Development of a practical and scalable synthesis of a potent selective dual antagonist for 5-HT2B and 5-HT7 receptors

Article abstract of DOI:10.1021/op200380z

Process research and development of a practical and scalable synthetic route toward compound (S)-1 and compound (R)-1, which are potent selective dual antagonists for 5-HT2B and 5-HT7 receptors, respectively, is described. The medicinal chemistry route and second generation route were also unattractive for large-scale use for a variety of reasons. The new synthetic method does not require any purification by column chromatography for all steps and highly exothermic reactions. Additionally, we developed an efficient method of optical resolution in which each carboxylic acid isomer was separated with chiral amine in high yield and high enantiopurity. This highly efficient and scalable process was successfully demonstrated in the large scale synthesis of compound (S)-1 and compound (R)-1 in high enantiopurity.

Full text of DOI:10.1021/op200380z

Article
pubs.acs.org/OPRD
Development of a Practical and Scalable Synthesis of a Potent
Selective Dual Antagonist for 5-HT2B and 5-HT7 Receptors
Shinya Yoshida,*,† Makoto Kasai,† Takenori Kimura,‡ Takahiro Akiba,† Takumi Takahashi,†
and Shuichi Sakamoto§
†Process Chemistry Laboratories, Astellas Pharma Inc., 160-2 Akahama, Takahagi-shi, Ibaraki 318-0001, Japan
‡Astellas Research Technologies Co., Ltd., 21 Miyukigaoka, Tsukuba-shi, Ibaraki 305-8585, Japan
§Astellas Pharma Europe B.V., Elisabethhof 19, 2353 EW Leiderdorp, The Netherlands
ABSTRACT: Process research and development of a practical and scalable synthetic route toward compound (S)-1 and
compound (R)-1, which are potent selective dual antagonists for 5-HT2B and 5-HT7 receptors, respectively, is described. The
medicinal chemistry route and second generation route were also unattractive for large-scale use for a variety of reasons. The new
synthetic method does not require any purification by column chromatography for all steps and highly exothermic reactions.
Additionally, we developed an efficient method of optical resolution in which each carboxylic acid isomer was separated with
chiral amine in high yield and high enantiopurity. This highly efficient and scalable process was successfully demonstrated in the
large scale synthesis of compound (S)-1 and compound (R)-1 in high enantiopurity.
There were several drawbacks in the medicinal chemistry
route (first generation route), and those are summarized below.
INTRODUCTION
■
(S)-N-(Diaminomethylidene)-4′,5′-dihydro-3′H-spiro[fluorene-
9,2′-furan]-2-carboxamide monohydrochloride (compound
(S)-1, Figure 1) has a potent medicament for IBS, wherein
• Methyl protection of carboxylic acid should be avoided
to produce a shorter synthetic route.
• The use of highly flammable and toxic BH3−THF
complex has a risk of provoking a dangerous operation
on a large scale.
• Heating conditions of H2O2 in THF will be an
unfavorable operation from the point of view of safety.
• The yield of optical resolution with cinchonidine was
very low (16%). It was including the recovery from
filtrate and following 8 times recrystallization of the
cinchonidine salt to meet >99% ee.
Figure 1. Structures of (S)-1 and (R)-1.
• The overall yield was very low (4.2%).
the selective dual antagonist for 5-HT2B and 5-HT7 receptors
comprises a selective dual antagonistic compound for 5-HT2B
and 5-HT7 receptors having selective binding affinity for both
5-HT2B and 5-HT7 receptors.1a An enantioisomer (compound
(R)-1, Figure 1) also has a potent same pharmacological
effect.1a Additionally, these compounds have an advantage in
that they have high affinity for serotonin receptor subtypes,
particularly for 5-HT2B receptor and 5-HT7 receptor, and
show excellent pharmacological effects in comparison with the
conventional compounds which have only one of the
antagonistic activities of 5-HT2B receptor and 5-HT7 receptor;
this is useful as a prophylactic antimigraine agent having high
safety and excellent effect.1b Herein, we report our efforts to
develop an efficient synthetic route capable of being operated in
a scale up synthesis of compound (S)-1 for the first GMP
delivery and compound (R)-1 for a preclinical study campaign.
• Purification using column chromatography in a final step
is not an efficient method, and the use of chlorinated
solvent in the final step is not favorable from the point of
view of the environment, as well.
Second Generation Synthetic Route (Three-Hundred-
Gram Scale Synthesis). Owing to the urgency with which we
were required to provide material for further preclinical studies,
we decided that chiral separation using the SMB method
(simulated moving bed chromatography) should be better than
an optical resolution method for the preparation of the 300 g
campaign (Scheme 2). In this second generation route, the
C−C bond formation method was improved; it was achieved
using a Grignard reaction instead of a hydroboration−oxidation
reaction, and the number of steps was shortened.
While the route was successfully used to prepare 350 g of
(S)-1 in good yield and high enantiopurity, it had many
unattractive issues with regard to scale up synthesis for the
20 kg campaign. These are listed below.
RESULT AND DISCUSSION
■
First Generation Synthetic Route (Medicinal Chem-
istry Synthetic Route). The medicinal chemistry synthetic
route (first generation route) is shown in Scheme 1.
Received: December 22, 2011
Published: April 5, 2012
© 2012 American Chemical Society
654
dx.doi.org/10.1021/op200380z | Org. Process Res. Dev. 2012, 16, 654−663

Organic Process Research & Development
Article
Scheme 1. First Generation Synthetic Route (Medicinal Chemistry Synthetic Route)
Chiral Separation Using SMB Chromatography2 in a Final
Stage. The chiral separation of racemate in the final stage was
successful on the 1 kg scale. We could prepare the 99.7% ee
sample for the precrinical study. The yield of the SMB chiral
separation of 10 was 38%. However, it was estimated that it
would be impossible to separate on a larger scale because of the
low solubility and too expensive to separate for a 20 kg scale
synthesis.3
Our efforts were focused on designing a safe, scalable, and
efficient synthesis of (S)-1 (Scheme 4) that incorporates the
following key points: (1) C−C bond formation using acetylide
ion6 with 9-oxo-9H-fluorene-2-carboxylic acid 2; (2) a scalable
synthetic method for spiro-ring formation under acidic condi-
tions; and (3) optical resolution of racemate with commercially
available chiral amine compound (establish a scalable optical
resolution method).
Control of Grignard Reaction. This Grignard reaction was
highly exothermic and was performed near or at the boiling
temperature of THF solvent. Sudden initiation of the reaction
occurred, and high heat release rate was observed, which could
lead to vigorous boiling and to flooding in the vapor tube with
risk of overpressure and explosion.4
We studied the new strategic approach to the 20 kg synthesis
campaign of (S)-1. In the following sections, we would like to
describe these results and discuss them.
C−C Bond Formation. To develop a suitable C−C bond
formation, the reaction between compound 2 and non-
protected propargyl alcohol 15 was attempted (Table 1).
However, the reaction did not complete. This can be attributed
to the higher acidity of the hydroxyl proton as compared to the
acetylenic proton. If an additional quantity of base was used,
another impurity formation was observed (HPLC). In this case,
it was difficult to prepare the desired compound in high yield
(entries 1−4). Next, we focused on a C−C bond formation
with commercially available and inexpensive THP protected
propagyl alcohol 17. The use of DMF as a solvent gave better
conversion compared with THF solvent. At 20 °C and above,
the reaction did not complete (entries 5 and 6). On the other
hand, addition of t-BuOK at −20 °C gave a good conversion
and 18 was prepared in 92.8% yield (entry 7). This method was
performed in a large scale synthesis campaign and gave 92.8 kg
of 18.
Inefficient Method for the Spiro-Ring Formation. The
second generation method for spiro-ring formation using an
O-leaving group5 was described in Scheme 3. The use of TsCl
(2.1 equiv) gave a good conversion, but there were issues for
scale up synthesis. For example, it was difficult to not only
control the production of impurity A (typically 5%) but also
prevent the formation of 14 that was caused by hydrolysis of
the tosylate during the treatment with aqueous NaOH
(typically 5%); both of these impurities were difficult to
remove without SiO2 column chromatography workup.
New Approach for 20 kg Scale Synthesis (Third-
Generation). Under such circumstances, there has been a brisk
demand for the development of an excellent production
process for 1 with a higher overall yield that does not require
column chromatography purification.
655
dx.doi.org/10.1021/op200380z | Org. Process Res. Dev. 2012, 16, 654−663

Organic Process Research & Development
Article
Scheme 2. Second Generation Synthetic Route (Three-Hundred-Gram Scale Synthesis)
Scheme 3. Spiro-ring Formation in the Second Generation Method
Reduction of Alkyne and Deprotection of Alcohol.
Several methods were investigated in reduction of the alkyne
step (Table 2). During this hydrogenation, we focused on the
formation of impurity B. The hydrogenation of 18 (with H2,
Pd/C, and Et3N in MeOH) gave 5% of impurity B (entry 1).
The reaction conditions without Et3N accelerated the
formation of impurity B (entry 2). In this case, the carboxylic
acid unit worked as an acid in this reaction. Next, catalytic
hydrogen transfer reductions using ammonium7 or potassium
formate were attempted (entries 3 and 4). The use of
potassium formate8 with Pd/C in MeOH gave the best result.
The level of impurity B was less than 2%. This might be due to
the reduction of the benzyl alcohol moiety being suppressed by
the high basicity of potassium formate.9
Intermediate 19 was used in the next step without
purification and isolation. The following deprotection of THP
656
dx.doi.org/10.1021/op200380z | Org. Process Res. Dev. 2012, 16, 654−663

Organic Process Research & Development
Article
Scheme 4. New Synthetic Strategy for Compound (S)-1
Table 1. C−C Bond Formation with Propagyl Substrates
b
entry
R (equiv)
base (equiv)/solvent
temp (°C)
HPLC yield of 16 or 18 (%)
1
2
3
4
5
6
H (1.0)
t-BuOK (2 + 2)/DMF
t-BuOK (2 + 2)/THF
KOH (2 + 2)/THF
NaH (2 + 2)/THF
t-BuOK (2.1)/THF
t-BuOK (2.1)/DMF
t-BuOK (2.1)/DMF
0−25
0−25
0−25
0−25
20−40
20−40
48
21
49
2
H (1.0)
H (1.0)
H (1.0)
THP (1.05)
THP (1.05)
THP (1.05)
74
94
99.5
a
7
−20 to 0
a
b
t-BuOK was added to a DMF solution of 2 and 17 at −20 °C. Isolated yield was 92.8%. Determined by HPLC methods A and B (see
Experimental Section).
a
Table 2. Reduction of Alkyne
c
c
entry
conditions
18/19
impurity B (%)
b
1
2
3
H2, Et3N (× 1 mol), 2 h
H2, Et3N (none), 1 h
ND/94
ND/92
ND/96
ND/98
5
8
4
b
ammonium formate (× 8 mol), 2 h
potassium formate (× 8 mol), 2 h
4
<2
a
b
c
Catalyst, 10%Pd/C(M), Kawaken Fine Chemicals Co., Ltd.; solvent, MeOH; reaction temp, 25 °C. Under atmospheric pressure. Determined by
HPLC methods A and B.
using a catalytic amount of 38 wt % HCl in MeOH gave 64.1 kg
of diol compound 14 in excellent yield in a first scale up
synthesis (88.7% yield in 2 steps from compound 18).
Spiro-ring Formation. In the initial medicinal chemistry
route, the spiro-ring formation under acidic conditions was
given1a,b (Scheme 5). However, in these cases, about 16% of
657
dx.doi.org/10.1021/op200380z | Org. Process Res. Dev. 2012, 16, 654−663

Organic Process Research & Development
Article
characteristics of (1) high water solubility and (2) high boiling
point aprotic solvent (no esterification).
Scheme 5. Spiro-ring Formation under Acidic Conditions
And so we selected diglyme as a solvent and tried the cycliza-
tion reaction (entry 11). As we expected, this condition worked
well; the level of impurity C was <1% in a reaction mixture.
After the completion of the reaction, the target intermediate 7
was isolated in 88−90% yield simply by adding water to the
reaction mixture. The purity of isolated compound 7 was 99%
(HPLC), and it contained <1% of impurity C (<0.5% E/Z each
other). Thereafter, 52.4 kg of compound 7 was prepared in a
first scale up synthesis (87.7% yield).
Optical Resolution of Racemate. For the 20 kg
manufacturing campaign, we challenged the optical resolution
with the chiral amine approach. Both enantiomers had already
been separated in high enantiopurity by medicinal chemists by
using chiral amine. (S)-Carboxylic acid 9 was separated as a
cinchonidine salt, and (R)- carboxylic acid 21 was separated as
a brucine salt (Scheme 6).
eliminated impurity C was observed, and it was difficult to
remove impurity C without SiO2 column chromatographic
purification. We thought that if we could prevent the formation
of eliminated impurity C, the spiro-ring synthesis under acidic
conditions10 would be better because of the difficulties of the
leaving group methods in a large scale synthesis as shown above.
As shown Table 3, the use of combination acid and a low-
polarity solvent such as toluene did not give good resluts;
impurity C was observed over 10% (entries 1−5). Also, AcOH or
HCO2H as a solvent did not give good results (entries 6 and 7).
On the other hand, a polar solvent such as aqueous i-PrOH
or t-BuOH could prevent the formation of impurity C, but
esterificated byproduct was detected (entries 8−10). To reduce
the level of impurity C, we focused on the solvent with
To optimize the optical resolution method to prepare
optically pure (S)-carboxylic acid 9, several solvents were
screened with cinchonidine (Table 4). The use of i-PrOH and
i-PrOAc did not give crystals (entries 3 and 5), and MIBK
Table 3. Spiro-ring Formation under Acidic Conditions
g
HPLC area %
a
entry
acid (equiv)
solvent
°C/h
14
7
impurity C
comment
1
TsOH·H2O (1.0)
TsOH·H2O (0.1)
TsOH·H2O (0.1)
PPTS (0.1)
toluene
100/3
100/3
100/3
100/3
100/3
80/3
<1
<1
82
<1
<1
30
<1
1
79
82
8
16
14
10
25
59
10
trace
5
2
toluene
3
1,4-dioxane
toluene
4
71
33
42
43
89
84
89
94
5
Amberlyst-15
AcOH
toluene
c
6
7
HCO2H
80/3
c
d
e
f
8
H2SO4 (0.6)
MsOH (0.6)
H2SO4 (0.6)
H2SO4 (0.6)
i-PrOH/water (50/50)
i-PrOH/water (50/50)
t-BuOH/water (57/43)
100/37
100/65
100/44
93/24
9
2
6
10
11
<1
<1
3
b
diglyme (6 v/w) water (5 v/w)
<1
a
b
c
d
Impurity C was calculated as an E/Z mixture. Diglyme = diethyleneglycol dimethyl ether. Complex mixture. 4% of isopropyl ester was observed.
e
f
g
6% of isopropyl ester was observed. 3% of tert-butyl ester was observed. Determined by HPLC methods A and B.
Scheme 6. Optical Resolution of the Racemate of the Medicinal Chemistry Method
658
dx.doi.org/10.1021/op200380z | Org. Process Res. Dev. 2012, 16, 654−663

Organic Process Research & Development
Article
a
performed using the filtrate of the brucine salt preparation for
(R)-carboxylic acid 21 was attempted. The filtrate was treated
with the EtOAc−aqueous hydrochloric acid, and brucine was
removed in an aqueous layer. The organic layer that included
9 was concentrated. In this stage, the level of enantioisomer
21 was about 4−5% by HPLC analysis. The residue was
dissolved in methyl propionate, and 0.5 equiv of cinchonidine
(vs racemate 7) was added. During the salt formation with
cinchonidine, addition of a small amount of water (0.5% v/v of
methyl propionate) gave a good result; the optical purity was
increased to 99.4% ee. When crystallization was performed
without water, optical purity was 97% ee and recrystallization
was required once more to meet the desired optical purity. We
guessed that the solubility of the optical isomer was increased by
the addition of water. The crystalline morphology of 8 appeared
to be the same compared with the sample derived from non-
water adding conditions by XRD analysis. Finally, we succeeded
in a first scale up for optical resolution and obtained 46.7 kg of 8
in high optical purity (99.3% ee, 42.6% yield from 7, Scheme 7).
Additionally, the crude brucine salt was purified in aqueous
DMF two times and gave the optically pure brucine salt 20
(99.6% ee, 42.3% yield from 7). The desired highly enantiopure
carboxylic acid compounds were prepared in an overall yield
of 85%.11
Table 4. Solvent Screening for Cinchonidine Salt
b
entry
solvent (v/w)
MEK (10)
temp (°C) yield (%) from 7 % ee
1
2
3
4
5
6
7
25
25
25
25
25
25
0
42
66
20
MIBK (15)
67
i-PrOH (10)
not crystallized
EtOAc (10)
61
30
i-PrOAc (10)
methyl propionate (10)
not crystallized
57
50
38
60
methyl propionate (20)
a
b
Cinchonidine: 1 equiv. Determined by HPLC method C (see the
Experimental Section).
and EtOAc gave low enantioselectivity (entries 2 and 4). The
use of methyl propionate gave the best result in this case (entry 7).
However, this method was not perfect; cinchonidine salt
formation and three times recrystallization of cinchonidine salt
were required to meet the desired optical purity (>98% ee),
and the yield was very low (24% from 7).
At that time, we were required to synthesize not only
compound 9 but also compound 21 for preclinical study. And
so, we focused on the medicinal chemistry methods.
(R)-Carboxylic acid 21 was separated as a brucine salt in
aqueous DMF solution in high optical purity, in good
sequences. On the basis of these results, a new approach that
cinchonidine salt formation for (S)-carboxylic acid 9 was
Completion of Synthesis. We describe here the
completion of the synthesis for the final product (Scheme 8).
Scheme 7. Improved Method of Optical Resolution
Scheme 8. Completion of the Synthesis
659
dx.doi.org/10.1021/op200380z | Org. Process Res. Dev. 2012, 16, 654−663

Organic Process Research & Development
Article
To minimize the formation of bis-form impurity D to the level
of 1%, the second generation synthetic method required use of a
14 vol/wt amount of DMF. After the completion of the reaction,
it was necessary to remove the DMF under high vacuum, at high
temperature evaporation for the isolation of desired compound
10. In the medicinal chemistry method, impurity D was removed
by column chromatography purification. Incidentally, the direct
crystallization and phase separation was difficult. Needless to say,
it was not a practical method in a large scale synthesis. To solve
these issues, we investigated a new procedure, slow addition
(5−8 h) of the carbonyl imidazole that was prepared from
carboxylic acid 9 and CDI (carbonyldiimidazole) in CH3CN to
the suspension of guanidine carbonate in NMP, which gave 1%
of impurity D in a reaction mixture. After the extraction with
EtOAc, the desired free base 10 was prepared in a good purity.
Compound 10 was used in the next step as a i-PrOH solution
without isolation. After that, HCl salt formation was achieved in
aqueous i-PrOH in a good procedure, and the crystal form was
controlled perfectly. The details are described in the
Experimental Section. The overall yield of (S)-1 from 2 was
22.9%. Additionally, enantioisomer (R)-1 (free base: 22) was
synthesized from 20 by using almost the same procedure, and
the overall yield of (R)-1 from 2 was 22.7%.
240 nm. 2: 7.4 min, 18: 10.5 min, DMF: 1.7 min, 14: 2.6 min,
7: 10 min, cinchonidine: 1.4 min, brucine: 1.6 min.
Method B. YMC-Pack ODS-A, 5 μm, 4.6 mm × 150 mm
column, elution 0.05% TFA aq/CH3CN = 4/6, over 30 min,
1.0 mL/min, at 25 °C, with UV detection at 240 nm. 2:
2.9 min, 18: 3.1 min, DMF: 1.6 min, double-bond intermediate:
3.4 min, 19: 3.2 min, 14: 1.8 min, 7: 3.4 min, cinchonidine:
1.2 min, brucine: 1.3 min.
Method C. DAICEL CHIRALPAK AD-H, 5 μm, 4.6 mm ×
250 mm column, elution n-hexane/EtOH/TFA = 2.55 L/
0.45 L/3.0 g, over 60 min, 0.5 mL/min, at 25 °C, with UV
detection at 240 nm. 9: 15 min, 21: 23 min.
Method D. YMC-Pack ODS-A, 5 μm, 4.6 mm × 150 mm
column, elution 0.01 M K2HPO4 (adjust pH 7.0 by aq
H3PO4)/CH3CN = 7/3, over 60 min, 1.0 mL/min, at 25 °C,
with UV detection at 240 nm. 9, 21: 3.3 min, 10, 22: 13 min,
carbonyl imidazole intermediate: 50 min.
Method E. DAICEL CHIRALCEL OJ-RH, 5 μm, 4.6 mm ×
150 mm column, elution 0.01 M K2HPO4 (adjust pH 7.0 by aq
H3PO4)/CH3CN = 75/25, over 30 min, 1.0 mL/min, at 40 °C,
with UV detection at 305 nm. 22: 10 min, 10: 15 min.
9-Hydroxy-9-[3-(tetrahydropyran-2-yloxy)prop-1-yn-
1-yl]-9H-fluorene-2-carboxylic Acid (18, Mixture of All
Stereoisomers). A solution of 9-oxo-9H-fluorene-2-carboxylic
acid (2) (61.5 kg, 274.3 mol) and (RS)-2-(prop-2-yn-1-
yloxy)tetrahydropyran (17) (40.4 kg, 288.2 mol) in DMF
(490 L) was cooled to −25 °C, and potassium tert-butoxide
(64.7 kg, 576.6 mol) was added at −25 to 5 °C. The reaction
mixture was aged for 0.5 h at 0 °C, after which time HPLC
analysis indicated <1% starting material 2 remained (HPLC
methods A and B). To the batch was added water (190 L), and
2 N aqueous HCl was added until pH 7. After that, EtOAc (750 L)
and 2 N aqueous HCl were added to the batch (total amount of
HCl was 1.1 equiv of potassium tert-butoxide). The separated
aqueous layer was re-extracted with EtOAc (430 L). And the
two organic layers were combined and washed with water (280 L)
two times. The organic layer was concentrated in vacuo. After
that, toluene (470 L) was added to the batch, and the resulting
mixture was concentrated in vacuo. Toluene (890 L) and EtOAc
(110 L) were added to the residue, and the solvent was distilled
at atmospheric pressure to evaporate 420 L. The resulting slurry
was cooled to 0−5 °C and aged for 12 h at 0−5 °C, and filtered
and washed with toluene. The resulting wet cake was dried
in vacuo at 70 °C to afford the desired 18 with a 98% purity, via
HPLC methods A and B (92.8 kg, 92.8% yield).
CONCLUSION
■
A new synthetic route has been developed for large-scale
manufacture of the candidate drug compound (S)-1. The
undesirable features of the medicinal chemistry route and the
second generation synthesis method were avoided by employing
a new C−C bond formation method, a new spiro-ring formation
approach, and a practical optical resolution procedure. From
9-oxo-9H-fluorene-2-carboxylic acid (2) was manufactured
21 kg of compound (S)-1 in 22.9% overall yield in safe
operation. Impurities B, C, and D were controlled well at low
levels during the hydrogenation of alkyne, spiro-ring formation,
and amidation with guanidine carbonate. As a consequence,
a highly optically pure drug substance (S)-1 was prepared for
GMP delivery and (R)-1 was prepared for a preclinical study
campaign.
EXPERIMENTAL SECTION
■
General. Starting materials, reagents, and solvents were
obtained from commercial suppliers and used without further
1
purification. H and 13C NMR spectra were recorded in the
1
FAB-MS (neg): 363.1. ESI-MS (pos): 365.2. H NMR (500
specified deuterated solvent. Chemical shifts of 1H NMR
spectra are reported in parts per million (ppm) on the δ scale
from an internal standard of residual solvent (DMSO-d6
2.50 ppm) or TMS. Data are reported as follows: chemical
shift, multiplicity (s = singlet, d = doublet, t = triplet, dd =
doublet doublet, m = multiplet, and br = broad), coupling
constant (Hz), and integration.
MHz, DMSO-d6): δ 13.0 (1H, br), 8.15−8.20 (1H, m), 8.00−
8.05 (1H, m), 7.85−7.92 (2H, m), 7.65−7.70 (1H, m), 7.42−
7.50 (2H, m), 6.75 (1H, s), 4.67 (1H, s), 4.24 (1H, dd, J = 15.9,
2.1 Hz), 4.17 (1H, dd, J = 15.9, 2.8 Hz), 3.62−3.69 (1H, m),
3.30−3.37 (1H, m), 1.56−1.68 (2H, m), 1.39−1.47 (4H, m).
13C NMR (100 MHz, DMSO-d6): δ 166.9, (148.6, 148.5),
(148.2, 148.1), (142.7, 142.6), (137.3, 137.2), (130.7, 130.5),
129.4, (128.8, 128.1), (125.2, 125.1), (124.6, 124.5), 121.1,
120.3, 96.3, (87.0, 86.9), (78.2, 78.1), 73.2, (61.2, 61.1), 53.8,
29.7, 24.8, 21.0, (18.8, 18.7). IR (KBr)/cm−1: 3343, 2941, 1686,
1667, 1611, 1261, 1031. Mp: 160.1 °C (by DSC).
9-Hydroxy-9-[3-(tetrahydropyran-2-yloxy)propyl]-9H-
fluorene-2-carboxylic acid (19, mixture of all stereo-
isomers). A solution of 18 (46.3 kg, 127.1 mol) in MeOH
(400 L) was added potassium formate (85.5 kg, 1016 mol) and
10% Pd on carbon (6.95 kg, 50%-wet, M-type purchased from
Kawaken Fine Chemicals Co., Ltd.) and the batch was stirring
Chemical shifts of proton-decoupled 13C NMR spectra are
reported in ppm from the central peak of DMSO-d6 (39.5 ppm)
on the δ scale. IR spectra were performed by the KBr disk
method (JP16). HPLC was performed using the HITACHI D-
2500 or D-7500 system. The HPLC methods are described
below. Glass-lined reactors were used in all steps, and a vacuum
pump was used for concentration of organic layers.
HPLC Methods. Method A. YMC-Pack ODS-A, 5 μm,
4.6 mm × 150 mm column, elution 0.05% TFA aq/CH3CN =
6/4, over 30 min, 1.0 mL/min, at 25 °C, with UV detection at
660
dx.doi.org/10.1021/op200380z | Org. Process Res. Dev. 2012, 16, 654−663

Organic Process Research & Development
Article
at 10 − 35 °C for 2 h, after which time HPLC analysis indicate
<0.5% starting material 18 and <0.5% double-bond intermediate
remained (HPLC method B). After that, the batch was added
water (560 L) and filtered. The reaction vessel and lines filter
were washed with MeOH (50 L, twice). The combined filtrate
was filtered, via a <0.6 μm filter, into another vessel. The filtrate
was distilled at atmospheric pressure to evaporate 585 L. After
cooled to 25 °C, EtOAc (510 L) was added to the residue and
then aqueous HCl (38 wt % HCl 79.2 kg in water 190 L) was
added to the batch below 15 °C. After the phase separation, the
organic layer was washed with water (93 L) twice. One more
batch was conducted in the same manner. The organic later of
first batch and second batch were combined and concentrated
in vacuo to afford desired 19, it was added MeOH (560 L). The
solution was used next step without purification. Analytical pure
19 was obtained by concentration in laboratories.
The resulting wet cake was dried in vacuo at 70 °C to afford the
desired 7 with a 99% purity (52.4 kg, 87.7% yield). The level of
each impurity C was <0.5% by HPLC analysis (HPLC methods
A and B).
1
FAB-MS (pos): 267.1. FAB-MS (neg.): 265.1. H NMR
(500 M Hz, DMSO-d6): δ 12.95 (1H, s), 8.00 (1H, s), 7.95−
7.98 (1H, m), 7.82−7.87 (2H, m), 7.54−7.56 (1H, m), 7.35−
7.44 (2H, m), 4.25 (2H, t, J = 6.6 Hz), 2.35−2.47 (2H, m),
2.27−2.33 (2H, m). 13C NMR (100 M Hz, DMSO-d6): δ
167.2, 150.3, 149.5, 143.3, 137.6, 130.4, 130.2, 129.2, 128.9,
124.3, 123.9, 120.9, 120.0, 88.9, 69.2, 37.1, 26.8. Anal. Calcd for
C17H14O3: C, 76.68; H, 5.30. Found: C, 76.65; H, 5.34. IR
(KBr)/cm−1: 2982, 2953, 2868, 2853, 1683, 1611, 1294, 1049.
Mp: 230.1 °C (by DSC).
(9R)-8a-Cinchonan-9-ol Mono[(S)-4′,5′-dihydro-3′H-
spiro[fluorene-9,2′-furan]-2-carboxylate] (8). To the
solution of 7 (52.0 kg, 195.3 mol) in DMF (330 L) was
added brucine dihydrate (92.5 kg, 214.9 mol). The batch was
heated to 60 °C, and to the resulting solution was added water
(42 L). After the addition of seed crystals 20 (21 g) and water
(68 L) at 60 °C, the resulting slurry was cooled to 0 °C and
aged for 12 h, filtered, and washed with DMF/water, 3:1 (89 L).
The wet cake of crude-20 was stored at 5 °C. To the filtrate was
added EtOAc (780 L), and the resulting mixture was cooled to
5 °C. After that, a solution of 38 wt % HCl (34.3 kg) in water
(150 L) was added to the batch below 15 °C. The separated
aqueous layer was re-extracted with EtOAc (730 L). And these
two organic layers were combined and washed with water (370 L,
twice). The organic layer was concentrated in vacuo to afford
desired crude-9. At this stage, the level of enamtioisomer 21 was
4.6% by HPLC analysis (method C). To the residue was added
EtOAc (220 L), and the resulting mixture was concentrated
in vacuo. To the resulting residue was added methyl propionate
(1040 L). To the solution was added water (5.2 L) and
cinchonidine (28.8 kg, 97.8 mol), and the resulting mixture was
heated to 75 °C. After that, the solution was cooled to 70 °C,
and seed crystals of 8 (26 g) were added. The batch was cooled
to 30 °C and aged for 12 h. The resulting slurry was filtered
and washed with methyl propionate. The wet cake was dried
in vacuo at 70 °C to afford the desired 8 with a 99% purity, via
HPLC methods A and B (46.7 kg, 42.6% yield from 7). The
level of enantioisomer 21 was 0.34% by HPLC analysis (HPLC
method C).
FAB-MS (neg.); 367.1, ESI-MS (neg.); 367.0. 1H NMR (500
MHz, DMSO-d6): δ 12.89 (1H, br), 8.04 (1H, s), 7.94 − 7.99
(1H, m), 7.80−7.88 (2H, m), 7.50 − 7.54 (1H, m), 7.36 − 7.42
(2H, m), 5.70 (1H, s), 4.36 (1H, d, J = 12.2 Hz), 3.58 − 3.62
(1H, m), 3.40 − 3.43 (1H, m), 3.31 − 3.38 (1H, m), 3.08 −
3.17 (1H, m), 2.05 − 2.10 (2H, m), 1.58 − 1.65 (1H, m), 1.50
− 1.55 (1H,m), 1.25 − 1.47 (4H, m), 1.00 − 1.08 (2H, m). 13
C
NMR (100 MHz, DMSO-d6): δ 167.3, (150.5, 150.4), (149.9,
149.8), 143.6, 138.0, (130.0, 129.8), (128.8, 128.6), (124.4,
124.3), 123.7, 120.8, 119.8, (97.8, 97.7), 80.8, (66.5, 66.4),
(61.2, 61.1), (36.7, 36.6), (30.2, 30.1), 24.9, (24.3, 24.2), 20.7,
(19.2, 19.1), 14.0. IR (KBr)/cm−1: 2944, 2870, 1714, 1685,
1613, 1261, 1021. Mp: 128.1 °C (by DSC).
(RS)-9-Hydroxy-9-(3-hydroxypropyl)-9H-fluorene-2-
carboxylic Acid (14). To the solution of 19 in MeOH was
added 38 wt % HCl (0.6 kg, 6.25 mol) at 20 °C, and the
resulting mixture was aged for 2 h, after which time HPLC
analysis indicate <1% starting material 19 remained (HPLC
method B). After the addition of seed crystals of 14 (5g) and
toluene (470 L) to the reaction mixture, the batch was concentrated
in vacuo. To the residue was added toluene (560 L) and EtOAc
(180 L), and the resulting slurry was stirred at 65 °C and then
cooled to 0 °C. The batch was aged for 2 h at 0 °C and filtered and
washed with toluene. The resulting wet cake was dried in vacuo at
70 °C to afford the desired 14 with a 98% purity, via HPLC
methods A and B (64.1 kg, 88.7% yield in 2 steps).
1
FAB-MS (neg): 283.0. ESI-MS (neg): 282.8. H NMR (500
MHz, DMSO-d6): δ 12.88 (1H, br), 8.02 (1H, d, J = 1.3 Hz),
7.96 (1H, dd, J = 8.0, 1.6 Hz), 7.80−7.87 (2H, m), 7.50−7.53
(1H, m), 7.32−7.42 (2H, m), 5.67 (1H, s), 4.23 (1H, t, J = 5.2
Hz), 3.19 (2H, dd, J = 11.6, 6.4 Hz), 2.03 (2H, dd, J = 16.5, 7.0
Hz), 0.91−0.99 (2H, m). 13C NMR (100 MHz, DMSO-d6): δ
167.4, 150.7, 150.0, 143.6, 138.0, 130.0, 129.7, 128.8, 128.5,
124.4, 123.8, 120.8, 119.8, 80.9, 60.8, 36.7, 27.4. Anal. Calcd for
C17H16O4: C, 71.82; H, 5.67. Found: C, 71.78; H, 5.65. IR
(KBr)/cm−1: 3347, 3069, 2954, 1890, 1612, 1263, 1070. Mp:
222.6 °C (by DSC).
(RS)-4′,5′-Dihydro-3′H-spiro[fluorene-9,2′-furan]-2-
carboxylic Acid (7). To a solution of 14 (63.8 kg, 224.4 mol)
in diethyleneglycol dimethyl ether (diglyme, 380 L) was added
water (319 L). After that, a 64% aqueous solution of sulfuric
acid (8.5 kg, 55.5 mol) was added, and the batch was heated to
93 °C and aged for 24 h at the same temperature, after which
time HPLC analysis indicate <2% starting material 14 remained
(HPLC methods A and B). After that, to the batch was added
water (829 L) at 80−90 °C. The resulting slurry was cooled
to 25 °C and aged for 12 h, filtered, and washed with water.
21: free base of 20. FAB-MS (pos): 295.1 (cinchonidine),
FAB-MS (neg); 265.2 (compound 9). H NMR (500 MHz,
1
DMSO-d6): δ 8.85 (1H, d, J = 4.6 Hz), 8.34 (1H, d, J = 8.6
Hz), 7.96−8.05 (3H, m), 7.82 (2H, t, J = 7.8 Hz), 7.73 (1H, t,
J = 7.7 Hz), 7.53−7.61 (3H, m), 7.41 (1H, t, J = 7.5 Hz), 7.36
(1H, t, J = 7.3 Hz), 5.80−5.90 (1H, m), 5.43 (1H, d, J = 6.1
Hz), 5.00 (1H, d, J = 17.1 Hz), 4.95 (1H, d, J = 10.3 Hz), 4.22
(2H, t, J = 6.6 Hz), 3.31 (1H, br), 3.17 (1H, dd, J = 15.0, 8.0
Hz), 2.96 (1H, dd, J = 13.1, 9.8 Hz), 2.56−2.60 (4H, m), 2.22−
2.43 (4H, m), 1.79 (1H, d, J = 3.1 Hz), 1.69−1.77 (3H, d, J =
8.0 Hz), 1.50 (1H, br). 13C NMR (100 MHz, DMSO-d6): δ
168.3, 150.2, 150.0, 149.5, 149.2, 147.8, 142.2, 141.5, 137.9,
132.9, 130.2, 129.7, 128.8, 126.3, 125.6, 124.3, 124.0, 123.8,
120.6, 119.6, 119.0, 114.6, 96.8, 88.9, 69.5, 69.3, 69.1, 60.2, 54.9,
41.8, 38.6, 37.1, 35.2, 27.2, 26.8, 26.3. IR (KBr)/cm−1: 3197,
3071, 2972, 2946, 1613, 1373, 1268. Mp: 143.3 °C (by DSC).
(S)-N-(Diaminomethylidene)-4′,5′-dihydro-3′H-spiro-
[fluorene-9,2′-furan]-2-carboxamide (10). A solution of 38
wt % HCl (16.5 kg,172 mol) in water (116 L) was added to a
stirred suspension of 8 (46.0 kg, 82.0 mol) in EtOAc (700 L),
661
dx.doi.org/10.1021/op200380z | Org. Process Res. Dev. 2012, 16, 654−663

Organic Process Research & Development
Article
keeping the temperature below 15 °C. The separated organic
layer was washed with water (180 L) and then concentrated
in vacuo. To the residue was added toluene (180 L), and the
resulting mixture was concentrated in vacuo. To the residue 9
was added CH3CN (180 L), and the resulting mixture was
concentrated in vacuo. To the residue was added CH3CN
(110 L) and CDI (20.0 kg, 123.3 mol) at 20 °C, and the batch
was stirred at 10−25 °C for 1 h, after which time HPLC analysis
indicated <1% starting material 9 remained and >98% desired
carbony imidazole intermediate (HPLC method D). After that,
the solution was added for 8 h to a stirred suspension of
guanigine carbonate (110.8 kg, 615 mol) in NMP (164 L) at
20 °C, followed by a wash of CH3CN (20 L). And the batch was
stirred at 20 °C for 12 h, after which time HPLC analysis
indicated <3% starting material 9 remained (HPLC method D).
After that, the batch was concentrated in vacuo to about 160 L.
After that, EtOAc (650 L) and water (650 L) were added to the
batch, the resulting aqueous layer was re-extracted with EtOAc
(650 L), and these two organic layers were combined and
washed with water (330 L) five times. The organic layer was
concentrated in vacuo. After that, 2-propanol (260 L) was added
to the residue and concentrated in vacuo to afford desired free
base 10, to which was added 2-propanol (710 L). The solution
was used in the next step without purification.
123.7, 121.2, 120.3, 88.9, 69.3, 37.0, 26.8. IR (KBr)/cm−1:
3323, 3144, 1686, 1611, 1567, 1270, 1053, 738, 656. Anal.
Calcd for C18H17N3O2·HCl. Calc: C (62.88%), H (5.28%), N
(12.22%), Cl (10.31%). Found: C (62.82%), H (5.38%), N
20
(12.16%), Cl (10.24%). Mp: 248.5 °C (by DSC). [α]D
−9.04° (solvent: MeOH, 0.20 g/20 mL, 100 mm cell).
=
Preparation of (R)-1. 2,3-Dimethoxystrychnidin-10-one
Mono[(R)-4′,5′-dihydro-3′H-spiro[fluorene-9,2′-furan]-2-
carboxylate] (20). To a solution of crude-20 (1.32 kg of dry
base, 2.0 mol, 83% ee) in DMF (9.24 L) was added water (1.32 L)
at 60 °C and then to the solution was added seed crystals of 20
(1.3 g) and water (3.3 L) at 60 °C. The resulting slurry was
aged at the same temperature for 0.5 h and cooled to 0 °C and
aged for 2 h. The slurry was filtered and washed with DMF/
water = 8/2 (2.64 L). This wet cake was dissolved in DMF
(8.6 L). The solution was warmed to 60 °C, water (0.86 L) was
added, and to the solution was added seed crystals 20 (1.3 g)
and water (1.32 L) at 60 °C. The resulting slurry was aged at
the same temperature and cooled to 0 °C and aged for 2 h. The
slurry was filtered and washed with DMF/water = 8/2 (1.32 L).
The wet cake was dried in vacuo at 70 °C to afford the desired
20 with a 99% purity, via HPLC methods A and B (1.1 kg,
42.3% yield from 7). The level of enantioisomer 9 was 0.2% by
HPLC analysis (HPLC method C).
Analytically pure 9 was obtained by concentration in our
FAB-MS (pos): 395.2 (brucine). FAB-MS (neg): 265.3
(compound 21). 1H NMR (500 MHz, DMSO-d6): δ 8.00 (1H, s),
7.98 (1H, d, J = 7.9 Hz), 7.81 (2H, t, J = 7.9 Hz), 7.65 (1H, s),
7.55 (1H, d, J = 7.3 Hz), 7.34−7.43 (2H, m), 6.99 (1H, s),
5.89 (1H, br), 4.27−4.31 (2H, m), 4.23 (2H, t, J = 6.6 Hz),
4.05 (2H, d, J = 6.4 Hz), 3.95 (1H, s), 3.78 (1H, d, J = 10.7
Hz), 3.74 (3H, s), 3.73 (3H, s), 3.65 (1H, d, J = 14.3 Hz),
3.10−3.16 (2H, m), 2.85−2.94 (1H, m), 2.59−2.80 (2H, m),
2.55−2.58 (1H, m), 2.25−2.48 (4H, m), 1.78−1.85 (2H, m),
1.34 (1H, d, J = 14.3 Hz), 1.24 (1H, d, J = 10.7 Hz). 13C NMR
(100 MHz, DMSO-d6): δ 168.5, 168.0, 150.2, 149.3, 148.6,
145.8, 142.6, 139.4, 137.8, 135.5, 132.0, 130.3, 129.0, 128.9,
128.0, 124.3, 123.9, 123.6, 120.7, 119.7, 107.0, 100.5, 88.9, 76.7,
69.2, 63.6, 59.8, 59.0, 56.0 (2 carbons), 55.7, 51.8, 51.4, 49.6,
47.3, 41.6, 37.1, 30.7, 26.9, 25.9. Mp: 205.4 °C (by DSC).
(R)-N-(Diaminomethylidene)-4′,5′-dihydro-3′H-spiro-
[fluorene-9,2′-furan]-2-carboxamide Monohydrochlo-
ride ((R)-1). A solution of 38 wt % HCl (0.334 kg, 3.48 mol)
in water (11 L) was added to a stirred suspension of 20 (1.1 kg,
1.665 mol) in EtOAc (14.1 L), keeping the temperature below
15 °C. The separated organic layer was washed with water
(4.4 L) and then concentrated in vacuo. To the residue was
added toluene (3.7 L), and the resulting mixture was con-
centrated in vacuo. To the residue (21) was added CH3CN
(3.7 L), and the resulting mixture was concentrated in vacuo.
To the residue was added CH3CN (2.2 L) and CDI (405 g,
2.5 mol) at 20 °C, and the batch was stirring at 10−25 °C for
1 h, after which time HPLC analysis indicated <1% starting
material 21 remained and >98% desired carbony imidazole
intermediate (HPLC method D). After that, the solution was
added for 8 h to a stirred suspension of guanigine carbonate
(2.25 kg, 12.5 mol) in NMP (3.3 L) at 20 °C, followed by a
wash of CH3CN (0.45 L). And the batch was stirring at 20 °C
for 12 h, after which time HPLC analysis indicated <3% starting
material 21 remained (HPLC method D). After that, the batch
was concentrated in vacuo to about 3.3 L. After that, EtOAc
(13 L) and water (13 L) were added to the batch and the
resulting aqueous layer was re-extracted with EtOAc (13 L).
And the two organic layers were combined and washed with
20
laboratories. Free carboxylic acid 9: [α]D = −4.91° (solvent:
MeOH, 0.20 g/20 mL, 100 mm cell). Mp: 192.5 °C (by DSC).
1
Free base 10: FAB-MS (pos); 308.2. H NMR (600 MHz,
DMSO-d6): δ 8.17 (1H, s), 8.11 (1H, d, J = 7.8 Hz), 8.05 (2H,
br), 7.76 (1H, d, J = 7.2 Hz), 7.73 (1H, d, J = 7.8 Hz), 7.53
(1H, d, J = 7.2 Hz), 7.38 (1H, t, J = 7.5 Hz), 7.32 (1H, t, J = 7.5
Hz), 6.70 (2H, br), 4.20−4.26 (2H, m), 2.34−2.44 (2H, m),
2.24−2.32 (2H, m). 13C NMR (150 MHz, DMSO-d6): δ 175.5,
163.0, 150.2, 148.6, 141.0, 139.1, 138.3, 129.4, 128.7, 128.4,
123.8, 123.7, 120.3, 119.0, 89.0, 60.0, 37.2, 26.8. IR (KBr)/
cm−1: 3335, 3198, 1634, 1567, 1272, 1610, 1045, 756, 658. Mp:
128.9 °C (by DSC).
(S)-N-(Diaminomethylidene)-4′,5′-dihydro-3′H-spiro-
[fluorene-9,2′-furan]-2-carboxamide Monohydrochlo-
ride ((S)-1). A solution of 10 in 2-propanol was added via a
0.6 μm filter, into another vessel, followed by a wash of 2-
propanol (50 L). After addition of water (8 L) to the solution,
seed crystals (S)-1 (26 g) were added to the batch at 35 °C. After
that, a solution of 38 wt % HCl (8.24 kg, 85.9 mol) in water (15 L)
was added via a 0.6 μm filter to the batch at 35 °C, and the resulting
slurry was aged for 1 h at 35 °C. Next, the batch was heated to
65 °C and aged for 1 h. After that, the slurry was cooled to 25 °C
and aged at the same temperature for 12 h. Before the filtration, the
crystal form was checked by XRD analysis to meet the desired form.
After that, the slurry was filtered, washed with 2-propanol, and dried
in vacuo at 60 °C to give 21.0 kg of (S)-1 having a purity of 99.1%
(HPLC method D). Optical isomer 22 was observed to be 0.14%
by Chiral HPLC analysis (HPLC method E), 74.4% yield from 8.
The overall yield was 22.9% from 9-oxo-9H-fluorene-2-
carboxylic acid (2).
1
FAB-MS (pos); 308.0. H NMR (600 MHz, DMSO-d6): δ
12.26 (1H, s), 8.93 (2H, br), 8.58 (2H, br), 8.43 (1H, d, J = 1.8
Hz), 8.18 (1H, dd, J = 7.8, 1.8 Hz), 7.97 (1H, d, J = 7.8 Hz),
7.88 (1H, d, J = 6.6 Hz), 7.58 (1H, d, J = 7.2 Hz), 7.39−7.46
(2H, m), 4.35 (1H, dd, J = 14.4, 8.4 Hz), 4.22 (1H, dd, J = 14.4,
6.6 Hz), 2.50−2.58 (1H, m), 2.34−2.44 (2H, m), 2.26−2.32
(1H, m). 13C NMR (150 MHz, DMSO-d6): δ 167.2, 155.8,
150.5, 149.9, 144.3, 137.2, 130.3, 130.0, 129.6, 129.0, 124.0,
662
dx.doi.org/10.1021/op200380z | Org. Process Res. Dev. 2012, 16, 654−663

Organic Process Research & Development
Article
Chem. 1981, 46, 3091. (d) Alonso, D.; Font, J.; Ortuno, R. M. J. Org.
Chem. 1991, 56, 5567.
(6) (a) Verkruijsse, H. D.; Graaf, W.; Brandsma, L. Synth. Commun.
1988, 18, 131. (b) Nakatsuji, S.; Nakazumi, H.; Fukuma, H.; Yahiro,
T.; Nakashima, K.; Iyoda, M.; Akiyama, S. J. Chem. Soc., Perkin Trans. 1
1991, 8, 1881.
water (6.7 L) five times. The organic layer was concentrated
in vacuo. After that, 2-propanol (5.1 L) was added to the
residue and concentrated in vacuo to the afford desired free
base 22, to which was added 2-propanol (14.3 L). The solution
was used in the next step without purification.
The solution of 22 in 2-propanol was added via a 1 μm filter
into another vessel, followed by a wash of 2-propanol (1.0 L).
After addition of water (162 mL) to the solution, the seed
crystals (R)-1 (0.51 g) were added to the batch at 35 °C. After
that, a solution of 38 wt % HCl (167.1 g, 1.74 mol) in water
(0.3 L) was added via a 1 μm filter to the solution at 35 °C, and
the resulting slurry was aged for 1 h at 35 °C. Next, the batch
was heated to 65 °C and aged for 1 h. After that, the slurry was
cooled to 25 °C and aged at the same temperature for 1 h.
Before the filtration, the crystal form was checked by XRD
analysis. After that, the slurry was filtered and washed with
2-propanol and dried in vacuo at 60 °C to give 0.425 kg (74.3%
yield from 20) of (R)-1 having a purity of 99.1% (HPLC
method D). Optical isomer 10 was observed, 0.2% by Chiral
HPLC analysis (HPLC method E). The overall yield was 22.7%
from 9-oxo-9H-fluorene-2-carboxylic acid 2.
(7) (a) Ranu, B. C.; Sakkar, A.; Sankar, K. G.; Ghosh, K. J. Indian
Chem. Soc. 1998, 75, 690. (b) Ram, S.; Ehrenkaufer, R. E. Synthesis
1988, 91. (c) Anwer, M. K.; Sherman, D. B.; Roney, J. G.; Spatola, A. F.
J. Org. Chem. 1989, 54, 1284.
(8) (a) Baidossi, M.; Joshi, A. V.; Mukhopadhyay, S.; Sasson, Y.
Synth. Commun. 2004, 34, 643. (b) Wiener, H.; Blum, J.; Sasson, Y.
J. Org. Chem. 1991, 56, 4481.
(9) (a) Hasegawa, K. Fine Chem. 2002, 31, 5. (b) Rylander, P. N.
Catalytic Hydrogenation over Platinum Metals; 1967, 449, Academic
Press. (c) Rylander, P. N. Catalytic Hydrogenation in organic synthesis;
1979, 271, Academic Press.
(10) (a) Parham, W. E.; Sayed, Y. A. Synthesis 1976, 116. (b) Eaton,
P. E.; Cooper, G. F.; Johnson, R. C.; Mueller, R. H. J. Org. Chem. 1972,
37, 1947. (c) Sutherland, I. O.; Ramsay, M. V. J. Tetrahedron 1965, 21,
3401. (d) Cannone, P.; Belanger, D.; Lemay, G. Synthesis 1980, 301.
(e) Garcia, D.; Foubelo, F.; Yus, M. Tetrahedron 2008, 64, 4275.
(f) Foubelo, F.; Moreno, B.; Yus, M. Tetrahedron 2004, 60, 4655.
(g) Yus, M.; Soler, T.; Foubelo, F. Tetrahedron 2002, 58, 7009.
(h) Yus, M.; Foubelo, F.; Ferrandez, J. V.; Bachki, A. Tetrahedron
2002, 58, 4907.
Free carboxylic acid 21 (analytically pure 21 was obtained by
20
concentration in our laboratories): [α]D = +5.28° (solvent:
MeOH, 0.20 g/20 mL, 100 mm cell).
(11) Compound 20 and 8 were easily converted to free carboxylic
acid by treatment with aqueous HCl solution and extracted with
EtOAc as shown in the Experimental Section. In a preliminary test,
brucine and cinchonidine were recovered after the addition of an
aqueous solution of NaOH to the resulting acidic aqueous layer.
20
Compound (R)-1. [α]D = +8.77° (solvent: MeOH, 0.20
g/20 mL, 100 mm cell).
The analytical data of 21, 22, and (R)-1 (mass, IR, 1H NMR
and 13C NMR spectra) were in agreement with the above data
(compounds 9, 10, and (S)-1).
AUTHOR INFORMATION
■
Corresponding Author
*Telephone: +81-293-23-5478. Fax: +81-293-24-2708. E-mail:
shinya.yoshida@astellas.com.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We are grateful to Dr. H. Yamada, Ms. A. Moritomo, and Ms.
T. Matsuzawa for helpful discussions about medicinal chemistry
synthetic routes to (S)-1 and (R)-1, and we thank Dr. T.
Tokunaga, Mr. H. Yamashita, Mr. Y. Awamura, and Ms. N.
Amenomiya for analytical study.
REFERENCES
■
(1) (a) Akuzawa, S.; Ito, H.; Watanabe, T.; Yamada, H.
WO2006085510. (b) Yamada, H.; Itahana, H.; Moritomo, A.;
Matsuzawa, T.; Harada, K.; Fujiyasu, J.; Koga, Y.; Oku, M.; Okazaki,
T.; Akuzawa, S.; Watanabe, T. WO2005080322.
(2) (a) Minoda, T. Chromatography 2010, 31 (1), 37. (b) Dahl,
A. C.; Mealy, M. J.; Nielsen, M. A.; Lyngsø, L. O.; Suteu, C. Org. Process
Res. Dev. 2008, 12, 429.
(3) The SMB separation of 14 and 7 was also attempted. But it did
not give efficient results in comparison with the separation of
compound 8.
(4) Actually, it had difficulty with controlling the reaction in the
hundred-gram scale synthesis. In step 2, magnesium was added to the
batch very carefully and slowly. It would be challenging to operate on
scale-up synthesis. Magnesium: MSDS, Kanto Chemical Co., Inc.;
mesh size, turning; not activated.
(5) (a) Canonne, P.; Foscolos, G. B.; Belanger, D. J. Org. Chem.
1980, 45, 1828. (b) Picard, P.; Moulines, J. Bull. Soc. Chim. Fr. 1974,
2256. (c) Canonne, P.; Belanger, D.; Lemay, G.; Foscolos, G. B. J. Org.
663
dx.doi.org/10.1021/op200380z | Org. Process Res. Dev. 2012, 16, 654−663