A practical chemoenzymatic process to access (R)-quinuclidin-3-ol on scale

Article abstract of DOI:10.1016/S0957-4166(03)00363-X

(±)-3-Butyryloxyquinuclidinium butyrate 6 (2 M, 571 g/L), prepared from (±)-quinuclidin-3-ol 1 and butyric anhydride, undergoes enantioselective hydrolysis by an Aspergillus melleus protease {1.0% (w/v)} in water in the presence of Ca(OH)₂ to keep the reaction at pH 7 and trap butyric acid that is introduced as part of (±)-6 and generated by the enzymatic hydrolysis. After a 24 h period, extraction with n-heptane provides (R)-quinuclidin-3-yl butyrate 5a, which, on methanolysis with Na₂CO₃, is converted into (R)-1, a common pharmacophore of neuromodulators acting on muscarinic receptors, in 96% ee and 42% overall yield from (±)-1. The unwanted antipode (S)-1, which is extracted into n-butanol and purified via its hydrochloride salt in 89% ee and 40% overall yield from (±)-1, can be racemized by the catalysis of Raney Co at 140°C under an atmosphere of H₂ (5 kg/cm²) to regenerate (±)-1 in 97% yield.

Full text of DOI:10.1016/S0957-4166(03)00363-X

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 14 (2003) 1871–1877
A practical chemoenzymatic process to access
(R)-quinuclidin-3-ol on scale
Fumiki Nomoto,* Yoshihiko Hirayama, Masaya Ikunaka,* Toru Inoue and Koutaro Otsuka
Research and Development Center, Nagase & Co., Ltd., 2-2-3 Murotani, Nishi-ku, Kobe 651-2241, Japan
Received 20 March 2003; accepted 15 April 2003
Abstract—( )-3-Butyryloxyquinuclidinium butyrate 6 (2 M, 571 g/L), prepared from ( )-quinuclidin-3-ol 1 and butyric anhydride,
undergoes enantioselective hydrolysis by an Aspergillus melleus protease {1.0% (w/v)} in water in the presence of Ca(OH)2 to keep
the reaction at pH 7 and trap butyric acid that is introduced as part of ( )-6 and generated by the enzymatic hydrolysis. After
a 24 h period, extraction with n-heptane provides (R)-quinuclidin-3-yl butyrate 5a, which, on methanolysis with Na2CO3, is
converted into (R)-1, a common pharmacophore of neuromodulators acting on muscarinic receptors, in 96% ee and 42% overall
yield from ( )-1. The unwanted antipode (S)-1, which is extracted into n-butanol and purified via its hydrochloride salt in 89%
ee and 40% overall yield from ( )-1, can be racemized by the catalysis of Raney Co at 140°C under an atmosphere of H2 (5
kg/cm2) to regenerate ( )-1 in 97% yield. © 2003 Elsevier Science Ltd. All rights reserved.
1. Introduction
such as talsaclidine 2 (M1 agonist; cognition enhancer),1
and revatropate 3 (M3 antagonist; anti-asthma agent),2
while its racemic form was embedded in cevimeline 4
(M1 and M3 agonist; Evoxac®)3 for the treatment of dry
mouth (xerostomia) in patients suffering from Sjo¨gren’s
syndrome (autoimmune, rheumatic disease).
The purpose of the present article is to disclose and
elaborate on our process R & D program that has
culminated in developing enzyme-catalyzed, enantiose-
lective hydrolysis of ( )-quinuclidin-3-yl ester into a
scalable, industrially viable process for (R)-quinuclidin-
3-ol 1 (Fig. 1). Representing a common pharmacophore
of neuromodulators acting on cholinergic muscarinic
receptors, (R)-quinuclidin-3-ol 1 has been incorporated
into therapeutic agents under clinical investigation,
To produce (R)-1 of such pharmacological importance,
five approaches have been developed to date, each,
however, suffering from a practical drawback as indi-
cated:4 (1) resolution via diastereomeric salt formation
Figure 1. Structures of (R)-quinuclidin-3-ol 1, talsaclidine 2, revatropate 3, and cevimeline 4.
* Corresponding authors. Tel.: +81 992 3164; fax: +81 992 1050; e-mail: fumiki.nomoto@nagase.co.jp; masaya.ikunaka@nagase.co.jp
0957-4166/03/$ - see front matter © 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0957-4166(03)00363-X

1872
F. Nomoto et al. / Tetrahedron: Asymmetry 14 (2003) 1871–1877
of ( )-1 with (1S)-(+)-camphorsulfonic acid {low over-
all yield and tedious repetition of fractional recrystal-
lization};5a (2) kinetic resolution via enantioselective
hydrolysis of (R)-quinuclidin-3-yl butyrate by horse-
serum butyrylcholine esterase {limited availability of
the enzyme};6 (3) kinetic resolution via enantioselective
hydrolysis of (S)-3-butyryloxyquinuclidinium butyrate
by subtilisin Carlsberg {low volume efficiency as indi-
cated by the substrate concentration of no more than
4.9% (w/v) in spite of high enantioselectivity leading to
(R)-1 of >95% ee};7 (4) kinetic resolution via enantiose-
lective acylation of (S)-1 with vinyl butyrate by Chiro-
CLEC™-BL, the cross-linked enzyme crystals of
subtilisin, available from Altus biologics, Inc. {limited
availability of the enzyme in a crystalline state suitable
for immobilization into the cross-linked crystal form
and a mediocre yield of 34% for (R)-1 in spite of its
high enantiomeric purity of 96.2% ee};8 (5) asymmetric
hydrogenation of 1-benzhydryl-3-oxoquinuclidinium
bromide by a chiral rhodium complex catalyst prepared
in situ from [Rh(COD)Cl]2 and (R)-1-[(S)-2-
(diphenylphosphino)-ferrocenyl]ethyl-di-tert-butylphos-
phine} in a 1:2 ratio {as low enantioselectivity as 58%
ee}.9
during which the reaction was analyzed by HPLC for
the conversion {CAPCELL PAK C-18 (Shiseido)} and
for the enantioselectivity {Chiralcel OD (Daicel)}. In
the latter chiral analysis, authentic (R)-5a was prepared
from (R)-1 available from Acros Organics and used as
reference to assign the absolute configuration of the
left-over ester 5a and hence that of the digested alcohol
1.
As can be seen from Table 1 summarizing the enzyme
screen, all the proteases tested showed (S)-selectivity,
leaving (R)-butyrate 5a unaffected (Scheme 1); and it
was an Aspergillus melleus protease that showed the
highest E value of 98 among the hydrolases evaluated.
In contrast, no significant hydrolysis was observed with
the lipase preparations tested except for the Pseu-
domonas sp. lipase (Sigma), the enantioselectivity of
which, however, was so poor as to leave (S)-5a of 6%
ee with E11=3 at 12% conversion (16 h). Thus, the A.
melleus protease was nominated for further
investigation.
2.2. Structure–selectivity relationship
Six aliphatic esters of ( )-1, ( )-5b (acetyl), ( )-5c
(valeryl), ( )-5d (caproyl), ( )-5e (capryl), ( )-5f (isobut-
yryl), and ( )-5g (pivaloyl), were prepared according to
the usual procedures5b and subjected to the A. melleus
protease-catalyzed hydrolysis to gain quick insight into
the approximate structure-selectivity relationship and
identify the fittest substrate for the enzyme: 5a–g, 0.2
M; the A. melleus protease, 0.1% (w/v); a 0.2 M potas-
sium phosphate buffer (pH 7.5); 25°C; 16 h (Table 2).
While ( )-acetate 5b (R=Ac), the shortest of the homo-
logues tested, was reluctant to hydrolyse, the A. melleus
protease could accommodate the straight higher homo-
logues up to ( )-caprylate 5e {R=n-C7H15; 40% con-
version and 94% ee for (R)-5e} without significant loss
of selectivity, showing substantial tolerance to the chain
length of the acyl moiety. In contrast, however, branch-
ing in the acyl moiety caused a deleterious effect on the
enzymatic catalysis as illustrated by ( )-isobutyrate 5f
{R=i-Pr; 28% conversion and 37% ee for (R)-5f} and
( )-pivalate 5g (R=tert-Bu; 2% conversion).
Thus, weighing the pros and cons of such prior
methodologies against industrial viability, we chose to
explore the enzyme-catalyzed enantioselective hydroly-
sis of esters of ( )-quinuclidin-3-ol and develop the
kinetic resolution processes for (R)-1 amenable to scale-
up.10
2. Results and discussion
2.1. Enzyme screen
( )-Quinuclidin-3-yl butyrate 5a was prepared from
( )-1 and butyric anhydride according to literature
precedents,5b,6 and seven commercially available hydro-
lases (three proteases and four lipases) were tested for
the ability to hydrolyze ( )-5a in an enantioselective
manner (Scheme 1). A mixture of ( )-5a (0.2 M), a
hydrolase preparation {approximate 0.1% (w/v); a
scoop of a microspatula} and a 0.2 M potassium phos-
phate buffer (pH 7.5) was agitated at 25°C for 16 h
For the ester homologues of a medium-sized straight
chain, the A. melleus protease exerted better enantiose-
Scheme 1. Enzyme-catalyzed enantioselective hydrolysis of ( )-quinuclidin-3-yl ester 5.

F. Nomoto et al. / Tetrahedron: Asymmetry 14 (2003) 1871–1877
1873
Table 1. Enzyme screening for enantioselective hydrolysis of ( )-quinuclidin-3-yl butyrate 5aa
Enzyme (Microbial source)
Conversion (%)b
Ee (%) for (R)-5ac
Ee (%) for (S)-1d
Ee
Protease (A. melleus)f
44
17
23
12
3
75
18
27
95
93
91
98
33
28
3
Protease (Bacillus subtilis)f
Protease (Aspergillus oryzae)f
Lipase (Pseudomonas sp.)g
Lipase (Candida rugosa)h
Lipase (Pseudomonas cepacia)j
Lipase (Rhizopus japonicus)f
6 for (S)-5a
48 for (R)-1
i
i
i
–
–
–
–
–
i
i
i
0
0
–
–
i
i
i
–
–
a [( )-5a]=0.2 M, [enzyme]=approximate 0.1% (w/v), a 0.2 M potassium phosphate buffer (pH 7.5), 25°C, 16 h.
b Determined by HPLC {CAPCELL PAK C-18 (Shiseido)} under the conditions detailed in Section 4.3.1.
c Determined by HPLC {Chiralcel OD (Daicel)} under the conditions detailed in Section 4.3.2.
d Calculated on the basis of conversion and ee values measured for (R)-5a.
e See Ref. 11.
f Available from Nagase ChemteX Corporation.
g Available from Sigma.
h Available from Meito Sangyo Co., Ltd.
i Not calculated.
j Available from Amano Enzyme, Ltd.
Table 2. Enantioselective hydrolysis of esters of ( )-quinu-
clidin-3-ol 5 by the A. melleus proteasea
sis at an industrially accepted substrate concentration
(not less than 1 M) without compromising the enan-
tioselectivity or conversion rate. To deal with these two
problems, butyric anhydride was reacted with neat ( )-
1 according to Muchmore’s procedures7 and, without
purification, the resulting ( )-3-butyryloxyquinucli-
dinium butyrate 6, ( )-5a·n-C3H7CO2H, was subjected
to the A. melleus protease-catalyzed hydrolysis at 25°C
where NaOH was used in place of a 0.2 M potassium
phosphate buffer to adjust the pH of the reaction to
7.5. With ( )-6 (1 M) and the A. melleus protease {0.5%
(w/v)}, the enzymatic hydrolysis proceeded with high
enantioselectivity at pH 7.5 (NaOH) and 25°C to afford
the left-over (R)-butyrate 5a of 94% ee at 50% conver-
sion after 24 h (Table 3). However, when the concentra-
tion of ( )-6 was increased beyond 2 M with a ratio of
[( )-6 (M)] to [A. melleus protease {% (w/v)}] kept
constant at 1:0.5, the enzymatic hydrolysis started to
suffer not only a decrease in the enantiomeric purity of
(R)-5a, but also a slight retardation in the progress: for
R for ( )-5
Conversion (%)b Ee (%) for (R)-5c Ed
a
b
c
d
e
f
n-Pr
Me
n-Bu
n-C5H11
n-C7H15
i-Pr
44
2
75
–
98
–
e
e
51
53
40
28
2
90
99
94
37
42
95
67
70
e
e
g
tert-Bu
–
–
a [( )-5]=0.2 M, [A. mellelus protease]=0.1% (w/v), a 0.2 M potas-
sium phosphate buffer (pH 7.5), 25°C, 16 h.
b Determined by HPLC {CAPCELL PAK C-18 (Shiseido)} under the
conditions detailed in Section 4.3.1.
c Determined by HPLC {Chiralcel OD (Daicel)} under the conditions
detailed in Section 4.3.2.
d See Ref. 11.
e Not calculated.
lection on ( )-butyrate 5a (R=n-Pr; E=98) than on
( )-valerate 5c (R=n-Bu; E=42) while it showed no
substantial difference in the enantioselectivity between
( )-5a and ( )-caproate 5d (R=n-C5H11; E=95) (Table
2). However, ( )-caproate 5d cost more to prepare than
( )-butyrate 5a since caproic anhydride was much more
expensive than butyric anhydride, caproic anhydride
and butyric anhydride being available from Aldrich at
the respective price of Yꢀ11,900/250 g and Yꢀ8,300/L, for
instance. Hence, taking into account not only enzymo-
logical but also economical aspects, ( )-butyrate 5a was
nominated for further investigation into scale-up.
Table 3. Effect of inorganic bases on the A. melleus
protease-catalyzed hydrolysis of ( )-6 under condensed
conditions
[( )-6] (M)
Inorganic base Conversion
(%)a
Ee (%) for
(R)-5ab
1.0c
2.0c
2.5c
3.0c
2.0d
2.0c
NaOH
NaOH
NaOH
NaOH
Ca(OH)2
Mg(OH)2
50
45
42
19
51
48
94
78
69
23
96
86
2.3. Enzymatic process on scale
a Determined by HPLC {CAPCELL PAK C-18 (Shiseido)} under the
conditions detailed in Section 4.3.1.
To adapt the A. melleus protease-catalyzed enantiose-
lective hydrolysis of ( )-5a for the scalable processes for
(R)-quinuclidin-3-ol 1, the following issues should be
addressed (1) to merge the enzymatic hydrolysis with
the preceding butyrate-forming step and establish a
through process; and (2) to run the enzymatic hydroly-
b Determined by HPLC {Chiralcel OD (Daicel)} under the conditions
detailed in the Experimental section 4.3.2.
c [( )-6 (M)]:[A. melleus protease {% (w/v)}]=1: 0.5, pH 7.5 (adjusted
with each inorganic base), 25°C, 24 h.
d [( )-6 (M)]=2 M, [A. melleus protease]=1% (w/v), pH 7.5, 25°C, 24
h.

1874
F. Nomoto et al. / Tetrahedron: Asymmetry 14 (2003) 1871–1877
the enzymatic hydrolysis running at a 2 M concentra-
tion of ( )-5a for 24 h, the enantiomeric purity of
(R)-5a went down to 78% ee with decrease in the
conversion to 45% (Table 3).
was kept confined in the aqueous phase. Once the
butyric acid in the aqueous residue was extracted into
n-heptane, (S)-1 could be recovered successfully by
extraction with n-butanol.
Under the above-mentioned conditions using ( )-6 at
concentrations exceeding 2 M, surplus butyric acid was
introduced as part of ( )-6 to the reaction medium,
where it built up further as the enzymatic hydrolysis
proceeded. Thus, such deleterious phenomena should
be ascribed to the unusually high concentration of
butyric acid causing product inhibition against the A.
melleus protease.12 Hence, to keep the effective concen-
tration of butyric acid below a harmless level, inorganic
bases, such as Ca(OH)2 and Mg(OH)2, were tested for
the ability to capture the butyric acid and squeeze it
from the reaction milieu.
In the final analysis, the optimum conditions for the
scalable kinetic resolution of ( )-quinuclidin-3-ol 1 via
the enzyme-catalyzed enantioselective hydrolysis of ( )-
5a were defined as follows (Scheme 2): On esterification
of ( )-1 with butyric anhydride, crude ( )-6 (2 M) was
treated with the A. melleus protease {1.0% (w/v)} in
water at 25°C for 24 h. Throughout the enzymatic
hydrolysis, solid Ca(OH) was added to keep the pH of
the reaction at 7.5. When the hydrolysis reached 51%
conversion, the reaction mixture was filtered to remove
precipitated Ca(OCOPrn)2. The filtrate was basified to
9.7 with 48% aqueous NaOH, and extracted with n-
heptane to recover the left-over (R)-butyrate 5a of 96%
ee, which, on Na2CO3-promoted methanolysis, pro-
vided (R)-1 of 96% ee {chiral GLC (b-DEX 120,
Sperco)} in an overall yield of 42% from ( )-1.13 The
aqueous residue was then made litmus-acidic with phos-
phoric acid, and extracted with n-heptane to remove
butyric acid. Finally, extraction with n-butanol recov-
ered crude (S)-alcohol 1, which was purified via crys-
talline-salt formation with HCl to afford (S)-1 of 89%
ee {chiral GLC (a-DEX 120, Sperco)} in an overall
yield of 40% from ( )-1.
As can be deduced from Table 3, Ca(OH)2 proved to be
an effective scavenger for butyric acid, alleviating its
inhibitory activity against the A. melleus protease, while
no such beneficial effect was observed with Mg(OH)2.
In fact, when all the butyric acid present in the reaction
medium was neutralized with Ca(OH)2 and the pH of
the reaction was kept at 7.5 throughout the enzymatic
hydrolysis {( )-6, 2.0 M; the A. melleus protease, 1.0%
(w/v); 25°C, 24 h}, not only the enantioselectivity but
also the conversion could be restored to the original
level: 96% ee for (R)-5a at 51% conversion.
Having succeeded in telescoping preparation of ( )-6
into its enzymatic hydrolysis at the concentration of 2
M (571 g/L), we directed our effort towards developing
a practical method to separate the digested (R)-
butyrate 5a from the left-over (S)-alcohol 1 by parti-
tion. After much experimentation based on
Muchmore’s work,7 the extractive separation method
was devised which took advantage of the difference in
hydrophobicity between (R)-5a and (S)-1: Extraction
with n-heptane at pH 9.7 recovered the less polar
(R)-5a selectively, while the more water-soluble (S)-1
2.4. Reuse of the unwanted antipode (S)-1 by
racemization
The enzyme-catalyzed enantioselective hydrolysis being
kinetic resolution by nature, its yield would not exceed
50% in theory. Thus, starting with ( )-quinuclidin-3-ol
1 that was costly due to its synthesis requiring multiple
steps,5a,14 the process would remain far from practical
unless the unwanted enantiomer (S)-1 could be reused.
Thus, we undertook investigation into controlled
racemization15 of (S)-1 to regenerate ( )-1.
Scheme 2. Chiral separation of ( )-quinuclidin-3-ol 1 by the catalysis of an A. melleus protease.

F. Nomoto et al. / Tetrahedron: Asymmetry 14 (2003) 1871–1877
1875
Its stereogenic center residing in an isolated secondary
carbinol carbon, (S)-1 should be difficult to racemize
by other method than simultaneous oxidation/reduc-
tion.16 Thus, Raney alloys (Ni, Co, and Cu) were
explored for the catalytic ability to oxidize (S)-1 to
quinuclidin-3-one 7 and reduce the latter to ( )-1 in a
concurrent fashion. As can be deduced from Table 4
summarizing the comparative study on catalyst and
solvent species, Raney Co was identified as the catalyst
of choice (Scheme 3).17 After much experimentation,
the optimum conditions were established as follows: An
o-xylene solution of (S)-1 was heated to 140°C with a
Raney Co catalyst in the same weight as that of (S)-1
under an atmosphere of H2 (5 kg/cm2). The racemiza-
tion went to completion in half an hour to give ( )-1 in
97% yield as monitored by chiral GLC {b-DEX 120
(Sperco)}.18
tive hydrolysis of ( )-quinuclidin-3-yl butyrate 5a was
developed successfully where the following tactical pro-
cedures were taken to increase the effective productivity
for the overall processes to a practical level: (1) To
allow the A. melleus protease to hydrolyze ( )-3-butyryl-
oxyquinuclidinium butyrate 6 at its concentration of 2
M (571 g/L), Ca(OH)2 was added to the reaction
medium to squeeze the butyric acid that was introduced
as part of ( )-6 and generated by the enzymatic hydrol-
ysis from there such that it would not cause inhibitory
effect on the protease. (2) The unwanted antipode (S)-1
was racemized by the catalysis of Raney Co under H2
atmosphere (5 kg/cm2) at 140°C to regenerate ( )-1 in
97% yield.19
4. Experimental
4.1. General
In the meantime, the redox mechanism was corrobo-
rated by the following control experiment: when a
mixture of (S)-1, excess Raney Co catalyst, and o-
xylene was heated at 142°C in the absence of H2 for 10
h, (S)-1 was converted to quinuclidin-3-one 7 in 97%
conversion as assessed by GLC (Scheme 3).
Melting points were measured on an Electrothermal
1A9100 melting point apparatus and are uncorrected.
1H NMR spectra were recorded at 400 MHz on a
Varian UNITY-400 spectrometer in a CDCl3 solution
with tetramethylsilane as an internal standard. FT-IR
spectra were recorded on a Nicolet AVATAR 360
spectrometer. Elemental analyses were performed on an
Elementar vario EL analyzer. Optical rotations were
measured on a Horiba SEPA-200 polarimeter.
3. Conclusion
To resolve ( )-quinuclidin-3-ol 1 into (R)-1 of 96% ee
on scale, the A. melleus protease-catalyzed enantioselec-
Table 4. Raney alloy-catalyzed racemization of (S)-1a
Raney alloy
Solvent
Pressure of H2 (kg/cm2)
Time (h)
Yield (%)
Cob
Cob
Cob
Cob
Nic
Xylene
Xylene
Cyclohexanol
Cyclohexanol
Cyclohexanol
Cyclohexanol
Atmospheric
5
Atmospheric
5
5
5
0.5
0.5
0.5
0.5
1.0
1.0
68
97
93
97
98
97
Cuc
a (S)-1, 6.4 g; solvent, 150 mL; 140°C.
b Catalyst weight, 6.4 g.
c Catalyst weight, 64 g.
Scheme 3. Racemization of (S)-1 via quinuclidin-3-one 7 by Raney Co catalysis.

1876
F. Nomoto et al. / Tetrahedron: Asymmetry 14 (2003) 1871–1877
4.2. ( )-Quinuclidin-3-yl esters 5a–g
with n-heptane (50 mL×1, 125 mL×1, 150 mL×1); a
portion (10 mL) of the combined n-heptane extracts was
analyzed by chiral HPLC for the enantiomeric purity of
(R)-quinuclidin-3-yl butyrate 5a under the conditions
described above: (R)-5a, 98%; (S)-5a, 2%. The n-hep-
tane solution was concentrated in vacuo until its weight
diminished to 8.8 g. To the residue were added MeOH
(25 mL) and Na2CO3 (5.0 g), and the mixture was
stirred and heated at 65°C for 16 h. The mixture was
allowed to cool to rt, and filtered to remove inorganic
solids. The filtrate was concentrated in vacuo to dry-
ness. To the residue was added CHCl3 (30 mL), and the
mixture was filtered to remove insoluble materials. The
filtrate was concentrated in vacuo to give (R)-1 (5.3 g,
42%), the enantiomeric purity of which was determined
to be 96% ee by chiral GLC: A portion (approximate
10 mg) of it was dissolved in CHCl3 (1 mL), and 5 mL
of the CHCl3 solution was injected to a chromatograph
running under the following conditions: column, b-
DEX 120 (Sperco), 0.25 mmf×30 m; carrier gas, He,
200 kPa; injection temperature, 220°C; column temper-
ature, 140°C; detection, FID, 220°C; tR 7.3 min for
(R)-1, 6.9 min for (S)-1. IR wmax (KBr) 3105, 2941,
2872, 1456, 1346, 1319, 1128, 1117, 1045, 991, 941, 818,
( )-Quinuclidin-3-ol 1 was treated with acid anhydride
(RCO)2O (R=n-Pr, Me, n-Bu, n-C5H11, n-C7H15, i-Pr,
and tert-Bu) under the usual conditions5b to give the
respective ester 5a–g uneventfully.
4.3. Enzyme screen and structure-selectivity relationship
for the enantioselective hydrolysis of ( )-quinuclidin-3-
yl esters 5a–g
An Eppendorf® microtube (2.0 mL) was charged with
( )-quinuclidin-3-yl ester 5a–g, an enzyme preparation,
0.2 M potassium phosphate buffer (pH 7.5) to set up
reaction conditions designed for the specific objective
(Tables 1–3). The reaction mixture was agitated at
25°C, and analyzed at appropriate intervals by HPLC
for the conversion and enantioselectivity.
4.3.1. HPLC analysis to determine the conversion. An
aliquot (40 mL) was taken from the reaction mixture
and dissolved in MeCN (1.96 mL). A portion (10 mL)
of the solution was injected to a chromatograph run-
ning under the following conditions: column, CAP-
CELL PAK C-18 (Shiseido), 0.46 cmf×25 cm; elution,
MeOH/2.5% aqueous NH3 (45:55), 1 mL/min; detec-
tion, UV at 210 nm; tR 5.21 min for ( )-1. Quantity of
( )-1 generated by the enzymatic hydrolysis was esti-
mated by comparing the peak area with those of stan-
dard solutions of ( )-1 (Aldrich) prepared at specific
concentrations.
1
795, 775, 633 cm−1; H NMR l 1.34 (m, 1H), 1.44 (m,
1H), 1.67 (m, 1H), 1.79 (m, 1H), 1.94 (m, 1H), 2.55–
2.66 (m, 2H), 2.70–2.79 (m, 2H), 2.89 (m, 1H), 3.08 (m,
1H), 3.79 (m, 1H), 4.42 (br s, 1H). A portion of (R)-1
thus obtained was further purified according to the
literature procedures7 to afford an analytical sample:
mp 219°C {lit.:6 219–220°C}; [h]2D5 −44.9 (c 2.0, 1 M
HCl) {lit.:6 [h]D25 −46 1 (c 2.0, 1 M HCl)}. Anal. found:
C, 66.0%; H, 10.3%; N, 11.0%. calcd for C7H13NO: C,
66.09%; H, 10.32%; N 11.01%.
4.3.2. HPLC analysis to determine the enantiomeric
composition of quinuclidin-3-yl ester 5. Authentic (R)-
5a–g were all prepared from (R)-1 (Acros Organics)
under the standard conditions5b,6 and used as reference
in the following analysis. An aliquot (0.2 mL) was
taken from the reaction mixture and extracted with
AcOEt (1.0 mL). A portion (1.0 mL) of the AcOEt
solution was injected to a chromatograph running
under the following conditions: column, Chiralcel OD
(Daicel), 0.46 cmf×25 cm; elution, n-hexane/i-PrOH/
CF3CO2H (90:10:0.1), 1 mL/min; detection, UV at 210
nm; tR 15.3 min for (R)-5a, 17.7 min for (S)-5a; 14.3
min for (R)-5c, 17.0 min for (S)-5c; 13.9 min for
(R)-5d, 17.9 min for (S)-5d; 13.7 min for (R)-5e, 15.8
min for (S)-5e; 15.2 min for (R)-5f, 17.2 min for (S)-5f.
4.5. (S)-Quinuclidin-3-ol 1
On extraction with n-heptane at pH 9.7 as described in
Section 4.4, H3PO4 (11 mL) was added to the aqueous
residue. The mixture was extracted with n-heptane (30
mL × 3). The aqueous residue was concentrated in
vacuo to remove remaining n-heptane. KOH (30 g) was
added, and the aqueous mixture was extracted with
n-butanol (45 mL × 3). To the combined n-butanol
extracts was added 35% HCl (7 mL). The mixture was
concentrated in vacuo to remove water azeotropically.
The residual solution was allowed to cool to 5°C, and
the precipitated solids were collected by filtration to
give wet (S)-1·HCl (13.2 g), to which was added 5 M
aqueous solution of NaOH (40 mL). The mixture was
extracted with CHCl3 (50 mL × 3). The combined
CHCl3 extracts were washed with 5 M aqueous solution
of NaOH (20 mL × 1), dried (Na2SO4), and concen-
trated in vacuo to give (S)-1 (5.1 g) in 89% ee and 40%
overall yield from ( )-1; the enantiomeric purity of
(S)-1 was determined by chiral GLC under the condi-
tions specified in Section 4.4.
4.4. (R)-Quinuclidin-3-ol 1
( )-Quinuclidin-3-ol 1 (12.7 g, 100 mmol; purchased
from Aldrich) was added to butyric anhydride (17.4 g,
110 mmol) in portions at rt. The resultant mixture was
stirred at 50°C for 2 h, and distilled water (22 mL) was
added. To the mixture was added solid Ca(OH)2 until
pH of the mixture reached 7.5. XP-488 (0.5 g; a prepa-
ration of an A. melleus protease available from Nagase
ChemteX Corporation) was added, and the mixture
was stirred at 25°C for 24 h during which Ca(OH)2 was
added to keep the pH of the mixture at 7.5. The
precipitated solids were filtered off. To the filtrate was
added 10 M aqueous solution of NaOH to adjust the
pH of the mixture to 9.7. The mixture was extracted
4.6. Racemization of (S)-1
A bomb (300 mL) was charged with (S)-1 (89% ee; 5.1
g, 40 mmol), o-xylene (120 mL), and Raney Co (5.1 g).
The mixture was stirred and heated at 140°C under an

F. Nomoto et al. / Tetrahedron: Asymmetry 14 (2003) 1871–1877
1877
atmosphere of H2 (5 kg/cm2) for 1 h, during which the
progress of the racemization was monitored by chiral
GLC: From the supernatant of the reaction mixture
was taken an aliquot (approximate 0.1 mL), to which
was added CHCl3 (0.2 mL). The mixture was filtered to
remove solid materials and 5 mL of the filtrate was
injected to a chromatograph running under the condi-
tions recorded in Section 4.4. The mixture was allowed
to cool to 90°C, and filtered to remove the catalyst. The
filtrate was further cooled to 0°C. The precipitated
solids were collected by filtration and air-dried to give
( )-1 (4.9 g, 97%). Its spectral and physicochemical
data were identical to those reported.5a
A.; Yeh, W.-K.; Zmijewski, M. J., Jr., Eds.; Cross-Linked
Enzyme Crystals: Biocatalysts for the Organic Chemist.
Marcel Dekker: New York, 2001; pp. 209–226.
9. Brieden, W. (Lonza, Ltd.). EP 785198, 1996; Jpn. Kokai
Tokkyo Koho 97 194,480, 1997.
10. (a) Nomoto, F.; Otsuka, K. (Nagase & Co., Ltd.). Jpn.
Kokai Tokkyo Koho 98 210,997, 1998; (b) Spec. Chem.
Mag. 2001, July/August, 14; (c) Ikunaka, M. Stereoselec-
tive Synthesis of Chiral Pharmaceutical Intermediates.
Proceedings of the Fourth International Conference on
Organic Process Research and Development; Hong
Kong, March 18–21, 2001; Scientific Update: Mayfield,
UK, 2001; (d) Chikusa, Y.; Hirayama, Y.; Ikunaka, M.;
Inoue, T.; Kamiyama, S.; Moriwaki, M.; Nishimoto, Y.;
Nomoto, F.; Ogawa, K.; Ohno, T.; Otsuka, K.; Sakota,
A. K.; Shirasaka, N.; Uzura, A.; Uzura, K. Org. Process
Res. Dev. 2003, 7, in press.
11. (a) Bornsheuer, U. T.; Kazlauskas, R. J. Hydrolases in
Organic Synthesis: Regio- and Stereoselective Biotransfor-
mations; Wiley-VCH: Weinheim, 1999; pp. 32–33; (b)
Wong, C.-H.; Whitesides, G. M. Enzymes in Synthetic
Organic Chemistry; Pergamon: Oxford, 1994; pp. 9–13.
12. For the inhibition caused not by a cleaved achiral
byproduct but by a resolved chiral product itself, see: (a)
Akeboshi, T.; Ohtsuka, Y.; Ishihara, T.; Sugai, T. Adv.
Synth. Catal. 2001, 343, 624; (b) Lundhaug, K.; Over-
beeke, P. L. A.; Jongejan, J. A.; Anthonsen, T. Tetra-
hedron: Asymmetry 1998, 9, 2851.
4.7. Oxidation of (S)-1 to quinuclidin-3-one 7
A mixture of (S)-1 (0.64 g, 5.03 mmol), Raney Co (1.59
g), and o-xylene (15 mL) was stirred and heated at
142°C. After 10 h, the reaction mixture was analyzed
by chiral GLC under the conditions recorded in Section
4.4: tR 5.8 min for 7 (96.6%), 6.9 min for (S)-1 (1.6%),
7.3 min for (R)-1 (1.7%); authentic 7 was obtained by
liberating it from the hydrochloride salt 7·HCl
(Aldrich) under the usual conditions.
Acknowledgements
13. When the reaction period was prolonged to 30 h and the
enzymatic hydrolysis was driven to 53% conversion, the
enantiomeric purity of (R)-5a could be increased to as
high as 99% ee.
Thanks are due to Mr. Masafumi Moriwaki, Director
of Research and Development Center, Nagase & Co.,
Ltd., for his consistent encouragement.
14. For the synthesis of quinuclidinon-3-one 7, an immediate
precursor to ( )-1, see: Daeniker, H. U.; Grob, C. A.
Organic Syntheses; Wiley & Sons: New York, 1973;
Collect. Vol. V, 989.
15. Ebbers, E. J.; Ariaans, G. J. A.; Houbiers, J. P. M.;
Bruggink, A.; Zwanenburg, B. Tetrahedron 1997, 53,
9417.
16. For a discrete oxidation/reduction approach to recycling
the unwanted (S)-1, see: Bjørsvik, H.-R.; Liguori, L.;
Costantino, F.; Minisci, F. Org. Process Res. Dev. 2002,
6, 197.
17. Nomoto, F.; Hirayama, Y.; Inoue, T. (Nagase & Co.,
Ltd.). Jpn. Tokkyo Kokai Koho 99 246,558, 1999.
18. For racemization of chiral primary amine over a Raney
Co catalyst, see: Fujii, A.; Fujima, Y.; Harada, H.;
Ikunaka, M.; Inoue, T.; Kato, S.; Matsuyama, K. Tetra-
hedron: Asymmetry 2001, 12, 3235.
References
1. (a) Leusch, A.; Tro¨ger, W.; Greischel, A.; Roth, W.
Xenobiotica 2000, 30, 797; (b) Drugs Fut. 1994, 19, 35.
2. Moore, B. A.; Keir, R. F.; Stuart, E. F.; Stobie, A.;
Wright, K. N. Eur. Respir. J. 1996, 9 (Suppl. 23), Abs.
PO261.
3. (a) Fisher, A.; Heldman, E.; Grunfeld, Y.; Karton, I.;
Levy, A. (Israel Institute for Biological Research). US
Patent 4,855,290, 1989; (b) Drugs Fut. 2000, 25, 558.
4. (a) For the resolution of (S)-quinuclidin-3-ol 1 into its
(S)-1 via diastereomeric salt formation of ( )-1-benzyl-3-
hydroxyquinuclidinium chloride with dibenzoyl
acid, see: Kalir, A.; Sali, E.; Shirin, E. Isr. J. Chem. 1971,
9, 267; (b) For the synthesis of (S)-1 from -glucose, see:
D-tartaric
D
Fleet, G. W. J.; James, K.; Lunn, R. J. Tetrahedron Lett.
1986, 27, 3053.
19. (a) In parallel with this study, we found Rhodotorula
rubra strains able to reduce 7 to (R)-1 with high enan-
tioselectivity: Nomoto, F.; Kuramura, A.; Otsuka, K.
(Nagase & Co., Ltd.). Jpn. Kokai Tokkyo Koho 99
196,890 1999. For other related approaches of enantiose-
lective microbial reduction, see: (b) Matsuyama, A.;
Hamatani, T. (Daicel Chemical Industries, Ltd.). Jpn.
Kokai Tokkyo Koho 98 243,795, 1998; (c) Goto, M.;
Ueda, M. (Mitsubishi Chemical Corporation). Jpn.
Kokai Tokkyo Koho 2000 245,495, 2000.
5. (a) Sternbach, L. H.; Kaiser, S. J. Am. Chem. Soc. 1952,
74, 2215; (b) Sternbach, L. H.; Kaiser, S. J. Am. Chem.
Soc. 1952, 74, 2219.
6. Rehavi, M.; Maayani, S.; Sokolovsky, M. Life Sci. 1977,
21, 1293.
7. Muchmore, D. C. (Bend Research, Inc.). US Patent
5,215,918, 1993.
8. (a) Bossard, P. (Lonza, Ltd.). Ger. Offen. DE 197 15 465,
1997; (b) Grim, M. D. In Enzyme Technologies for Phar-
maceutical and Biotechnological Applications; Kirst, H.