillan reports using chiral GC to determine the ee of the
fluoropiperidinone (ref 14). Without access to a chiral GC instrument,
we investigated chiral HPLC; however, we observed several peaks in
the UV trace independent of the means of synthesis of the
fluoropiperidinone. We were thus not able to make an accurate
determination of enantiopurity at the fluoropiperidinone stage;
however, both the cis alcohol and the corresponding tosylate could
be determined using chiral HPLC. See the Supporting Information.
(18) The free base 6 was purified by column chromatography
without improving the enantioselectivity.
(19) MacMillan investigated the cinchonine-derived catalyst with
cyclohexanone as substrate and saw lower levels of ee relative to the
quinidine-derived catalyst; see ref 14.
(20) Sidorowicz, L.; Skarzewski, J. Synthesis 2011, 708−710.
(21) For a review, see: Xu, L.-W.; Luo, J.; Lu, Y. Chem. Commun.
2009, 1807−1821.
(22) Chai, Z.; Zhao, G. Catal. Sci. Technol. 2012, 2, 29−41.
F
dx.doi.org/10.1021/jo401352z | J. Org. Chem. XXXX, XXX, XXX−XXX