Development of two diastereoselective routes towards trans-4-aminomethyl-piperidin-3-ol building blocks

Article abstract of DOI:10.1016/j.tet.2007.12.055

Two diastereoselective, scaleable routes towards trans-3,4-disubstituted piperidines with a 4-hydroxymethyl-3-hydroxy or 4-aminomethyl-3-hydroxy substitution pattern are being described. In the first route, the 3,4-trans configuration was introduced regio- and diastereoselectively via a hydroboration/oxidation sequence starting from 4-hydroxymethylpyridine. In the second route, regioselective epoxide ring opening of N-benzyl-3,4-epoxy-piperidine was achieved with LiCN, in situ generated from acetocyanohydrin and LiNH₂. The regioselectivity of both the hydroboration and the epoxide ring opening was positively influenced by the presence of the basic piperidine nitrogen. Both routes have been optimized to be performed at large scale.

Full text of DOI:10.1016/j.tet.2007.12.055

Available online at www.sciencedirect.com
Tetrahedron 64 (2008) 2456e2464
www.elsevier.com/locate/tet
Development of two diastereoselective routes towards
trans-4-aminomethyl-piperidin-3-ol building blocks
a
a
a
Harrie J.M. Gijsen a, , Michel J.A. De Cleyn , Christopher J. Love , Michel Surkyn ,
*
Sven F.A. Van Brandt a, Marc G.C. Verdonck a, Luc Moens b,
Jef Cuypers b, Jean-Paul R.M.A. Bosmans a,
*
a Medicinal Chemistry Department, Johnson & Johnson Pharmaceutical Research & Development, Turnhoutseweg 30, 2340 Beerse, Belgium
b Chemical DevelopmentdProcess Research, Johnson & Johnson Pharmaceutical Research & Development, Turnhoutseweg 30, 2340 Beerse, Belgium
Received 28 September 2007; received in revised form 19 December 2007; accepted 20 December 2007
Available online 4 January 2008
Abstract
Two diastereoselective, scaleable routes towards trans-3,4-disubstituted piperidines with a 4-hydroxymethyl-3-hydroxy or 4-aminomethyl-3-
hydroxy substitution pattern are being described. In the first route, the 3,4-trans configuration was introduced regio- and diastereoselectively via
a hydroboration/oxidation sequence starting from 4-hydroxymethylpyridine. In the second route, regioselective epoxide ring opening of
N-benzyl-3,4-epoxy-piperidine was achieved with LiCN, in situ generated from acetocyanohydrin and LiNH2. The regioselectivity of both
the hydroboration and the epoxide ring opening was positively influenced by the presence of the basic piperidine nitrogen. Both routes have
been optimized to be performed at large scale.
Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction
piperidineshavebeensurprisinglylimited,withonlyoneexample
of a 4-aminomethyl substituted piperidine, having a cis rela-
tionship to a 3-methoxy substituent.3 The synthesis of trans-3,4-
disubstituted piperidines with similar functional groups has
been described for 4-amino-piperidin-3-ols,4 but the methodol-
ogies used cannot be easily translated into the synthesis of elon-
gated 4-aminomethyl analogues. In this report we describe our
evolving efforts towards the synthesis of trans-4-amino-
methyl-piperidin-3-ol and trans-4-hydroxymethyl-piperidin-3-
ol containing building blocks 1 and 2, respectively, resulting
in two diastereoselective and scaleable routes towards 1.
In a medicinal chemistry program directed towards 5-HT4
receptor ligands, we have discovered several potent agonists
with excellent gastro-prokinetic activity, which have now pro-
gressed to clinical trials, as exemplified by R149402 and
R199715 (Fig. 1).1 Both compounds contain a trans-4-amino-
methyl-piperidin-3-ol moiety and the large scale preparation
of these clinical candidates or their derivatives required a scale-
able synthetic route towards trans-3,4-disubstituted piperidine
building blocks 1 and 2.
Functionalized piperidines constitute one of the most
common fragments present in biologically active compounds
of both natural and synthetic origin. This has resulted in a wealth
of synthetic methodology for their preparation and incorpora-
tion in more complex structures, which has been extensively
reviewed.2 Reports on the synthesis of 3,4-disubstituted
OH
OH
O
O
Cl
R'
N
H
N
O
N
NH2
P
R
1 R' = NH2
2 R' = OH
P = protecting group
R
R = H, R199715
R = Me, R149402
* Corresponding authors. Tel.: þ32 14 606830; fax: þ32 14 605344.
E-mail addresses: hgijsen@prdbe.jnj.com (H.J.M. Gijsen), jbosmans@
prdbe.jnj.com (J.-P.R.M.A. Bosmans).
Figure 1.
0040-4020/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2007.12.055

H.J.M. Gijsen et al. / Tetrahedron 64 (2008) 2456e2464
2457
2. Results and discussion
hydroboration/oxidation sequence, led to a an isolated yield of
83% of 9. Overreduced 10 was only observed in trace amounts,
according to GCeMS analysis. Due to the cost of the required
TBS-chloride and the extra protection/deprotection steps, this
improvement in yield compared to the reaction of 5 to 6 was
eventually not implemented in the scaled-up process.
The synthesis of N-benzyl protected trans-2 (compound 6,
Scheme 1) has been described via reduction of commercially
available N-benzyl-3-oxo-piperidin-4-carboxylate esters.3,5
Due to the enolic nature of the 3,4-piperidine bond in these
b-keto-esters, invariably a mixture of cis and trans diastereo-
mers is formed with the cis diastereomer as the major product.
Although this route provided access to gram quantities of 2,
the need for larger quantities of both 1 and 2 made it worth-
while to investigate alternative, more diastereoselective routes.
In the subsequent sections two different diastereoselective
routes will be described.
To obtain the corresponding trans-4-aminomethyl-piperi-
din-3-ol building blocks 1, the conversion of the 4-hydroxy-
methyl-group into a 4-aminomethyl-group comprised some
straightforward functional group transformations.3 Transfor-
mation of the primary hydroxyl group in 6 into a suitable leav-
ing group, such as a tosylate or mesylate, resulted in very
unstable compounds, reacting inter- or intramolecularly with
the tertiary piperidine nitrogen atom. Therefore, the benzylic
protection group in 6 was replaced with a tert-butyl carbamate
(Boc) group to give 11 in nearly quantitative yield (Scheme 2).
Now, tosylation of 11 proceeded cleanly to give 12. Subse-
quent displacement of the tosylate with a suitable nitrogen
nucleophile would then lead to the required 4-aminomethyl
building block. Initially benzylamine was used as the nitrogen
nucleophile to give 13 (Scheme 3). This again required a de-
protection step to obtain 14. Similarly, the use of sodium azide
would require an extra reduction step to arrive at 14, and was
not considered as an alternative for benzylamine due to its
hazardous nature. To avoid the last deprotection step, some ex-
periments were run to directly replace the tosylate with ammo-
nia. After some optimization, this proceeded in high yield via
the use of excess ammonia at elevated temperatures in a pres-
sure reactor.
2.1. Diastereoselective route A
The preparation of piperidines via reduction of pyridines is
a commonly used method.2 The quaternization of the pyridine
nitrogen allows for mild reduction conditions, which can be
halted at the 3,4-dehydropiperidine (1,2,3,6-tetrahydropyri-
dine) stage.2,6 The double bond can subsequently be used as
a handle for the introduction of substituents on the 3- and/or
4-position. According to this strategy, we started with the
quaternization of 4-hydroxymethylpyridine 3 with benzyl-
chloride, followed by reduction of 4 to 3,4-dehydropiperidine
5 (Scheme 1).
Introduction of a 3-substituent in a trans diastereoselective
fashion would require a trans anti-Markovnikov addition to
the double bond, a requirement fulfilled via a borane addition,
followed by oxidation with retention.7 Indeed, treatment of 5
with an excess of borane, followed by oxidation gave the desired
trans-3,4-disubstituted piperidine 6 with complete diastereose-
lectivity in 50% (kilogram scale) to 64% (gram scale) yield after
crystallization. According to GCeMS analysis, product 7 was
formed as the main side-product in a w2:1 ratio of 6:7, which
was easily removed during work-up via crystallization.
The presence of a basic piperidine and a free hydroxyl group
required the use of at least 2 equiv of borane for the reaction
to go to completion. This excess of borane could be the cause
of the reduction of 5 to 7. Therefore some borane addition exper-
iments were carried out with a protected hydroxyl group. tert-
Butyl dimethylsilyl (TBS) protection of 5, followed by the
Despite the direct conversion of the tosylate 12 into amino-
methylpiperidine 14, the synthesis of 14 still required a labori-
ous four-step sequence starting from 6. This prompted the
development of an alternative route towards trans-4-amino-
methyl-piperidin-3-ol building block 2.
2.2. Diastereoselective route B
Since an epoxide ring opening also would result in a 3,4-
trans relationship between the resulting hydroxyl group and
the incoming nucleophile, we started to investigate the forma-
tion and opening of 3,4-epoxy-piperidines. Cyanide was cho-
sen as a suitable nucleophile and precursor towards the
OH
i) BH3, THF, -30 °C
ii) H2O2, NaOH
BnCl
NaBH4
MeOH
81%
OH
OH
OH
OH
OH
MeCN, Δ
N+
+
N
N
N
N
95%
Cl–
Bn
Bn
Bn
Bn
3
4
5
6
7
50-64% (isolated)
~20-30% (GCMS)
TBSCl, Et3N
DMAP, CH2Cl2
95%
OH
OTBS
i) BH3, THF, -30 °C
ii) H2O2, NaOH
+
OTBS
OTBS
N
N
N
Bn
Bn
Bn
8
9 83%
10 trace
Scheme 1.

2458
H.J.M. Gijsen et al. / Tetrahedron 64 (2008) 2456e2464
OH
OH
conditions from NaCN in protic solvents to LiCN in THF
i) H2, Pd/C
ii) (Boc)2O
TsCl
would result in the formation of a more covalent lithiume
oxygen bond and possibly suppress the epimerization at the
piperidine 4-position.11 Since LiCN has limited commercial
availability and only as a solution in DMF, it was generated
in situ by the combination of LiH and acetocyanohydrin.12
Addition of the epoxide to the LiCN/THF mixture resulted
in a fast epoxide ring opening without any observed formation
of cis isomers. Unfortunately, still a 4:1 mixture of trans
regioisomers 20 and 21 was observed.
At this point we returned to the benzyl protected piperidines
(Table 1, Scheme 4). The neighbouring group participation of
the piperidine nitrogen with its lone pair in 3,4-epoxy-piperidine
24 could make nucleophilic attack at the 4-position even more
preferred than for epoxide 19.
OH
pyridine
OH
N
N
N
>98%
>98%
Bn
Boc
Boc
11
6
OH
OH
BnNH2
THF, 125 °C
NHR
OTs
N
Boc
13 R = Bn
14 R = H
12
H2, Pd/C
76% (2 steps)
NH3
THF, 125 °C
>98%
Scheme 2.
To investigate the epoxide ring opening in 24, we required
a suitable method for the epoxidation of 17. The use of stan-
dard oxidizing agents such as m-CPBA led to overoxidation
of the piperidine nitrogen. After several attempts a successful
epoxidation method was developed with N-bromosuccinimide
followed by base treatment (Table 1).
mCPBA
i) H2, Pt/C
ii) MsCl, Et3N
iii) KOtBu
or
O
NBS,
aq. NaOH
80%
O
toluene
N
N
P
N
>90%
EtOOC
P
19
15 P = Bn
17 P = Bn
16 P = COOEt
18 P = COOEt
Table 1
Optimization experiments for the epoxidation of 20 to 27
CN
NC
OH
HO
CN
OH
NC
NaCN
HO
Br
aq. EtOH
O
or
Br
+
+
+
N
N
N
N
NC OH
+
Bn
N
N
N
COOEt
COOEt
COOEt
22
COOEt
23
Bn
Bn
LiH, THF
25
20
21
24
17
Scheme 3.
Reaction conditionsa
Product ratio
(%, GC)
desired aminomethyl functionality.8 The syntheses of the 3,4-
dehydropiperidine intermediates 176,9 and 1810 have been
described and are mostly high-yielding two-step sequences
starting from readily available starting materials. In our hands,
access to these compounds was achieved on large scale, start-
ing from the readily available ketones 15 and 16, respectively
(Scheme 3). Catalytic hydrogenation of the ketones in toluene
to the corresponding alcohols, followed by dehydration pro-
vided the 3,4-dehydropiperidines 17 and 18 in high yields.
The low levels of impurities combined with the absence of
a solvent switch made this the favoured reaction pathway,
especially to tetrahydropyridine 17.
The epoxidation of the double bond in 17 was anticipated
to be difficult with the tertiary amino functionality present.
Therefore the first experiments were performed on N-ethoxy-
carbonyl protected epoxide 19, easily accessible from 18.9
Epoxide opening of 19 with NaCN in ethanol/water proceeded
smoothly, and resulted in a 4:1 mixture of regioisomers 20 and
21, with the desired major product 20 arising from nucleo-
philic attack at the less sterically hindered 4-position. In
addition, some epimerization took place, presumably at the
a-nitrilic position, resulting in traces of most likely cis isomers
22 and 23 (not isolated). We reasoned that the observed epime-
rization was most likely due to the formation of a basic alkox-
ide, formed after epoxide ring opening. A change in reaction
i
NBS (1 equiv), aqueous dioxane
NBS (2 equiv), aqueous dioxane
NBS (3 equiv), aqueous dioxane
0
0
ii
iii
iv
v
0
0
53
68
TFA (1 equiv), NBS (1 equiv), aqueous dioxane
TFA (1 equiv), NBS (1 equiv), H2O
<5
>95
>90
0
a
Product ratio determined after treatment with aqueous NaOH.
OH
CN
LiNH2
THF
NC OH
N
LiCN
+
Bn
Bn
26 >95%
90%
+
one-pot from 17
in toluene:
51%
i) TFA, NCS
H2O
CN
O
OH
ii) NaOH
89%
N
N
N
Bn
Bn
24
17
27 <5%
LiCN, then
Me2SO4
OMe
Me2SO4
NaH
CN
26
76%
N
Bn
28
Scheme 4.

H.J.M. Gijsen et al. / Tetrahedron 64 (2008) 2456e2464
2459
OH
Reaction of 17 with one or more equivalents of N-bromosuc-
OH
H2, MeOH/NH3
Raney-Ni
CN
cinimide in aqueous dioxane only led to the formation of in-
creasing amounts of dibromide 25 (Table 1, entries ieiii).
Two subsequent changes in the reaction conditions were found
to be critical for success (Table 1, entries iv and v). Protonation
of the basic piperidine nitrogen with trifluoroacetic acid, prior to
the addition of N-bromosuccinimide, resulted in the formation
of epoxide 24 for the first time, albeit in trace amounts.13
Finally, removal of the organic solvent led to the exclusive for-
mation of epoxide 24, with traces of unreacted alkene 17 as the
only visible impurity. During further optimization for upscaling
purposes, N-bromosuccinimide was replaced by N-chlorosucci-
nimide, which suppressed the dihalogenation reaction even fur-
ther, and thus resulted in an improved reproducibility of the
reaction.
NH2
N
N
81%
R
R
26 R = Bn
29 R = alkyl
31 R = Bn
32 R = alkyl
H2
Pd/C
>98%
alkyl-X
ref 16
30 R = H
33 R = H
(Boc)2O
92%
HCl, iPrOH
98%
OH
OH
H2, MeOH/NH3
Raney-Ni
CN
NH2
N
N
>98%
Boc
Boc
34
14
Subsequent reaction of epoxide 24 with in situ generated
LiCN in THF gave product 26 in 90% yield (Scheme 4).
Only traces of the regioisomer 27 were present according to
GCeMS analysis, thus confirming the positive contribution
of the basic piperidine nitrogen to the regioselectivity of the
epoxide opening.
The use of the flammable and moisture sensitive LiH is not
feasible for large scale applications. For the in situ generation
of LiCN from acetocyanohydrin it has been successfully re-
placed by alkyllithiums,14 but these have their own safety
drawbacks. For production scale purposes, we replaced LiH
with the much more convenient to handle lithium amide. Prior
to addition of the epoxide, the reaction mixture was heated to
expel the residual ammonia.
Scheme 5.
of a large number of potent 5-HT4 ligands, including the large
scale synthesis of R149402 and R199715.15 As mentioned be-
fore, immediate quaternization of pyridine with the ultimately
desired alkyl group may shorten the total synthesis of a final
compound further via intermediates 29 and 32 by avoiding
all protecting group manipulations. Alternatively, fully depro-
tected 33 can be selectively alkylated at the secondary N-
piperidine amino group as described before.16 The most active
compounds in our medicinal chemistry program were found to
have the 3S,4S absolute configuration. Further optimisation is
currently directed towards the introduction of enantioselectiv-
ity in the synthetic route towards 14.17
Most conveniently, on >100 g scale, the epoxide 24 was
not isolated, but the toluene extraction layer, containing 24,
was directly treated with the in situ generated LiCN, to give
after crystallization the desired 26 in >97% purity and 51%
yield from 17.
3. Conclusion
Although most of our final compounds incorporated the
free 3-hydroxyl group, we were also interested in making 3-
alkoxy piperidine analogues. This could be done by deproto-
nation of 26 with sodium hydride followed by reaction of
the alkoxide with dimethylsulfate. Since the lithium 3-alkoxy-
4-cyano piperidine salt is a likely intermediate during the ep-
oxide ring opening of 24 to 26, we tried to combine this ring
opening with a subsequent alkylation in a one-pot procedure.
Indeed, this gave the 3-methoxy piperidine 28 as the main
product, although always 10e30% of non-alkylated 26 re-
mained present.
Product 26 was subjected to hydrogenolysis (Scheme 5).
Depending on the conditions and the catalyst used, either the
benzyl group could be selectively removed to give 30, or the
nitrile group could be reduced to compound 31. N-Boc protec-
tion of 30, followed by extensive catalytic hydrogenation led
to 14 in excellent yield, and this was found to be the most con-
venient route towards this intermediate.
In order to prepare large quantities of trans-4-hydroxy-
methyl- and trans-4-aminomethyl-piperidin-3-ol containing
building blocks, we have developed two diastereoselective
and scaleable routes. Starting from N-benzyl 1,2,3,6-tetrahy-
dropyridines, the double bond is used as a handle for the intro-
duction of the trans-3,4-substituents via, respectively,
a hydroboration/oxidation sequence (route A, Scheme 1) and
a regioselective epoxide ring opening (route B, Schemes 4
and 5). The regioselectivity of both the hydroboration and
the epoxide ring opening was positively influenced by the
presence of the basic piperidine nitrogen. In route B new con-
ditions have been developed for the epoxidation of a double
bond in the presence of a basic piperidine nitrogen by proton-
ation of the amino group and using water as the only solvent.
An improved, scaleable method for the in situ formation of
LiCN has been developed via the reaction of acetocyanohydrin
with LiNH2.
Both routes A and B have been performed at kilogram scale
and are optimal for access to either trans-4-hydroxymethyl- or
trans-4-aminomethyl piperidin-3-ols. Depending on the
amounts required of these valuable building blocks, or their
precursors, these routes can find application for both small
and large scale purposes.
The second diastereoselective route described in this paper
has now been carried out on kilogram quantities. Starting from
building block 15, trans-4-aminomethyl-piperidin-3-ol 14 has
been prepared in a seven-step sequence in 60% overall yield.
Building block 14 has been used successfully in the synthesis

2460
H.J.M. Gijsen et al. / Tetrahedron 64 (2008) 2456e2464
4. Experimental
4.1. General
ꢀ30 ꢁC. After the addition, the reaction mixture was stirred
for 4 h, allowed to warm up to room temperature and stirred
at room temperature for 18 h. The reaction mixture was cooled
to ꢀ10 ꢁC and water (25 mL) was added dropwise. Then, si-
multaneously, NaOH (3 M in water, 70 mL) and the hydrogen
peroxide (30% solution in water, 63.3 mL) were added drop-
wise while the reaction mixture was kept at a temperature of
ꢀ10 ꢁC. Again NaOH (50% in water, 140 mL) was added.
The reaction mixture was stirred at reflux for 4 h and then
cooled and filtered. The filtrate was concentrated under re-
duced pressure. The resulting precipitate was dissolved in
water (500 mL) and saturated with K2CO3. The product was
extracted with CH2Cl2. The resulting solution was dried over
MgSO4 and evaporated. The residue, containing predomi-
nantly a mixture of 6 and 7 in a 2:1 ratio according to
GCeMS, was crystallized from diisopropyl ether/MeCN
Silica gel column chromatography was performed with
Kiesel gel 60 (0.063e0.200 mm) (E. Merck, AG Darmstadt,
Germany). 1H NMR spectra were recorded on 360 and
400 MHz spectrometers (Bruker) and chemical shifts (d) are
expressed in parts per million (ppm) with TMS as internal
standard. Elemental analysis were carried out on a CarloErba
EA1110. Melting points were determined on a Mettler, a Buchi
545, and a Kofler apparatus and are uncorrected. Mass spectral
¨
data were obtained from by GCeMS analysis, using an Agi-
lent 6890 series GC, combined with a 5973N Mass Selective
Detector. Commercial solvents were used without any pre-
treatment, including anhydrous solvents where required.
¨
1
to yield 56 g (50%) of 6 as a white solid, mp 113 ꢁC. H
4.2. 1-(Phenylmethyl)-4-pyridinemethanol chloride (4)
NMR (360 MHz, CDCl3) d ppm 1.26 (1H, qd, J¼13.0,
3.7 Hz), 1.47e1.58 (2H, m), 1.85 (1H, t, J¼10.3 Hz), 1.96
(1H, td, J¼11.7, 2.8 Hz), 2.81 (1H, br d, J¼11.4 Hz), 2.9
(1H, br s), 2.97 (1H, ddd, J¼10.7, 4.4, 1.5 Hz), 3.3 (1H, br
s), 3.48 (1H, d, J¼12.9 Hz), 3.54 (1H, d, J¼12.9 Hz), 3.59e
3.75 (3H, m), 7.22e7.33 (5H, m). Anal. Calcd for
C13H19NO2: C, 70.56; H, 8.65; N, 6.33. Found: C, 70.44; H,
8.83; N, 6.33.
A solution of 4-pyridinemethanol 3 (200 g, 1.84 mol) in
MeCN (1 L) was added to a solution of benzylchloride
(279 g, 2.2 mol) in MeCN (1 L) and the reaction mixture
was refluxed for 3 h, cooled to room temperature and concen-
trated under reduced pressure. The residue was suspended in
diethyl ether, filtered and dried, yielding crude 4 (411 g,
1
97%). H NMR (360 MHz, DMSO-d6) d ppm 4.79 (2H, s),
5.86 (2H, s), 7.39e7.48 (3H, m), 7.51e7.59 (2H, m), 8.06
(2H, d, J¼6.7 Hz), 9.17 (2H, d, J¼6.7 Hz).
4.5. 1-(Phenylmethyl)-4-(tert-butyl-dimethyl-silanyl-
oxymethyl)-1,2,3,6-tetrahydropyridine (8)
4.3. 1-(Phenylmethyl)-4-hydroxymethyl-1,2,3,6-
tetrahydro-4-pyridine (5)
To a solution of 5 (10 g, 0.05 mol) in CH2Cl2 (250 mL) were
added Et3N (16 mL, 0.11 mol), and a catalytic amount of
DMAP. A solution of tert-butyl dimethylsilyl chloride (8.3 g,
0.055 mol) in CH2Cl2 (125 mL) was added dropwise and the re-
action mixture was stirred at room temperature for 20 h. The
mixture was washed with a 1 M aqueous K2CO3 solution and
the organic layer was dried on MgSO4, filtered and concentrated
under reduced pressure. The residue was purified by silica gel
column chromatography (elution CH2Cl2/MeOH(NH3) 95:5),
yielding 15 g (95%) of 8 as an oil. 1H NMR (360 MHz,
CDCl3) d ppm 0.05 (6H, s), 0.90 (9H, s), 2.06e2.13 (2H, m),
2.58 (2H, t, J¼5.9 Hz), 2.96e3.01 (2H, m), 3.58 (2H, s),
4.01e4.05 (2H, m), 5.59e5.64 (1H, m), 7.22e7.37 (5H, m).
MS (EIþ) m/z (rel intensity) 317 (8%, Mþ), 316 (8%), 302
(4%), 260 (5%), 185 (11%), 172 (100%), 91 (43%).
A solution of 4 (174 g, 0.87 mol) in methanol (2.2 L) was
cooled to ꢀ20 ꢁC. Sodium borohydride (66 g, 1.75 mol) was
added portionwise under a nitrogen atmosphere. The reaction
mixture was stirred for 30 min and water (200 mL) was added
dropwise. The reaction mixture was partially concentrated
under reduced pressure, water was added and the reaction
mixture was extracted with CH2Cl2. The organic layer was
separated, dried on MgSO4, filtered and concentrated under re-
duced pressure. The residue was purified by silica gel column
chromatography (elution CH2Cl2), yielding 155 g (87%) of 5.
This could be used as such in the next step. Crystallization
from diisopropyl ether afforded pure 5 as a light brown solid,
1
mp 64 ꢁC. H NMR (360 MHz, DMSO-d6) d ppm 1.96e2.03
(2H, m), 2.49 (2H, t, J¼5.9 Hz), 2.83e2.88 (2H, m), 3.51 (2H,
s), 3.79 (2H, s), 4.67 (1H, br s), 5.50e5.54 (1H, m), 7.21e7.27
(1H, m), 7.27e7.34 (4H, m). Anal. Calcd for C13H17NO: C,
76.81; H, 8.43; N, 6.89. Found: C, 76.56; H, 8.20; N, 6.83.
4.6. (ꢂ)-1-(Phenylmethyl)-(trans)-4-(tert-butyl-dimethyl-
silanyloxymethyl)-piperidin-3-ol (9)
To a solution of 8 (11 g, 0.036 mol) in THF (300 mL),
cooled to ꢀ30 ꢁC, was added dropwise a 1 M solution of bo-
rane in THF (75 mL, 0.075 mol) under a nitrogen atmosphere
while the reaction mixture was kept at a temperature between
ꢀ20 and ꢀ30 ꢁC. After the addition, the reaction mixture was
stirred for 3 h, allowed to warm up to room temperature and
stirred at room temperature for 20 h. The reaction mixture
was cooled to ꢀ10 ꢁC and water (1.2 mL) was added dropwise.
Then, simultaneously, NaOH (3 M in water, 5 ml) and
4.4. (ꢂ)-1-(Phenylmethyl)-(trans)-4-hydroxymethyl-
piperidin-3-ol (6)
A solution of 5 (102 g, 0.5 mol) in THF (1 L) was cooled to
ꢀ30 ꢁC and was added dropwise under a nitrogen atmosphere
to a 1 M solution of borane in THF (1 L, 1 mol) while the re-
action mixture was kept at a temperature between ꢀ20 and

H.J.M. Gijsen et al. / Tetrahedron 64 (2008) 2456e2464
2461
hydrogen peroxide (30% solution in water, 4.6 ml) were added
dropwise while the reaction mixture was kept at a temperature
of ꢀ10 ꢁC. Again NaOH (50% in water, 9 ml) was added, fol-
lowed by water (30 mL). The reaction mixture was stirred at re-
flux for 3 h and then cooled. The organic layer was separated,
dried on MgSO4, and filtered. The filtrate was concentrated un-
der reduced pressure and the residue was purified by silica gel
column chromatography (elution CH2Cl2/MeOH(NH3) 95:5),
yielding 10 g (83%) of 9 as an oil. 1H NMR (400 MHz,
CDCl3) d ppm 0.05 (3H, s), 0.06 (3H, s), 0.87 (9H, s), 1.18
(1H, qd, J¼12.4, 4.0 Hz), 1.43e1.58 (2H, m), 1.82 (1H, t,
J¼10.2 Hz), 1.92 (1H, td, J¼11.6, 2.7 Hz), 2.77 (1H, br d,
J¼11.3 Hz), 3.01 (1H, ddd, J¼10.6, 4.6, 1.6 Hz), 3.50 (2H,
s), 3.57e3.68 (2H, m), 3.72 (1H, dd, J¼9.9, 4.0 Hz), 4.03
(1H, s), 7.17e7.24 (1H, m), 7.24e7.30 (4H, m).
was dried on MgSO4 and concentrated under reduced pressure.
Toluene was added to the residue, and the mixture was again
concentrated under reduced pressure to give 12 as an oil,
which could be used as such in the next step. Crystallization
from diisopropyl ether and a small amount of MeCN gave
1
a white solid, mp 111 ꢁC. H NMR (360 MHz, DMSO-d6)
d ppm 1.08 (1H, qd, J¼13.0, 4.6 Hz), 1.37 (9H, s), 1.53e
1.63 (2H, m), 2.36 (1H, br s), 2.42 (3H, s), 2.57 (1H, br s),
3.02e3.13 (1H, m), 3.83 (1H, br s), 3.99 (1H, dd, J¼9.5,
7.0 Hz), 3.96 (1H, br s), 4.14 (1H, dd, J¼9.3, 2.7 Hz), 5.11
(1H, d, J¼5.1 Hz), 7.48 (2H, d, J¼8.0 Hz), 7.78 (2H, d,
J¼8.0 Hz). Anal. Calcd for C11H21NO4: C, 56.09; H, 7.06;
N, 3.63. Found: C, 56.27; H, 7.18; N, 3.50.
4.9. (ꢂ)-1,1-Dimethylethyl trans-4-(aminomethyl)-3-hydroxy-
1-piperidinecarboxylate (14)
4.7. (ꢂ)-1,1-Dimethylethyl trans-3-hydroxy-4-
(hydroxymethyl)-1-piperidinecarboxylate (11)
4.9.1. Two-step procedure from 12
A mixture of 12 (0.22 mol) and benzylamine (0.084 mol) in
THF (100 mL) was stirred for 16 h at 125 ꢁC (autoclave). The
reaction mixture was cooled and the solvent was evaporated
under reduced pressure. The residue was partitioned between
CH2Cl2 and an aqueous K2CO3 solution. The organic layer
was separated, dried on MgSO4, filtered and concentrated un-
der reduced pressure, yielding 15.4 g of 1,1-dimethylethyl
(trans)-3-hydroxy-4-[[(phenylmethyl)amino]methyl]-1-piperi-
dinecarboxylate 13 together with remaining benzylamine. In
order to obtain a pure sample of 13, after a similar reaction
the excess benzylamine was removed via high vacuum distil-
lation, followed by crystallization of the residue in diisopropyl
A solution of 6 (486 g, 2.2 mol) in methanol (3 L) was hydro-
genated under a hydrogen atmosphere, at 50 ꢁC, with palladium
on activated carbon (10%) (25 g) as a catalyst. After uptake of
1 equivofH2, the catalyst was filtered off and thefiltratewasevap-
orated, giving 290 g of (ꢂ)-trans-4-hydroxymethyl-piperidin-
3-ol (quantitative yield), which was used in the next reaction
step,withoutfurtherpurification.Crystallizationfromdiisopropyl
ether afforded pure (ꢂ)-trans-4-hydroxymethyl-piperidin-3-ol as
awhite solid, mp104 ꢁC. 1HNMR(360 MHz,CDCl3)dppm1.12
(1H, qd, J¼12.3, 4.3 Hz), 1.55e1.71 (2H, m), 2.44 (1H, dd,
J¼11.6, 10.1 Hz), 2.58 (1H, dt, J¼12.2, 2.5 Hz), 3.0 (1H, br d,
J¼12.2 Hz), 3.18 (1H, dd, J¼11.6, 4.7 Hz), 3.56 (1H, dt, J¼10,
4.7 Hz), 3.68 (1H, dd, J¼10.5, 8.6 Hz), 3.74 (1H, dd, J¼10.5,
4.0 Hz). Anal. Calcd for C6H13NO2: C, 54.94; H, 9.99; N,
10.68. Found: C, 54.71; H, 10.25; N, 10.57.
A solution of (Boc)2O (464 g, 2.13 mol) in CH2Cl2 (0.5 L)
was added dropwise to a solution of the crude intermediate
(280 g, 2.13 mol) in CH2Cl2 (3 L) at room temperature, while
cooling on a waterbath. After the addition, methanol (1 L) was
added and the resulting reaction solution was stirred for 1 h at
room temperature. The solvent was evaporated to yield 492 g
(quantitative yield) of 11. This could be crystallized from diiso-
propyl ether, mp 88 ꢁC. 1H NMR (360 MHz, CDCl3) d ppm 1.16
(1H, qd, J¼12.7, 4.4 Hz), 1.44 (9H, s), 1.52e1.71 (2H, m),
2.43e2.58 (1H, m), 2.59e2.77 (1H, m), 3.45e3.56 (1H, m),
3.66 (1H, dd, J¼10.3, 8.4 Hz), 3.74 (1H, br d, J¼10.3 Hz),
3.89e4.33 (2H, m). Anal. Calcd for C11H21NO4: C, 57.12; H,
9.15; N, 6.06. Found: C, 57.14; H, 9.21; N, 6.01.
1
ether, to give pure 13 as a white solid, mp 91 ꢁC. H NMR
(360 MHz, CDCl3) d ppm 1.11 (1H, qd, J¼13.2, 4.4 Hz),
1.45 (9H, s), 1.46e1.58 (2H, m), 2.49 (1H, br t, J¼11.3 Hz),
2.61 (1H, t, J¼11.5 Hz), 2.56e2.7 (1H, m), 2.92 (1H, dd,
J¼12.1, 2.6 Hz), 3.47 (1H, td, J¼9.7, 5.1 Hz), 3.75 (1H, d,
J¼13.0 Hz), 3.87 (1H, d, J¼13.0 Hz), 4.06e4.31 (2H, m),
7.25e7.37 (5H, m).
The crude 13 was dissolved in methanol (100 mL) and hydro-
genated with palladium on carbon (10%, 1 g) as a catalyst at
50 ꢁC under 1 atm hydrogen pressure. After uptake of 1 equiv
of H2, the catalyst was filtered off and the filtrate was concen-
trated under reduced pressure. The residue was crystallized
from diisopropyl etherþMeCN, filtered off and dried (vacuum,
40 ꢁC), yielding 4 g (76% over two steps) of 14 as a white solid,
mp 178 ꢁC. 1H NMR (360 MHz, DMSO-d6) d ppm 0.98 (1H, qd,
J¼12.6, 4.6 Hz), 1.19e1.32 (1H, m), 1.38 (9H, s), 1.59 (1H, dq,
J¼13.2, 2.8 Hz), 2.25e2.5 (1H, m), 2.53e2.67 (1H, m), 2.59
(1H, dd, J¼12.4, 5.5 Hz), 2.67 (1H, dd, J¼12.4, 6.6 Hz), 3.15
(1H, td, J¼9.9, 4.8 Hz), 3.79e4.00 (2H, m). Anal. Calcd for
C11H22N2O3: C, 57.37; H, 9.63; N, 12.16. Found: C, 57.61; H,
9.80; N, 11.92.
4.8. (ꢂ)-1,1-Dimethylethyl trans-3-hydroxy-4-[[(4-methyl-
phenyl)sulfonyl]oxymethyl]-1-piperidinecarboxylate (12)
To a cooled (ꢀ10 ꢁC) solution of 11 (117 g, 0.51 mol) in
pyridine (200 mL) was added tosyl chloride (110 g,
0.54 mol) portionwise. The reaction mixture was stirred at
room temperature for 1 h. Then CH2Cl2 (1.5 L) was added
and the mixture was washed with 5% aqueous HCl solution,
followed by an aqueous NaHCO3 solution. The organic layer
4.9.2. One-step procedure from 12
A solution of 12 (35 g, 0.09 mol) in THF (250 mL) was
treated with liquid NH3 in an autoclave at 125 ꢁC during
16 h. The reaction mixture was filtered and the filtrate was
concentrated under reduced pressure. The residue was

2462
H.J.M. Gijsen et al. / Tetrahedron 64 (2008) 2456e2464
partitioned between a 5% aqueous NaOH solution and
CH2Cl2. The organic layer was separated, dried on MgSO4, fil-
tered and the solvent was concentrated under reduced pres-
sure, yielding 14 (16 g, quantitative yield).
4.13. Ethyl (ꢂ)-(trans)-4-cyano-3-hydroxy-
piperidinecarboxylate (20) and ethyl (ꢂ)-(trans)-
3-cyano-4-hydroxy-piperidinecarboxylate (21)
To a cooled (0e5 ꢁC) suspension of LiH (0.72 g, 0.09 mol)
in THF (75 mL) was added a solution of acetocyanohydrin
(8.2 mL, 0.09 mol) in THF (15 mL) under a N2 atmosphere.
The reaction mixture was stirred for 2 h at room temperature,
and then a solution of 19 (12.8 g, 0.075 mol) in THF (30 mL)
was added dropwise. The reaction mixture was subsequently
refluxed for 1 h. After cooling to room temperature, water
(200 mL) was added, and the mixture was extracted with
CH2Cl2 (3ꢃ100 mL). The combined organic layers were dried
on MgSO4, filtered, and concentrated under reduced pressure.
The residue was purified by silica gel column chromatography
(elution CH2Cl2/MeOH(NH3) 96:4), yielding 12 g (80%) of
4.10. 1-(Phenylmethyl)-1,2,3,6-tetrahydropyridine (17)
A solution of 15 (70 g, 0.37 mol) in toluene (300 mL) was
hydrogenated overnight at 40 ꢁC with 5% Pd/C (8.4 g) as
a catalyst. After GC analysis showed the complete conver-
sion of 15 to the corresponding alcohol, the reaction mixture
was filtered over dicalite and the dicalite washed with tolu-
ene (100 mL). To this filtrate, a solution of triethylamine
(41 g, 0.405 mol) in toluene (400 mL) was added. The result-
ing mixture was cooled to 5 ꢁC and mesylchloride (46.5 g,
0.405 mol) was added dropwise over a 20 min period, during
which time the temperature increased to 25 ꢁC and a precipi-
tate was formed. The mixture was stirred for an additional
1.5 h at room temperature, and then washed with water
(2ꢃ200 mL). The organic layer was partially concentrated
under reduced pressure to remove 200 mL of solvent. The re-
maining solution containing the mesylate was diluted with
toluene (600 mL) and DMA (400 mL) and KOtBu (54 g,
0.48 mol) was added. The reaction mixture was stirred for
1.5 h at room temperature, and subsequently washed with
water (400 mL) and concentrated under reduced pressure to
give 63 g of crude 17. The residue was distilled (75 ꢁC at
0.15 mbar) to give 17 as an oil (41 g, 64% from 15). The ana-
lytical data were in agreement with those reported in the
literature.9
1
a 4:1 mixture of 20 and regioisomer 21. Compound 20 H
NMR (360 MHz, DMSO-d6) d ppm 1.17 (3H, t, J¼7.0 Hz),
1.54e1.66 (1H, dddd, J¼13.2, 11.5, 11.5, 4.4 Hz), 1.99 (1H,
dq, J¼13.2, 3.4 Hz), 2.65 (1H, br s), 2.72 (1H, ddd, J¼11.3,
9.5, 4.0 Hz), 2.82 (1H, br t, J¼10.2 Hz), 3.46e3.56 (1H, m),
3.80 (1H, br d, J¼13.5 Hz), 3.94 (1H, br s), 4.02 (2H, q,
J¼7.0 Hz), 5.79 (1H, d, J¼5.5 Hz). Compound 21 (identifiable
1
peaks) H NMR (360 MHz, DMSO-d6) d ppm 1.18 (3H, t,
J¼7.0 Hz), 1.27e1.41 (1H, m), 1.74e1.86 (1H, m), 3.14
(1H, ddd, J¼13.5, 9.5, 3.3 Hz), 3.34 (1H, br s), 3.64 (1H, br
d, J¼13.2 Hz), 5.58 (1H, d, J¼5.1 Hz).
4.14. 3-(Phenylmethyl)-7-oxa-3-aza-bicyclo-
[4.1.0]heptane (24)
A mixture of trifluoroacetic acid (89 mL, 1.15 mol) in water
(2 L) was stirred at room temperature. Compound 17 (200 g,
1.15 mol) was added dropwise to the mixture and the resulting
suspension was stirred at room temperature for 15 min. N-Bro-
mosuccinimide (250 g, 1.4 mol) was added portionwise over
1 h, during which time the temperature increased to 30e
35 ꢁC. The reaction mixture was stirred for an additional
30 min. Again N-bromosuccinimide (15 g, 0.085 mol) was
added portionwise (temperature to 35 ꢁC). The reaction mixture
was stirred overnight at room temperature and then decanted and
added dropwise to a 20% aqueous NaOH solution (2 L). The
mixture was stirred overnight at room temperature. The product
was extracted with CH2Cl2 (3ꢃ600 mL). The combined organic
layers were dried on MgSO4, filtered, and concentrated under re-
duced pressure to yield crude 24 (193 g, 89%), which could be
used as such in the next step, or purified further via distillation
under reduced pressure (bp 60 ꢁC at 0.1 mmHg). The analytical
data were consistent with those reported in the literature.13 MS
(EIþ) m/z (rel intensity) 189 (34%, Mþ), 133 (33%), 91 (100%).
4.11. Ethyl 1,2,3,6-tetrahydropyridinecarboxylate (18)
Compound 18 was prepared from 16 using the same proce-
dure as described for 17, with analytical data consistent with
those reported in the literature.10
4.12. Ethyl 7-oxa-3-azabicyclo[4.1.0]heptane-3-
carboxylate (19)
To a solution of 18 (217 g, 1.39 mol) in water (1 L) and di-
oxane (1 L) was added N-bromosuccinimide (261 g, 1.47 mol)
portionwise at room temperature. During the addition, the
temperature slowly increased to 35 ꢁC. After addition, the re-
action was stirred 20 h at room temperature and subsequently
extracted with CH2Cl2 (3ꢃ500 mL). The combined organic
layers were washed with brine, dried on MgSO4 and concen-
trated under reduced pressure. The residue was dissolved in
THF (700 mL), and a 5% aqueous NaOH solution (3,5 L)
was added. The reaction mixture was stirred for 1 h at room
temperature, and the product extracted with CH2Cl2
(3ꢃ300 mL). The combined organic layers were dried on
MgSO4 and concentrated under reduced pressure. The residue
was distilled under reduced pressure (bp 65 ꢁC at 0.1 mmHg)
to yield 23 (189 g, 80%). The analytical data were consistent
with those reported in the literature.9,10
4.15. (ꢂ)-1-(Phenylmethyl)-(trans)-4-cyano-
piperidin-3-ol (26)
To a suspension of LiNH2 (15.1 g, 0.66 mol) in THF
(600 mL) was added a solution of acetocyanohydrin (60 mL,
0.66 mol) in THF (150 mL) at 40 ꢁC under a N2 atmosphere.

H.J.M. Gijsen et al. / Tetrahedron 64 (2008) 2456e2464
2463
The reaction mixture was stirred for 0.5 h at 40 ꢁC, cooled to
room temperature, and then a solution of 24 (114 g, 0.6 mol)
in THF (250 mL) was added dropwise. After complete addition,
the reaction mixture was stirred and refluxed for 4 h, then stirred
overnight at room temperature. Water was added (2 L) and the
mixture was extracted with CH2Cl2 (2ꢃ1 L). The combined or-
ganic layers were dried on MgSO4, filtered, and concentrated
under reduced pressure, yielding 26 (128 g, 90%), which could
be used as such in the next step, or purified further via crystalli-
zation from diisopropyl ether to give a white solid, mp 85 ꢁC. 1H
NMR (360 MHz, DMSO-d6) d ppm 7.23e7.35 (5H, m), 5.51
(1H, d, J¼6.0 Hz), 3.52 (1H, d, J¼13.3 Hz), 3.48e3.57 (1H,
m), 3.42 (1H, d, J¼13.2 Hz), 2.86 (1H, ddd, J¼10.9, 4.4,
1.6 Hz), 2.72 (1H, d, J¼11.5 Hz), 2.44 (1H, ddd, J¼12.2,
10.0, 4.1 Hz), 1.98 (1H, dq, J¼12.8, 3.3 Hz), 1.88 (1H, td,
J¼11.6, 2.6 Hz), 1.71 (1H, dd, J¼11.0, 9.7 Hz), 1.64 (1H, qd,
J¼12.5, 4.1 Hz). Anal. Calcd for C13H16N2O: C, 72.22; H,
7.74; N, 12.82. Found: C, 72.19; H, 7.46; N, 12.45.
at room temperature under a N2 atmosphere. The mixture was
stirred for 30 min at room temperature. A solution of dime-
thylsulfate (41.7 mL, 0.44 mol) in THF (650 mL) was added
dropwise to the mixture. The reaction mixture was refluxed
for 3 h under a N2 atmosphere, and then allowed to stand over-
night at room temperature. Water (2 L) was added, and the
mixture was extracted with CH2Cl2 (2ꢃ1 L). The combined
organic layers were dried on MgSO4, filtered, and concen-
trated under reduced pressure. The residue was crystallized
as from isopropanol/MeCN and oxalic acid, yielding the oxa-
late of 28 (97 g, 76%), mp 152 ꢁC. CHN calcd/found C 54.99/
54.45, H 6.29/6.30, N 8.74/8.50. The oxalate salt was con-
verted into the free base by adding water and Na2CO3 fol-
lowed by extraction of the aqueous layer with CH2Cl2. The
organic layers were dried on MgSO4, filtered, and concen-
trated under reduced pressure. The residue was co-evaporated
1
with toluene (2ꢃ) to yield 28 (70 g, 76%) as an oil. H NMR
(360 MHz, CDCl3) d ppm 1.77e2.09 (4H, m), 2.42 (1H, ddd,
J¼11.3, 9.2, 4.4 Hz), 2.71e2.81 (1H, m), 3.10 (1H, dd,
J¼10.8, 3.5 Hz), 3.45 (3H, s), 3.44 (1H, dt, J¼9.1, 4.2 Hz),
3.51 (1H, d, J¼13.2 Hz), 3.57 (1H, d, J¼13.2 Hz), 7.24e
7.35 (5H, m).
4.15.1. One-pot procedure from 17
To a stirred suspension of 17 (132 g, 0.76 mol) in water
(950 mL) was added trifluoroacidic acid (87 g, 0.76 mol) drop-
wise over a 15 min period. To the resulting clear solution, N-
chlorosuccinimide (122.5 g, 0.92 mol) was added portionwise,
and the reaction mixture was heated to 40 ꢁC overnight. Sub-
sequently, the reaction mixture was cooled to room tempera-
ture and a 50% aqueous NaOH solution (245 g, 3 mol) was
added dropwise over 15 min, while cooling on a waterbath.
During the addition, the temperature rose to 40 ꢁC, and the re-
action mixture was subsequently stirred at this temperature for
2 h. Then, the reaction mixture was extracted with toluene
(1.15 L), and the organic layer was partially concentrated
under reduced pressure to remove 400 mL of solvent to give
a remaining solution of epoxide 24 in toluene.
A suspension of LiNH2 (21.1 g, 0.92 mol) in THF
(750 mL) was heated to 40 ꢁC for 15 min. Then acetocyanohy-
drin (78.1 g, 0.92 mol) was added over 15 min and the result-
ing suspension was stirred at 40 ꢁC for an additional 25 min.
To this suspension the solution of 24 in toluene was added, fol-
lowed by an additional 375 mL of toluene, and the mixture
was heated at reflux for 1.5 h. The reaction mixture was cooled
to room temperature and washed with two 375-mL portions of
water. The organic layer was concentrated under reduced pres-
sure to give crude 26 (137 g). Part of this residue (117 g) was
heated to 50 ꢁC and diisopropyl ether (233 mL) was added
dropwise over 15 min. Then the homogeneous mixture was
heated at reflux for 30 min, cooled to 40 ꢁC and the solution
was seeded with crystalline 26. The mixture was cooled fur-
ther to room temperature overnight, and the solid product col-
lected via filtration to afford 26 as a beige solid (72 g, 51%
overall yield from 17).
4.16.1. One-pot procedure from 24
To a suspension of LiH (0.19 g, 0.024 mol) in THF
(100 mL) was added
a solution of acetocyanohydrin
(2.2 mL, 0.024 mol) in THF (6 mL) under a N2 atmosphere.
The reaction mixture was stirred for 2 h at room temperature,
and then concentrated under reduced pressure to afford LiCN
as a white solid, which was immediately dissolved in THF
(100 mL). To this LiCN solution was added a solution of 24
(3.8 g, 0.02 mol) in THF (5 mL) dropwise under a N2 atmo-
sphere. Subsequently, the reaction mixture was refluxed for
2.5 h, and then allowed to cool to 40 ꢁC. To this mixture a so-
lution of dimethylsulfate (2.1 mL, 0.022 mol) in THF (5 mL)
was added dropwise, and the reaction mixture was refluxed
for 3 h under a N2 atmosphere. Aqueous work-up as above
provided a mixture of 28 and 26 in a 9:1 ratio plus traces of
remaining epoxide 24.
4.17. (ꢂ)-trans-4-Cyano-piperidin-3-ol (30)
A solution of 26 (59 g, 27 mol) in methanol (500 mL) was
hydrogenated at 40 ꢁC under a starting pressure of 7 atm hy-
drogen with 10% Pd/C (5 g) as a catalyst. After the reaction
had proceeded for about 60% according to LCeMS analysis,
further uptake of hydrogen stopped, and the reaction mixture
was filtered over dicalite and concentrated under reduced pres-
sure. Then the residue was again subjected to the same hydro-
genation conditions, until uptake of 1 equiv of hydrogen in
total. Again, the reaction mixture was filtered over dicalite
and concentrated under reduced pressure. After co-evaporation
with toluene, 35 g (quantitative yield) of 30 was obtained as an
oil, which was used as such in the next reaction step. 1H NMR
(360 MHz, DMSO-d6) d ppm 1.51 (1H, qd, J¼12.1, 4.2 Hz),
1.92 (1H, dq, J¼12.8, 3.2 Hz), 2.16 (1H, dd, J¼12.1,
9.5 Hz), 2.30 (1H, td, J¼12.2, 2.7 Hz), 2.46e2.54 (1H, m),
4.16. (ꢂ)-1-(Phenylmethyl)-(trans)-4-cyano-3-methoxy-
piperidine (28)
To a solution of 26 (85.6 g, 0.4 mol) in THF (1.4 L) was
added NaH (60% in mineral oil, 17.6 g, 0.44 mol) portionwise

2464
H.J.M. Gijsen et al. / Tetrahedron 64 (2008) 2456e2464
2.77 (1H, dt, J¼12.5, 3.4 Hz), 2.94 (1H, dd, J¼12.1, 4.4 Hz),
3.38 (1H, td, J¼9.7, 4.4 Hz). MS (EIþ) m/z (rel intensity) 126
(51%, Mþ), 97 (11%), 72 (14%), 57 (100%).
2.70 (1H, ddd, J¼11.5, 9.5, 3.8 Hz), 2.75 (1H, m), 3.46 (1H,
tt, J¼9.6, 5.0 Hz), 3.77 (1H, d, J¼13.5 Hz), 3.91 (1H, br s),
5.79 (1H, d, J¼5.5 Hz).
4.18. (ꢂ)-1-(Phenylmethyl)-(trans)-4-(aminomethyl)-
piperidin-3-ol (31)
4.21. (ꢂ)-1,1-Dimethylethyl trans-4-(aminomethyl)-3-
hydroxy-1-piperidinecarboxylate (14)
Compound 26 (130 g, 0.6 mol) in a 7 N NH3/MeOH solu-
tion (1.5 l) was hydrogenated at 14 ꢁC with Raney nickel as
a catalyst (20 g). After uptake of 2 equiv of hydrogen, the cat-
alyst was filtered off and the solvent was evaporated. The res-
idue was dissolved in MeCN and filtered to remove remaining
solids. The filtrate was concentrated again under reduced pres-
sure and the residue was treated with diisopropyl ether to af-
ford crystalline 31 (107 g, 81%), mp 74 ꢁC. 1H NMR
(360 MHz, CDCl3) d ppm 1.21 (1H, qd, J¼12.4, 4.0 Hz),
1.25e1.39 (1H, m), 1.51 (1H, dq, J¼12.7, 2.8 Hz), 1.85
(1H, t, J¼10.1 Hz), 1.95 (1H, td, J¼11.4, 2.7 Hz), 2.68 (1H,
dd, J¼12.1, 10.6 Hz), 2.82 (1H, br d, J¼11.3 Hz), 2.96 (2H,
br s), 3.02e3.09 (2H, m), 3.54 (2H, s), 3.68 (1H, td, J¼9.3,
4.4 Hz), 7.20e7.28 (1H, m), 7.28e7.34 (4H, m). Anal. Calcd
for C13H20N2O: C, 70.87; H, 9.15; N, 12.72. Found: C, 70.53;
H, 9.05; N, 12.46.
Nitrile 34 could be reduced via catalytic hydrogenation to
14 using the same procedure as described for 31 in quantitative
yield. The analytical data were identical to the previously pre-
pared 14 (vide supra).
References and notes
1. De Maeyer, J. H.; Prins, N. H.; Schuurkes, J. A. J.; Levebvre, R. A.
J. Pharmacol. Exp. Ther. 2006, 317, 955e964.
2. (a) Buffat, M. G. P. Tetrahedron 2004, 60, 1701e1729; (b) Weintraub,
P. M.; Sabol, J. S.; Kane, J. M.; Borcherding, D. R. Tetrahedron 2003,
59, 2953e2989 and references cited in these reviews.
3. Sonda, S.; Kawahara, T.; Ktayama, K.; Sato, N.; Asano, K. Bioorg. Med.
Chem. 2005, 13, 3295e3308.
4. For example: (a) Efange, S. M. N.; Khare, A. B.; Foulon, C.; Akella, S. K.;
Parsons, S. M. J. Med. Chem. 1994, 37, 2574e2582; (b) Naito, T.;
Nakagawa, K.; Nakamura, T.; Kasei, A.; Ninomiya, I.; Kiguchi, T.
J. Org. Chem. 1999, 64, 2003e2009; (c) Langlois, N.; Calvez, O. Synth.
Commun. 1998, 28, 4471e4478; (d) Marquis, R. W.; Yamashita, D. S.;
Ru, Y.; LoCastro, S. M.; Oh, H.-J.; Erhard, K. F.; DesJarlais, R. L.;
Head, M. S.; Smith, W. W.; Zhao, B.; Janson, C. A.; Abdel-Mgeguid,
S. S.; Tomaszed, T. A.; Levy, M. A.; Veber, D. F. J. Med. Chem. 1998,
41, 3563e3567; (e) Kubota, D.; Ishikawa, M.; Ishikawa, M.; Yahata,
N.; Murakami, S.; Fujishima, K.; Kitakaze, M.; Ajito, K. Bioorg. Med.
Chem. 2006, 14, 4158e4181.
4.19. (ꢂ)-trans-4-(Aminomethyl)-piperidin-3-ol (33)
To a solution of 14 (HCl salt, 58 g, 0.22 mol) in isopropanol
(2.2 L) was added a 6 N HCl solution in isopropanol (220 mL)
and the mixture was refluxed for 30 min. The reaction mixture
was concentrated under reduced pressure and the residue co-
evaporated with toluene (500 mL). The solid residue was sus-
pended in diisopropyl ether. The solid was filtered off and
dried in vacuo to give 33 as a double HCl salt (43 g, 98%),
5. (a) Lewis, N. J.; Barker, K. K.; Fox, R. M.; Mertes, M. P. J. Med. Chem.
1973, 16, 156e159; (b) Furegati, S.; Ganci, W.; Gorla, F.; Ringeisen, U.;
Rueedi, P. Helv. Chim. Acta 2004, 87, 2629e2661.
6. (a) Petrow, V.; Stephenson, O. J. Pharm. Pharmacol. 1962, 14, 306e314;
(b) Moriconi, E. J.; Misner, R. E. J. Org. Chem. 1969, 34, 3672e3674.
7. Brown, H. C. Hydroboration; W.A. Benjamin: New York, NY, 1962.
8. (a) Kergomard, A.; Veschambre, H. Tetrahedron Lett. 1976, 17, 4069e
4072; (b) Golodova, K. G.; Yakimovitch, S. I. Zh. Org. Khim. 1972, 8,
2481e2489.
1
mp 237 ꢁC. H NMR (360 MHz, DMSO-d6) d ppm 1.37e
1.53 (1H, m), 1.72e1.85 (1H, m), 1.99 (1H, dq, J¼14.3,
3.3 Hz), 2.56e2.69 (2H, m), 2.75 (1H, br t, J¼11.5 Hz),
3.08 (1H, d, J¼9.9 Hz), 3.12e3.23 (2H, m), 3.53e3.66 (1H,
m), 5.84 (1H, d, J¼4.8 Hz), 8.16 (3H, br s), 9.38 (2H, br s).
Anal. Calcd for C6H14N2O$2HCl: C, 35.48; H, 7.94; N,
13.79. Found: C, 35.46; H, 7.81; N, 14.33.
9. Takemura, S.; Miki, Y.; Uono, M. i; Yoshimura, K.; Kuroda, M.; Suzuki,
A. Chem. Pharm. Bull. 1981, 29, 3026e3032.
10. Imanishi, T.; Shin, H.; Hanaoka, M.; Momose, T.; Imanishi, I. Chem.
Pharm. Bull. 1982, 30, 3617e3623.
11. Ciaccio, J. A.; Stanescu, C.; Bontemps, J. Tetrahedron Lett. 1992, 33,
1431e1434.
4.20. (ꢂ)-1,1-Dimethylethyl trans-4-cyano-3-hydroxy-1-
piperidinecarboxylate (34)
12. Livinghouse, T. Org. Synth. Coll. 1990, 7, 517e522.
13. Recently, the epoxidation of 17 to 24, using similar conditions as
described here, was published: Grishina, G. V.; Borisenko, A. A.; Veselov,
I. S.; Petrenko, A. M. Russ. J. Org. Chem. 2005, 41, 272e278.
14. Ciaccio, J. A.; Smrtka, M.; Maio, W. A.; Rucando, D. Tetrahedron Lett.
2004, 45, 7201e7204.
To a solution of 30 (32.8 g, 0.26 mol) in methanol
(290 mL) was added (Boc)2O (69 g, 0.32 mol) at room tem-
perature and the reaction mixture was stirred at room temper-
ature overnight. Water (150 mL) was added and the mixture
was stirred for an additional 4 h at room temperature. The
methanol was evaporated under reduced pressure at room tem-
perature and the resulting mixture was extracted with CH2Cl2
(450 mL). The organic layer was dried on Na2SO4 and con-
centrated under reduced pressure to give 34 (54 g, 92%) as
an oil, which could be used as such in the next step. 1H
NMR (360 MHz, DMSO-d6) d ppm 1.39 (9H, s), 1.50e1.62
(1H, m), 1.98 (1H, dq, J¼13.4, 3.2 Hz), 2.57 (1H, br s),
15. (a) Bosmans, J.-.P. R. M. A.; De Cleyn, M. A. J.; Surkyn, M. PCT Int.
Appl. WO199902156, 1999; Chem. Abstr. 1999, 130, 139258; (b)
Bosmans, J.-.P. R. M. A.; Gijsen, H. J. M.; Mevellec, L. A. PCT Int.
Appl. WO2005000838, 2005; Chem. Abstr. 2005, 142, 114101.
16. Laduron, F.; Tamborowski, V.; Moens, L.; Horvath, A.; De Smaele, D.;
Leurs, S. Org. Process Res. Dev. 2005, 9, 102e104.
17. Racemic 14 could be resolved with S-(þ)-mandelic acid in isopropanol to
give enantiopure 3S,4S-14 as the S-(þ)-mandelic acid salt. Standard acide
base extraction then provided 3S,4S-14: [a]D þ4.4 (c 0.5, MeOH), with
further analytical data in agreement with the racemic 14.