Process research and large-scale synthesis of a novel 5,6-dihydro-(9H)-pyrazolo[3,4-c]-1,2,4-triazolo[4,3-a]pyridine PDE-IV inhibitor

Article abstract of DOI:10.1021/op010222x

An efficient synthesis of the PDE IV inhibitor, 9H-cyclopentyl-7-ethyl-3-(thiophen-2-yl)-pyrazolo[3,4-c]-1,2,4-triazolo-5,6- dihydro-[4,3-a]pyridine 1 is described. Starting from commercially available γ-caprolactone, the synthesis was carried out in 10 steps. Key transformations were the selective O-methylation of diketone, 3-hydroxy-1-(4-methoxybenzyl)-4-propionyl-5,6-dihydro-1H-pyridin-2-one, with dimethyl sulfate and cesium carbonate in dimethylformamide, a one-pot pyrazole formation with subsequent acidic deprotection to provide lactam, 1-cyclopentyl-3-ethyl-1,4,5,6-tetrahydropyrazolo[3,4-c]pyridin-7-one, and finally the utilization of imidate, 1-cyclopentyl-7-ethoxy-3-ethyl-4,5-dihydro-1H-pyrazolo[3,4-c]pyridine for the introduction of the triazole moiety. This process avoided the use of harsh reaction conditions, undesirable reagents and overcame the environmental concerns in the original synthesis.

Full text of DOI:10.1021/op010222x

Organic Process Research & Development 2001, 5, 575−580
Process Research and Large-Scale Synthesis of a Novel
5,6-Dihydro-(9H)-pyrazolo[3,4-c]-1,2,4-triazolo[4,3-a]pyridine PDE-IV Inhibitor
Frank J. Urban,* Bruce G. Anderson,1 Susan L. Orrill, and Peter J. Daniels
Pfizer Global Research and DeVelopment, Chemical Research and DeVelopment Department, Eastern Point Road,
Groton, Connecticut 06340, U.S.A.
Abstract:
concerns for the initial scale up. Step one was the reaction
of 4-iodoanisole and 2-pyrrolidinone with copper powder and
potassium carbonate in the absence of solvent with heating
to 150 °C. This reaction had shown an exotherm when scaled
up to 200 g. Addition of ethyl acetate to the crude melt at
elevated temperature was necessary to prevent formation of
a solid mass upon cooling. Reaction two was the greater con-
cern. The mono-addition of ethylmagnesium bromide to pyr-
rolidinone lactam 3 occurred cleanly in ethyl ether, but was
complex in tetrahydrofuran. The desired ethyl (4-methoxy-
phenylamino)propyl ketone 4 was not stable to storage. Other
process concerns were the use of methyl p-tolyltriazine 7 to
O-methylate the intermediate â-diketone 6, the synthesis and
purification of cyclopentyl hydrazine 9, the ceric ammonium
nitrate (CAN) deprotection of p-methoxyphenyl amide 2b,
and finally, the use of thiolactam chemistry to introduce the
triazole moiety. A more detailed discussion of these reactions
is given in the discussion of the preferred process described
below.
Our first approach was to intersect with the Duplantier
synthesis at oxalamide 5. To avoid the Grignard reaction
for the introduction of an ethyl group, commercially available
γ-caprolactone 11 was chosen as the starting material, and
p-anisidine was the source of the nitrogen atom. The se-
quence was demonstrated on laboratory scale with the yields
shown in Scheme 2. Initial reduction of lactone 11 with Di-
BAL provided lactol 12 as a mixture of diastereomers. Lactol
12 was used without purification in the reductive alkylation
to provide amino alcohol 13 in good yield. Amino alcohol
13 was converted uneventfully to the desired oxalamide 5.
The problem with this approach was discovered on initial
scale up to kilogram quantities. Some batches of lactol 12
underwent the reductive alkylation in yields similar to those
seen in the lab, but more often the lactol was found to de-
compose, possibly through self-condensation, resulting in
complex reaction mixtures after the reductive alkylation step.
Also, since the 4-methoxyphenyl moiety was serving as the
nitrogen-protecting group, this would require the use of CAN
to carry out the deprotection later in the synthesis. With these
factors in mind, a different source for the nitrogen atom was
chosen.
An efficient synthesis of the PDE IV inhibitor, 9H-cyclopentyl-
7-ethyl-3-(thiophen-2-yl)-pyrazolo[3,4-c]-1,2,4-triazolo-5,6-dihy-
dro-[4,3-a]pyridine 1 is described. Starting from commercially
available γ-caprolactone, the synthesis was carried out in 10
steps. Key transformations were the selective O-methylation of
diketone, 3-hydroxy-1-(4-methoxybenzyl)-4-propionyl-5,6-di-
hydro-1H-pyridin-2-one, with dimethyl sulfate and cesium car-
bonate in dimethylformamide, a one-pot pyrazole formation
with subsequent acidic deprotection to provide lactam, 1-cy-
clopentyl-3-ethyl-1,4,5,6-tetrahydropyrazolo[3,4-c]pyridin-7-
one, and finally the utilization of imidate, 1-cyclopentyl-7-
ethoxy-3-ethyl-4,5-dihydro-1H-pyrazolo[3,4-c]pyridine for the
introduction of the triazole moiety. This process avoided the
use of harsh reaction conditions, undesirable reagents and
overcame the environmental concerns in the original synthesis.
Recently, Allen Duplantier et al. reported on the synthesis
and selective phosphodiesterase activity of a series of 7-oxo-
4,5,6,7-tetrahydro-1H-pyrazolo[3,4-c]pyridines including ana-
logue 2b.2 In their further optimization of the desired
biological activity, a key intermediate was the unsubstituted
lactam 2a which was available from 2b through removal of
the 4-methoxyphenyl moiety with ceric ammonium nitrate
(CAN). Additional molecular modification resulted in a new
series of compounds which featured an additional triazole
ring and led to the scale-up of the thiophene analogue 1 for
further study.3 In this paper, we describe the process research
and large-scale preparation of multikilogram quantities of 1
to support these studies. In approaching this target, an
efficient synthesis was developed which paralleled the
original Duplantier route, but which addressed the problems
of harsh reaction conditions, multiple chromatographies, and
undesirable reagents for large-scale work.
The Duplantier synthesis of compound 1 is shown in
Scheme 1. The first two steps of the sequence were major
(1) Current address: Ilex Oncology, 14785 Omicron Drive, # 201, San Antonio,
TX 78245-3221.
(2) Duplantier, A. J.; Andresen, C. J.; Cheng, J. B.; Cohan, V. L.; Decker, C.;
DiCapua, F. M.; Kraus, K. G.; Johnson, K. L.; Turner, C. R.; UmLand, J.
P.; Watson, J. W.; Wester, R. T.; Williams, A. S.; Williams, J. A., J. Med.
Chem. 1998, 41, 2268.
The synthesis that was carried out at large scale starting
with 28.75 kilograms of γ-caprolactone is shown in Scheme
3. 4-Methoxy-benzylamine was heated with caprolactone 11
without solvent to provide amide 15 as a crystalline solid.
This group allowed more options for deprotection. A similar
(3) Duplantier, A. J.; Cooper, K. U.S. Patent 6,004,974, 1999.
10.1021/op010222x CCC: $20.00 © 2001 American Chemical Society and The Royal Society of Chemistry
Published on Web 09/25/2001
Vol. 5, No. 6, 2001 / Organic Process Research & Development
•
575

Scheme 1a
a a) Cu, 150 °C, 83%; b) EtMgBr, Et2O, 0 °C, 60-80%; c) ethyl oxalyl chloride, NaOH, benzene; d) NaOEt, EtOH, 41% over c and d; e) methyl-4-tolyltriazine
7, CH2Cl2, 70%; f) cyclopentyl hydrazine hydrochloride, 120 °C, 75%; g) CAN, acetonitrile/water, 65%; h) P4S10, dioxane, 80%; i) hydrazine, pyridine; j)
2-thiophenecarbonyl chloride, pyridine followed by reflux in DMF, 64%.
Scheme 2a
Scheme 3a
a a) DiBAL, THF, -78 °C, 85%; b) p-anisidine, sodium triacetoxyborohy-
dride, 70%; c) ethyl oxalyl chloride, triethylamine, CH2Cl2, 90%; d) TEMPO,
sodium hypochlorite, CH2Cl2, 80%.
reaction of lactone 11 with p-anisidine was not successful
since the aniline was less nucleophilic.
Amide 15 could be reduced to the amine 16 by a variety
of reducing agents including borane-THF, sodium borohy-
dride with boron trifluoroetherate or lithium aluminum
hydride. For the large-scale batch, sodium borohydride in
tetrahydrofuran with addition of acetic acid was used.4 This
procedure scaled up well although there was foaming during
the addition of the acetic acid. Amine 16 was isolated as a
solution in ethyl acetate for use directly in the acylation with
ethyl oxalyl chloride under Schotten-Baumann conditions.
Once again, crude oxalamide 17 was used as is for the
oxidation of the secondary alcohol. In the lab, the oxidation
of alcohol 17 was done with either chromium trioxide or
a a) 4-Methoxybenzylamine, neat, 73%; b) NaBH4, acetic acid, THF, 79%;
c) ethyl oxalyl chloride, 86%; d) TEMPO, sodium hypochlorite, CH2Cl2, 85%;
e) KOt-Bu, THF, 73%; f) cesium carbonate, dimethyl sulfate, DMF, 89%.
TEMPO with sodium hypochlorite.5 The hypochlorite pro-
cedure worked well in the Pilot Plant provided that the
sodium hypochlorite was freshly prepared from calcium
hypochlorite and sodium carbonate. The product ketone was
an oil and was used in the next reaction without further
purification. The Dieckmann cyclization was done in tet-
(4) Umino, N.; Iwakuma, T.; Itoh, N. Tetrahedron Lett. 1976, 763.
(5) Anelli, P. L.; Montanari, F.; Quici, S. Org. Synth. 1990, 69, 212.
576
•
Vol. 5, No. 6, 2001 / Organic Process Research & Development

Scheme 4a
Scheme 5a
a a) Ph3P, THF; b) concentrated HCl, 69%.
rahydrofuran with a slight excess of potassium tert-butoxide.
Lactam 19 was isolated as a solid by precipitation from the
aqueous layer in the work-up by addition of 6 N HCl. This
combination of extraction into an aqueous solution and
precipitation of solid 19 was the second actual purification
in the process and provided material of high quality.
In the next step, it was necessary to introduce a methyl
group onto the enolic oxygen in 19. Duplantier had used
methyl p-tolyltriazine 7 to avoid the side reaction of carbon
alkylation.6 Reaction of 19 with methyl iodide and potassium
carbonate in acetone gave a mixture of C- and O-alkylation.
Conditions for exclusive O-alkylation were found using
dimethyl sulfate and cesium carbonate in dimethylformamide.
Once these conditions were found, other reagents were not
explored.
The next issue was the preparation of cyclopentyl hydra-
zine 9. Duplantier had used a literature procedure which
started with cyclopentanone and tert-butylcarbazate to give
an alkylidine carbazate. This was reduced with borane-THF,
and the tert-butyl ester was hydrolyzed with hydrochloric
acid to afford hydrazine 9.7 The hydrazine prepared in this
manner in our lab was contaminated with boric acid which
was difficult to remove without loss of yield. Instead, a one-
pot procedure was developed which relied on a Mitsunobu
reaction between cyclopentanol and di-tert-butyl azodicar-
boxylate.8 (Scheme 4) Our variation of this procedure was
to add 6 N HCl to the tetrahydrofuran reaction mixture after
the Mitsunobu reaction was complete to remove the tert-
butyl esters and isolate the hydrazine hydrochloride directly
as a crystalline solid. On 7-kg scale, the dihydrochloride salt
of 9 was isolated, while in the lab the mono hydrochloride
salt was isolated on one occasion. This may have been due
to a more complete azeotropic drying with 2-propanol. The
identity of the salt and the stoichiometry for the next reaction
was based on a chloride analysis of the isolated material.
The remainder of the process is shown in Scheme 5. The
pyrazole ring was formed in the same manner as the
procedure described by Duplantier. A mixture of the enol
ether 20 and cyclopentyl hydrazine dihydrochloride 9 was
suspended in tetrahydrofuran and heated to reflux. The
solvent was distilled off, and the resulting melt was heated
to 85-90 °C under a sweep of nitrogen to help remove
tetrahydrofuran, methanol, and excess HCl. Crude pyrazole
23 was an oil, which was used without isolation in a one-
pot procedure for this scale up, but could be isolated as a
crystalline salt with either p-toluenesulfonic acid or benz-
a a) 9, 88 °C, THF; b) CF3CO2H, CH3SO3H, 70 °C, 71% (two steps); c)
Et3OBF4, CH2Cl2, 93%; d) 25, n-butanol, 90°, 67%.
enesulfonic acid from ethyl acetate. Removal of the p-
methoxybenzyl moiety was done by treating crude 23 with
trifluoroacetic acid at 50-60 °C followed by addition of
methanesulfonic acid. These conditions for solvolysis of the
protecting group had been described previously by Martin.9
The resulting mixture was heated for 2 h at 70 °C to complete
the deprotection. Extractive work up provided the key lactam
2a in 71% yield over the two steps.
In the Duplantier synthesis, lactam 2a was converted to
the thiolactam 10 with phosphorus pentasulfide. This was
reacted with anhydrous hydrazine to give an amidrazone
derivative, which was acylated with 2-thiophenecarbonyl
chloride to provide 26 in situ. The reaction solution was
heated to cyclize 26 to triazole 1. The thiolactam procedure
seemed to be the preferred method in the literature for
conversion of a lactam to a fused triazole.10 The environ-
mental problems associated with the scale up chemistry
involving phosphorus pentasulfide, hydrogen sulfide, and
anhydrous hydrazine caused us to seek an alternative.
Considerable effort was expended to prepare an imidoyl
chloride derivation from 2a, but the best yield of 1 from a
sequence in which 2a was converted to an imidoyl chloride
intermediate with phosphorus pentachloride followed by
treatment with 2-thiophenecarboxylic hydrazide 25 was ca.
20%. The next approach was to generate the imidate of 24.
The intermediacy of imidates for the preparation of triazoles
was seen mostly in cases where the triazole was not fused
to another ring.11 Reaction of lactam 2a with a freshly
prepared solution of triethyloxonium tetrafluoroborate in
methylene chloride gave the desired imidate 24 in almost
quantitative yield.12 Imidate 24 was treated with 2-thiophen-
ecarboxylic hydrazide in refluxing n-butanol to provide the
desired triazole 1. The intermediate acyl amidrazone 26 can
be seen in the reaction and was monitored by HPLC until
the conversion to 1 was complete. The isolated yield from
2a to 1 over the three steps was 67% and was not optimized
at that time.
(6) Vyas, D. M.; Benigni, D.; Partyka, R. A.; Doyle, T. W. J. Org. Chem.
1986, 51, 4307.
(9) Martin, S. F.; Oalmann, C. J.; Liras, S. Tetrahedron 1993, 49, 3521.
(10) Hester, J. B., Jr.; Rudzik, A. D.; Kamdar, B. V. J. Med. Chem. 1971, 14,
1078.
(7) Ghali, N. I.; Venton, D. L.; Hung, S. C.; Le Breton, G. C. J. Org. Chem.
1981, 46, 5413.
(8) Dow, R. L.; Kelly, R. C.; Schletter, I.; Wierenga, W. Synth. Commun. 1981,
11, 43.
(11) Poonian, M. S.; Nowoswiat, E. F. J. Org. Chem. 1980, 45, 203.
(12) Smith, M. B.; Menezes, R. Synth. Commun. 1988, 18, 1625.
Vol. 5, No. 6, 2001 / Organic Process Research & Development
•
577

In conclusion, an efficient process for the preparation of
a clinical candidate was presented which was carried out
successfully to produce 7.9 kg of lactam 2a. A 3.7 kg portion
of 2a was taken on to compound 1. The final process
minimized environmental concerns with the Duplantier route
such as generation of sulfurous volatile side-products and
the use of anhydrous hyrazine. In addition, the need for
multiple chromatographic purifications was avoided. The
facile nature of the imidate chemistry in particular should
allow others to replace some of the thiolactam chemistry
which is still the primary method for the conversion of a
lactam to a fused triazole.13
min at 20-25 °C, and then amide 15 (45.75 kg, 182 mol)
was added as a solid. After 30 min, the reaction was cooled
to 5-10 °C, and over a 4-8 h period a solution of acetic
acid (34.4 L) in tetrahydrofuran (45.4 L) was added, keeping
the temperature at 0-10 °C. A slight nitrogen bleed was
kept on the tank to help remove the hydrogen. When the
addition was complete, the reaction was warmed to 20-25
°C and stirred for 1 h. The temperature of the reaction was
slowly increased to a gentle reflux (∼66 °C) and held there
for 16 h. The reaction was quenched by the addition of 1 N
HCl, keeping the temperature <25 °C. Excess tetrahydro-
furan was removed by atmospheric distillation. Ethyl acetate
was added to the resulting aqueous solution to extract
unreacted amide. The acidic aqueous layer was then brought
to pH 11 to allow the product amine 16 to be extracted into
ethyl acetate and held for use in the next step. An aliquot of
the ethyl acetate solution of product was worked up to project
final yield and concentration. The yield for the Pilot Plant
run was 55% which was less than for the lab experiments
(78.8%). The large batch had 12.8% unreduced amide
starting material 15 after the quench, which accounted in
part for the lower yield.
Experimental Section
Melting points were determined on a Thomas-Hoover
capillary melting point apparatus and were uncorrected. NMR
spectra were obtained on either a Brucker WM 300 (300
MHz) or a Varian Unity 400 (400 MHz) spectrometer in
deuteriochloroform or dimethyl sulfoxide-d6. Infrared spectra
were taken in KBr by diffuse reflectance infrared Fourier
transform spectroscopy (DRIFTS). Mass spectra were de-
termined with a Finnigan 4510 mass spectrometer using fast
atom bombardment (FAB). Elemental analyses were per-
formed by Schwarzkopf Microanalytical Laboratory, Wood-
side, NY. The reactions were monitored by TLC on silica
gel plates, and isolated compounds were analyzed by HPLC.
2-Thiophenecarboxylic hydrazide was purchased from
Aldrich Chemical Co., and crystalline triethyloxonium tet-
rafluoroborate, from Fluka Chemicals.
4-Hydroxyhexanoic Acid 4-Methoxybenzylamide (15).
γ-Caprolactone 11 (28.75 kg, 251.8 mol) and 4-methoxy-
benzylamine (38.0 kg, 277 mol) were placed in a 100-gal
glass-lined tank. The solution was heated to 80-85 °C and
held at that temperature for 16 h. TLC on silica gel plates
showed the reaction was complete. (TLC system: ethyl
acetate with detection at 254 nm). Ethyl acetate (68 L) was
slowly charged to the pot after cooling to 60 °C. Hexanes
(68 L) were added until a haze was achieved. After 30 min,
to allow crystallization to start, the remainder of the hexanes
were added. The slurry was cooled to 25 °C and granulated
for 3 h. The solid was collected by filtration and washed
with a 1:1 mixture of ethyl acetate and hexanes. The wet
cake was vacuum-dried with no additional heat to produce
46.05 kg (72.8%) of the desired amide 15: mp 81-82 °C.
1HMR (CDCl3, 300 MHz) δ 7.18 (d, 2), 6.84 (d, 2), 6.27
(bs, 1), 4.32 (d, 2), 3.79 (s, 3), 3.50 (m, 1), 3.19 (bs, 1),
2.35 (t, 2), 1.85 (m, 1), 1.67 (m, 1), 1.49 (m, 2), 0.92 (t, 3).
Anal. Calcd for C14H21NO3: C, 66.91; H, 8.42; N, 5.57.
Found: C, 67.26; H, 8.71; N, 5.55.
1HMR (CDCl3, 300 MHz) δ 7.21 (d, 2), 6.83 (d, 2), 3.78
(s, 3), 3.69 (s, 2), 3.41 (m, 2), 2.78 (m, 1), 2.58 (m, 1), 1.71
(m, 2), 1.45 (m, 4), 0.95 (t, 3). GC mass spectrum: m/e 237
(M+).
HPLC: Inertsil, C8, acetonitrile, 20%/0.01M K2HPO3 @
pH ) 7.0 with H3PO4.
N-(4-Hydroxyhexyl)-N-(4-methoxybenzyl)oxalamic Acid
Ethyl Ester (17). 6-(4-Methoxybenzylamino)hexan-3-ol 16
(24 kg, 101.1 mol) in ethyl acetate (598 L) was charged to
a clean and dry, nitrogen-purged 500-gal glass-lined tank.
This solution was cooled to 0-5 °C, and then a solution of
sodium bicarbonate (16.99 kg, 202.2 mol in water (193 L)
was added, maintaining a temperature of 0 to 5 °C. A solution
of ethyl oxalyl chloride (16.57 kg, 121.4 mol) in ethyl acetate
(75.7 L) was added, while maintaining a temperature of 0-5
°C over a time period of about 25 min. The reaction was
allowed to warm to 20-25 °C at which point it was complete
by HPLC. The reaction was stirred for an additional 16 h to
allow any residual ethyl oxalyl chloride to decompose. The
lower aqueous layer was discarded, and the ethyl acetate
solution was washed with water (185.5 L). The layers were
separated. The remaining ethyl acetate was washed with 2
N HCl (128 L). The remaining ethyl acetate was vacuum
stripped to obtain the crude product amide 17 as an oil, 29.3
1
kg (85.9% theory). HMR (CDCl3, 400 MHz) δ 7.18 (m,
2), 6.83 (m, 2), 4.41 (m, 1), 4.31 (m, 3), 3.76 (d, 3), 3.43
(m, 1), 3.25 (m, 1), 3.13 (t, 1), 2.00 (bs, 1), 1.80-1.26 (m,
8), 0.87 (t, 3). IR (neat) 3456, 1739, 1654, 1513 cm-1. 13CMR
(CDCl3, 400 MHz) δ 163.5, 162.1, 159.7, 159.0, 129.6,
129.2, 128.0, 127.1, 114.2, 114.1, 72.6, 72.5, 62.1, 55.3, 50.9,
47.0, 46.3, 43.6, 33.5, 33.4, 30.3, 30.2, 24.3, 22.9, 14.0, 9.9.
GC mass spectrum: m/e, 338 (M+).
HPLC: Inertsil, C8, acetonitrile, 20%/0.01M K2HPO3 @
pH ) 7.0 with H3PO4.
N-(4-Methoxybenzyl)-N-(4-oxo-hexyl)oxalamic Acid Eth-
yl Ester (18). Potassium bromide (593 g, 5 mol) was
HPLC: Inertsil, C8, acetonitrile, 20%/0.01M K2HPO3 @
pH ) 7.0 with H3PO4.
6-(4-Methoxybenzylamino)hexan-3-ol (16). Tetrahydro-
furan (458 L) and sodium borohydride (22.15 kg, 585.6 mol)
were charged to a clean and dry nitrogen-purged 500-gal
glass lined tank. The suspension was allowed to stir for 30
(13) After the completion of our work, another example of using imidates to
prepare 1,2,4-triazolo[4,3-a]pyridines appeared. Lawson, E. C.; Maryanoff,
B. E.; Hoekstra, W. J. Tetrahedron Lett. 2000, 41, 4533.
578
•
Vol. 5, No. 6, 2001 / Organic Process Research & Development

dissolved in water (18.9 L) in a 100-gal glass-lined tank. A
solution of oxalamide alcohol 17 (33.62 kg, 99.6 mol) in
methylene chloride (128.7 L) was added. The TEMPO
catalyst (150 g) was added, and the reaction was cooled to
0-5 °C. Fresh sodium hypochlorite solution (prepared from
calcium hypochlorite (12.11 kg) and sodium carbonate (17.96
kg) in water (378.5 L) adjusted to pH 9.5 with sodium
bicarbonate (1.7 kg) and filtered to remove calcium carbon-
ate) was added slowly, keeping the temperature at 10-15
°C. When the reaction was complete, the layers were
separated, and the aqueous layer was extracted with ad-
ditional methylene chloride (30 L). The combined organic
layers were washed with a solution of concentrated HCl (1.4
gal, 5.4 L) and potassium iodide (331 g) in water (14.5 L).
The organic layer was then washed with a solution of sodium
thiosulfate (1.2 kg) in water (20 L). The methylene chloride
was washed with water (37.9 L) and then stripped without
vacuum to an oil. The oil was stripped further after being
transferred to the 50-L reactor. A yield of 33.41 kg of product
was obtained, but this material contained 15 wt % methylene
chloride (by NMR). The corrected yield was 28.4 kg (85%
theory). 1HMR (CDCl3, 300 MHz) δ 7.18 (dd, 2), 6.82 (dd,
2), 4.49 (s, 1), 4.27 (m, 3), 3.74 (d, 3), 3.22 (t, 1), 3.10 (t,
1), 2.34 (m, 4), 1.77 (m, 2), 1.29 (m, 3), 0.98 (t, 3). 13CMR
(CDCl3, 300 MHz) δ 163.2, 163.1, 162.3, 159.5, 159.2,
145.6, 129.7, 129.2, 127.96, 127.1, 114.2, 114.1, 112.1, 62.1,
55.2, 50.6, 46.1, 46.1, 42.7, 39.0, 38.1, 35.8, 21.5, 20.6, 13.9,
7.7. GC mass spectrum: m/e 335 (M+).
HPLC: Inertsil, C8, acetonitrile, 45%/water, 55% with
0.1 vol % H3PO4 and 0.2 vol % triethylamine; 18, 10.8 min.
3-Hydroxy-1-(4-methoxybenzyl)-4-propionyl-5,6-dihy-
dro-1H-pyridin-2-one (19). Oxalamide ketone 18 (28.3 kg,
84.4 mol) was dissolved in dry tetrahydrofuran (106 L) in a
100-gal glass-lined tank. This solution was added to a
solution of potassium tert-butoxide (10.4 kg) in tetrahydro-
furan (159 L) in a 300-gal glass-lined tank over a 30 min
period, keeping the temperature <35 °C. After 1 h at 20-
25 °C, the reaction was complete by HPLC. Water (371 L)
was added to the reaction, followed by isopropyl ether (91
L). The layers were separated, and the aqueous layer
containing the product as its potassium salt was washed a
second time with isopropyl ether. The aqueous layer was
partially evaporated in vacuo to remove any residual THF
and acidified to pH 2.1 with 6 N HCl (15.1 L). The resulting
slurry was filtered, and the solids were washed with water.
Product 19 was air-dried at 50 °C to provide 17.9 kg
(73%): mp 102-103 °C. 1HMR (CDCl3, 300 MHz) δ 7.20
(d, 2), 6.86 (d, 2), 4.60 (s, 2), 3.70 (s, 3), 3.33 (t, 2), 2.69 (q,
2), 2.56 (t, 2), 1.13 (t, 3).
Dimethyl sulfate (8.55 kg, 67.8 mol) was added neat over a
period of 30 min, keeping the temperature 20-25 °C. When
the charge was complete, the addition funnel was rinsed into
the tank with additional DMF (500 mL). The reaction was
allowed to stir at 20-25 °C for 16 h. The reaction was diluted
with ethyl acetate (409 L) and was washed with water (4 ×
83.3 L). The ethyl acetate solution was washed with a
solution made with 6.94 L of 50% sodium hydroxide in 83.3
L of water, followed by washing with a solution made up of
6.94 L of 35% HCl in 83.3 L of water. The organic solution
was dried by washing with brine (53 L). The ethyl acetate
was vacuum-stripped to give an oil which was suitable for
use in the next step. The estimated yield based on NMR
analysis of residual solvent was 89%. A small sample was
1
isolated for characterization. HMR (CDCl3, 300 MHz) δ
7.14 (d, 2), 6.78 (d, 2), 4.51 (s, 2), 3.88 (s, 3), 3.71 (s, 3),
3.2 (t, 2), 2.81 (q, 2), 2.42 (t, 2), 1.02 (t, 3). 13CMR (CDCl3,
300 MHz) δ 201.8, 159.1, 145.6, 129.3, 128.7, 126.5, 114.1,
60.2, 55.2, 49.6, 43.8, 37.0, 22.8, 8.1. GC mass spectrum:
m/e 303 (M+).
HPLC: Inertsil, C8, acetonitrile, 45%/water, 55% with
0.1 vol % H3PO4 and 0.2 vol % triethylamine; 19, 6.1 min;
20, 8.0 min.
Cyclopentylhydrazine Dihydrochloride (9). Cyclopen-
tanol (6.13 kg, 71.1 mol) and triphenylphosphine (18.67 kg,
71.25 mol) were dissolved in tetrahydrofuran (151 L) in a
nitrogen-purged 100-gal glass-lined tank and was cooled to
5 °C. A solution of di-tert-butyl azodicarboxylate 21 (14.9
kg, 64.7 mol) in tetrahydrofuran (36 L) was added over about
2 h, keeping the temperature <6 °C. The reaction was
allowed to stir for 5 h as the temperature was allowed to
slowly increase to 20-25 °C. A solution 6 N HCl (26.5 L)
was added to the reaction at 20 °C. The reaction was allowed
to stir for 24 h at 20-25 °C at which point the starting
material had been consumed. Water (37.85 L) was added,
and the tetrahydrofuran was removed by vacuum distillation.
During the concentration, triphenylphosphine oxide precipi-
tated, and an additional 75.7 L of water was added. The
reaction was cooled, and methylene chloride (113.6 L) was
added. The layers were separated, and the aqueous layer was
extracted twice more with methylene chloride (37.85 L) The
aqueous was concentrated by distillation. As the volume was
reduced, 2-propanol (3 × 75.7 L) was added to azeotrope
the residual water. The resulting slurry was filtered and the
solids were dried under vacuum at rt to give 7.68 kg (68.6%
theory) over multiple crops. This material was characterized
1
to be the dihydrochloride salt: mp 189-194 °C. HMR
(DMSO-d6, 300 MHz) d 3.48 (m, 1), 1.79 (m, 2), 1.64 (m,
4), 1.49 (m, 2).
HPLC: Inertsil, C8, acetonitrile, 45%/water, 55% with
0.1 vol % H3PO4 and 0.2 vol % triethylamine; retention
time: 18, 10.8 min; 19, 6.1 min.
3-Methoxy-1-(4-methoxybenzyl)-4-propionyl-5,6-dihy-
dro-1H-pyridin-2-one (20). 3-Hydroxy-1-(4-methoxyben-
zyl)-4-propionyl-5,6-dihydro-1H-pyridin-2-one 19 (17.35 kg,
60 mol) and cesium carbonate (22.13 kg, 67.9 mol) were
added to dry dimethylformamide (90.8 L) in a 100-gal tank.
The suspension was stirred for 30 min to ensure dispersion.
1-Cyclopentyl-3-ethyl-6-(4-methoxybenzyl)-1,4,5,6-tet-
rahydropyrazolo[3,4-c]-pyridin-7-one (23). 3-Methoxy-1-
(4-methoxybenzyl)-4-propionyl-5,6-dihydro-1H-pyridin-2-
one 20 (14.47 kg, 47.76 mol) was dissolved in tetrahydrofuran
(39.7 L) in a 100-gal glass-lined tank. Cyclopentylhydrazine
dihydrochloride 9 (7.66 kg, 44.3 mol) was added, and the
reaction was warmed slowly to ∼88 °C while nitrogen swept
over the reaction to remove methanol, THF, and HCl. The
reaction was monitored by HPLC until the conversion was
Vol. 5, No. 6, 2001 / Organic Process Research & Development
•
579

complete, which required heating overnight in most cases.
The product was a thick dark oil. A sample of 1-cyclopentyl-
3-ethyl-6-(4-methoxybenzyl)-1,4,5,6-tetrahydropyrazolo[3,4-
c]pyridin-7-one 23 was isolated for characterization. 1HMR
(CDCl3, 300 MHz) δ 7.23 (d, 2), 6.85 (d, 2), 5.72 (m, 1),
4.62 (s, 2), 3.77 (s, 3), 3.44 (t, 2), 2.62 (t and q, 4), 2.06 (m,
4), 1.89 (m, 2), 1.67 (m, 2), 1.17 (t, 3). 13CMR (CDCl3, 300
MHz) δ 159.5, 159.0, 148.0, 145.6, 129.6, 129.3, 118.5,
114.0, 112.9, 60.4, 55.2, 48.6, 47.2, 32.7, 24.4, 20.2, 19.9,
13.8. GC mass spectrum: m/e 353 (M+). This was used
directly in the next step. Purification as a p-toluenesulfonic
acid or benzenesulfonic acid salt is described.
Preparation of the p-Toluenesulfonic Acid and Ben-
zenesulfonic Acid Salts of 1-Cyclopentyl-3-ethyl-6-(4-
methoxybenzyl)-1,4,5,6-tetrahydropyrazolo-[3,4-c]pyridin-
7-one (23). Crude lactam 23 (1 g, 2.8 mmol) was dissolved
in ethyl acetate (5 mL) and treated with a solution of
anhydrous p-toluenesulfonic acid (0.487 g, 2.8 mmol) in ethyl
acetate (2 mL). The salt crystallized from the mixture which
was then cooled and filtered to provide 1.21 g of pure tosylate
salt as a white solid in 81% yield: mp 110-113.8 °C. Anal.
Calcd for C28H35N3O5S: C, 63.98; H, 6.71; N, 7.99; S, 6.10.
Found: C, 63.83; H, 6.69; N, 8.02; S, 6.14.
3.51 (dt, 2), 2.72 (t, 2), 2.62 (q, 2), 2.08 (m, 4), 1.90 (m, 2),
1.65 (m, 2), 1.40 (t, 3).
HPLC: Inertsil, C8, acetonitrile, 65%/water, 35%; reten-
tion time: 2a, 4.7 min; 20, 5.1 min; 23, 21.0 min.
1-Cyclopentyl-7-ethoxy-3-ethyl-4,5-dihydro-1H-pyra-
zolo[3,4-c]pyridine (24). A freshly prepared solution of
triethyloxonium tetrafluoroborate (3.37 kg, 17.74 mol) in
methylene chloride (10.8 L) was added slowly to a suspen-
sion of lactam 2a (3.6 kg, 15.43 mol) in methylene chloride
(7.2 L) over a period of about 40 min. The solution was
then allowed to react for about 21 h at 18-22 °C. After the
reaction was complete, the organic solution was washed with
aqueous 10% sodium carbonate (36 L) and evaporated to
an oil which was used directly in the next step. The yield
1
for this step was 92.9%. HMR (CDCl3, 300 MHz) δ 5.14
(quintet, 1), 4.25 (q, 2), 3.62 (t, 2). 2.58 (m, 4), 2.07 (m, 4),
1.88 (m, 2), 1.61 (m, 2), 1.35 (t, 3), 1.19 (t, 3). GC mass
spectrum: m/e 261 (M+).
HPLC: Waters Symmetry-C8, 550 water/250 acetonitrile/
200 methanol/2 perchloric acid; retention time: 24, 6.2 min;
2a, 8.8 min.
9H-Cyclopentyl-7-ethyl-3-(thiophen-2-yl)-pyrazolo[3,4-
c]-1,2,4-triazolo-5,6-dihydro-[4,3-a]pyridine (1). A solution
of 1-cyclopentyl-7-ethoxy-3-ethyl-4,5-dihydro-1H-pyrazolo-
[3,4-c]pyridine 24 (3.74 kg, 14.3 mol) and 2-thiophenecar-
boxylic hydrazide 25 (2.24 kg, 15.8 mol) was heated in
1-butanol (37 L) to ∼90 °C in a 50-gal glass-lined tank for
48 h. At this point, approximately 3 L of 1-butanol was
distilled off to remove water and to help drive the reaction
to completion. The reaction was concentrated to a 5-L
volume, and methylene chloride (15 L) was added. The
organics were washed twice with 1 N HCl (30.3 L) and
concentrated by distillation to 5 L. 2-Propanol (16 L) was
added to the concentrate, and the resulting slurry was cooled
and stirred for 2 h. The product was collected by filtration
and vacuum oven dried at 40 °C. The yield was 3.25 kg
The benzenesulfonic acid salt was formed in the same
manner: mp 126.6-131.4 °C. Anal. Calcd for C27H33N3O5S:
C, 63.38; H, 6.50; N, 8.21. Found: C, 63.09; H, 6.48; N,
8.21.
Either of these crystalline salts can be used in the
deprotection reaction with trifluoroacetic acid and methane-
sulfonic acid.
1-Cyclopentyl-3-ethyl-1,4,5,6-tetrahydropyrazolo[3,4-
c]pyridin-7-one (2a). The reaction mixture from the previous
preparation containing 23 was cooled to 55 °C, and trifluo-
roacetic acid (87.3 kg, 764 mol) was added slowly, while
keeping the temperature between 50 and 60 °C. The first
one-third of the charge was exothermic and required external
cooling. Methanesulfonic acid (6.34 L, 97.7 mol) was added,
and the reaction was warmed to ∼70 °C for 2 h. The reaction
was cooled to 20-25 °C, and methylene chloride (64 L)
was added, followed by the slow addition of water (64 L).
The layers were separated, and the aqueous layer was diluted
further with water (22.7 L) and then re-extracted with
methylene chloride (22.7 L). The combined methylene
chloride layers were mixed with water (110 L) and then
brought to pH 7.0 by the addition of saturated sodium
bicarbonate (ca. 170 L). The layers were separated, and the
methylene chloride was atmospherically distilled to about
35 L. Ethyl acetate (49 L) was added, and the reaction was
distilled to about 35 L. The resulting slurry was cooled to
room temperature with stirring. The solids were collected
by filtration, washed with ethyl acetate, and vacuum-dried
at 40 °C under full vacuum. The yield was 7.91 kg, 71.2%:
1
(67%) of a white solid: mp 126 °C. HMR (CDCl3, 300
MHz) δ 7.51 (m, 2), 7.28 (s, 1), 7.20 (dd, 1), 5.61 (m, 1),
4.35 (t, 2), 3.00 (t, 2), 2.70 (q, 2), 2.18 (m, 4), 1.97 (m, 2),
1.62 (m, 2), 1.29 (t, 3).
Anal. Calcd for C18H21N5S: C, 63.69; H, 6.24; N, 20.63.
Found: C, 63.82; H, 6.30; N, 20.77.
HPLC: Waters Symmetry-C8, 550 water/250 acetonitrile/
200 methanol/2 perchloric acid; retention time: 1, 15.3 min;
24, 6.2 min; 2a, 8.8 min.
Acknowledgment
We thank Clive Mason, Pfizer Central Research, Sand-
wich, UK, for the procedures for the preparation of the salts
of pyrazole 23.
Received for review June 30, 2001.
OP010222X
1
mp 152-153 °C. HMR (CDCl3, 300 MHz) δ 5.61 (m, 2),
580
•
Vol. 5, No. 6, 2001 / Organic Process Research & Development