Synthesis of a muscarinic receptor antagonist via a diastereoselective Michael reaction, selective deoxyfluorination and aromatic metal-halogen exchange reaction

Article abstract of DOI:10.1021/jo0157425

An efficient synthesis of a structurally unique, novel M₃ antagonist 1 is described. Compound 1 is conveniently disconnected retrosynthetically at the amide bond to reveal the acid portion 2 and the amine fragment 3. The synthesis of key intermediate 2 is highlighted by a ZnCl₂-MAEP complex 19 catalyzed diastereoselective Michael reaction of dioxolane 7 with 2-cyclopenten-1-one (5) to establish the contiguous quaternary-tertiary chiral centers and a subsequent geminal difluorination of ketone 17 using Deoxofluor in the presence of catalytic BF₃·OEt₂. The synthesis of the amine moiety 3 is highlighted by the discovery of a novel n-Bu₃MgLi magnesium-halogen exchange reaction for selective functionalization of 2,6-dibromopyridine. This new and practical metalation protocol obviated cryogenic conditions and upon quenching with DMF gave 6-bromo-2-formylpyridine (26) in excellent yield. Further transformations afforded the amine fragment 3 via reductive amination with 35, Pd-catalyzed aromatic amination, and deprotection. Finally, the highly convergent synthesis of 1 was accomplished by coupling of the two fragments. This synthesis has been used to prepare multi-kilogram quantities of the bulk drug.

Full text of DOI:10.1021/jo0157425

J . Org. Chem. 2001, 66, 6775-6786
6775
Syn th esis of a Mu sca r in ic Recep tor An ta gon ist via a
Dia ster eoselective Mich a el Rea ction , Selective Deoxyflu or in a tion
a n d Ar om a tic Meta l-Ha logen Exch a n ge Rea ction
Toshiaki Mase,*,† Ioannis N. Houpis,*,‡ Atsushi Akao,† Ilias Dorziotis,‡ Khateeta Emerson,‡
Thoa Hoang,‡ Takehiko Iida,† Takahiro Itoh,† Keisuke Kamei,† Shinji Kato,† Yoshiaki Kato,†
Masashi Kawasaki,† Fengrui Lang,‡ J aemoon Lee,‡ J oseph Lynch,‡ Peter Maligres,‡
Audrey Molina,‡ Takayuki Nemoto,† Shigemitsu Okada,† Robert Reamer,‡ J ake Z. Song,‡
David Tschaen,‡ Toshihiro Wada,† Daniel Zewge,‡ R. P. Volante,‡ Paul J . Reider,‡ and
Koji Tomimoto†
Process R & D, Laboratories for Technology Development, Banyu Pharmaceutical Co. Ltd.,
Kamimutsuna 3-Chome-9-1, Okazaki, Aichi 444-0858, J apan, and Merck Research Laboratories,
Department of Process Chemistry, P.O. Box 2000, Rahway, New J ersey 07065
masetk@banyu.co.jp
Received May 9, 2001
An efficient synthesis of a structurally unique, novel M3 antagonist 1 is described. Compound 1 is
conveniently disconnected retrosynthetically at the amide bond to reveal the acid portion 2 and
the amine fragment 3. The synthesis of key intermediate 2 is highlighted by a ZnCl2-MAEP complex
19 catalyzed diastereoselective Michael reaction of dioxolane 7 with 2-cyclopenten-1-one (5) to
establish the contiguous quaternary-tertiary chiral centers and a subsequent geminal difluorination
of ketone 17 using Deoxofluor in the presence of catalytic BF3‚OEt2. The synthesis of the amine
moiety 3 is highlighted by the discovery of a novel n-Bu3MgLi magnesium-halogen exchange
reaction for selective functionalization of 2,6-dibromopyridine. This new and practical metalation
protocol obviated cryogenic conditions and upon quenching with DMF gave 6-bromo-2-formylpyridine
(26) in excellent yield. Further transformations afforded the amine fragment 3 via reductive
amination with 35, Pd-catalyzed aromatic amination, and deprotection. Finally, the highly
convergent synthesis of 1 was accomplished by coupling of the two fragments. This synthesis has
been used to prepare multi-kilogram quantities of the bulk drug.
In tr od u ction
In addition to the interesting biological profiles, 1 poses
a number of interesting synthetic challenges including
the contiguous quaternary-tertiary chiral centers as well
as a gem-difluoro functionality. The heterocyclic amine
portion contains no chiral centers; however, an economi-
cal and elegant synthesis is still a significant task. In
this paper we describe an efficient and scalable synthesis
of the target molecule 1 by virtue of the following
discoveries: (a) a Zn-triamine complex catalyzed, highly
diastereoselective Michael reaction that establishes both
chiral centers in one operation, (b) a chemoselective
deoxyfluorination reaction using the newly discovered
reagent Deoxofluor and (c) a novel magnesium ate
complex-induced metal-halogen exchange reaction which
enables facile generation of a new, stable metallopyridine
species at higher temperature.
Over the past century classical muscarinic antagonists
have been widely used for the treatment of certain
diseases. However, their therapeutic applicability was
limited, due to side effects in both the peripheral and
central nervous system. In the past decade, this field has
clearly been moving into the limelight due to the discov-
ery of muscarinic receptors of distinctive subtypes, sug-
gesting new therapeutic utility if ligands could be de-
signed which selectively bind to each subtype. Five
receptor subtypes have been so far identified and cloned
(m1 to m5),1 and among them, four receptor subtypes (m1
to m4) have been pharmacologically classified as M1-
M4 on the basis of their response to selective antagonists.
Our recent drug candidate (1) is a highly potent, orally
active, long acting, selective M3 antagonist and as such
is being investigated for the treatment of chronic obstruc-
tive pulmonary diseases and urinary incontinence.2
* To whom correspodence should be addressed. FAX: +81 564 51
7086.
† Banyu Pharmaceutical Co., Ltd.
‡ Merck Research Laboratory
(1) (a) Kubo, T.; Fukuda, K.; Mikami, A.; Maeda, A.; Takahashi, H.;
Mishina, M.; Haga, T.; Haga, K.; Ichiyama, A.; Kanagawa, K.; Kojiima,
M.; Matsuo, H.; hirose, T.; Numa, S. Nature 1986, 323, 411-416. (b)
Kubo, T.; Maeda, A.; Sugimoto, K.; Akiba, I.; Mikami, A.; Takahashi,
H.; Haga, T.; Haga, K.; Ichiyama, A.; Kanagawa, K.; Matsuo, H.;
Hirose, T.; Numa, S. FEBS Lett. 1986, 209, 367-372. (c) Peralta, E.
G.; Ashkenawi, A.; Winslow, J . W.; Smith, D. H.; Ramachandran, J .;
Capon, D. J . EMBO J . 1987, 6, 3923-3929. (d) Bonner, T. I.; Buckley,
N.; Young, A. C.; Brann, M. R. Science 1987, 237, 527-532. (e) Bonner,
T. I.; Young, A. C.; Brann, M. R.; Buckley, N. J . Neuron 1988, 1, 403-
410.
I. Retr osyn th etic An a lysis
The final target compound 1 can be conveniently
dissected at the amide bond to give two segments, the
carbocyclic mandelic acid derivative 2 and the hetero-
cyclic polyamine 3. Typically, introduction of the gem-
10.1021/jo0157425 CCC: $20.00 © 2001 American Chemical Society
Published on Web 09/07/2001

6776 J . Org. Chem., Vol. 66, No. 20, 2001
Mase et al.
Sch em e 1
Sch em e 2
difluoro functionality could be envisioned by one of the
following three ways:3 (1) a fluorinated building block,
(2) an electrophilic fluorination process,4 or (3) a nucleo-
philic fluorination of a carbonyl group. We envisioned the
latter approach to be the most efficient, thus generating
the advanced ketone intermediate 4, with its contiguous
quaternary and tertiary chiral centers. Application of a
retro-Michael transformation to ketone 4 generates the
optically active mandelic acid (6) and the commercially
available 2-cyclopenten-1-one (5) starting materials. In
the forward, synthetic direction this analysis suggests
the use of the Seebach chiral dioxolane 75 as an appropri-
ate functionalized Michael partner with 2-cyclopenten-
1-one (5) (Scheme 1).
Our retrosynthetic analysis (Scheme 2) of the amine
segment 3 reveals two key potential end game discon-
nections, (a) the C-N bond or (b) the C-C bond. The
former suggests alkylation of the protected aminopiperi-
dine 9 with a properly functionalized pyridine 8 by either
reductive amination (with an aldehyde, R ) CHO) or a
simple N-alkylation type displacement (R ) CH2X). The
C-C bond formation approach suggests a sequential
homologation of 9, for example via initial formation of
the iminium cation 11 followed by treatment with a
metalated pyridine 106 (i.e., the Petasis protocol (M )
B(OH)2).7 The aminopyridine moieties 8 or 10 could, in
turn be made from the readily available and inexpensive
2,6-dibromopyridine (13) or 2-amino-6-picoline (14). In
the case of 13, the amino group on the pyridine ring could
potentially be introduced by a transition metal catalyzed
aromatic amination.
II. Syn th esis of th e Ma n d elic Acid F r a gm en t 2
A. Syn th esis of th e Ma n d elic Acid Aceta l 7. The
preparation of acetal 7 from mandelic was readily ac-
complished in the laboratory at kilogram scale according
to the literature procedure.8 (R)-Mandelic acid (6), piv-
alaldehyde, and a catalytic amount of p-toluenesulfonic
acid were heated at reflux in pentane while water was
azeotropically removed by means of a Dean-Stark ap-
paratus. The resulting solid was filtered and triturated
in pentane at ambient temperature to afford the product
(6) For classical iminium approaches: review, see (a) Kleinman, E.
F. Comprehensive Organic Synthesis; Trost B. M.; Fleming, I., Eds.;
Pergamon Press: Oxford, 1991; vol. 2, pp 893-951. (b) Heaney, H.
Comprehensive Organic Synthesis; Trost B. M.; Fleming, I., Eds.;
Pergamon Press: Oxford, 1991; vol. 2, pp 953-973. (c) Kleinman, E.
F.; Volkmann, R. A. Comprehensive Organic Synthesis; Trost B. M.;
Fleming, I., Eds.; Pergamon Press: Oxford, 1991; vol. 2, pp 975-1006.
(7) (a) Petasis, N. A.; Akritopoulou, I. Tetrahedron Lett. 1993, 34,
583-586. (b) Petasis, N. A.; Zavialov, I. J . Am. Chem. Soc. 1997, 119,
445-446. (c) Petasis, N. A.; Zavialov, I. J . Am. Chem. Soc. 1998, 120,
11798-11799. (d) Harwood, L. M.; Currie, G. S.; Drew, M. G. B.; Luke,
R. W. A. Chem. Commun. 1996, 1953-1954.
(8) Several methods exist in the literature for the synthesis of lactic
and mandelic acid dioxolanones: (a) Seebach, D.; Naef, R. Helv. Chim.
Acta. 1981, 64, 2704-2710 and references therein. (b) Ortholand, J .
Y.; Vicart, N.; Greiner, A. J . Org. Chem. 1995, 60, 1880-1884. (c) Ott,
J .; Ramos Tombo, G. M.; Schmid, B.; Venanzi, L. M.; Wang, G.; Ward,
T. R. Tetrahedron Lett. 1989, 30, 6151-6154.
(2) (a) Mitsuya, M.; Kobayashi, K.; Kawakami, K.; Satoh, A.; Ogino,
Y.; Kakikawa, T.; Ohtake, N.; Kimura, T.; Hirose, H.; Sato, A.;
Numazawa, T.; Hasegawa, T.; Noguchi, K.; Mase, T. J . Med. Chem.
2000, 43, 5017-5029. (b) Mitsuya, M.; Ogino, Y.; Kawakami, K.;
Uchiyama, M.; Kimura, T.; Numazawa, T.; Hasegawa, T.; Ohtake, N.;
Noguchi, K.; Mase, T. Bioorg. Med. Chem. 2000, 8, 825-832. (c)
Mitsuya, M.; Mase, T.; Tsuchiya, Y.; Kawakami, K.; Hattori, H.;
Kobayashi, K.; Ogino, Y.; Fujikawa, T.; Satoh, A.; Kimura, T.; Noguchi,
K.; Ohtake, N.; Tomimoto, K. Bioorg. Med. Chem. 1999, 7, 2555-2567.
(d) Mitsuya, M.; Kawakami, K.; Ogino, Y.; Miura, K.; Mase, T. Bioorg.
Med. Chem. Lett. 1999, 9, 2037-2038.
(3) Tozer, M. J .; Herpin, T. F. Tetrahedron. 1996, 52, 8619-8683.
(4) (a) Konas, D. W.; Coward, J . K. Org. Lett. 1999, 1, 2105-2107.
(b) Lal, G. S.; Pez, G. P.; Syvret, R. G. Chem. Rev. 1996, 96, 1737-
1755.
(5) Seebach, D.; Naef, R.; Carderari, G. Tetrahedron 1984, 40, 1313-
1324.

Synthesis of a Muscarinic Receptor Antagonist
J . Org. Chem., Vol. 66, No. 20, 2001 6777
Sch em e 3
in 85% yield as a single diastereomer. Unfortunately,
safety considerations precluded the use of pentane at
pilot plant scale and mandated the discovery of new
methodology for this condensation. Unfortunately, simple
replacement of pentane with safer and higher boiling
solvents gave rather poor ratios of 7:16, making the
isolation of the desired diastereomer challenging. There-
fore, we explored a chemical rather than azeotropic
means for removal of water in the acetal formation.
Reaction of (R)-mandelic acid (6) with triisopropyl
orthoformate in toluene at ambient temperature afforded
intermediate 15 as an equilibrium mixture with the
starting acid (15:6 ) 4:1). Distillation under vacuum
effectively removed i-PrOH, thus driving the reaction to
completion and producing 15 as a mixture of anomeric
isomers. Addition of pivalaldehyde and a catalytic amount
of TsOH-H2O (6 mol %) at ambient temperature afforded
7 and 16 in ratios exceeding 120:1 in favor of the desired
isomer. Trituration from n-heptane then gave 7 as a pure
diastereomer in 92% yield (Scheme 3). This impressive
stereochemistry appears to be kinetic in origin and will
be discussed in more detail elsewhere.
B. Syn th esis of th e Mich a el Ad d u ct 17. Over the
last two decades, study of the Michael reaction has been
intense since appropriate design and implementation of
the reaction affords a C-C bond and potentially two
contiguous chiral centers.9 The Heathcock group10 and
others,11 in a series of elegant papers, have systematically
addressed the mechanistic characteristics for the stereo-
chemical course. However, the methodologies to construct
contiguous stereogenic centers enantio- and diastereo-
selectively in one operation are limited.9,14 Furthermore,
the general strategy for complex substrates targeting
pharmaceutical drugs and natural products is still to be
established due to the fact that the stereoselectivity
significantly depends on effects of enolate counterions,
additives, solvents, and temperature. Therefore, we first
screened representative metal enolates to find an ap-
propriate metal counterion for our system (Table 1). As
illustrated in Table 1, in all cases, as expected, the facial
selectivity at the quarternary center was excellent (less
than 10% 17b was generated in all reactions).12 It was
also encouraging that the competing 1,2-addition to
cyclopentenone 5 was not observed. However, the intrin-
sic diastereoselectivity of the reaction, as observed with
Li, Na, K, Zr enolates, all favored the undesired isomer
17a .13 In an attempt to investigate the effects of nonco-
ordinating or coordinatively saturated metals, the silyl
ketene acetal (en ol-7; M ) SiR3) was synthesized and
reacted with 6 in the presence of Lewis acids.10a,14
Interestingly, the results of this approach were unsatis-
factory (entries 8-11), as the maximum ratio obtained
was a marginal 2:1 favoring the desired compound. The
best selectivity (17:17a ) was achieved by the use of a Zn-
enolate (entry 7), which gave a 3:1 ratio, favoring the
desired product 17.15 It is intriguing to compare the
difference between entries 4 and 5, both of which arise
from Zr-enolates. With plain ZrCl4, an unfavorable ratio
of 1:3.2 was obtained, and with the use of Cp2ZrCl2, the
ratio improved to 2.8:1. While it is not surprising that a
conformational change in the Zr-enolate, by the introduc-
tion of the Cp ligand, could result in a different diaste-
reoselectivity, the magnitude of the difference (ca. 9-fold)
is quite significant. Thus, since the Zn-enolate gave the
best selectivity in the screening experiments, we focused
our efforts on the use additive ligands to further improve
the selectivity.
(9) (a) Schmalz, H.-G. Comprehensive Organic Synthesis; Semmel-
hach, M.; Trost, B. M., Eds.; Pergamon Press: Oxford, 1991; vol. 4, pp
199-236. (c) Evans, D. A.; Bilodeau, M. T.; Somers, T. C.; Clardy, J .;
Cherry, D.; Kato, Y. J . Org. Chem. 1991, 56, 5750-5752. (d) Payne, S.
G.; Dong, G.; Skelton, G. W.; White, A. H. J . Org. Chem. 1997, 62,
2337-2343. (e) Yasuda, K.; Shindo, M.; Koga, K. Tetrahedron Lett.
1997, 38, 3531-3534. (f) Aitken, R. A.; Thomas, A. W. Synlett 1998,
102-104. (g) Yoshizaki, H.; Tanaka, T.; Yoshii, E.; Koizumi, T.; Takeda,
K. Tetrahedron Lett. 1998, 39, 47-50. (h) Corey, E. J .; Noe, M. C.; Xu,
F. Tetrahedron Lett. 1998, 39, 5347-5350. (i) Corey, E. J .; Houpis, I.
N. Tetrahedron Lett. 1993, 34, 2421-2424.
The Zn-enolate was generated by addition of a sieve-
dried solution of ZnCl2 in THF to the Li-enolate at <-35
°C (eq 1). Addition of 1 equiv of TMEDA had no effect on
(12) For a recent example see: Visser, T. J .; Vanwaarde, A.; J ansen,
T. J . H.; Visser, G. M.; Vandermark, T. W.; Kraan, J .; Ensing, K.;
Vaalburg, W. J . Med. Chem. 1997, 40, 117-124.
(13) Stereochemical assignments for 17 and 17a : see refs 3b and
31. For 17b, it was determined by the fact that the free hydroxy acid
derived from 17b was identical with 17 in 1H, 13C, MS, IR spectral
data. The diastereomer of 17b, which should be defined as 17c, appears
at a very small level (below 1%). It was not discussed unnecessarily.
(14) For recent work on Mukaiyama-Michael reaction between silyl
enol ethers and R,â-unsaturated ketones: see (a) Mukaiyama, T.;
Tamura, M.; Kobayashi, S. Chem. Lett. 1986, 1017-1020. (b) Evans,
D. A.; Willis, M. C.; J ohnston, J . N. Org. Lett. 1991, 1, 865-868. (c)
Berl, V.; Helmchen, G.; Preston, S. Tetrahedron Lett. 1994, 35, 233-
236.
(10) Heathcock, C. H.; Norman, M. H.; Uehling, D. E. J . Am. Chem.
Soc. 1985, 107, 2797-2799. (b) Heathcock, C. H.; Oare, D. A. J . Org.
Chem. 1985, 50, 3022-3024. (c) Heathcock, C. H.; Henderson, M. A.;
Oare, D. A.; Sanner, M. A. J . Org. Chem. 1985, 50, 3019-3022. (d)
Oare, D. A.; Henderson, M. A.; Sanner, M. A.; Heathcock, C. H. J . Org.
Chem. 1990, 55, 132-157. (e) Oare, D. A.; Heathcock, C. H. J . Org.
Chem. 1990, 55, 157-172. (f) Stafford, J . A.; Heathcock, C. H. J . Org.
Chem., 1990, 55, 5433-5434.
(11) For
a
theoretical study see: Bernardi, A.; Capelli, A. M.;
(15) Unfortunately in this case there is some erosion of selectivity
at the quaternary center and 17b is produced in substantial amounts
(ca. 20%).
Comotti, A.; Gennari, C.; Scolastico, C. Tetrahedron Lett. 1991, 32,
823-826.

6778 J . Org. Chem., Vol. 66, No. 20, 2001
Ta ble 1. Rea ction of En ola tes of Ma n d elic Acid Aceta l 7 w ith 2-Cyclop en ten -1-on e (5)
Mase et al.
entrya
M
Lewis acid
solvent
THF
THF
THF/PhCH3
THF
THF
THF
THF
ratio of 17:17a f
yield, %f
1
2
3
4
5
6
7
8
Li
Na
K
Li
Li
Li
Li
-
-
-
1:1.5
1:2
89
92
92
90
88
90
75
93
21
98
93
1:2.5
1:3.2
2.8:1
1.5:1
3:1
2.1:1
1.9:1
1.9:1
1.3:1
d
ZrCl4
Cp2ZrCl2
Cp2TiCl2
ZnCl2
e
TBSc
TBS
TBS
TBS
TiCl4
CH2Cl2
CH2Cl2
THF
9b
10
11
Cp2ZrCl2
ZrCl2
ZrCl4
CH2Cl2
a
All reactions were carried out at -78 °C unless otherwise mentioned. bThe reaction was carried out at 25 °C. cThe tert-butyl dimethyl
silyl (TBS) derivatives were generated from the Li enolate in THF followed by addition of TBS-Cl. The TMS derivetives were also preparared
and gave similar results. dMetal enolates were prepared by using 1.2 equiv of additives to the lithium enolate. eThe enone and Lewis acid
were premixed in the appropriate solvent at the reaction temperature, followed by addition of a solution of the preformed silyl ketene
f
acetal. Ratios and yields were determined by HPLC.
the selectivity of the reaction; however, additional amounts
of TMEDA showed a substantial increase in the selectiv-
ity with a maximum ratio of 17:17a ) 8.6:1 in the
presence of 20 equiv of TMEDA.16 Additional TMEDA (50
equiv) did not further enhance the selectivity. The fact
that as much as 20 equiv of TMEDA was needed to secure
the optimal diastereoselectivity strongly suggested sol-
vent participation in the reaction. Since it is well-known
that THF can coordinate strongly with Zn-enolate, it was
postulated that a less-coordinating solvent might allow
reduction of the necessary amount of ligand with en-
hancement of the selectivity. Therefore, we successively
focused on optimization of the solvent system. A number
of polar, nonpolar, ethereal, and hydrocarbon solvents
were examined and interestingly, our findings suggested
that a substantial proportion of a nonchelating hydro-
carbon solvent to a chelating solvent was necessary for
demonstration of the additive effect of the ligand. The
optimal combination was found to be DME:toluene (1:1,
wherein the maximum selectivity was obtained with only
4 equiv of TMEDA (vs 20 equiv in THF, 17:17a ) 11:1,
17:17a +17b ) 8.4:1; reaction yield ) 95%).
Finally, our optimization studies focused on the struc-
ture of the ligands (Table 2). Initial attempts with
polydentate amine ligands, diol alkoxides,17 aminoalkox-
ides,18 and bis-amide and bis-sulfonamide anions were
not effective, and low yields of 17 and poor ratios (17:
17a +17b) were obtained. However, to our great delight,
unsymmetrical diamine ligands significantly increased
the diastereoselectivity of this reaction. Indeed, merely
substitution of one of the dimethyl amino groups in
TMEDA with the diethyl amino moiety resulted in a
significant increase (entry 3) in the diastereomer ratio.
Furthermore enclosing the nitrogen in a ring (entries
4-6) further increased the selectivity in favor of the
desired isomer. It is also worth noting that the largest
effect observed is on the facial selectivity of the cyclo-
pentenone (as expressed in the ratio of 17:17a ) while the
face selectivity with respect to the enolate remains fairly
constant (17+17a :17b). Thus, with TMEDA the ratio of
17:17a is 11:1 while use of the piperazinyl-derivative 18
(entry 5) resulted in an increased ratio of 17:17a to ca.
45:1. Interestingly, a symmetrical open-chain analogue
did not perform satisfactorily in this reaction (entry 7).
From the results in Table 2 we chose 1-(2-dimethylami-
noethyl)-4-methylpiperazine (18, MAEP) as the preferred
ligand for this reaction. Fortunately, this ligand was also
the cheapest and most readily available.19
(16) With 5 equiv of TMEDA, the ratio is 5:1 and with 10 equiv of
TMEDA, the ratio is 7:1.
(17) (a) Sasai, H.; Arai, T.; Shibasaki, M. J . Am. Chem. Soc. 1994,
116, 1571-1572. (b) Bolin, C.; Ewald, M. Tetrahedron Lett. 1990, 31,
5011-5012. (c) Asano, Y.; Iida, A.; Tomioka, K. Tetrahedron Lett. 1997,
38, 8973-8976.
(18) (a) Huffman, M. A.; Yasuda, N.; DeCamp, A. E.; Grabowski, E.
J . J . J . Org. Chem. 1995, 60, 1590-1594. (b) Thompson, A. S.; Corley,
E. G.; Huntington, M. F.; Grabowski, E. J . J . Tetrahedron Lett. 1995,
36, 8937-8940.
With satisfactory stereoselectivity for the Michael
addition in hand, we modified this reaction condition to
fit for the pilot plant implementation. Anhydrous ZnCl2
is very hygroscopic and is not soluble in either DME or
(19) Ligand 18 is commercially available from TOSOH Co.

Synthesis of a Muscarinic Receptor Antagonist
J . Org. Chem., Vol. 66, No. 20, 2001 6779
Ta ble 2. Effect of th e Str u ctu r e of th e Zn Liga n d on th e Dia ster eoselectivity of th e Mich a el Rea ction of 7 a n d
2-cyclop en ten -1-on e (eq 1)
toluene. Further experimentation revealed that prefor-
mation of the complex 1920 is operationally advantageous
as the complex is not hygroscopic and thus can be used
directly in the water sensitive Michael reaction.21 How-
ever, the reaction with 19 was heterogeneous and sig-
nificant amount of solids were observed, and the solid
was found to be the unreacted 19 itself.
reaction of the lithium enolate of 7 with cyclopentenone
5 was a very fast reaction (complete in e3 min). To
explain these results, a detailed kinetic study was
initiated (the full data to be presented elsewhere) with
preliminary indications showing that the “Zn enolate”
reacts 7-10 times faster (in the presence of 18) than the
lithium enolate (with or without 18).
Our typical procedure for this reaction is shown in
Scheme 4. A mixture of LDA and 18 (4 equiv) in DME at
-15 °C was treated with a solution of 7 at -15 to -10
°C. The resulting enolate, which is stable for several days
at this temperature, was treated with 19 (15 mol %),23
and toluene was added. The resulting heterogeneous
mixture was allowed to stand for 1 h at 0 °C upon which
time 19 completely dissolved to give a homogeneous
reaction mixture. It should be noted that this last
operation is critical, and failure to perform it results in
diminished selectivity (17:17a + 17b e 4:1). The mixture
was then cooled to -78 °C (where it remained homoge-
neous), and cyclopentenone was added over 1 h. Acid
quench and extractive workup afforded 17 in 83% assay
yield. Finally, the product was isolated from EtOH:H2O
(1:1) in 74% isolated yield with excellent chemical and
diastereomeric purity on a 25 kg scale.
C. Diflu or in a tion of th e Mich a el Ad d u ct 17. In our
initial attempts to convert 17 to the gem-difluoro com-
pound 21, we employed several literature deoxyfluori-
nation procedures,24 oxidative fluorination of ketone
derivatives25 and SF4-HF.26 Although some success was
realized with these methods, safety and reproducibility
concerns prompted us to investigate alternative ap-
proaches.27
We therefore examined the possibility of using less
than 1 equiv of the ZnCl2-MAEP complex 19. Interest-
ingly, only 15 mol % of 19 was needed to achieve an equal
stereoselectivity and yield. This very interesting finding
translated to a reduction in the quantities of 19 employed
in the reaction which in turn produced a homogeneous
reaction mixture, reduced raw materials cost, facilitated
reaction workup, and provided a more environmentally
friendly waste stream. Further detailed investigation
indicated that the lithium enolate of 7 was significantly
more stable in the presence of 18,22 and the Zn-MAEP
complex 19 could act as an catalyst. In contrast, the
(20) The complex 19 was formed easily by adding 18 (1.2 equiv) to
a solution of ZnCl2 in THF (Aldrich) at 0 °C. A precipitate formed
immediately and was filtered and worked with THF and dried at 40
°C in vacuo to afford 19 in 85-90% yield. An X-ray structure of 19
revealed that 18 acts as a bidentate ligand with the distal ring nitrogen
not bound to Zn.
(21) In cases when ZnCl2 in THF was used, the solution was dried
overnight over 4 Å molecular sieves. Studies showed that under these
conditions there is substantial change in the chloride titration indicat-
ing some instability of this solution on storage.
Fortuitously, Air Products Inc. had just announced the
commercialization of a new, safer, and more reactive
(22) The lithium enolate in the presence of MAEP (t1/2 ) 24 h at
-10 °C is more stable than the lithium enolate itself (t1/2 ) 12 h at
-35 °C) in DME-toluene.
(23) The ratio of 17:17a :17b ) 90:2:8 is the same when 1 equiv of
19 is used.
(24) Middleton, W. J . J . Org. Chem. 1975, 40, 574-578.

6780 J . Org. Chem., Vol. 66, No. 20, 2001
Mase et al.
Sch em e 4
Ta ble 3. Effect of Ad d itives to th e Diflu or in a tion
Rea ction of 17 w ith Deoxoflu or
reagent for the conversion of carbonyls to gem-difluoro
derivatives: Deoxofluor (20). Differential scanning cal-
yield (%)
temp (°C) time (h)c 21 22 23 24
entry
additives
1
2
3
4
none1
40
40
40
55
24
16
14
36
80 1.5 1.5 5
H2O (20%)a
85 0.7 0.7 2.0
83 0.7 0.7 2.5
92 0.3 0.3 0.4
CF3CO2H (20%)a
BF3‚OEt2 (5%)b
a
2.5 equiv of Deoxofluor were required for these reaction
conditions. b1.4 equiv of Deoxofluor were used. cTime to >95%
conversion.
these impurities could not be easily removed at this or
later stages in the synthesis.
To overcome these problems, we investigated reports
in the literature that protic acids accelerate the deoxy-
fluorination reaction and thus may allow reduction of
equivalents of Deoxofluor required. Hopefully, the in-
crease in the rate of the desired reaction would also
generate less of the impurities 22-24. Addition of H2O
(20 mol %) to the reaction mixture, to generate HF in
situ, or an anhydrous protonic acid (trifluoroacetic acid,
20 mol %) produced only a marginal rate increase (Table
3, entries 2, 3). More importantly these additives reduced
the level of impurities 22-24 formed to approximately
half of the amounts produced in the original reaction.
To our surprise, even small amounts of (5-10 mol %)
of Lewis acids such as BF3‚OEt229 substantially retarded
the rate of the reaction (70% conversion in 36 h at 40 °C
and 95% conversion in 36 hat 55 °C). However, the yield
of 21 improved by 10-12% (entry 4) and only trace
amounts of 22-24 were formed; moreover, 1.4 equiv of
Deoxofluor were sufficient to drive the reaction to comple-
tion. A typical procedure (eq 3) involved slow addition of
BF3‚OEt2 to a cold (0 °C) solution of 20 in toluene followed
by allowing the mixture to stand for 1 h at 0 °C. A
solution of 17 in toluene was added over 3 h, and the
mixture was heated to 55 °C for 36 h. Careful quench
orimetry data indicated this compound might be safer
than DAST or morpholino-DAST.28 Moreover, when 17
was reacted with 20 (2.5 equiv) in 1,2-dichloroethane at
40 °C, complete conversion to the difluoroacetal 21 was
realized in ca. 24 h (80% assay yield). In comparison,
temperatures of ca. 90 °C were required with DAST in
order to obtain similar results. Despite the good yield of
21, two issues remained problematic. First the require-
ment for 2.5 equiv dramatically increased the cost of this
step, and second, three major impurities were generated
(22-24).
For the reaction described above (eq 2), 1.5% each of
22 and 23 as well as ca. 5% of 24 were formed. Moreover,
(25) (a) Kuroboshi, M.; Hiyama, T. Synlett. 1991, 909-910. (b)
Sondej, S. C.; Katzenellenbogen, I. A. J . Org. Chem. 1986, 51, 3508-
3513. (c) Motherwell, W. B.; Wilkinson, J . A. Synlett. 1991, 191-192.
(d) Rozen, S.; Brand, M.; Zamir, D.; Hebel, D. J . Am. Chem. Soc. 1987,
109, 896-897. (e) Rozen, S.; Zamir, D. J . Org. Chem. 1991, 56, 4695-
4700. (f) York, C.; Prakash, G. K. S.; Wang, A.; Olah, G. A. Synlett
1994, 425-426.
(26) Hammond, G. B.; Plevey, R. G. J . Fluorine Chem. 1993, 63, 13
and references therein.
(27) Messina, P. A.; Mange, K. C.; Middleton, W. J . J . Fluorine
Chem. 1989, 42, 137-143.
(28) Lal, G. S.; Pez, G. P.; Pesaresi, R. J .; Prozonic, F. M.; Cheng,
H. J . Org. Chem. 1999, 64, 7048-7054.

Synthesis of a Muscarinic Receptor Antagonist
J . Org. Chem., Vol. 66, No. 20, 2001 6781
into a NaOH solution, while constantly maintaining basic
pH, followed by extractive workup afforded 21 in ca. 92%
assay yield.
prepare the bromomethyl pyridine (8: R ) CH2Br) from
2-amino-6-picoline (14), radical bromination of 2-Boc-
amino-6-picoline with NBS/AIBN was examined.32 How-
ever, considerable amounts of the corresponding dibromo
side product were obtained, and it was difficult to
suppress the undesired path even after many attempts.
We therefore turned our attention to selective formylation
of 2,6-dibromopyridine (13) for the reductive amination
protocol with a protected amino piperidine.
A. Syn t h esis of 6-Br om o-2-for m ylp yr id in e (26).
Selective monometalation of 2,6-dibromopyridine (13) is
a useful way to produce diversely functionalized pyridine
derivatives and thus has received considerable attention.
Lithium-bromide exchange of 13 with 1.00 equiv of BuLi
proceeded cleanly in THF at -78 °C to give the mono-
lithiated compound, and addition of DMF followed by
aqueous workup afforded the product 26 in excellent yield
(eq 5).33
D. Hyd r olysis of th e Aceta l 21. Hydrolysis of crude
21 to give the key acid fragment 2 was accomplished
conveniently under basic conditions (eq 4). Treatment of
However, cryogenic temperatures (e-70 °C) and pin-
point accuracy of the BuLi charge were essential for the
success of the above metalation/formylation reaction.33
The latter was particularly crucial and was also the most
difficult parameter to control on scale-up (Scheme 5). The
need for perfect control of the BuLi charge lies in the fact
that exposure of the monolithiated species 27 to excess
BuLi readily affords the bis-lithiated species 28 which
eventually produces 29. On the other hand, a slight
deficiency in BuLi results in reaction of 27 with unreacted
13 to give the ortho-lithiated product 30 and 2-bromo-
pyridine (31). Reaction of 30 with DMF affords 32.
the deep brown solution of 21 in MeOH with 5 equiv of
aqueous LiOH (1 M) at 45 °C effected a rapid metha-
nolysis of the acetal to produce the methyl ester and a
subsequent slow hydrolysis (45 °C, 20 h) to give the
desired hydroxy acid 2. Compound 2 was treated with
dicyclohexylamine (DCHA) to give the crude DCHA salt
25 which was recrystallized from MeOH-H2O (1:1) to
give pure 25 in 75% yield from 21.30 The purity of this
compound by HPLC analysis was g99%.
It is noteworthy that we have been unable to effect the
hydrolysis of 21 under acidic conditions even though the
starting acetal 7 was fairly easily hydrolyzed in acidic
methanol solutions. Heating of 21 with H2SO4 (2 M) in
MeOH or EtOH at reflux resulted in recovery of the
starting material (97%).31
34
Interestingly, replacement of THF with CH2Cl2 or
35
PhCH3 allowed the use a slight excess of BuLi (1.05-
1.1 equiv). However, cryogenic conditions were still
necessary as intermediate 27 was not stable at temper-
atures above -50 °C.
To prepare a more stable intermediate, we next focused
on the magnesium-bromine exchange reaction. Magne-
sium-halogen exchange reactions are also well estab-
lished,36 and Que´guiner et al have recently reported the
magnesium-bromine exchange of 2,6-dibromopyridine
(13) using i-PrMgCl.37 On the basis of this procedure, we
aggressively refined the reaction conditions. However,
(32) (a) Maris, V. P.; Goswami, S.; Hamilton, A. D. J . Heterocycl.
Chem. 1995, 32, 675-681. (b) Goswami, S.; Hamilton, A. D.; Engen,
D. V. J . Am. Chem. Soc. 1989, 111, 3425-3426.
(33) Cai, D.; Hughes, D. L.; Verhoeven, T. R. Tetrahedron Lett. 1996,
37, 2537-2540.
II. Am in e Syn th esis
Since a number of attempts6,7 to effect coupling of the
iminium cation species with metallopyridines gave un-
satisfactory results (Scheme 2), we focused our attention
on the C-N retrosynthesis path (b), Scheme 2. To
(34) Peterson, M. A.; Mitchell, J . R. J . Org. Chem. 1997, 62, 8237-
8239.
(35) Unpublished result from our laboratory: The exchange reaction
with 2,6-dibromopyridine occurred very cleanly below -40 °C.
(36) (a) Furukawa, N.; Shibutani, T.; Fujihara, H. Tetrahedron Lett.
1987, 28, 5845-5848. (b) Nishiyama, H.; Isaka, K.; Itoh, K.; Ohno, K.;
Nagase, H.; Matsumoto, K.; Yoshiwara, H. J . Org. Chem. 1992, 57,
407-410. (c) Shinokubo, H.; Miki, H.; Yokoo, T.; Oshima, K.; Utimoto,
K. Tetrahedron 1995, 11681-11692. (d) Be´rillon, L.; Lepreˆtre, A.;
Turck, A.; Ple´, N.; Que´guiner, G.; Cahiez, G.; Knochel, P. Synlett 1998,
1359-1360. (e) Boymond, L.; Rottla¨nder, M.; Cahiez, G.; Knochel, P.
Angew. Chem., Int. Ed. 1998, 37, 1701-1703. (f) Rottla¨nder, M.;
Boymond, L.; Cahiez, G.; Knochel, P. J . Org. Chem. 1999, 64, 1080-
1081. (g) Abarbri, M.; Knochel, P. Synlett. 1999, 1577-1578.
(29) Other Lewis acids studied include AlCl3; EtAlCl2; Et2AlCl,
Ti(OPri)4; and GaCl3. BF3‚OEt2 was optimal for reduction of impurities
22-24.
(30) The dicyclohexylamine salt was optimal for isolation.
(31) For earlier synthetic work from our medicinal chemistry
division, see: Mitsuya, M.; Ogino, Y.; Ohtake, N.; Mase, T. Tetrahedron
2000, 56, 9901-9907.

6782 J . Org. Chem., Vol. 66, No. 20, 2001
Mase et al.
Sch em e 5
even after optimization, the reaction was relatively
sluggish and required excess i-PrMgCl for completion.
In addition, formation of the side product 2-formyl-6-
isopropylpyridine after treatment with DMF could not
be ignored (more than 5%). Interestingly, the metalated
mono-bromopyridine intermediate from i-PrMgCl was
quite stable at 20 °C even after standing for a long time
(>24 h). It should also be noted that Knochel et al. have
reported the i-Pr2Mg-induced exchange reaction of 2,6-
dibromopyridine (13) (albeit a sluggish reaction).38
These results prompted us to devise a new metalation
protocol that addressed both issues of low temperature
and stoichiometry. Ideally, this solution requires that the
desired metalated species should have stability similar
to the RMgX and R2Mg species; however, its reactivity
should be greater. With this in mind, we envisioned that
a magnesium ate complex such as R3MgLi, would possess
the necessary reactivity difference, between BuLi and
RMgX/R2Mg, and might also exhibit the requisite stabil-
ity profile. To our best knowledge, no report on a
magnesium ate complex induced metal-halogen ex-
change reaction had been published at that time, al-
though several examples with other metal-based ate
complexes are known.39 Furthermore, in all cases re-
ported, only one alkyl on the metal was active in the
exchange reaction.
convenient and safer alternative method. Indeed, the
reagent prepared from mixing n-BuMgCl (0.35 equiv to
13) and n-BuLi (0.70 equiv to 13) in a ratio of 1:2 in
toluene at -10 °C showed similar reactivity to the one
prepared from n-Bu2Mg. Treatment of the above pre-
pared, R3MgLi species with 13, in toluene, resulted in
complete consumption of the starting dibromopyridine,
indicating conversion to the desired mono-bromopyridine
derived ate complex 33 and upon subsequent quenching
yielded 2-bromopyridine (31), quantitatively. Analo-
gously, when the reaction mixture was treated with DMF
at -10 °C, the desired 6-bromo-2-formylpyridine (26) was
formed in 94-97% assay yield (eq 6).
After several attempts, we were pleased to find that
the halogen-metal exchange reaction of 13 cleanly pro-
ceeded at -10 °C in toluene with only 0.35 equiv of
n-Bu3MgLi (prepared by mixing n-Bu2Mg and n-BuLi
according to the literatures).40 However, since it is known
that n-Bu2Mg can be prepared from n-BuMgCl and
n-BuLi,41 we chose to prepare the speculated ate complex
by direct reaction of n-BuMgCl and n-BuLi as a more
This remarkable metal-bromine exchange at higher
temperature did not occur cleanly with either of the
components alone. Extensive decomposition occurred
with n-BuLi at -10 °C, and n-BuMgCl and n-Bu2Mg were
much less active. In addition, this exchange is not
sensitive to the accuracy of the reagent charge, unlike
the original lithium-bromine exchange. The combination
of various amounts of n-BuMgCl and n-BuLi were
comparable to the typical procedure (0.35/0.7equiv) and
afforded 26 in 96% yield.
(37) Tre´court, F.; Breton, G.; Bonnet, V.; Mongin, F.; Marsais, F.;
Que´guiner, G. Tetrahedron Lett. 1999, 40, 4339-4342.
(38) Abarbri, M.; Dehmel, F.; Knochel, P. Tetrahedron Lett. 1999,
40, 7499-7453.
(39) Ate complex-induced halogen-metal exchange reactions: (a)
Harada, T.; Katsuhira, T.; R. Hattori, K.; Oku, A. J . Org. Chem. 1993,
58, 2958-2965. (b) Harada, T.; Katsuhira, T.; Hara, D.; Kotani, Y.;
Maejima, K.; Kaji, R.; Oku, A. J . Org. Chem. 1993, 58, 2958-2965. (c)
Uchiyama, M.; Koike, M.; Kameda, M.; Kondo, Y.; Sakamoto, T. J . Am.
Chem. Soc. 1996, 118, 8733-8734. (d) Kondo, Y.; Matsudaira, T.; Sato,
J .; Murata, N.; Sakamoto, T. Angew. Chem., Int. Ed. Engl. 1996, 35,
736-738. (e) Inoue, R.; Shinokubo, H.; Oshima, K. Tetrahedron Lett.
1996, 37, 5377-5380. (f) Inoue, R.; Shinokubo, H.; Oshima, K. J . Org.
Chem. 1998, 63, 910-911. (g) Hojo, M.; Harada, H.; Ito, H.; Hosomi,
A. J . Am. Chem. Soc. 1997, 119, 5459-5460. (h) Hojo, M.; Hosomi, A.
Chem. Commun. 1997, 2077-2078.
(40) (a) Malpass, D. B.; Eastham, J . F. J . Org. Chem. 1973, 38, 3718.
(b) Ashby, E. C.; Chao, L.-C.; Laemmle, J . J . Org. Chem. 1974, 39,
3258.
(41) For preparation of n-Bu2Mg: see Wakefield, B. J . Organomag-
nesium Methods in Organic Synthesis; Academic Press: New York,
1995; pp 61-71.
In conclusion, this novel, n-Bu3MgLi metal-halogen
exchange reaction provides a robust, scaleable process
for selective functionlization of 2,6-dibromopyridine.42
Detailed studies using n-Bu3MgLi and related complexes
will be reported elsewhere.
(42) A patent application was filed: Iida, T.; Wada, T.; Mase T.,
J apan Application No. J P 2000-024613 20000202. During the prepara-
tion of this manuscript, an analogous metal-halogen exchange using
R3MgLi has been published: Kitagawa, K.; Inoue, A.; Shinokubo, H.;
Oshima, K. Angew. Chem., Int. Ed. 2000, 39, 2481-2483. They used
3 equiv of the organometallic reagents in most cases.

Synthesis of a Muscarinic Receptor Antagonist
J . Org. Chem., Vol. 66, No. 20, 2001 6783
Sch em e 6
B. Red u ctive Am in a tion Rea ction to for m 36
including classical copper-catalyzed amination with am-
monia exsist, they are mechanistically ill defined and
generally employ harsh or inconvenient conditions.46,47
Several ingenious procedures have been developed to
introduce variably substituted amine surrogates.48 Among
them, benzophenone imine (37) appears to be optimal due
to its facile hydrolysis to NH2,49 despite economic and
volumetric productivity issues. We examined several
more available ammonia equivalents using Pd, Ni, or Cu
based catalysts without much success. Consequently,
benzophenone imine (37) was used for the amination
(Scheme 6).
The preparation of the key bromopyridine intermediate
36 was accomplished by reductive amination of aldehyde
26 with readily available 4-Boc-aminopiperidine 3543 (eq
7). The aldehyde and piperidine acetic acid salt were
44
mixed in toluene:THF(1:1), and NaHB(OAc)3 was in-
troduced slowly as a DMSO solution over 3 h. The crude
reaction mixture was partitioned between toluene and
water and crystallized to give 36 in 82% overall yield
(from 13).
Typically, bromopyridine 36 in toluene was treated
with 0.25 mol % of Pd(OAc)2 and 0.5 mol % of diphen-
ylphosphino ferrocene (dppf) followed by 1.1 equiv of
t-BuONa and the imine 37. The mixture was heated to
80 °C for 1 h, n-Bu3P was added,50 and the toluene
solution was stirred with aqueous citric acid to effect
hydrolysis of the imine of 38 while leaving the Boc
protecting group intact. The benzophenone byproduct and
most of the palladium remain in the organic layer and
are easily removed. The product 39 was isolated from the
aqueous layer by basification and extraction. Crystal-
lization from i-PrOAc/n-heptane (1:2.5) afforded 39 in
86% yield.
D. Boc-Dep r otection Rea ction To F or m 3. Depro-
tection of the Boc group was readily achieved by treat-
ment of 39 with excess anhydrous (or aqueous concen-
trated) HCl in methanol at 40 °C for 4 h. The tris-HCl
salt thus produced was isolated by removal of the volatile
components followed by crystallization from MeOH:t-
BuOMe (1:2) in 92% yield and 99.9% purity (eq 8).51
C. Ar om a tic Am in a tion To F or m 39. Amination of
aromatic halides under transition metal catalysis has
received considerable attention in the past few years,45
and a bigger challenge involves the introduction of the
unsubstituted “NH2” group. Although direct methods
(43) The piperidine moiety was readily synthesized from com-
mercially available N-benzyl amino piperidine (12). Protection of the
primary amine as the carbamate was accomplished by brief exposure
(1 h) of 12 to Boc2O in THF/MeOH at 10-15 °C to give the intermediate
34 in nearly quantitative yield.
III. F in a l Cou p lin g
Heterogeneous catalytic hydrogenolysis of the benzyl group in 34 at
the free amine stage proved problematic requiring 50 wt % of Pd/C
and long reaction times. However, addition of 1 equiv of acetic acid to
the crude solution of 34 (in MeOH/THF) followed by hydrogenation
with 5 wt % of Pd/C at 40 psi H2, afforded 35 in >95% crude yield.
After filtration through Solka-Floc, to remove the catalyst, and
evaporation of the solvent, 35 was isolated in pure enough form to be
used directly in the next step.
A streamlined protocol for coupling of the dicyclohexyl-
amine salt 25 was also developed. A solution of 25 in H2O
(44) Abdel-Magid, A. F.; Carson, K. G.; Harris, B. D.; Maryanoff, C.
A.; Shah, R. D. J . Org. Chem. 1996, 61, 3849-3862.

6784 J . Org. Chem., Vol. 66, No. 20, 2001
Mase et al.
is treated with 3 equiv of NaOH and extracted with
heptane. The organic phase, containing free dicyclohexy-
lamine, is removed, and the remaining aqueous solution
of the sodium salt of 2 (an extra 2 equiv of base) is treated
with 3 in CH3CN. The excess NaOH neutralizes two
equiv of acid from 3, and the mixture is treated with
HOBT and EDC. The resulting homogeneous mixture is
stirred for 16 h at ambient temperature to produce 1 in
80-85% yield after extractive workup and i-PrOAc/n-
heptane crystallization (eq 9).
novel zinc-MAEP complex was demonstrated to catalyze
a selective lithium enolate Michael reaction (which was
itself inactivated by MAEP) to produce the desired adduct
with high diastereoselectivity (this finding is in contrast
to the usual case where a transition-metal catalytically
activates inactive Zn-containing organometallics), (3)
Deoxofluor/BF3‚OEt2 was demonstrated as a highly ef-
ficient and selective reagent for the conversion of func-
tionalized ketones to their corresponding geminally
difluorinated derivatives (this reagent proved to be
superior to DAST in suppression of vinyl fluoride side
product formation), (4) n-Bu3MgLi was shown to be a
highly effective reagent for selective metal-halogen
exchange of 2,6-dihalopyridines and obviated the need
for cryogenic reaction conditions (as required with the
alkyllithium reagents). This approach should be widely
applicable to other aryl or vinyl halides. This work
enabled us to provide multi-kilograms of highly pure bulk
drug without the use of chromatographic purification.
Con clu sion
A highly efficient and convergent synthesis of drug
candidate 1 was accomplished in excellent overall yield.
The synthesis is highlighted by the following transforma-
tions: (1) a triisopropyl orthoformate-induced acetaliza-
tion was used as an alternative to the usual azeotropic
approach to give the desired mandelic acid derived
dioxolane with unprecedented stereoselectivity, (2) a
Exp er im en ta l Section
Aceta l F or m a tion : (2R,5R)-2-ter t-Bu tyl-5-p h en yl-1,3-
d ioxola n -4-on e (7). A slurry of (R)-mandelic acid (6) (7.00
kg) in toluene (70 L) was treated with triisopropyl orthofor-
mate (10.51 kg) added dropwise at 24 °C. The batch was stirred
at 24 °C for 2 h, until NMR monitoring showed a ratio of 15:6
of 82:18). Toluene (70 L) was added (to 170 L total volume),
and the batch was distilled maintaining a constant volume of
170 L by gradual addition of toluene. NMR analysis showed
complete consumption of 6 at this time. The batch was
concentrated to 70 L at 16 °C under vacuum. Gas chromato-
graphic analysis showed <1.5% of i-PrOH. p-Toluenesulfonic
acid (6 mol %, 525 g) and a solution of pivalaldehyde (5151 g)
in toluene (35 L) were added at 25-28 °C over a period of 50
min, and the reaction mixture was stirred at 28 °C for 30 min.
After confirming the completion of the reaction, by HPLC
analysis, the batch was diluted with toluene (90 L) to 200 L
and was washed with 5% aqueous sodium bicarbonate solution
(70 L) and 30% aqueous NaCl solution. The organic layer was
concentrated in vacuo to a minimum stirrable volume, n-
heptane was added (50 L), and the batch was further concen-
trated to a volume of 25 L. The same procedure was repeated
three times to switch the solvent to n-heptane. The resulting
slurry was filtered, washed with n-heptane (2 × 25 L), and
dried under reduced pressure to give the desired compound 7
(9.3 kg, 92% yield, 98.8% purity).5,53
Mich a el Ad d ition : (2R,5R)-2-ter t-Bu tyl-5-[(1R)-4-oxo-
2-cyclop en tyl]-5-p h en yl-1,3-d ioxola n -4-on e (17). A 3000
mL three-neck flask was equipped with mechanical stirrer and
temperature probe. Dimethoxyethane (DME, 350 mL) and 18
(176 mL) were mixed at room temperature and cooled to -20
°C. A solution of LDA (2.0 M, 159 mL) was added over 10 min
while keeping the temperature below -10 °C. The resulting
solution was cooled to -20 °C, and a solution of 7 (50.0 g) in
DME (470 mL, heating to 35 °C is required for complete
dissolution) was added dropwise over 45 min while keeping
the temperature below -15 °C. The resulting solution was
allowed to stand for 15 min at -15 °C. The ZnCl2-MAEP
complex 19 (20.9 g) and toluene (820 mL) were added sequen-
tially. The temperature was raised to 0 °C, and the mixture
was allowed to stand for 2 h. The reaction mixture was then
cooled to -78 °C, and a solution of 2-cyclopenten-1-one (20.9
mL) in toluene (82 mL) was added dropwise over 50 min while
(45) Yang, B. H.; Buchwald, S. L. J . Organomet. Chem. 1999, 576,
125-146. (b) Hartwig, J . F. Angew. Chem., Int. Ed. 1998, 37, 2046-
2067. (c) Wolfe, J . P.; Wagaw, S.; Marcoux, J .-F.; Buchwald, S. L. Acc.
Chem. Res. 1998, 31, 805-818. (d) Hartwig, J . F. Acc. Chem. Res. 1998,
31, 852-860. (e) Hartwig, J . F. Synlett. 1997, 329-340.
(46) (a) Vedejs, E.; Trapencieris, P.; Suna, E. J . Org. Chem. 1999,
64, 6724-6729. (b) Lindley, J . Tetrahedron 1984, 40, 1433-1456. (c)
Paine, A. J . J . Am. Chem. Soc. 1987, 109, 1496-1502. (d) Ott, H.;
Hardtmann, G.; Denzer, M.; Frey, A. J .; Gogerty, J . H.; Leslie, G. H.;
Trapold, J . H. J . Med. Chem. 1968, 11, 777-787.
(47) Hori, K.; Mori, M. J . Am. Chem. Soc. 1998, 120, 7651-7652.
(48) (a) Wolfe, J . P.; A° hman, J .; Sadighi, J . P.; Singer, R. A.;
Buchwald, S. L. Tetrahedron Lett. 1997, 38, 6367-6370. (b) Mann, G.;
Hartwig, J . F.; Driver, M. S.; Ferna´ndez, -R. C. J . Am. Chem. Soc.
1998, 120, 827-828. (c) J aime-Figueroa, S.; Liu, Y.; Muchowski, J .
M.; Putman, D. G. Tetrahedron Lett, 1998, 39, 1313-1316. (d) Hartwig,
J . F.; Kawatsura, M.; Hauck, S. I.; Shaughnessy, K. H.; Alcazar-Roman,
L. M. J . Org. Chem. 1999, 64, 5575-5580.
(49) Use of allylamine or diallylamine as a surrogate was abandoned
due to sluggish cleavage and its toxicity. For the deallylations: see
(a) Garro-Helion, F.; Merzouk, A.; Guibe´, F. J . Org. Chem. 1993, 58,
6109-6113. (b) Lemaire-Audoire, S.; Savignac, M.; Dupuis, C.; Geneˆt,
J . P. Bull. Chim. Soc. Fr. 1995, 132, 1157-1166. (c) Lemaire-Audoire,
S.; Savignac, M.; Geneˆt, J . P.; Bernard, J .-M. Tetrahedron Lett. 1995,
36, 1267-1270. (d) Honda, M.; Morita, H.; Nagakura, I. J . Org. Chem.
1997, 62, 8932-8936. (e) Alonso, E.; Ramon, D. J .; Yus, M. Tetra-
hedron. 1997, 53, 14355-14366.
(50) Larsen, R. D.; King, A. O.; Chen, C. Y.; Corley, E. G.; Foster,
B. S.; Roberts, F. E.; Yang, C.; Lieberman, D. R.; Reamer, R. A.;
Tschaen, D. M.; Verhoeven, T. R.; Reider, P. J . J . Org. Chem. 1994,
59, 6391-6394.
(51) Alternatively, amine moiety 3 could also be prepared from
commercially available 2-amino-6-bromopyridine (40). Conversion to
pivalamide 41 (quantitative) and metal-halogen exchange with i-
PrMgCl followed by quenching with DMF provided the aldehyde 42
(87-90%). Reductive amination with N-acyl-protected piperidine 35b
gave 43 which was readily deprotected in the presence of acid to furnish
the desired tris-HCl salt amine coupling partner 3 (86%).
(52) Mach, R. H.; Luedtke, R. R.; Unsworth, C. D.; Boundy, V. A.;
Nowak, P. A.; Scripko, J . G.; Elder, S. T.; J ackson, J . R.; Hoffman, P.
L.; Evora, P. H.; Rao, A. V.; Molinoff, P. B.; Childers, S. R.; Ehren-
kaufer, R. L. J . Med. Chem. 1993, 36, 3707-3720.
(53) All spectral data (1H, 13C NMR, IR, HR-MS, elemental analysis,
melting point and optical rotation) obtained were identical with those
of authentic samples.

Synthesis of a Muscarinic Receptor Antagonist
J . Org. Chem., Vol. 66, No. 20, 2001 6785
keeping the temperature between -75 to -70 °C. After 20 min
HPLC analysis indicated that the reaction was complete. H2-
SO4 (4 M, 25 mL) was added over 5 min (the temperature of
the reaction mixture increased to about -50 °C), and the
reaction mixture was allowed to warm to -20 °C and trans-
ferred slowly into H2SO4 (4 M, 475 mL, precooled to -10 °C)
while controlling the temperature below 10 °C. The two-phase
mixture was allowed to warm to room temperature and was
stirred for 20 min. The layers were separated (the pH of
aqueous layer was between 1 and 3), and the organic layer
was washed sequentially with 250 mL water, 500 mL 5%
NaHCO3 and 2 × 250 mL water. The final aqueous wash pH
was 5 to 7.
The worked-up solution from above was solvent-switched
to EtOH by repeated concentrations and additions of EtOH,
and the concentration of the solution was adjusted to ca. 5
mL/g (ca. 275 mL). The solvent amount was checked by NMR
(EtOH:Michael adduct molar ratio ) 20:1.). The mixture was
heated to 50 °C to effect complete dissolution and then cooled
to 40 °C and seeded (2-5 wt %). The resulting slurry was
allowed to cool to 18 °C and 20 mL/g EtOH:MeOH:H2O (1:1:2)
(275 mL:275 mL:550 mL) was added over 2 h. The slurry was
allowed to stand at ambient temperature for 18 h and then
cooled to 0 °C for 6 h. The slurry was then filtered at 0 °C,
and the filter cake washed with 3 mL/g of cold (0 °C) EtOH:
MeOH:H2O (2:1:2) (165 mL × 2) and dried in vacuo under a
N2 stream. The yield of 17 was 50.8 g (74%).2b,53
the end of this time, the batch was filtered through an Estrella
filter (prepacked with 10 kg of solka floc) and the cake was
washed with additional i-PrOAc (50 L). The resulting light
yellow solution was subjected to azeotropic distillation at
constant volume to lower the water content of the mixture to
160 µg/mL. This dry toluene solution of 2 was treated with
dicyclohexylamine (DCHA, 1.4 kg) at ambient temperature.
After 1 h, initial crystallization was observed and the remain-
der of the DCHA (5.6 kg) was added over 110 min. The
resulting slurry was allowed to stand for 17 h. The slurry was
filtered through a DeDeitrich pressure filter, and the filter cake
was washed with fresh i-PrOAc (10 kg). The wet cake was
strained under nitrogen for 60 min and then dried for 15 h at
28-30 °C, to give crude 25 (86% recovery, 93.6 A% purity).
This material and 99 kg of methanol were charged to a 100
gallon glass-lined vessel. The resulting slurry was heated to
60 °C for 1 h to give a homogeneous solution. Water (33 kg)
was added over 30 min, maintaining the temperature between
58 and 62 °C. The batch was seeded with 100 g of pure DCHA
salt and allowed to stand for 2 h at 60 °C. Water was charged
in the following sequence: 15 kg over 4.5 h, 35 kg over 3 h
and finally 42 kg over 1.5 h. The slurry was then cooled to 20
°C over 3 h and allowed to stand at this temperature for
another 8 h. After this age, the batch was filtered via a
DeDeitrich pressure filter and washed with 23 kg of 1:1 )
MeOH:H2O. The wet cake was strained under nitrogen for 30
min and dried for 67 h at 20-25 °C, under nitrogen. 8.7 kg of
99.5% pure 25 were obtained (87% yield).2a,31,53
Diflu or in a tion : (2R,5R)-2-ter t-Bu tyl-5-[(1R)-3,3-d iflu o-
r ocyclop en tyl]-5-p h en yl-1,3-d ioxola n -4-on e (21). Toluene
(27 L) was charged to a Teflon-coated 50 gallon reactor,
followed by a Deoxofluor 20 (10.25 kg, 8.5 L). The batch
temperature was lowered to 5 °C and boron trifluoride diethyl
etherate (0.47 kg, 0.4 L) was charged into the reactor as a
solution in toluene (3.5 L). The resulting mixture was allowed
to stand for 2 h at 5 °C. [Th e a d d ition of Bor on tr iflu or id e
to Deoxoflu or is exoth er m ic, so ca r e m u st be ta k en to
k eep th e tem p er a tu r e betw een 5-15 °C]. The Michael
adduct 17 (10.0 kg) was charged to the vessel as a toluene
solution (25 L). The temperature of the batch was gradually
raised to 55 °C, and the batch was allowed to stand for 36 h
at this temperature. HPLC analysis indicated 3% residual
starting material remained. A separate 200 gal Teflon-lined
vessel was charged with 2 N NaOH (200 L) and toluene (190
L), and the mixture was cooled to 0 °C with agitation. The
reaction mixture was added slowly to the basic solution while
maintaining the temperature below 18 °C. After the transfer
was complete, the mixture was agitated for 30 min. The two
layers were allowed to settle, and the aqueous layer was
separated. The remaining organic layer was washed with 2 N
NaOH (100 L) and deionized water (100 L). HPLC assay of
the organic layer showed 90% yield.2a,53
Hyd r olysis of 21 a n d Sa lt F or m a tion : (2R)-2-[(1R)-3,3-
Diflu or ocyclopen tyl]-2-h ydr oxy-2-ph en ylacetic Acid (25).
The solution of 21 in toluene (from the previous step) was
charged into a 100 gallon reactor, where the batch was distilled
to a minimum volume (in vacuo). Methanol was added and
distilled until the toluene content was 2-5%. The volume of
the mixture was adjusted to 110 L. To that methanolic
solution, 2 N LiOH (81 L) was added, and the reaction mixture
was heated to 40 °C for 20 h. The mixture was cooled to
ambient temperature, and heptane (100 L) was added. The
batch was agitated for 30 min and allowed to settle. The
aqueous layer (ca. 120 L) was transferred to a 200 gallon
reactor where it was cooled to 7 °C, and i-PrOAc was added.
The chilled, agitated mixture was treated with 2 N HCl (75
L, added at such a rate so that the temperature did not
exceeding 15 °C). The pH of the aqueous layer was checked to
ensure it was 2.5-3.0. The layers were separated, and the
organic phase was washed with deionized water (100 L). HPLC
assay of the organic layer verified the formation of 2 (7.55 kg,
99%).
6-Br om o-2-for m ylp yr id in e (26). Dry toluene (45.3kg) and
n-BuLi (1.63 M in hexane, 1230 g, 2.97 mol) were charged to
a 800 L reactor and cooled to -10 °C. n-BuMgCl (1.95 M in
THF, 26.9 kg, 54.6 mol) was added over 30 min, while
maintaining the temperature at -10 to 0 °C, and the mixture
was stirred at -10 °C for 30 min. A solution of 2,6-dibromo-
pyridine (34.92 kg, 144.8 mol) in toluene (30.2 kg, KF ) 69
ppm) was added dropwise over a period of 1 h while keeping
the temperature of the mixture below -5 °C. The resulting
suspension was stirred at -10 °C for 2.5 h. The mixture was
transferred via cannula to a cooled solution (-10 °C) of DMF
(14 kg, 188.9 mol) in toluene (43.2 kg) while maintaining the
temperature below 10 °C. The solution was allowed to stand
at -5 to -10 °C for 30 min and then transferred via a Teflon
cannula to an aqueous citric acid solution (56.6 kg in 105 L of
water) while maintaining the temperature of the mixture
below 20 °C. After stirring the mixture below 20 °C for 10 min,
the organic layer was separated and washed with water (105
L). The organic layer was concentrated to ca. 130 L, in vacuo,
and was used in the next step. HPLC analysis showed that
the desired product was obtained in 91% assay yield (25.0
kg).32,33
4-(ter t-Bu toxyca r bon yla m in o)-1-(6-br om op yr id in -2-yl-
m eth yl)p ip er id in e (36). The toluene solution of 6-bromo-2-
formylpyridine (26), obtained from previous step, was added
to a suspension of 4-Boc-aminopiperidine AcOH salt 3552 (38.55
kg, 4.07 mol) in THF (46.6 kg) at ambient temperature, and
the mixture was stirred at 15 to 20 °C for 30 min. A solution
of NaBH(OAc)3 (33.04 kg, mol) in DMSO (100 kg) was added
at 15 to 20 °C over a period of 3 h, and the mixture was further
stirred at 15 to 20 °C for 2.5 h. After cooling the reaction
mixture to 10 °C, an aqueous NaOH solution (1 N, 8 L) was
added while maintaining the temperature below 25 °C. A
solution of phthalic anhydride (4.18 kg, mol) in DMSO (20 L)
was added at 15 to 20 °C, and the mixture was further stirred
at 15 to 20 °C for 1 h. After cooling to 10 °C, aqueous NaOH
solution (1 N, 291 L) was added while maintaining the
temperature below 25 °C, and the mixture was warmed to
ambient temperature. The organic layer was separated and
washed with brine (20%, 146 L). The resulting solution was
concentrated to ca. 80 L at 60 °C (bath temperature) in vacuo,
and then n-heptane (569.7 kg) was added at 70 °C over a period
of 1 h. The mixture was allowed to cool to ambient temperature
and stand overnight at 0 °C for 2 h. The resulting slurry was
filtered, and the cake was washed with n-heptane (47.44 kg)
twice. Drying the white crystalline solid under reduced pres-
sure with nitrogen stream at ambient temperature gave the
The above brown solution of 2 in i-PrOAc (containing 5%
H2O and 5% MeOH, as determined by gas chromatographic
assay) was treated with Darco G60 (5 kg), and the resulting
suspension was stirred at room temperature for 60 min. At

6786 J . Org. Chem., Vol. 66, No. 20, 2001
Mase et al.
desired product 36 (43.5 kg, 80% yield). 1H NMR(CDCl3, ppm)
δ 1.44 (s, 9H), 1.38-1.53 (m, 2H), 1.91-1.94 (m, 3H), 2.21 (td,
J ) 2.0, 11.2 Hz, 2H), 2.78-2.83 (m, 2H), 3.62 (s, 2H), 4.45
(bs, 1H), 7.35 (d, J ) 7.6 Hz, 1H), 7.41 (d, J ) 7.3 Hz, 1H),
7.52 (dd, J ) 7.3 and 7.6 Hz, 1H); 13C NMR (CDCl3, ppm) δ
28.4, 32.5, 52.4, 63.7, 121.5, 126.2, 138.7, 141.3, 160.8. Anal.
Calcd for C16H24BrN3O2: C, 51.90; H, 6.53; N, 11.35. Found:
C, 51.96; H, 7.09; N, 11.37.
slurry over a period of 30 min. After allowing to stand at
ambient temperature overnight the resulting slurry was
filtered. The filter cake was washed with methanol- methyl
tert-butyl ether (2:1) (14.2 L) and then dried at room temper-
ature under reduced pressure with a nitrogen sweep for 12 h
to give 3 (6580 g, 90% yield) as white crystals (mp 258 °C).2a
1H NMR (DSS in D2O, ppm) δ 0.60-0.66 (m, 1H), 1.74-1.79
(m, 1H), 1.96-2.11 (m, 2H), 2.39 (d, J ) 13.5 Hz, 2H), 2.88-
2.94 (m, 1H), 3.30-3.39 (m, 2H), 3.76 (d, J ) 12.2 Hz, 2H),
4.49 (s, 2H), 4.80 (bs, 2H), 7.14 (d, J ) 5.9 Hz, 1H), 7.14 (d, J
) 6.6 Hz, 1H), 7.95 (dd, J ) 5.9, 6.6 Hz, 1H); 13C NMR (DSS
in D2O, ppm) δ 21.8, 29.4, 47.7, 53.7, 57.0, 58.7, 118.5, 119.4,
119.4, 138.1, 146.6. Anal. Calcd for C11H21Cl3N4: C, 41.85; H,
6.71; N, 17.75. Found: C, 41.96; H, 6.98; N, 17.83.
4-(ter t-Bu toxyca r bon yla m in o)-1-(6-a m in op yr id in -2-yl-
m eth yl)p ip er id in e (39). Dry toluene (50.4 L) was introduced
into a 80 L reactor, and the mixture was degassed (vaccum/
N2 cycle, three times). Mono-bromopyridine 36 (5290 g, 95 wt
%), dppf (75.5 g), NaOtBu (1570 g), Pd(OAc)2 (15.3 g), and
benzophenone imine (3021 g) were charged, and the mixture
was heated at such a rate that the inner temperature rose to
78-82 °C within 30-60 min. The reaction mixture was
agitated for 1h at 78-82 °C. The mixture was allowed to cool
to ambient temperature -15-20 °C) and then allowed to stand
for 10 min. Tri n-butylphosphine (306 g) was added to the
mixture which was allowed to stand for 10 min. The toluene
solution was then poured into 5% aq citric acid (91 L) and
allowed to stand for 30 min at ambient temperature. The
aqueous layer was washed with i-PrOAc (25.2L), and the
organic layer was discarded (after HPLC analysis to ensure
no product loss to this layer). The citric acid layer was treated
with i-PrOAc (62.6 L) and 5 N NaOH (12.1 L) at 15-20 °C.
The organic layer was washed with 7% aq NaCl (10.1 L) and
then dried over anhydrous sodium sulfate (6000 g). The
resulting solution was pretreated with activated carbon (417
g, Shirasagi P) for 1 h at ambient temperature and then
filtered through Celite to remove activated carbon, and the
filter cake was washed with isopropyl acetate (9 L). The filtrate
was concentrated to ca. 25 L under reduced pressure. Crystal-
lization was induced by seeding, and the resultant slurry was
allowed to stand for 1 h at 20-25 °C. n-Heptane (50.1L) was
added over a period of 30 min, and the mixture was agitated
for 1 h at 15-20 °C. The slurry was filtered through a glass
filter, and the filter cake was washed with n-heptane/i-PrOAc
) 2.5/1 (12.5 L) and dried at 40 °C under reduced pressure
with a nitrogen stream for 20 h to give 3600 g of 39 (86% yield)
as white crystals (mp 139.5 °C). 1H NMR (CDCl3, ppm) δ 1.44
(s, 9H), 1.40-1.58 (m, 2H), 1.87-1.93 (m, 3H), 2.14 (td, J )
2.3, 11.6 Hz, 2H), 2.81-2.86 (m, 2H), 3.44 (s, 2H), 3.46 (bs,
1H), 4.44 (bs, 2H), 6.37 (d, J ) 7.9 Hz, 1H), 6.69 (d, J ) 7.3
Hz, 1H), 7.38 (dd, J ) 7.3, 7.9 Hz, 1H); 13C NMR (CDCl3, ppm)
F in a l Cou p lin g: (2R)-N-[1-(6-Am in op yr id in -2-ylm eth -
yl)p ip er id in -4-yl]-2-[(1R)-3,3-d iflu or ocyclop en tyl]-2-h y-
d r oxy-2-p h en yla cta m id e (1). In a separation vessel, the
DCHA salt 25 (3102 g, 8.15 mol) was mixed with 1.02 N NaOH
(20.3 L, 2.54 equiv) and n-heptane (15.5 L). The two layers
were separated, and the aqueous layer was re-extracted with
n-heptane (15.5 L). The aqueous layer was mixed with CH3-
CN (20.2 L) and HOBT (958 g), and amine trihydrochloride 3
(2350 g) was added sequentially. After all of the solids were
dissolved, EDC (1631 g) was added. The resulting homoge-
neous solution had a pH of about 6-6.5. The batch was allowed
to stand at 30 °C for 6 h until HPLC indicated <2 area % of
the acid 2 remained. The batch was cooled to e20 °C and
charged with 5 N NaOH (4.3 L, 3 equiv) and MTBE (20.2 L),
and the two layers were separated. The organic layer was
washed with 1 N NaOH (12.4 L) and 2 N HCl (14.1 L, 3 equiv),
and the layers were separated. The organic layer was dis-
carded. The aqueous layer was mixed with MTBE (20.2 L) and
cooled to e20 °C, and then 50% NaOH solution (2.4 kg) was
added. The pH of the aqueous layer was g11. The two layers
were separated, the organic layer was washed with brine (12.4
L), solvent was switched with i-PrOAc and treated with
activated charcoal Darco G-60 (155 g), and the mixture was
stirred for 1 h. Filtration through Celite afforded a clear
solution. The Celite cake was washed with additional i-PrOAc
(3.1 L). The filtrate was concentrated in vacuo to about 23.5 L
total volume. The resulting slurry was heated to 60 °C, to
dissolve the solid, and then n-heptane was added while
keeping the temperature of the batch at 60 °C. After n-heptane
(1.6 L) addition, the batch was seeded and allowed to stand
for 30 min until crystallization initiated. The remainder of the
n-heptane (total of 44.9 L, 2× of the i-PrOAc volume) was
added over 30 min. The batch was allowed to cool to ambient
temperature (20-25 °C) and allowed to stand overnight. The
solid product was collected by filtration and the cake washed
with 2:1 n-heptane/i-PrOAc (9.3 L) and then n-heptane (7.8
L). The product was dried at in a vacuum oven under nitrogen
sweep at 45 °C overnight to afford 1 (2612 g, 83% from 25).2a
δ 28.4, 32.5, 52.6, 64.6, 106.9, 138.0. Anal. Calcd for C16H26
-
N4O2: C, 62.72; H, 8.55; N, 18.29. Found: C, 62.84; H, 8.76;
N, 18.25.
4-Am in o-1-(6-a m in op yr id in -2-ylm eth yl)p ip er id in e Tr i-
h yd r och lor id e Sa lt (3). To a solution of 39 (7100 g, 23.17
mol) in methanol (24.9 L) was added a mixture of concentrated
HCl (10.2 L, 115.86 mol) and methanol (10.2 L), over a period
of 30 min, while maintaining the temperature of the mixture
below 30 °C. The mixture was warmed to 40 °C over ca. 30
min and stirred at 40 °C for 4 h. The resulting solution was
cooled to ambient temperature and then concentrated to ca.
9.5 L. MeOH (11.8 L) was added, and the solution was again
concentrated to 11 L. After being seeded, the solution was
allowed to stand at ambient temperature for 1 h to make a
seed bed, and methyl tert-butyl ether (7.1 L) was added to the
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and spectroscopic data of compounds 34, 35, 41, 42,
35b, 43, and 3 (from alternative synthesis). This material is
available free of charge via the Internet at http://pubs.acs.org.
J O0157425