Development of a manufacturing process for 1-(1-pyridin-2-yl methyl-piperidin-4-yl)-1H-indole: A key intermediate for protein kinase C inhibitor LY317615

Article abstract of DOI:10.1021/op060068k

This contribution describes process development leading to the production of 1-(1-pyridin-2-yl methyl-piperidin-4-yl)-1H-indole (11). The title compound 11 was produced via a Leimgruber-Batcho indole synthesis using key intermediates 2-(2,2-dimethoxyethyl

Full text of DOI:10.1021/op060068k

Organic Process Research & Development 2006, 10, 1205−1211
Development of a Manufacturing Process for 1-(1-Pyridin-2-yl
methyl-piperidin-4-yl)-1H-indole: A Key Intermediate for Protein Kinase C
Inhibitor LY317615
†
,†
†
‡
‡
Sathish Boini, Kenneth P. Moder,* Radhe K. Vaid, Micheal Kopach, and M. Kobierski
Chemical Product Research and DeVelopment, Eli Lilly and Company, Tippecanoe Laboratories, 1650 Lilly Road
Shadeland, Indiana 47909, U.S.A., and Chemical Process Research and DeVelopment, Eli Lilly and Company, Lilly
Corporate Center, Indianapolis, Indiana 46285, U.S.A.
3
Abstract:
sulfonate (9) and subsequent coupling with 6 were identified
2
This contribution describes process development leading to the
production of 1-(1-pyridin-2-yl methyl-piperidin-4-yl)-1H-indole
as key steps which required optimization. As part of the
overall process development strategy, each step in the
synthesis was examined with respect to optimum conditions
and special considerations needed for reagents, reaction
conditions, robustness, and minimization of the environmen-
tal impact.
Amino Acetal 6 Synthesis. Enamines 3 and 4. Genera-
tion of enamine (E)-1-[2-(2-nitrophenyl)ethenyl]pyrrolidine
3 using DMF-dimethylacetal in DMF and pyrrolidine at 80
(11). The title compound 11 was produced via a Leimgruber-
Batcho indole synthesis using key intermediates 2-(2,2-dimethoxy-
ethyl)benzenamine (6) and, 1-(2-pyridinylmethyl)-4-piperidi-
none camphor sulfonate (9). Direct crystallization of 11 from
IPA or ethanol-water was developed to provide (11) with <1%
impurities and high yield (78%). The combined process leads
to a five-step synthesis of 11 that was efficient and reflected
Eli Lilly and Company’s commitment to implementation of
environmental friendly processes whenever feasible.
2
°C has been described previously. It was later determined
by H NMR that this reaction results in a mixture of enamines
1
3
and (E)-1-[2-(2-nitrophenyl)ethenyl]dimethylamine 4 in a
ratio of 85:15 as shown in Scheme 1. Impediments to full
scale production of 3 and 4 were identified to be as follows:
Introduction
4
(1) purification of 3 and 4 by high vacuum distillation would
Most solid tumors increase in mass though proliferation
of malignant and stromal cells leading to formation of tumor
vasculatures. LY317615, an investigatory new drug, is an
inhibitor for protein kinase C which is presently under
evaluation in the clinic for the treatment of glioblastoma.
The mode of action of LY317615 is to prevent angiogenesis
not be feasible due to thermal instability; (2) conducting the
conversion of 2 to 3 and 4 at elevated temperatures bordered
4
on the thermal stability of the reaction mixture; (3) the need
for the aqueous extractions so as to remove DMF which
impedes conversion of 3 and 4 to 1-(2,2-dimethoxyethyl)-
2-nitrobenzene (5); and (4) complications encountered during
1
by cancer cells during tumor growth.
the aqueous extractive workup to remove DMF in the form
of tary black-on-black layer separations coupled with the
propensity to form stable emulsions. The key to the solution
was DMF. Could this solvent be eliminated and the process
conducted so as to afford 3 and 4 safely and in a form capable
with the next step?
Our goal was to evolve the operative small scale synthesis2
for the key starting material 1-(1-pyridin-2-yl methyl-
piperidin-4-yl)-1H-indole (11) (Scheme 1) into a scalable,
robust manufacturing process capable of producing multiton
quantities of high quality 11 in a cost-effective and environ-
ment friendly manner where feasible.
Eliminating the DMF permitted the facile removal, by
simple atmospheric distillation, of the methanol at temper-
atures between 79 and 85 °C. The DMF cosolvent used in
the initial protocol impeded the distillation of the MeOH
byproduct thus prolonging the reaction time. Removing
DMF, permitting ready removal of methanol, resulted in
accelerating the conversion rate from 24 to 10 h. Eliminating
DMF also eliminated the need for the subsequent problematic
extractions. This change translated into several production
oriented bonuses. First, the new process required a minimum
equipment set. Second, the process cycle time was improved
from ∼36 h to ∼15 h. Third, productivity would be improved
as the batch size could be increased by 100%. Fourth, by
Results and Discussion
Faul et al. developed a synthetic protocol that was
sufficient for lab scale and small pilot scale production of
2
-(2,2-dimethoxyethyl)benzamine (6) from o-nitrotoluene 1
2
(Scheme 1). As the project proceeded towards full scale
production, 1-(2-pyridinylmethyl)-4-piperidinone camphor
*
To whom correspondence should be addressed. E-mail: moder_kenneth_p@
lilly.com.
†
Chemical Product Research and Development.
Chemical Process Research and Development.
‡
(
1) Graff, J. R.; McNulty, A. M.; Hanna, K. R.; Konicek, B. W.; Lynch, R. L.;
Bailey, S. N.; Banks, C.; Capen, A.; Goode, R.; Lewis, J. E.; Sams, L.;
Huss, K. L.; Campbell, R. M.; Iversen, P. W.; Neubauer, B. L.; Brown, T.
J.; Musib, L.; Geeganage, S.; Thornton, D. Cancer Res. 2005, 65 (16),
7
462-9.
(3) Faul, M. M.; Kobierski, M. E.; Kopach, M. E. J. Org. Chem. 2003, 68
(14), 5739-5741.
(4) Accelerated rate calorimetry (ARC) data indicated an uncontrolled exotherm
starting at 120 to 130 °C that negated higher temperatures as an option.
(
2) Faul, M. M.; J. L.; Kobierski, M. E.; Kopach, M. E.; Krumrich, C. A.;
Staszak, M. A.; Udodong, U.; Vicenzi, J. T.; Sullivan, K. A. Tetrahedron
2
003, 59, 7215-7229.
1
0.1021/op060068k CCC: $33.50 © 2006 American Chemical Society
Vol. 10, No. 6, 2006 / Organic Process Research & Development
•
1205
Published on Web 10/24/2006

Scheme 1. Lab and pilot scale route to 11a
a
Reagents and conditions: (a) (i) DMF, pyrrolidine, 80 °C; (ii) MTBE/water; (iii) brine; (iv) MgSO4; (b) (i) TMSCl, MeOH; (ii) distill MeOH; (iii) EtOAc/water;
(iv) concentrate; (v) MgSO4; (vi) heptane/EtOAc, silica gel, concentrate; (c) 5% Pd/C, MeOH, H
2 2 3
, and concentration; (d) (i) Na CO /ACN; (ii) EtOAc, CSA; (e) (i)
THF/C CO H, NaBH(OAc) NaBH ; (f) (i) TFA/EtOAc; (ii) concentrate; (iii) IPA.
2
H
5
2
3
4
that retarded the conversion of 3 and 4 to 5. At issue with
the original TMSCl method was the need for an extensive
extractive workup, silica gel treatment, and finally a tritu-
ration with hot heptane to remove impurities. A literature
survey revealed that several methods with alternate acids
6
were known to accomplish the desired transformation. Bulk
commodity sulfonic acids such as p-toluenesulfonic acid
monohydrate and sulfuric acid in methanol were found to
work well. These sulfonic acids were not sensitive to residual
DMF and were compatible with the new process enamines.
In our hands, the p-TsOH process was more robust than the
sulfuric acid modification with respect to reaction rate and
impurity formation as confirmed by HPLC (area%). The
sulfuric acid option was prone to generation of additional
impurities, as detected by HPLC analysis. Our initial concern
Figure 1. LY317615.
eliminating the DMF one eliminated the problematic extrac-
tions so that the final product to waste ratio decreased from
was whether residual acid could lead to reversion to 3 and
1:51 g/mL to 1:1.6 g/mL without sacrificing yield, 95-99%.
7
4
as predicted by Gallo et al. Upon extending the distillation
The final unexpected bonus was ARC data on the new
reaction mixture that indicated that the thermal decomposi-
tion onset temperature was now 140 vs 120 °C. The
temperature safety gap was now ∼60 °C that represents a
huge safety improvement for production operations. The
resulting MeOH stream could either be recovered or used
as primary fuel for a waste incineration unit.
time for the removal of methanol postconversion to acetal,
reversion was detected by HPLC analysis. Extensive heating
resulted in near-complete reversion of 5 to enamines. The
portent for reversion would be of concern in a production
environment. The excess acid that not neutralized by the
displaced pyrollidine was addressed in the original process
after removal of the MeOH and dilution with EtOAc. The
excess acid and salts were removed using aqueous bicarbon-
ate or carbonate extractions. The need was to introduce a
base prior to distillation of the methanol to prevent product
reversion.
The improvements to the process for 3 and 4 would be
for naught if one could not convert the new process material
to 5. Of concern was the impact, on step 2, of the step 1
reaction byproducts, N-formylpyrrolidine and corresponding
5
orthoamide, as confirmed by HPLC retention time matching
Potential options included liquid organic bases, deemed
too expensive, or introduction of a cost-effective inorganic
and LC/MS data. Laboratory trials with the new process for
3
and 4 indicated no interference in the conversion to 5.
Acetal 5. Enamines 3 and 4 were initially converted to 5
(6) (a) Vinogradov, V. M.; Dalinger, I. L.; Starosotnikov, A. M.; Shevelev, S.
A. MendeleeV Commun. 2000, (4), 140-141. (b) Todd, M. H. Org. Letters
via in situ generated anhydrous HCl using TMSCl in
methanol. The TMSCl method was sensitive to residual DMF
1
999, 1 (8), 1149-1151. (c) Chem. Pharm. Bull. 1981, 29 (3), 726-738.
(d) Cornell, L.; Richards, I. C. Schering Agrochemical LTD UK, patent,
WO9107385, 1991.
(
5) Wawer, I.; Osek, J. J. Chem. Soc., Perkin Trans. II 1985, (10), 1669-
(7) Bianchetti, G.; Pocar, D.; Croce, P. D.; Gallo, G. G.; Vigevani, A.
Tetrahedron Lett. 1966, 15, 1637-1642.
1
671.
1206
•
Vol. 10, No. 6, 2006 / Organic Process Research & Development

Table 1. Stability of 5 postdistillation of methanol
reaction mixture. Granulated or powdered Na was
CO
2
3
areaa
% areaa
% areaa
preferred over calcium carbonate because Na CO repro-
2
3
%
(
3 and 4) impurities
(5)
ducibly inhibited reversion and degradation under normal
and stressed conditions. The deciding factor was that the
sodium p-toluenesulfonate byproduct was readily removed
by filtration, whereas the calcium salt filtered poorly. The
1
initial rxn mixture
<0.1
Flash Distillation
0.7
99.3
2
3
4
5
6
control 1 (no base added)
control 2 (amb. 24 h stir)
0.5
1.1
0
0
0
1.3
2.3
1.4
0.9
1.6
98.7
97.7
98.6
99.1
98.4
2 3
use of p-TsOH in this step followed by use of Na CO prior
to the vacuum distillation of methanol afforded 5 in >98%
yield and >97% by HPLC (area %). Following an extractive
water workup, the product was either isolated as an oil by
removal of the extraction solvent toluene or held as a toluene
solution. Elimination of the extensive workup protocol of
the original process improved the product to waste ratio from
1:84 g/mL to 1:23 g/mL. Ultimately, the use of a toluene
solution of 5 may lead directly into the reduction process
for 6. Further optimization of equivalents of acids should
further improve the performance of this process.
K
2
CO
CaCO
Na CO
3
3
2
3
Stressed: 55-60 °C for 30 min
7
8
9
1
control 3 (no base added)
3.72
52.8
54.5
0.9
47.2
45.5
99.1
98.8
K
2
CO
CaCO
Na CO
3
0
0
0
3
0
2
3
1.2
a
Values reflect area percent without the contribution of p-TsOH where
applicable.
Amino Acetal 6. Alcohol solvents were employed in the
low H pressure catalytic reduction of 5 to 6. Use of alcohol
2
base. The latter option was explored further. Addition of
inorganic bases such as Na CO , K CO , and CaCO , prior
2 3 2 3 3
solvents resulted in either leaching of the palladium from
the carbon support or partial digestion of the carbon support.
In either case, the amino acetal 6 was contaminated with
black solids after solvent removal. Alternate solvents, MTBE,
THF, EtOAc, and toluene, were evaluated with toluene being
found superior with regards to lack of impurity generation,
ease of catalyst removal, and limited catalyst support
digestion or palladium leaching. Using toluene as the
reduction solvent, yields > 95% were achieved with assays
between 92 and 96% by HPLC (area %). The major impurity
was indole.
to the distillation of methanol, was hoped to be effective at
preventing the reversion reaction and other degradation
pathways. Solid base was chosen so as not to introduce water
and risk product hydrolysis. Solids addition to a flammable
solvent mixture in production is not optimum; however the
operation was feasible in production. The relatively small
molar proportion of solid base did not give rise to concerns
about off gassing, as the solids addition would be controlled.
Laboratory trials were conducted to mimic the worse case,
rapid charging of base. Considering the large heat sink, the
volume of chilled MeOH (see experimental), the heat of
neutralization should be dissipated quickly. Equally, the small
As amino acetal 6 was to be a key intermediate, product
quality was paramount. The standard release assay for 6
needed to be >98% by HPLC (area %). The technical
material produced by the process described fell just short of
this goal. To achieve an assay of >98% by HPLC (area %),
the crude product oil was vacuum distilled. The overall yields
with distillation were 80 to 85% having assays of >98.5%
by HPLC (area %). The major impurity, indole, was
volume of CO
2
expected to be released should be soluble in
the reaction mixture. These expectations would be confirmed
by laboratory and 7.5 L scale trials. The question was which
base to use?
Multiple replicate trials were conducted to test the three
prospective bases under control and stress conditions. Entry
1
in Table 1 reflected the reaction mixture prior to solvent
1
confirmed by HPLC retention time match and H NMR and
removal, i.e., T . Entries 2, 4, 5, and 6 indicated the amount
o
comprised the bulk of the precut. The product to waste ratio
remained consistent at 1:5 g/mL. The above process sequence
has been run at a 7.4 L scale achieving the stated yields and
assays. The combined product to waste ratio, 2 thru 6, was
1:30 kg/L for the modified process vs the initial process ratio
average of 1:140 kg/L.
of impurities generated with and without base addition under
rapid solvent removal. To test robustness, the sample in entry
1
was allowed to stir for 24 h prior to rapid solvent removal,
and that in entry 2 was to test reaction mixture stability. The
level of impurities nearly doubled. In a production environ-
ment, solvent removal, either atmospheric or under a vacuum,
would require heating the vessel and thus subjecting material
at the interface to elevated skin temperatures. A worse case
production scenario, the entire reaction mixture was subjected
to elevated temperatures for a short period of time, entries
Synthesis of 11. The parallel piece of the synthesis
8
involved the alkylation of 8 with 2-picolyl chloride 7 and
salt formation to generate 1-(2-pyridinylmethyl)-4-piperidi-
none camphor sulfonate (9). The process involved condensa-
tion of 6 with 9 followed by reductive amination to 10, ring
closure of 10 under acidic conditions to form 11, and
7
-10, Table 1. With no base added, entry 7, extensive
degradation was observed with some product reversion to
starting material as determined by HPLC. Addition of K
CO , entry 8, prevented the reversion reaction but not the
degradation. The potassium salt of p-TsOH was soluble under
the test conditions. Addition of either Na CO or CaCO
2
,3
2
-
isolation and purification by crystallization. While es-
sentially following Faul’s procedure, subtle changes were
incorporated to improve process reliability while maintaining
the high level of product quality needed for downstream
processing.
3
2
3
3
afforded excellent product stability, which were further
evaluated as options for the production process. Both the
sodium and calcium salts of p-TsOH were insoluble in the
(8) Picolyl chloride HCl had a isomeric purity of >99% by HPLC (area %).
Vol. 10, No. 6, 2006 / Organic Process Research & Development
•
1207

These subtle, but important, process modifications began
with the preparation of 9. Conducting the alkylation of 8
with 7 under a 2-fold increase in concentration coupled with
increasing the temperature of the CSA salt formation step
from 55 °C to 70 °C permitted isolation of 9 under controlled
crystallization conditions vs precipitation. The original
°C reduced the cyclization time from 15 to 24 h to 4-5 h
without any decrease in yield or product quality.
The technical product was isolated in a similar manner
as that described for 10 with the exception that the pH range
was lowered to 8-9. Final solvent exchange to IPA followed
by slow cooling permitted the controlled crystallization of
11 in 75-80% yield with assays >99.5% by HPLC (area
%). Isolated yields could be improved substituting MeOH/
water (86%) or EtOH/water (82%) for IPA; however the
quality suffered respectively (96% and 98%). As 11 was a
key starting material, product quality was paramount and IPA
was retained as the isolation solvent.
process for conversion of 9 to 10 was accomplished with
9
the reagent NaBH(OAc)
3
at 50 °C with yields in the range
50-60%. These conditions, using acetic acid as solvent,
caused formation of acid-catalyzed cyclization of 6 to indole
and the formation of 1-acetylindole. Reducing the reaction
temperature to 20-25 °C improved yields to 80% and
suppressed impurity formation. Drawbacks to using the
triacetoxy borohydride reagent were instability of the reagent
upon storage and solidification of the reagent below 18 °C.
Conclusions
The work disclosed here has provided an efficient and
reagent frugal process for the manufacture of acetal 6. The
synthesis of 6 involved formation of 3 and 4 under solvent-
free conditions. The conversion of 3 and 4 to 5 using p-TsOH
in methanol followed by carbonate neutralization of excess
acid avoided multiple solvent extractions and both a silica
gel plug filtration and subsequent heptane trituration. The
toluene solution of 5 thus generated was reduced to 6 in using
Pd/C with low-pressure hydrogen. Purification of crude 6
by vacuum distillation provided high quality product >98.5%
by HPLC (area %). The product to waste ratio from the initial
process (1:140 kg/L) was improved to (1:30 kg/L) with
improved product quality, reductions in cycle times, and no
yield erosion. Modifications were described concerning the
synthesis of 9 and improvements to the borohydride reduction
10
Alternate acyloxborohydrides have been used for similar
conversions. Generation of a 1 M solution of NaBH-
2 5 3
(OCOC H ) in THF vs propionic acid performed admirably
and exhibited excellent stability for several weeks when
refrigerated; see Experimental Section. Using the alternate
borohydride reagent with a propionic acid solution of 9 at
-
20 to -10 °C afforded in situ yields of 10 of 85-90%.
Attempts at inverse addition or substitution of propionic acid
with EtOAc, THF, MTBE, or toluene resulted in elevated
levels (25-77% vs <8%) of the alcohol impurity 12.
2 5 3
reagent, NaBH(OCOC H ) . Conversion of 10 using a
modified TFA cyclization followed by direct recrystallization
from IPA provided a facile route to our target 1-(1-pyridin-
2
-yl methyl-piperidin-4-yl)-1H-indole, 11. Modifications in
the conversion of 7 to 11 did not afford any product to waste
improvements. However the modifications improved cycle
times, process reliability, and, in the case of the synthesis
of 9, batch size enhancement. The revised process route is
shown in Scheme 2.
Eli Lilly and Company successfully completed a 7.5 L
scale-up of the above-described chemistry with multiple lots
at each stage from 2 through 11. The 7.5 L scale-up mirrored
the laboratory results. Eli Lilly and Company chose to use
a contracted manufacturer for the supply of 11 and provided
all required technical documents. The contract manufacture
successfully translated the chemistry as described on labora-
tory scale and then more recently completed a successful
The original workup of the reduction reaction involved
addition of EtOAc followed by pH adjustment to 8.5-10 at
7-55 °C with aq. NaOH to neutralize excess acid. The
4
elevated temperatures were to prevent crystallization of the
sodium salts. The subtle change from aq. NaOH to aq. KOH
permitted the extractions to be completed at ambient tem-
perature thus avoiding issues with salt crystallization. The
potassium acid salt was more soluble than the sodium salt
and also avoided potential hydrolysis of EtOAc.
Final conversion of 10 to 11 was accomplished by first
concentrating the EtOAc layer followed by acid cyclization.
Of the following acids tested, acetic, formic, trifluoroacetic
5
-10 lot per step campaign at a greater than 1000 L scale
(
TFA), and chloroacetic acids, TFA was preferred. Selection
using the described chemistry with the claimed results.
of TFA was based upon reaction time, impurity profile by
HPLC (area %), and extent of conversion as the desired
quality attributes. Initial use of TFA (2.5 equiv) at 45-50
Experimental Section
°
C afforded clean conversion, 88-93% within 15 to 24 h.
General Methods. Reaction completion and product
purity for all steps were evaluated by HPLC using the
following RP-HPLC conditions: (Initial Route) Zorbax C-8
Increasing the charge of TFA to 4.9 equiv with temperature-
controlled addition (<30 °C) followed by heating to 44-48
2
5 cm × 4.6 mm, flow 1.0 mL/min; wavelength ) 250 nm;
(
9) Abdel-Magid, A. F.; Carson, K. G.; Harris, B. D.; Maryanoff, C. A.; Shah,
R. D. J. Org. Chem. 1996, 61, 3849.
temperature 23 °C; injection volume: 20 µL of a ca. 0.05%
solution in ACN/water 1:1 v/v; eluent A ) ACN, B ) (1.5
mL of Et
(
10) (a) McGill, J. M.; Labell, E. S.; Williams, M. A. Tetrahedron Lett. 1996,
3
7, 3977. (b) McGill, J. M.; Burks, J. E.; Espinosa, L.; Labell, E. S.; Ritter,
3 3 4 2
N/1.5 mL of H PO /1 L of H O) pH 3.0; and
A. R.; Speakman, J. L.; Williams, M. A.; Bradley, D. A.; Haenl, M. G.;
Schmid, C. R. Org. Process Res. DeV. 1997, 1, 198.
gradient: (0 min) A ) 20%, B ) 80%; (1 min) A ) 20%,
1208
•
Vol. 10, No. 6, 2006 / Organic Process Research & Development

Scheme 2. Refined process for synthesis of 11a
a
Reagents and conditions: (a) (i) Pyrrolidine, 80 °C, distill volatiles; (b) (i) p-TsOH, MeOH; (ii) Na
2
CO
3
, distill MeOH; (iii) toluene/water; (iv) concentrate; (c)
CO H, NaBH ; (f) (i) C CO H, solution
5
% Pd/C, toluene, H
2
, and concentration; (ii) vac. distill product; (d) (i) Na
2
CO
3
/ACN; (ii) EtOAc, CSA; (e) (i) THF/C
2
H
5
2
4
H
2 5
2
from (e); (ii) EtOAc, aq. NaOH; (iii) TFA/EtOAc; (iv) water, aq. NaOH; (v) concentrate; (vi) IPA.
B ) 80%; (35 min) A ) 80%, B ) 20%; (36 min) A )
0%, B ) 20%; (45 min) A ) 20%, B ) 80%. (Production
Route) Zorbax SB-C8 25 cm × 4.6 mm, flow 1.0 mL/min;
wavelength ) 250 nm; temperature 30 °C; injection vol-
ume: 20 µL of a ca. 0.05% solution in ACN; A ) 0.05%
trifluoroacetic (v/v) in ACN and B ) 0.05% trifluoroacetic
may result in detonation of the oil due to thermal instability
starting at 120 °C.
1-(2,2-Dimethoxyethyl)-2-nitrobenzene, 5. To a 2 L
three-necked flask equipped with a dropping funnel, ther-
8
2
mocouple, stirrer, condenser, and N inlet was charged crude
3 and 4 (150 g, 0.599 mol) followed by methanol (1.0 L).
To this solution was added a solution of p-toluenesulfonic
acid monohydrate (136 g, 0.716 mol) in methanol (1 L). The
reaction mixture experienced a mild exotherm during the
addition from 23 °C to 31 °C. The reaction mixture lightened
after the addition of acid but darkened somewhat during the
reaction. The resulting solution was heated to reflux 64-68 °C
for 3 h. The solution was cooled to ambient temperature,
(v/v) in Milli-Q water; gradient: (0 min) A ) 10%, B )
9
2
0%; (25 min) A ) 80, B ) 20%; (26 min) A ) 80%, B )
0%; (27 min) A ) 10%, B ) 90%; (35 min) A ) 40%, B
)
60%; (40 min) A ) 10%, B ) 90%. All reagents were
commercially available except where indicated. Melting
points were measured in open capillary tubes and are
1
uncorrected. H NMR spectra were measured in CDCl
3
unless otherwise indicated, and IR spectra were taken using
a KBr salt pellet.
2 3
and solid Na CO (20 g, 0.189 mol) was carefully added
and no exotherm or gas evolution was observed. The slurry
was stirred for 15 min at ambient temperature. Excess
methanol (1,290 mL) was removed by vacuum distillation.
To the resulting slurry after methanol distillation was added
agitation water (500 mL) followed by toluene (800 mL). The
mixture was agitated for ∼15 min to permit dissolution of
the salts. The agitation was ceased, and layers were allowed
to separate. The lower aqueous layer was back-extracted with
toluene (1 × 400 mL then 1 × 200 mL). The water layer
was discarded (805 mL, pH 11.0), and the toluene layers
were combined. The combined toluene layers were filtered
though a pad (5 cm × 2 cm) of filter aid to remove solids
and a small amount of entrained water. The toluene was then
removed by vacuum to form an oil. Some toluene remained
in the oil. Weight of product: 139 g. % Yield > 98%, based
upon HPLC analysis. A sample was purified by placing a
(E)-1-[2-(2-Nitrophenyl)ethenyl]pyrrolidine, 3, and (E)-
1-[2-(2-nitrophenyl)ethenyl]dimethylamine, 4. To a 1 L
three-necked flask equipped with a magnetic stir bar,
thermocouple, and distillation head was charged 2-nitrotolu-
ene (300.0 g, 2.19 mol) followed by pyrrolidine (186.5 g,
2.63 mol) and DMF dimethylacetal (313 g, 2.63 mol). The
resulting reddish/orange solution was heated to 80 ( 2 °C
for 18-24 h. The reactor was set to provide for distillation
of low boiling byproducts while maintaining the reactor set
point. During this period the reaction changed color from a
reddish/orange solution to dark red. The reaction was
monitored by HPLC. The final distillate volume was 294
mL. The crude product oil weighed, 561 g; theoretical, 100%;
weight, 477 g. The red oil 3 and 4 was 85% pure by HPLC
1
(
area %). H NMR (300 MHz, CDCl
3
) δ ) 1.94 (m, 4H),
3
1
7
1
.34 (m, 4H), 3.50 (s, 6H, N(CH
3
)
2
), 5.83 (d, J ) 13.4 Hz,
sample under a high vacuum (1 Torr) over 64 h at ambient
1
H), 6.93 (m, 1H), 7.23 (d, J ) 13.7 Hz, 1H), 7.31 (m, 1H),
.44 (dd, J ) 8.2, 1.2 Hz, 1H), 7.83 (dd, J ) 8.2, 1.2 Hz,
H). The resonance at 3.50 (s, 6H) corresponded to 4.
temperature. H NMR (300 MHz, CDCl
3
) δ ) 3.25 (d, J )
5.3 Hz, 2H), 3.38 (s, 6H), 4.60 (d, J ) 5.3 Hz, 1H), 7.42
(m, 2H), 7.56 (m, 1H), 7.92 (dd, J ) 7.9, 1.3 Hz, 1H).
2-(2,2-Dimethoxyethyl)benzamine, 6. To an inerted 1
L Parr bomb was charged wetted 5% Pd/C (6 g on a dry
Caution: The product oil contained pyrrolidine byproducts,
and attempts to further purify the product oil by distillation
Vol. 10, No. 6, 2006 / Organic Process Research & Development
•
1209

basis) followed by a solution of 5 (139 g) in toluene (600
mL). The Parr bomb was inerted 3× using a vacuum/nitrogen
procedure. The resulting inerted slurry was hydrogenated at
four-necked round bottom flask equipped with a mechanical
agitator, thermometer, and nitrogen purge was charged
sodium borohydride 119 (30 g, 0.793 mol) and THF (0.745
L, 21.5 vol). The suspension was cooled in an ice bath to
10 °C. Propionic acid [0.195 L, (193.44 g) 2.61 mol] was
added via syringe drive over 2 h 40 min at a rate of 4.0
mL/min. That exhibited an exotherm. The maximum tem-
perature reached during this exotherm was about 25 °C. At
the end of the addition, some solids remained so the reactor
contents were allowed to warmed to 25 °C with agitation
overnight. A homogeneous to slightly hazy solution was
2
20 to 30 °C and 30 to 40 psi of H pressure for 4 to 8 h or
until hydrogen uptake ceased. The system was purged of
hydrogen, repeating the vacuum/nitrogen procedure 3 times.
The reaction mixture was filtered across a (5 cm × 2 cm)
filter aid pad, and the reactor and filter pad were washed
with toluene (400 mL) followed by water (50 mL). The water
layer was separated and discarded. The toluene was removed
under a vacuum to afford an amber colored oil. Weight of
product: 120 g. The technical product 6 (120 g) was purified
by vacuum distillation. The oil was heated 120 to 130 °C at
2 5 3
obtained. The 1.0 M NaBH(OCOC H ) obtained was stored
in a refrigerator for future use.
1
to 5 mbar of pressure. A precut was taken at a vapor
1-(1-Pyridin-2-yl methyl-piperidin-4-yl)-1H-indole, 11.
temperature of 104 to 108 °C that was indole. The main cut
was then taken from 108 to 110 °C. Weight: 112.3 g.
To a 2 gal glass reactor equipped with an agitator, a
thermocouple, a condenser, and a N inlet was charged
2
(
93.6%) Assay: 98.9 by HPLC (area %) analysis of a clear
propionic acid (0.61 L) followed by addition of 9 (213.12
g, 0.504 mol) and 6 (91.52 g, 0.504 mol). The mixture was
agitated at 20 to 25 °C until the contents dissolved (ap-
proximately 15 to 30 min). Once the solids dissolved, the
reaction mixture was cooled to -12 to -8 °C. To the chilled
solution at -12 to -8 °C was charged, over 2.5 h, the 1.0
colorless oil. Alternately, one may decant the Parr reactor’s
contents through a (5 cm × 2 cm) filter aid pad, and the
filtrate is worked up as above. The Pd/C catalyst adhered to
the walls of the reactor due to the water generated via
reduction of the nitro group. The Parr reaction may then be
recharged with a toluene solution of 6, inerted, and subjected
to hydrogenation without any loss of catalytic activity.
2 5 3
M NaBH(OCOC H ) solution previously prepared. Once the
addition of the reducing agent was complete, the reaction
was monitored by HPLC for reaction completeness (<1%
starting materials or intermediate). The reaction mixture was
maintained at -12 to -8 °C. Once the reaction met the
reaction completeness criteria, the reaction mixture was
warmed to 8 to 10 °C, and EtOAc (1.28 L) was added. Using
40% w/v aq. NaOH, the pH of the reaction mixture was
adjusted to 9.8 to 10.5. The reaction mixture was allowed
to warm during the pH adjustment to 48 to 53 °C. Once the
pH range of 9.8 to 10.5 was achieved, the mixture was
agitated for 5 to 15 min, after which agitation was stopped
to allow the phases to separate while maintaining a solution
temperature of 48 to 53 °C. (Note: if the solution temperature
drops to below 45 °C, sodium salts will begin to crystallize.)
The lower aqueous phase was separated to a second reactor.
The aqueous phase was back-extracted with EtOAc (0.64
L) at 48 to 53 °C. The organic layers were combined and
washed with a 20% aq. brine solution (0.96 L) at 48 to 53
°C. The lower brine phase was discarded. The organic phase
was then placed under a vacuum, and EtOAc was distilled,
1.07 to 1.17 L (target volume: 1.12 L) with a maximum
jacket temperature of 55 °C. Once the distillation was
complete, the reactor contents were cooled to 20 to 25 °C
and returned to atmospheric pressure. To the reactor contents
at ambient temperature, trifluoroacetic (0.192 L, 2.58 mol)
was charged over 45 min keeping the maximum temperature
below 30 °C. Once the addition of trifluoroacetic was
complete, the reactor contents were warmed to 46 °C and
maintained at this temperature (1 °C for 6 h. Progress of
the cyclization was monitored by HPLC. When the conden-
sation product from the above protocol was <1% relative
to 11, the reactor contents were cooled to ambient temper-
ature. To the reactor was charged EtOAc (0.96 L) and water
(0.16 L). The pH of the reactor contents was carefully
adjusted to 8.3 to 8.8 with 40% w/v NaOH solution, while
1
-(2-Pyridinylmethyl)-4-piperidinone Camphor Sul-
fonate, 9. To a 2 gal glass reactor equipped with an overhead
stirrer, a condenser, and a nitrogen purge was charged
-picolyl chloride hydrochloride (300 g, 1.83 mol), 4-pip-
2
eridone monohydrate hydrochloride (294.96 g, 1.92 mol),
powdered sodium carbonate (775.32 g, 7.32 mol), followed
by ACN (1.62 L). The contents were carefully heated to 70
to 72 °C over a period of about 2 h to minimize foaming
due to off gassing. The reaction mixture changed from an
off-white slurry to light orange. The reaction mixture was
maintained at 70 to 72 °C for 5 h. The reaction mixture was
cooled to ambient temperature and filtered, and the solids
were rinsed with EtOAc (2 × 1.20 L). The filtrate was
reduced in volume to ∼1 L under reduced pressure at 50 to
5
5 °C. The solution was cooled to 30 to 35 °C and held for
the next step. To a second 2 gal reaction vessel equipped
with an overhead stirrer and a nitrogen purge was charged
camphor sulfonic acid (425 g, 1.83 mol) followed by EtOAc
3.3 L). The contents were heated 68 to 72 °C and stirred
until the CSA dissolved. To the warm solution of CSA at
8 to 72 °C was charged the solution from the previous
reaction above over a period of 20 to 30 min. The first reactor
was rinsed with EtOAc (100 mL), and the rinse was added
to the reactor containing CSA. The reactor contents were
maintained at 68 to 72 °C for about 30 min to permit
crystallization to begin. Once crystals were observed, the
reactor was cooled to 58 to 62 °C over 2 h. The thickening
crystal slurry was held at 58 to 62 °C for 1 h. The crystal
slurry was cooled to ambient temperature over 4 h and stirred
for 3 h. The crystals were filtered and washed with EtOAc
0.9 L). The crystals were dried in a 45 to 55 °C vacuum
oven overnight to afford 700 g of 9. (90.5%)
Preparation of 1.0 M NaBH(OCOC
tion. Caution: This reaction generates hydrogen. Into a 2 L
(
6
(
2
5 3
H ) THF Solu-
1210
•
Vol. 10, No. 6, 2006 / Organic Process Research & Development

allowing the contents to warm to 48 to 53 °C. To the reactor
contents was added 20% w/v aq. brine while maintaining a
temperature range of 48 to 53 °C to prevent crystallization
of salts. The layers were allowed to settle for 10 to 20 min
before separation. The lower aqueous layer was then back-
extracted with EtOAc (0.64 L) at 48 to 53 °C, the lower
aqueous layer was discarded, and the organic layers were
combined. With a jacket set point of not more than 40-44
was removed via vacuum distillation with a temperature
range of 40-50 °C. To the slurry was added IPA (0.638 L),
and the vacuum distillation was continued at a temperature
range of 40-50 °C until ∼630 mL were removed. The slurry
was held at 43-45 °C for 3 h and was then cooled to -5
°C over 3-4 h. The slurry was filtered and washed with
cold, 0-5 °C, IPA (2 × 0.25 L). The product was dried
under a vacuum at 50 °C to a constant weight. The weight
of the product was 114.7 g (78%).
°
C solvent was removed by distillation until the reaction
mixture was a thick paste.
Acknowledgment
Crystallization of 11 from EtOH/Water. Ethanol (2.6
L) was charged to the reactor to dissolve the solids. With a
jacket set point of NMT 55 °C, the volume of the reaction
was reduced by ∼1.25 L to yield a slurry. The mixture was
heated to reflux to effect dissolution of the solids and was
then cooled to 55 °C ( 2 °C. The solution temperature was
maintained at 55 °C ( 2 °C while adding water (3 L) over
The authors are thankful to Curtis Miller, Cindy Hammill,
Cindy McCurry, Dave Myers, and John Stafford for their
invaluable assistance with chemical analysis. We wish to
acknowledge the efforts of Ron Stockberger and Bill Henning
for their assistance in carrying out 2 gal experiments. We
wish to thank Jim Hermiller for physical characterizations
and Prof. William Roush for helpful suggestions and
discussions.
2.5 h. Once the water addition was complete, the resulting
slurry was maintained at 55 °C ( 2 °C for 1 h. The slurry
was then cooled to 20 to 24 °C over 4 h followed by further
cooling to -7 to -4 °C over 2.5 h. The agitated slurry was
maintained at -5 ( 2 °C for 2.5 h. The resulting slurry was
filtered, and the reactor and cake were washed with chilled
Supporting Information Available
Table of cosolvent effects on reductive amination with
2 5 3
NaBH(OCOC H ) and optimization of cyclization reaction
table. This material is available free of charge via the Internet
at http://pubs.acs.org.
40% aq. EtOH (0.16 L). The cake was dried under a vacuum
at 70 to 77 °C for at least 16 h to yield 124.4 g, 84.6% yield.
NMR and HRMS data compared to literature.2
Crystallization of 11 from IPA. Isopropanol (1.91 L)
was charged to the reactor contents, and ∼1.1 L of solvent
Received for Review March 23, 2006.
OP060068K
Vol. 10, No. 6, 2006 / Organic Process Research & Development
•
1211