Process development of a platelet aggregation inhibitor

Article abstract of DOI:10.1021/op980050c

A practical and efficient method for N-amination of piperazine via a nitrosoamine, suitable for a large scale synthesis, is described. This method involved the temporary transformation of an in situ prepared aminopiperazine to a hydrazone, allowing efficient separation of zinc salt byproducts from the system. Acylation and deprotection with hydroxylamine directly afforded FR062732 in satisfactory quality for pharmacological evaluation. These methods solved the operational problems usually inherent in zinc reduction of nitrosoamines.

Full text of DOI:10.1021/op980050c

Organic Process Research & Development 1998, 2, 418−421
Process Development of a Platelet Aggregation Inhibitor
Atsuhiko Zanka
Technological DeVelopment Laboratories, Fujisawa Pharmaceutical Co. Ltd., 2-1-6 Kashima,
Yodogawa-ku, Osaka 532, Japan
Abstract:
A practical and efficient method for N-amination of piperazine
via a nitrosoamine, suitable for a large scale synthesis, is
described. This method involved the temporary transformation
of an in situ prepared aminopiperazine to a hydrazone, allowing
efficient separation of zinc salt byproducts from the system.
Acylation and deprotection with hydroxylamine directly af-
forded FR062732 in satisfactory quality for pharmacological
evaluation. These methods solved the operational problems
usually inherent in zinc reduction of nitrosoamines.
Figure 1. Structures of Tiaramide and FR062732.
Scheme 1. Original route to FR062732 (1)a
Results and Discussion
Classical nonsteroidal antiinflammatory drugs (NSAIDs)
such as indomethacin and aspirin have been reported to have
1
a marked inhibitory action on platelet aggregation. Tiara-
mide (2, Figure 1), developed at Fujisawa Pharmaceutical
Co. Ltd., is also a widely used NSAID, used for treatment
of inflammatory diseases.2 This drug has also been shown
to have potent effects on platelet aggregation.3 This dis-
covery prompted efforts to find novel, selective platelet
aggregation inhibitors. During investigations into Tiaramide
derivatives, it was found that platelet aggregation inhibi-
tory activity mainly depends on the piperazine moiety,
leading to the discovery of 3-[(4-amino-1-piperazinyl)car-
bonylmethyl]-5-chloro-2(3H)-benzothiazolone hydrochloride
(
FR062732, 1, Figure 1), which was a crucial lead showing
4
good activity.
a
Reagents and conditions: (1) NaOH, H
2 4
O-MeOH; (2) piperazine, TiCl ,
The original method for the preparation of 1 in gram
quantities, shown in Scheme 1, was used to support the initial
pharmacological screening.4 The synthesis started with
hydrolysis of 3-(methoxycarbonylmethyl)-5-chloro-2(3H)-
benzothiazolone (3), an intermediate in the Tiaramide
production process, and involved coupling with piperazine
THF; (3) (CH
3
)
2
CHCH CH ONO, CH Cl ; (4) (a) Zn, AcOH, (b) HCl.
2
2
2
2
significantly reduced yield. Additionally, since nitroso
derivatives usually have carcinogenic activity, isolation of
was problematic from the viewpoint of personal hazards,
6
especially on a large scale.
4
in the presence of TiCl to give 5, which was then converted
In accordance with project progress, we were faced with
having to supply 1 in large quantities for toxicological and
biological evaluation. To develop a more practical and
efficient synthesis of 1, we investigated several synthetic
methods. 1 may be prepared most readily by the amination
to 6 with isoamyl nitrite. 6 was then reduced with zinc
powder to afford the desired product 1. At this stage, several
severe complications were encountered. The major draw-
back to this method was that zinc salt byproducts contami-
nated the final product. Tedious and multistep purification
procedures, including recrystallization and/or column chro-
matography, were required in order to meet acceptable limits
of amounts of residual metal (<1 ppm), resulting in a
2 2 3
of 5 with NH Cl or NH OSO H. However, these reagents
are unavailable in large quantities, are difficult to handle in
pilot plants with ordinary equipment, and cannot be stored
for long periods of time. Furthermore, since these methods
produce several byproducts, unacceptably low yields (∼50%)
were obtained, and this strategy was not pursued further.
We thus turned our attention to the preparation of
aminopiperazine in situ and coupling with 4 (Scheme 2).
To our knowledge, practical methods suitable for a large
(
(
1) O’Brien, J. R.; Finch, W.; Clark, E. J. Clin. Pathol. 1970, 23, 522.
2) Tsurumi, K.; Hiramatsu, Y.; Nozaki, M.; Hayashi, M.; Fujimura, H.
Arzneim-Forsch. 1972, 22, 716.
(
(
3) Takano, S.; Sakurai, K.; Suzuki, T. Jpn. J. Pharmacol. 1980, 30, 905.
4) Ueda, I.; Matsuo, M.; Tsuji, K.; Okumura; H. Nakaguchi, O. Eur. Pat. 0
2
32 740 A1.
4
18
•
Vol. 2, No. 6, 1998 / Organic Process Research & Development
10.1021/op980050c CCC: $15.00 © 1998 American Chemical Society and Royal Society of Chemistry
Published on Web 09/12/1998

Scheme 2. New synthetic route to FR062732 (1)a
step, affording a mixture of piperazine, monoaminopipera-
zine, and diaminopiperazine. Simple addition of benzalde-
hyde to the reaction mixture led to the formation of the
corresponding hydrazones. Whilst toluene had poor extrac-
tion efficacy and CH
zone, ethyl acetate was able to selectively extract 7. After
addition of aqueous NH Cl, followed by adjusting the pH
2 2
Cl extracted both 7 and the dihydra-
4
to 9.5 with aqueous NaOH, the layers were separated. In
this sequence, piperazine was completely removed in the
aqueous layer. Quantitative HPLC analysis showed that 2
molar equiv of piperazine was converted to 7 in 65% yield.
Since piperazine is inexpensive ($2.50/lb in bulk) and also
available in large quantities, excess use does not significantly
1
0
increase the cost.
Hydrolysis of 3 smoothly proceeded in the presence of
excess NaOH in a mixture of water and MeOH. After the
reaction was judged complete by HPLC analysis, water was
added to the reaction mixture in order to improve the
filtration procedure, followed by hydrochloric acid. Activa-
a
Reagents and conditions: (1) (a) NaNO
2
, AcOH, H
2
O; (b) zinc powder; (c)
CN-H O.
PhCHO; (2) Et
3
N, CH Cl ; (3) NH OH‚HCl, HCl, CH
2
2
2
3
2
scale synthesis are not known in the literature. Aminopip-
erazine may be prepared by nitrosation of piperazine fol-
lowed by reduction with reagents such as Na, LiAlH
and Zn. Despite the usefulness of reactions with Na or
4
LiAlH on a laboratory scale, it was decided to investigate
alternative methods owing to the pyrophoric hazards associ-
ated with these reagents. Thus, we selected zinc powder as
reagent in the presence of acetic acid. However, this method
also had serious shortcomings in the scale-up procedures that
needed to be solved before the required results could be
realized. One problem is that nitrosopiperazine, which
5
5,6
tion of 4 was conducted with excess amounts of SOCl (5
,
2
4
7
equiv), requiring removal of the remaining SOCl2 by
distillation; however, the Vilsmier method (SOCl
2
(1.02
equiv), DMF, CH Cl ) was more suitable for a large scale
2
2
synthesis. The subsequent acylation proceeded quantitatively
in the presence of an excess amount of 7 on a laboratory
scale; however, this reaction did not give reproducible yields
in the pilot plant campaign. A ∼13 °C exotherm (from 0 to
13 °C) was seen, and the yield was only 68%. Further
studies indicated that the yield is impacted by reaction
temperature, since the acyl chloride of 4 partially decomposes
at >10 °C in basic solution to give unknown byproducts.
Pure 8 was isolated by precipitation from ethyl acetate and
MeOH, with a small amount of dihydrazone remaining in
the mother liquor.
In early studies, the deprotection of 8 was conducted in
the presence of a large excess of hydrogen chloride.
However, the reaction did not proceed smoothly, and
hydrolysis of 8 to 4 was predominant and yielded several
unknown byproducts under higher temperature conditions.
As a result, both methods resulted in 1 in low yield (<10%).
Since 8 and 1 are in an equilibrium state, removal of
benzaldehyde from the system by converting it to an oxime
with hydroxylamine was investigated. Analysis of the
isolated 1 following deprotection of 8 with 5 molar equiv of
hydroxylammonium chloride revealed a small amount (1-
3%) of the starting material. However, HPLC analysis of
the wet cake immediately after filtration confirmed that the
product did not contain any starting material. The reason
for the contamination is not known, although it is possible
that small amounts of benzaldehyde in the filtrate reacted
with 1 after removal of excess hydrogen chloride. The
problem was easily solved by using a large excess of
hydroxylammonium chloride (10 equiv) to ensure complete
removal of the benzaldehyde. The wet cake was triturated
to remove excess hydroxylammonium chloride prior to the
product being dried under reduced pressure. The isolated
8
possesses marked carcinogenic activity, should be im-
mediately reduced with zinc powder without isolation. As
a result, the products are usually a mixture of piperazine,
monoaminopiperazine, and diaminopiperazine. Another
problem is that, even when nitrosopiperazine was purified
by extraction and/or column chromatography, the reduction
with zinc powder also involved deamination as a side
reaction, resulting in contamination by piperazine in the
product. A more fundamental problem is that piperazine
derivatives are strong bases and the zinc salt byproducts,
precipitated after adjusting the aqueous layer to pH 7.0, could
not be separated by filtration without significant product loss.
During our investigation, the most practical solution to these
problems was the temporary transformation of the aminopip-
erazine to a hydrazone, followed by extraction into an organic
phase.
Nitrosation of piperazine was conducted with isoamyl
nitrite, but NaNO in the presence of acetic acid was more
2
efficient for a large scale synthesis. Nitrosopiperazine was
immediately reduced with zinc powder in the presence of
acetic acid, which was already present from the nitrosation
9
(
5) Zimmer, H.; Audrieth, L. F.; Zimmer, M.; Rowe, R. A. J. Am. Chem. Soc.
1
955, 77, 790.
(
(
6) Hedrick, R. J.; Major, R. T. J. Org. Chem. 1964, 29, 2486.
7) (a) Hatt, H. H. Organic Syntheses; Wiley: New York, 1943; Collect. Vol.
II, p 211. (b) Hartman, W. W.; Roll, L. J. Organic Syntheses; Wiley: New
York, 1943; Collect. Vol. II, p 418.
8) Sweet, D. V. Registry of Toxic Effects of Chemical Substances (RTECS);
Department of Health and Human Services: Washington, DC, 1985-86;
Vol. 4, p 3551.
(
(
9) Fridman, A. L.; Mukhametshin, F. M.; Novikov, S. S. Russ. Chem. ReV.
1
971, 40, 34.
(10) Chem. Mark. Rep. 1996, 250 (11), 29 (step 9).
Vol. 2, No. 6, 1998 / Organic Process Research & Development
•
419

yield of 1 was about 62% from 8, regardless of the scale,
mainly due to significant loss in the mother liquor (19%)
and partitioning in the aqueous layer (5%), and contained
less than 0.4% total impurities and no zinc salt byproducts.
Several alternative isolation methods were investigated with
the aim of decreasing the mechanical loss, but these only
resulted in product of unsatisfactory quality and requiring
further purification procedures. Therefore, the methods
described were the best at this point when process economics
and operation were considered.
3
ClNO S: C, 44.36; H, 2.48; N, 5.75. Found: C, 44.33; H,
2.40; N, 5.68.
3-[(4-Benzylideneamino-1-piperazinyl)carbonylmethyl]-
5-chloro-2(3H)-benzothiazolone (8). Piperazine (86.14 g,
1
mol) was dissolved in a mixture of acetic acid (300 g, 5
mol) and water (430 mL). To this solution was added
dropwise a solution of sodium nitrite (69 g, 1 mol) in water
(172 mL) at 10 °C, and the reaction mixture was stirred for
1
h at ambient temperature. Zinc powder (172 g, 2.63 mol)
was then added to the solution in portions, maintaining the
temperature below 40 °C, and the reaction mixture was
stirred for 1 h. The precipitate was filtered off, and the cake
of zinc salts was washed with water (172 mL). Benzalde-
hyde (106 g, 1 mol) in 424 mL of ethanol was added to the
combined solutions of the filtrate and washings, and after
the mixture was stirred for 1 h at room temperature, ethyl
acetate (1034 mL) and ammonium chloride (281 g, 5.25 mol)
were added, and the pH was adjusted to 9.5 with 24% sodium
hydroxide in water. The layers were separated, and the
aqueous layer was re-extracted with ethyl acetate (690 mL).
The insoluble materials in the combined organic layer were
filtered off. The filtrate was dried over magnesium sulfate,
and, after filtration, the magnesium sulfate was washed with
methylene chloride (300 mL). To the combined filtrate was
added triethylamine (126.5 g, 1.25 mol), followed by cooling
below 0 °C. In another vessel, thionyl chloride (61 g, 0.51
mol) was added to 4 (122 g, 0.5 mol) in a mixture of
methylene chloride (610 mL) and dimethylformamide (61
mL). The reaction mixture was refluxed for 1 h and then
cooled to 20 °C. This solution was added dropwise to the
previous solution, maintaining the temperature below 0 °C,
and the reaction was stirred for 1 h at ambient temperature.
The reaction mixture was concentrated to ∼910 mL, and
methanol (1.2 L) was added. After the mixture was stirred
for 0.5 h at 0 °C, the precipitate was filtered off, washed
with methanol (366 mL), and dried to afford 8 (189 g, 91%
Conclusion
In conclusion, we have established a practical and facile
synthesis of 3-[(4-amino-1-piperazinyl)carbonylmethyl]-5-
chloro-2(3H)-benzothiazolone hydrochloride (FR062732, 1).
The conversion of a hydrazine to a hydrazone resolved
several operational problems and allowed the development
of a workable process. The procedures described in this
paper should be useful for the efficient synthesis of hydra-
zines on a large scale. The use of an optimum amount of
hydroxylammonium chloride as a scavenger for benzalde-
hyde resulted in a significant improvement in yield for the
deprotection of hydrazone, and these methods were used to
produce approximately 14 kg of 1.
Experimental Section
Melting points were measured on a Thomas-Hoover
apparatus and are uncorrected. IR spectra were recorded on
a HORIBA FT-210 spectrometer. NMR spectra were
1
measured on a Bruker AC200P ( H, 200 MHz). Chemical
shifts are given in parts per million, and tetramethylsilane
was used as the internal standard. Mass spectra were
measured on a Hitachi model M-80 mass spectrometer using
EI for ionization. Elemental analyses were carried out on a
Perkin-Elmer 2400 CHN elemental analyzer. HPLC analyses
were performed using a YMC GEL ODS 120-Å S-7 column
3
and a 0.1% TFA in CH CN and water mobile phase. The
1
yield) as a white solid: mp 213-215 °C; H NMR (200 MHz,
amount of zinc salt byproducts was measured on a Hitachi
Zeemann type atomic absorption spectrophotometer 180-80.
Reagents and solvents were used as obtained from com-
mercial suppliers without further purification.
DMSO-d
6
) δ 3.15-3.29 (m, 4H), 3.65-3.76 (m, 4H), 5.01
(
s, 2H), 7.23-7.74 (m, 9H); IR (KBr) 1695, 1683, 1647,
-
1
+
1
588, 1570, 1472 cm ; MS (EI) m/z 415 (M + H) . Anal.
S: C, 57.90; H, 4.62; N, 13.50.
Found C, 57.76; H, 4.56; N, 13.25.
-[(4-Amino-1-piperazinyl)carbonylmethyl]-5-chloro-
(3H)-benzothiazolone hydrochloride (1, FR062732). To
4 2
Calcd for C20H19ClN O
3-Carboxymethyl-5-chloro-2(3H)-benzothiazolone (4).
To a mixture of 3-(methoxycarbonylmethyl)-5-chloro-2(3H)-
benzothiazolone (3) (200 g, 0.776 mol) in MeOH (500 mL)
was added dropwise 8% NaOH in water (500 mL, 1 mol).
The reaction mixture was stirred at 25-30 °C over 3 h,
followed by addition of water (5.6 L). After the reaction
mixture was cooled to 10 °C, 36% HCl in water (107 mL)
was added. On complete addition, the resulting mixture was
stirred at ambient temperature for 1 h and the precipitate
filtered off and washed with water (133 mL). Drying under
3
2
a solution of hydroxylamonium chloride (168 g, 2.42 mol)
in a mixture of acetonitrile (2.0 L) and 17.5% hydrochloric
acid (1.0 L) was added 8 (100 g, 241 mmol). The reaction
mixture was stirred for 4 h at 47-50 °C and then cooled to
0
°C. The precipitate was filtered off and washed succes-
sively with water (200 mL) and methylene chloride (500
mL). Trituration of the crude cake with water (200 mL)
afforded 1 (54.3 g, 62% yield) as a white solid. The product
was confirmed as containing no zinc salt byproducts with
reduced pressure afforded 4 (182 g, 96% yield) as a white
1
solid: mp 242-243 °C; H NMR (200 MHz, DMSO-d
6
) δ
4.77 (s, 2H), 7.27 (dd, 1H, J ) 8.4, 2.0 Hz), 7.57 (d, 1H, J
)
2.0 Hz), 7.71 (d, 1H, J ) 8.4 Hz), 13.34 (brs, 1H); IR
the atomic absorption spectrophotometer: mp 261-263 °C;
-
1
1
(KBr) 1745, 1687, 1678, 1590, 1575, 1474 cm ; MS (EI)
2
H NMR (200 MHz, D O) δ 3.13-3.28 (m, 4H), 3.78-3.87
+
m/z 244 (M + H) , 170, 75. Anal. Calcd for C
9
H
6
-
(m, 4H), 4.98 (s, 2H), 7.16 (d, 1H, J ) 1.9 Hz), 7.27 (dd,
420
•
Vol. 2, No. 6, 1998 / Organic Process Research & Development

1
1
H, J ) 8.4, 1.9 Hz), 7.53 (d, 1H, J ) 8.4 Hz); IR (KBr)
Chemistry Research Laboratories, Fujisawa Pharmaceutical
Co. Ltd., for his interest and ongoing advice in this work.
-
1
677, 1667, 1642, 1591, 1533, 1480, 1467 cm ; MS (EI)
S: C,
2.98; H, 4.44; N, 15.42. Found C, 43.03; H, 4.43; N, 15.36.
2 4 2
m/z 327, 306, 73. Anal. Calcd for C13H16Cl N O
4
Received for review June 15, 1998.
OP980050C
Acknowledgment
I especially wish to thank Dr. David Barrett, Medicinal
Vol. 2, No. 6, 1998 / Organic Process Research & Development
•
421