Investigation of practical routes for the kilogram-scale production of cis-3-methylamino-4-methylpiperidines

Article abstract of DOI:10.1021/op049808k

Two routes for the synthesis of cis-N-protected-3-methylamino-4- methylpiperidine (3) were examined: a route hinging on the electrochemical oxidation of carbamate 1 to install a ketone at the 3 position of the piperidine followed by reductive animation (disconnection A), and a route involving the hydrogenation of an appropriately functionalized pyridine (disconnection B). While both routes to the desired compound were ultimately successful, the pyridine hydrogenation approach proved to be more amenable to kilogram-scale preparations due to the crystallinity and purity of intermediates in that route.

Full text of DOI:10.1021/op049808k

Organic Process Research & Development 2005, 9, 51−56
Investigation of Practical Routes for the Kilogram-Scale Production of
cis-3-Methylamino-4-methylpiperidines
Weiling Cai, James L. Colony,† Heather Frost, James P. Hudspeth,† Peter M. Kendall,‡ Ashwin M. Krishnan,†
Teresa Makowski, Duane J. Mazur,‡ James Phillips, David H. Brown Ripin,* Sally Gut Ruggeri, Jay F. Stearns,† and
Timothy D. White
Chemical Research and DeVelopment, Pfizer Global Research DiVision, Pfizer Inc., Eastern Point Road,
Groton, Connecticut 06340, U.S.A.
Abstract:
in nature. A Dieckmann condensation approach to 3-ketopi-
peridines has been reported,3 but more intriguing to us was
a strategy to install the ketone oxidatively at the unactivated
3 position.
Two routes for the synthesis of cis-N-protected-3-methylamino-
4-methylpiperidine (3) were examined: a route hinging on the
electrochemical oxidation of carbamate 1 to install a ketone at
the 3 position of the piperidine followed by reductive amination
(disconnection A), and a route involving the hydrogenation of
an appropriately functionalized pyridine (disconnection B).
While both routes to the desired compound were ultimately
successful, the pyridine hydrogenation approach proved to be
more amenable to kilogram-scale preparations due to the
crystallinity and purity of intermediates in that route.
Electrochemical Approach to 3
Shono has reported the electrochemical oxidation of
carbamate-protected piperidines directly to 2,3-diacetate4 10,
which can be converted to ketone5 12 (Scheme 3). The
oxidation of piperidines can be achieved without the use of
electrochemistry as well,6 but systems relevant to the target
at hand did not perform well under the reported conditions.
The electrochemical oxidation of carbamate-protected
piperidines reported by Shono is a close model for the
production of the desired ketone.4 The oxidation is run in
acetic acid with potassium acetate as the electrolyte, and a
close analogue to the desired system reported by Shono was
high yielding.
Introduction
cis-N-Protected-3-methylamino-4-methylpiperidine 3 was
identified as a useful intermediate in the synthesis of a
clinical drug candidate, necessitating its production on
kilogram scale. The route originally used to produce this
compound relied on a hydroboration-oxidation sequence
starting from tetrahydropyridine 6 followed by oxidation and
reductive amination to install the methylamino functionality1
(Scheme 1).
Although the route depicted above was able to provide
10 kg quantities of the desired piperidine, identifying an
alternate route was desirable1a for the following reasons. The
sulfur-based oxidations, although far preferable to the
chromium oxidations they replaced, leave the product
contaminated with dimethyl sulfide leading to handling
problems. The hydroboration reaction requires the use of
large volumes of solvent during the workup and isolation.
The overall route required five steps to complete.
From a retrosynthetic perspective, the methylamino
functionality could either originate from the ketone via
reductive amination or could be installed by hydrogenation
of an aminopyridine (Scheme 2). Strategies where the
methylamine is installed via addition of a nitrogen nucleo-
phile to an activated C3 do not appear to be viable due to a
facile ring contraction2 when the ring nitrogen is nucleophilic
The oxidation of the methyl and tert-butyl carbamates of
4-methylpiperidine proceeded to provide the desired prod-
(2) Fukuyama, T.; Jow, C.-K.; Cheung, M. Tetrahedron Lett. 1995, 36, 6373-
6374. On attempting to invert the alcohol stereocenter using N-methyl-p-
nitrobenzenesulfonamide under Mitsunobu conditions, the five-membered
ring product was isolated. This was presumably formed via displacement
of the activated alcohol by the piperidine nitrogen followed by opening of
the aziridine at the less hindered position. This instability also made the
epoxide intermediate that would be derived from compound 6 unattractive
synthetically.
(3) Use of an intramolecular Dieckmann reaction: Knight, D. W.; Lewis, N.;
Share, A. C.; Haigh, D. J. Chem. Soc., Perkin Trans. 1 1998, 22, 3673-
3684. Alkylation of methyl N-benzylglycine with methyl 4-chloropentanoate
provided the Dieckmann precursor. Clean cyclization was not achieved,
and the prohibitive cost of the alkylating agent led to the discontinuation
of this route.
(4) (a) Shono, T.; Matsumura, Y.; Onomura, O.; Kanazawa, T.; Habuka, M.
Chem. Lett. 1984, 1101-1104. (b) Shono, T.; Matsumura, Y.; Onomura,
O.; Ogaki, M.; Kanazawa, T. J. Org. Chem. 1987, 52, 536-541. (c) Shono,
T.; Matsumura, Y.; Tsubata, K.; Sugihara, Y.; Yamane, S.; Kanazawa, T.
J. Am. Chem. Soc. 1982, 104, 6697-6703. (d) Shono, T.; Matsumura, Y.;
Tsubata, K. Org. Synth. 1985, 63, 206-213.
(5) Matsumura, Y.; Takeshima, Y.; Okita, H. Bull. Chem. Soc. Jpn. 1994, 67,
304-306.
(6) (a) Scully, F. E., Jr. J. Org. Chem. Soc. 1980, 45, 1515-1517. (b) Scully,
F. E., Jr.; Davis, R. C. J. Org. Chem. Soc. 1978, 43, 1467-1468. (c) Quick,
J.; Oterson, R. Synthesis 1976, 745-746.
* Corresponding author: Fax: (860) 441-3630. E-mail: David_B_Ripin@
groton.pfizer.com.
† Seres Laboratories, 3331-B Industrial Drive, Santa Rosa, CA 95403.
‡ Electrosynthesis Co., Inc., 72 Ward Road, Lancaster, NY 14086-9779.
(1) (a) Ripin, D. H. B.; Abele, S.; Cai, W.; Blumenkopf, T.; Casavant, J. M.;
Doty, J. L.; Flanagan, M.; Koecher, C.; Laue, K. W.; McCarthy, K.; Meltz,
C.; Munchhoff, M.; Pouwer, K.; Shah, B.; Sun, J.; Teixeira, J.; Vries, T.;
Whipple, D. A.; Wilcox, G. Org. Process Res. DeV. 2003, 7, 873-878. (b)
Iorio, M. A.; Ciuffa, P.; Damia, G. Tetrahedron 1970, 26, 5519-5527.
10.1021/op049808k CCC: $30.25 © 2005 American Chemical Society
Published on Web 01/05/2005
Vol. 9, No. 1, 2005 / Organic Process Research & Development
•
51

Scheme 1. Hydroboration approach to 3a
a Reaction conditions: a) BnCl, acetone, 55 °C, 73%; b) NaBH4, EtOH, 15 °C, 73%; c) i) BF3‚OEt2, BH3•THF, THF; ii) HCl, H2O2; iii) TsOH, 88%; d) SO3‚pyridine,
DMSO, Et3N; e) MeNH2, NaHB(OAc)3, toluene, EtOH, THF, HOAc; f) HCl, EtOH, EtOAc, 53%, 3 steps.
Scheme 2. Retrosynthesis of N-protected 3
Scheme 3. Reported electrochemical route to analogue of 3a
a Reaction conditions: a) [O], HOAc, KOAc, 87%; b) AcOH, ∆, 75%; c) MeOH, NaHCO3, 95%.
Scheme 4. Electrochemical route to 3a
a Reaction conditions: a) [O], HOAc, KOAc, 45%; b) Ac2O, ∆, 90%; c) MeOH, NaHCO3, Na2CO3 85%.
ucts, although in lower yield than the example reported by
Shono (Scheme 4). The methyl carbamate 1 provided a
higher yield with fewer byproducts. At best, approximately
45% of 13 could be obtained from the reaction as a mixture
of diastereomers and with partial cleavage of the anomeric
acetate. The effects of varying current density, rigorously
drying the reaction mixture, and adding water scavengers
such as acetic anhydride were examined but did not improve
the yield of the reaction or the complexity of the product
mixture.7
chromatography. Although the overall yield for the three-
step process was about 35%, the lack of scaleable cleanups
led us to examine alternate routes. Additionally, the confor-
mational change in the six-membered ring imparted by the
sp2-hybridized nitrogen (a change from the benzyl-protected
system depicted in Scheme 1) resulted in low selectivity in
attempted reductive aminations to install the methylamino
group.
Pyridine Reduction Approach to 3
The electrochemical oxidation reaction mixture could be
carried directly into a thermal elimination of the anomeric
acetate run in acetic anhydride.5,8 This provided vinyl acetate
14 in good in situ yield, and cleavage of the acetate in a
sodium carbonate/sodium bicarbonate buffer solution and
methanol finally provided the desired ketone.5 None of these
intermediates were solids, nor do they contain a basic
nitrogen functionality from which a salt could be formed.
As a consequence, the material had to be carried through
the sequence without purification and provided material of
low purity. Distillation failed to provide material of accept-
able purity (50%), and the compound had to be purified by
An alternative to the reductive amination to install the
methylamino functionality would be the reduction of the
appropriately functionalized pyridine9 such as 15 (Scheme
5).10 Hydrogenation should provide the desired product 16
with the required stereochemistry.11 The commercially avail-
able 4-methyl-3-aminopyridine 4 is a fully functionalized
precursor to the desired piperidine and requires only reduc-
tion, protection of the piperidine nitrogen, and installation
of the methyl group to provide the desired product.
The methyl carbamate 15 of amine 4 was prepared as a
hydrogenation substrate (KOtBu, MeOCO2Me, THF, 61%).
The methyl carbamate was selected over the formamide, as
the formamide was hydrolysis-prone under some of the
(7) Current densities of 50-100 mA/cm2, starting material concentrations around
0.5 M, and 20 faradays/mol of starting material were investigated. Smaller-
scale reactions (5-7 g) were run in a 100 mL undivided glass cell with a
20 cm2 Pt mesh anode; larger reaction in the range of 30-40 g were run in
a 500 mL undivided glass cell with a 60 cm2 Pt mesh anode.
(8) Use of acetic anhydride as solvent in place of acetic acid led to clean
conversion to vinyl acetate 14 as opposed to a mixture of vinyl acetate 14
and anomeric alcohols 13, RdH in acetic acid.
(9) Keay, J. G. in ComprehensiVe Organic Synthesis; Trost, B. M., Ed.;
Pergamon: New York, 1991; Vol. 8, Chapter 3.6.
(10) Other protecting groups were investigated such as a formamide in the hope
of obtaining the methylaminopiperidine directly from the hydrogenation.
These protecting groups did not provide as clean a reaction profile as the
methyl carbamate, with solvolysis of the protecting group in the solvents
necessary to hydrogenate the pyridine ring as the main side reaction.
52
•
Vol. 9, No. 1, 2005 / Organic Process Research & Development

Scheme 5. Pyridine hydrogenation route to 3a
a Reaction conditions: a) KOtBu, (MeO2C)2O, THF, 61%; b) 5% Rh/C (type 23), AcOH, H2, 75%; c) PhCHO, NaHB(OAc)3, CH2Cl2, HOAc, 68%; d) LiAlH4,
THF; e) HCl, EtOH, 47%, 2 steps.
hydrogenation conditions in nucleophilic solvents. The
carbamate also provides the carbon of the methylamine
functionality in the final product.
purification point and allows the de of the material to be
upgraded to >97:3.1a For the five-step process, a 15% overall
yield of desired piperidine was achieved on 50 g scale. This
route was selected for further optimization for large-scale
manufacture.
The hydrogenation was initially investigated using Rh/
alumina and PtO2. The cis:trans selectivity of the reduction
1
was measured by HNMR of the isolated product of the
hydrogenation. Under the conditions originally investigated,
a disappointing 3:1 ratio of cis:trans isomers was obtained
from the process (Rh/Al2O3, ethanol, 100 psi, 80 °C; or PtO2,
acetic acid, 100 psi, 60 °C). Although not optimal, the purity
and de of the desired isomer could be upgraded downstream
as the HCl salt of the final product.1a The high levels of trans
isomer observed are indicative of a stepwise reduction with
the C-4 methyl bearing stereocenter scrambling via an imine
intermediate or of a decomplexation-recomplexation of the
metal catalyst prior to complete reduction of the ring.
A number of catalysts (PtO2, Rh/C (several types), RuO2,
Rh/Al2O3, Ru/C (several types), Lindlar’s catalyst, and
Wilkinson’s catalyst) and solvents (AcOH, iPrOH, EtOH,
MeOH/NH4OH, MeCN, THF, cyclohexane, heptanes, PhMe,
DMF, water) were screened for the hydrogenation. Ulti-
mately, hydrogenation with 5%Rh/C in acetic acid was most
effective for the transformation, providing the product in 11:1
selectivity. The product could be isolated as a solid by
precipitating the hydrochloride salt, but on larger scale it
was more practical to carry the material forward in solution
or as an oil since the free base is required for the next step.
Several of the other conditions screened did provide complete
hydrogenation of the pyridine, but with lower diastereose-
lectivity.
Optimization of the Pyridine Approach
Optimization of the route began with the carbamate
formation. In the laboratory, the aminopyridine and 3.0 equiv
of base were slurried in THF, cooled, then reacted by
portionwise addition of 1.5 equiv of dimethyl carbonate. This
order of addition generated intense colors that sometimes
carried through to the isolated product and left a sticky
residue in the reaction vessel. After a water quench, four
ethyl acetate washes were required to extract the product
out of the aqueous layer. The organics were concentrated,
and the product was crystallized from an ethyl acetate/
diisopropyl ether (IPE) mixture. The reaction appeared to
proceed very smoothly, but significant yield losses were
encountered during the isolation and purification, giving the
carbamate in a modest 61% yield. One surprising finding
was the high aqueous solubility of this and all the other
intermediates in the route. Care had to be taken to minimize
the volume of any necessary aqueous wash, or the recovery
of product was highly diminished. To address this issue, a
solvent screen was carried out. The ethers were superior
solvents for the reaction,12 and 2-methyltetrahydrofuran (2-
MeTHF) was selected because of its low miscibility with
water. After the reaction was complete, minimal water was
added to remove the residual salts, and no back-extractions
were required. The color generated in the reaction could be
avoided by changing the order of addition and adding a
solution of 4 last to the reaction mixture. Displacement of
the 2-MeTHF with toluene gave the desired product in 87%
yield. During the displacement it was important to azeotro-
pically remove any residual water to very low levels since
it has a negative impact on recovery. Additionally, the
presence of water during the crystallization results in
formation of material with very poor filtration properties
(Scheme 6).
With the hydrogenation of the pyridine complete, the
benzylamine was installed by reductive amination with
benzaldehyde and sodium triacetoxyborohydride to provide
the benzylamine in 68% yield. This material could then be
carried crude into the methyl carbamate reduction using
lithium aluminum hydride. After workup, the free base
diamine could be isolated as the HCl salt in 47% yield for
the reduction and salt formation. The HCl salt is an effective
(11) (a) Moon, M. W.; Morris, J. K.; Heier, R. F.; Childester, C. G.; Hoffmann,
W. E.; Piercey, M. F.; Althaus, J. S.; VonVoigtlander, P. F.; Evans, D. L.;
Figur, L. M.; Lahti, R. A. J. Med. Chem. 1992, 35, 1076-1092. (b)
Reitsema, R. H.; Hunter, J. H. J. Am. Chem. Soc. 1949, 71, 1680-1682.
(c) Ebnoether, A.; Jucker, E.; Rissi, E.; Rutschmann, J.; Shreier, E.; Steiner,
R.; Suess, R.; Vogel, A. HelV. Chim. Acta 1959, 42, 918-955. (d) Biel, J.
H.; Hoya, W. K.; Leiser, H. A. J. Am. Chem. Soc. 1959, 81, 2527-2532.
(e) Cooper, G. H.; Rickard, R. L. J. Chem. Soc. C 1971, 772-776. (f) Crider,
A. M.; Lamey, R.; Floss, H. G.; Cassady, J. M.; Bradner, W. J. J. Med.
Chem. 1980, 23, 848-851. (g) Armour, D. R.; Chung, K. M. L.; Congreve,
M.; Evans, B.; Guntrip, S.; Hubbard, T.; Kay, C.; Middlemiss, D.; Mordaunt,
J. E.; Pegg, N. A.; Vinader, M. V.; Ward, P.; Watson, S. P. Bioorg. Med.
Chem. Lett. 1996, 6, 1015-1020. (h) Nienburg, H. Chem. Ber. 1937, 70,
635-638. (i) Glennon, R. A.; Jacyno, J. M.; Salley, J. J.; J. Med. Chem.
1982, 25, 68-70. (j) Ledoussal, B.; Almstead, J.-I. K.; Gray, J. L.; Hu, X.
E. World Patent WO99/14214, 1999.
The original goal during optimization of this route was
to combine the pyridine reduction and reductive amination
into one step. Several catalysts were examined for their
ability to effect both transformations. The only catalyst that
gave promising results was Rh2(COD)2Cl2. However, during
the reactions, the catalyst plated out on the reaction vessel
(12) A number of precautions are used when selecting ether solvents: distillations
and solvent displacements are never run to near dryness, inhibitors are added
immediately to the distillates, and oxidation reactions are avoided in these
solvents. Laird, T. Org. Process Res. DeV. 2004, 8, 815.
Vol. 9, No. 1, 2005 / Organic Process Research & Development
•
53

Scheme 6. Optimized conditions of the pyridine reduction route to 3a
a Reaction conditions: KOtBu, (MeO2C)2O, 2-MeTHF; toluene, 87%; b) 5% Rh/C (JM type C101023-5), AcOH, H2; c) PhCHO, NaHB(OAc)3, toluene, 73%, 2
steps; d) LiAlH4, THF; HCl, IPO, 87%.
surfaces, which would not be acceptable for scale-up. Several
additives (phosphine ligands, Celite) were tried to avoid the
deposition or adsorb the catalyst as it precipitated, but all of
them retarded the pyridine reduction significantly and caused
erosion of the diastereoselectivity. Therefore, it was decided
to optimize the two-step sequence.
followed by portionwise addition of sodium triacetoxyboro-
hydride (STAB). After reaction completion, the mixture was
quenched with saturated NaHCO3 and extracted three times
with EtOAc. The solvent was replaced with IPE, and the
HCl salt was precipitated by bubbling gaseous HCl into the
solution. The salt was then recrystallized from EtOH to
obtain the product in 68% over the two steps. For environ-
mental reasons and ease of handling, one of the first
objectives was to replace the methylene chloride with a more
benign solvent. Toluene was found to be a convenient
replacement and had the added benefit of facilitating the
displacement of the acetic acid from the previous step.
Extensive displacement of the acetic acid led to the precipita-
tion of the acetate salt of the product. However, the salt was
thixotropic and the yield was low, so the displacement was
stopped once there were fewer than four equiv of acetic acid
remaining. This gave a convenient solution for the reductive
amination and minimized the amount of acid to quench at
the end of the reaction. A solution of the piperidine and
benzaldehyde was added to a slurry of STAB in toluene,
which avoided the need to add solids portionwise to a reactor.
The excess STAB and acetic acid were quenched with
aqueous sodium hydroxide. It was necessary to control the
pH of the quench to avoid the formation of thick emulsion
layers, and a range of 6-7 was found to be optimal. Addition
of concentrated HCl to the toluene solution precipitated the
desired product as its salt. The salt formation was best
performed at elevated temperatures to purge the trans
diastereomer and to grow larger crystals for faster filtrations.
After cooling and filtration, the N-benzylamine was isolated
in 73% yield.
The final step to form the target amine was reduction of
the methyl carbamate. Originally, the HCl salt of the starting
material was broken prior to addition of the LAH solution,
extracted into EtOAc, which was then displaced with THF.
The reaction was quenched with sodium sulfate decahydrate
at a rate that controlled the hydrogen evolution. After solvent
removal, the product was dissolved in IPE and the bis(HCl)
salt was formed with gaseous HCl. Recrystallization from
aqueous EtOH completed the synthesis of 3 in modest yield
(47%). During optimization, the salt break was found to be
unnecessary since it could be effected by addition of an
excess of LAH. Addition of 2.25 mol equiv of water (relative
to aluminum) diluted with THF to control the exotherm
during the quench successfully precipitated the aluminum
salts without gumming. After filtration of the salts, the THF
was displaced with 2-propanol for formation of the bis(HCl)
salt with concentrated HCl. The salt formation could be
carried out directly from the THF solution, but the solids
formed filtered very poorly and did not purge any of the
The laboratory procedure for the pyridine reduction
required the use of 5% Rh/C (Johnson-Matthey type C101023-
5) in acetic acid. The reduction proceeded with reasonable
diastereoselectivity, giving an 11:1 ratio of cis:trans isomers.
After reaction completion, the catalyst was removed by
filtration, and the majority of the acetic acid was removed
under reduced pressure. Residual acid was quenched with
saturated NaHCO3, and the product was extracted into
EtOAc. A large portion of the product still remained in the
aqueous layer after three extractions; thus, the aqueous layer
was stripped to dryness and re-extracted. The combined lots
could either be carried into the next step, or the intermediate
could be isolated as its HCl salt (all the intermediates after
the carbamate are oils, but all form solid salts readily). During
the optimization, which made use of different sized and
configured reactors, it was noted that the ratio of diastere-
omers sometimes improved to approximately 20:1. It is
believed that the final intermediate prior to the stereochemi-
cally defining step may be the tetra-substituted olefin. In
setups in which the hydrogen available for reaction was
plentiful (low fill volumes, good mixing), the reduction
proceeded with little stereochemical leakage. In those situ-
ations in which the reaction may have been hydrogen-starved
(high fill volumes, poor mixing), there was some analytical
evidence13 to suggest that olefin migration prior to reduction
was a competing event, which led to eroded diastereomeric
ratios.
In the optimized conditions, the reduction was run in 10
volumes of acetic acid. The catalyst loading could be lowered
from the initial 30 wt % to as little as 10 wt %, but was kept
at 20 wt % on large scale to ensure complete reduction. After
the reaction was complete, the catalyst was removed by
filtration, and the majority of the acetic acid was removed
under vacuum prior to proceeding to the reductive amination.
Since the procedure for the next step uses acetic acid, it was
decided that the quench and extractions were unnecessary
and the acid solution could be carried forward directly into
the reductive amination. Lab pilots confirmed that there was
no negative impact on the purity profile of the reaction by
omitting the aqueous washes.
The reductive amination was originally carried out by
mixing the amine and benzaldehyde in methylene chloride
(13) HPLC-MS indicates impurity peaks with a mass of M-2, suggesting
incomplete hydrogenation.
54
•
Vol. 9, No. 1, 2005 / Organic Process Research & Development

trans diastereomer. The salt formation in 2-propanol gave
larger crystals that filtered well and purged the trans
diastereomer to low levels. During the salt formation, it was
important to remove the water introduced with the HCl to
maximize the recovery, since even small levels of residual
water diminished the yield markedly. The yield for the final
step was 84-87%, and the level of the trans diastereomer
was <0.8%.
was added to aid stirring. The mixture was stirred between
20 and 35 °C for at least 2 h. The reaction was sampled and
checked for completion by GC, then cooled to 15-20 °C.
Water (255 L) was added at a rate to maintain the temper-
ature below 25 °C. The mixture was stirred for at least 30
min then allowed to settle for at least 60 min. The phases
were separated, and 2-methyltetrahydrofuran (255 L) was
added to the aqueous layer. The reaction mixture was allowed
to stir for at least 60 min and then allowed to settle for at
least 60 min. The phases were separated, and the 2-meth-
yltetrahydrofuran layers were combined. Darco KBB (10.0
kg) was added to the organic layer and allowed to stir for at
least 30 min. The Darco slurry was filtered through a bed of
Celite, and the cake was washed with 2-methyltetrahydro-
furan (51 L). The 2-methyltetrahydrofuran was displaced
with toluene under vacuum to a final volume of 480-540
L, then the mixture was cooled to 23-27 °C over at least
90 min. After sampling to ensure that the water content was
<1%, the slurry was stirred for at least 12 h. The resulting
solids were filtered and washed with toluene (212 L). After
drying under vacuum at 40-50 °C for at least 12 h with a
slight nitrogen bleed, 66.5 kg (400 mol) of (4-methyl-pyridin-
3-yl)carbamic acid methyl ester was isolated in 86.7% yield
Conclusion
Two new routes to cis-N-protected-3-methylamino-4-
methylpiperidine 3 have been identified. While the electro-
chemical approach to the molecule worked in reasonable
yield (35%) from inexpensive starting materials, purity and
isolation difficulties would make it difficult to use on large
scale. The pyridine reduction route did prove to be a practical
alternative to the hydroboration route previously reported and
could be utilized for the preparation of multikilogram
quantities of 3 in 55% overall yield.
Experimental Section
General. All materials were purchased from commercial
suppliers and used without further purification. All reactions
were conducted under an atmosphere of nitrogen unless noted
otherwise. All reactors were glass-lined steel vessels with
the exception of those used for catalytic hydrogenations,
which were Hastelloy. Reactions were monitored for comple-
tion by removing a small sample from the reaction mixture
and analyzing the sample by TLC, GC, or HPLC. HPLC
analyses were performed using one of the following sys-
tems: Waters Symmetry Shield RP18 column (4.6 m × 75
mm, 3.5 µm) and an acetonitrile/water (0.5% trifluoroacetic
acid) mobile phase, a Zorbax SB-CN 4.6 m × 150 mm
column or a Symmetry C-8 3.9 m × 150 mm column, and
a mobile phase consisting of acetonitrile and either 0.5%
perchloric acid or 0.2% phosphoric acid. GC analyses were
performed on a DBWaxEtr (15 m × 0.25 mm × 0.25 µm
film). Proton and carbon NMR spectroscopies were per-
formed on a Bruker-Spectrospin Avance 400 MHz instru-
ment. Karl Fischer measurements were obtained on a Mettler
Toledo DL38 Volumetric Karl Fischer apparatus. Melting
points were determined on a Buchi B-545 melting point
apparatus. Elemental analyses were performed by Quantita-
tive Technologies Inc. of Whitehouse, NJ.
3-Methoxycarbonylamino-4-methylpyridine (15). To a
clean, dry, nitrogen-purged 800 L reactor were charged
3-amino-4-methylpyridine 4 (49.9 kg, 461 mol) and 2-me-
thyltetrahydrofuran (383 L). The reaction was heated to 32-
38 °C for at least 90 min. To a clean, dry, nitrogen-purged
2000 L reactor were charged potassium tert-butoxide (104.0
kg, 927 mol) and 2-methyltetrahydrofuran (255 L). The
reaction was stirred at 23-27 °C for at least 30 min to break
up the potassium tert-butoxide. Dimethyl carbonate (49.9 kg,
554 mol) was added to the slurry of potassium tert-butoxide
at a rate that maintained the temperature below 35 °C. The
3-amino-4-methylpyridine solution was added to the 2000
L reactor at a rate that maintained the temperature from 20
to 35 °C. An additional 115 L of 2-methyltetrahydrofuran
1
(99.89% purity by HPLC) mp: 115.3-116.6 °C. H NMR
(400 MHz, DMSO): δ 9.12 (bs, 1H), 8.46 (s, 1H), 8.18 (d,
J ) 4.9 Hz, 1H), 7.20 (d, J ) 4.9 Hz, 1H), 3.64 (s, 3H),
2.19 (s, 3H). 13C NMR (DMSO): 155.6, 146.5, 146.3, 141.4,
134.2, 125.9, 52.6, 17.8. Anal. Calcd for C8H10N2O2: C,
57.82; H, 6.07; N, 16.86. Found: C, 57.71; H, 5.80; N, 16.85.
cis-3-Methoxycarbonylamino-4-methylpiperidine (16).
To a clean, nitrogen-purged 1200 L reactor were charged
Darco KBB14 (6.6 kg), (4-methyl-pyridin-3-yl)carbamic acid
methyl ester (66.5 kg) and acetic acid (677 L). The mixture
was stirred for at least 30 min at 20-30 °C and filtered
through a bed of Celite, and the cake was washed with acetic
acid (135 L). To a clean, nitrogen-purged 2000 L hydrogena-
tion reactor were charged 5% Rh/C (Johnson-Matthey type
C101023-5, 16.5 kg) and the (4-methyl-pyridin-3-yl)carbamic
acid methyl ester acetic acid solution. The reaction was
stirred for at least 15 min and then purged sequentially with
nitrogen and hydrogen. The reaction was heated to 72-78
°C and then pressurized with hydrogen gas at 70-80 psig.
The reaction was allowed to stir under these conditions until
hydrogen uptake ceased. A sample was obtained for reaction
completion check by GC. The reaction was purged with
nitrogen, and the catalyst was filtered on a water-wet Celite-
coated filter. The cake was washed with toluene (212 gal),
and the filtrates were combined. The solution of (4-methyl-
pipridin-3-yl)carbamic acid methyl ester (92.7% cis by GC)
was used in the next step without further purification. In the
lab, an aliquot (0.50 g) of the reduced product was purified
via silica gel chromatography (20% EtOAc-hexanes) to
yield purified product (0.38 g). The purified product was
dissolved in IPE (10 mL) and treated with bubbling HCl
gas. The white solid was filtered and dried under vacuum to
yield a white solid (11.9 g, 75%) mp 199-200.5 °C. Rf
(14) The additional Darco treatment removes residual catalyst poison present in
some lots of material.
Vol. 9, No. 1, 2005 / Organic Process Research & Development
•
55

(freebase) ) 0.21 (5:1 CH2Cl2-CH3OH). Anal. Calcd for
C8H17ClN2O2: C,46.04; H, 8.21; N, 13.42. Found: C, 46.93;
H, 8.84; N, 12.29. 1H NMR (HCl salt) (400 MHz, CD3OD):
δ 6.94 (bs, 1H), 3.95 (m, 1H), 3.66 (m, 4H), 3.38 (dd, J )
13.27, 4.15, 1H), 3.28 (m, 1H), 3.14 (dd, J ) 13.27, 3.11,
1H), 3.05 (ddd, J ) 12.85, 9.95, 4.56 1H), 2.07 (m, 1H),
1.74 (m, 2H), 1.01 (d, J ) 7.05, 3H). 13C NMR (300 MHz,
CD3OD): δ 158.0, 51.6, 48.7, 46.7, 42.9, 31.1, 25.6, 15.1.
cis-N-Benzyl-3-methoxycarbonylamino-4-methylpipe-
ridine Hydrochloride (17). To a clean, nitrogen-purged 2000
L reactor was charged (4-methyl-pipridin-3-yl)carbamic acid
methyl ester as the crude acetic acid and toluene solution
from the previous step. The acetic acid was displaced with
toluene under vacuum to a final volume of 500-560 L. The
reaction was cooled and a sample pulled to check for an
times with nitrogen and allowed to stir for at least 45 min at
20-30 °C to break up the solids. A sample was pulled to
ensure that the water content was <0.2% water. The reaction
was cooled to between -15 to 5 °C, and a 1.0 M solution
of lithium aluminum hydride in tetrahydrofuran (181.4 kg,
200 mol) was added at a rate to maintain the temperature
from -15 to 5 °C. The charge line was rinsed with
tetrahydrofuran (19 gal), and the reaction was heated and
held at reflux for at least 2.5 h. After cooling to 20-30 °C,
the reaction was sampled and checked for completion by
GC. The reaction was cooled to between -10 and 5 °C and
a chilled (-10 to 5 °C) solution of tetrahydrofuran (43 L)
and water (16.3 kg) was added at a rate to maintain the
temperature at -10 to 5 °C with slight nitrogen bleed. The
reaction was then heated to 20-25 °C over at least 60 min
and purged with nitrogen to remove any traces of hydrogen.
The resulting aluminum solids were filtered and washed with
tetrahydrofuran (2 × 86 L). The tetrahydrofuran was
displaced with 2-propanol until a temperature of at least 78
°C and a reaction volume was of 460-540 L were achieved.
The reaction was cooled to 65-75 °C, and concentrated HCl
(28.9 kg, 294 mol) was added over at least 60 min. The
displacement was continued until additional 2-propanol (1060
L) was added and the final temp was at least 81 °C and the
final volume was 390-460 L. The reaction was cooled to
between 65 and 75 °C and granulated for at least 2 h. The
reaction was cooled to between 15 and 25 °C, and a sample
was pulled to ensure that the water content was <1%. After
stirring at least 2 h at 15-25 °C, the solids were filtered
and washed with 2-propanol (85 L). After drying under
vacuum for at least 12 h at 40-50 °C with a slight nitrogen
bleed (cis-N-benzyl-3-methylamino-4-methypiperidine bis-
(hydrochloride) (35.5 kg, 122 mol) was isolated in 87.1%
yield (99.3% cis and 0.70% trans by GC) mp: 261.3-267.0
1
acetic acid content of <4% as determined by H NMR.
Benzaldehyde (46.4 kg, 437 mol) was added to the reaction
at 20-30 °C and stirred for at least 30 min. To a clean, dry,
nitrogen-purged reactor were charged sodium triacetoxy-
borohydride (92.6 kg, 437 mol) and toluene (472 L). The
mixture was allowed to stir for at least 60 min at 20-30 °C.
The benzaldehyde solution was transferred to the sodium
triacetoxyborohydride reactor at a rate that maintained the
temperature from 20 to 30 °C, then the reaction was stirred
for at least 2 h. The reaction was sampled and checked for
completion by GC. It was quenched by addition of a solution
of 50% aqueous sodium hydroxide (158.9 kg diluted with
352 gal of water), maintaining the temperature at 20-30 °C,
until a pH between 6 and 7 was achieved. The reaction was
then stirred for at least 60 min and allowed to settle for at
least 60 min, and the phases were separated. The toluene
layer was heated to 70-80 °C, and concentrated HCl (47.0
kg, 477 mol) was added over at least 30 min. The reaction
was held from 70 to 80 °C for at least 60 min. The reaction
was cooled to 15-25 °C over at least 60 min and held for
at least 2 h. The resulting solids were filtered, and the cake
was washed with toluene (190 L). After drying under vacuum
at 40-50 °C for at least 12 h with a slight nitrogen bleed,
cis-N-benzyl-3-methoxycarbonylamino-4-methylpiperidine hy-
drochloride (86.8 kg, 290 mol) was isolated in 73.1% yield
over two steps (96.6% cis and 2.5% trans by GC) mp:
187.3-191.4 °C. 1H NMR (400 MHz, CDCl3): δ 12.11 (bs,
1H), 7.57-7.52 (m, 3H), 7.43-7.39 (m, 3H), 4.26 (dd, J )
13.3, 4.6 Hz, 1H), 4.16 (bd, J ) 9.9 Hz, 1H), 4.07 (dd, J )
13.3, 5.8 Hz, 1H), 3.60 (s, 3H), 3.49 (bd, J ) 12.0 Hz, 1H),
3.29 (dd, J ) 12.9, 2.1 Hz, 1H), 2.89 (ddd, J ) 12.9, 10.8,
3.3 Hz, 1H), 2.74 (dddd, J ) 12.4, 12.0, 9.5, 2.9 Hz, 1H),
2.22 (dddd, J ) 13.3, 13.0, 12.4, 3.9 Hz, 1H), 1.82-1.74
(m, 1H), 1.64 (bd, J ) 13.0 Hz, 1H), 0.94 (d, J ) 6.6 Hz,
3H). 13C NMR (CDCl3): 157.8, 131.8, 130.5, 129.6, 127.9,
61.3, 56.8, 53.1, 52.5, 48.6, 32.6, 26.2, 17.2. Anal. Calcd
for C15H23ClN2O2: C, 60.29; H, 7.76; N, 9.38. Found: C,
60.21; H, 7.83; N, 9.29.
1
°C. H NMR (400 MHz, 1:1 CD3CN:D2O): δ 7.51-7.43
(m, 5H), 4.36-4.27 (m, 2H), 3.62-3.59 (m, 2H), 3.24-
3.13 (m, 3H), 2.65 (s, 3H), 2. (m, 1H), 1.95-1.91 (m, 1H),
1.82-1.75 (m, 1H), 1.03 (d, J ) 7.5 Hz, 3H). 13C NMR
(1:1 CD3CN:D2O): 131.8, 130.5, 129.6, 128.6, 118.7, 61.0,
55.8, 46.4, 31.3, 27.5, 26.0, 10.0. Anal. Calcd for C14H24-
Cl2N2: C, 57.73; H, 8.31; N, 9.62. Found: C, 57.77; H, 8.30;
N, 9.60.
Acknowledgment
We thank Ste´phane Caron, Robert Dugger, John Ragan,
and Stephen Ley for helpful discussions. We thank Grace
Vandal, Anthony Mahan, and Paul Concannon for analytical
support and Michael Bright, Steven Guinness, Eric Dias, and
David R. Bill for safety and engineering assistance. We also
thank Jeff Adler, Eric Allen, Thomas Behan, Joseph Bella-
vance, Thomas Brooks, Keith Brouwer, Randy Church,
Wayne Crandall, Johnathan Dykema, Thomas Limani, Larry
Lumbert, Brian Morgan, Alden Peckham, Steven Stott, Ralph
Scheck, III, Scott Schelkly, Walden Smolen, and Elizabeth
Storms for their execution of the large-scale procedures.
cis-N-Benzyl-3-methylamino-4-methypiperidine Bis-
(hydrochloride) (3). To a clean, dry, nitrogen-purged 2000
L reactor were charged (1-benzyl-4-methyl-piperidin-3-yl)-
carbamic acid methyl ester hydrochloride (41.9 kg, 140 mol)
and tetrahydrofuran (685 L). The reactor was purged three
Received for review October 27, 2004.
OP049808K
56
•
Vol. 9, No. 1, 2005 / Organic Process Research & Development