Development summary towards a manufacturable process for R 83842 [(S)-6-[(4-chlorophenyl) (1H-1,2,4-triazol-1-yl)methyl]-1-methyl-1H-benzotriazole]

Article abstract of DOI:10.1021/op990081n

A scalable process to produce enantiomeric R 83842, (S)-6-[(4-chlorophenyl)(1H-1,2,4-triazol-1-yl)methyl]-1-methyl-1H-benzotriazole, is developed and described as a lecture * transcript. Cheap and safe reagents have been used. A typical procedure for oxidative destruction of aqueous cyanide waste, and stability data on N-acetyl hydrazine are provided. Special focus is on cost analysis as an important tool in developing performant synthetic methods. The method consists of preparing a chiral monosubstituted hydrazine which is ring closed to the chiral 1-alkylated 1,2,4-triazole title compound.

Full text of DOI:10.1021/op990081n

Organic Process Research & Development 2000, 4, 162−166
Development Summary towards a Manufacturable Process for R 83842
[(S)-6-[(4-chlorophenyl)
(1H-1,2,4-triazol-1-yl)methyl]-1-methyl-1H-benzotriazole]†
A. G. M. De Knaep,* A. M. J. Vandendriessche, D. J. E. Daemen, J. J. Dingenen, K. D. Laenen, R. L. Nijs,
F. L. J. Pauwels, D. F. Van den Heuvel, F. J. Van der Eycken, R. W. E. Vanierschot, G. M. L. W. van Laar,
W. L. A. Verstappen, and B. L. A. Willemsens
Janssen Pharmaceutica N.V.,Turnhoutseweg 30, B-2340 Beerse, Belgium
Scheme 1
Abstract:
A scalable process to produce enantiomeric R 83842, (S)-6-[(4-
chlorophenyl) (1H-1,2,4-triazol-1-yl)methyl]-1-methyl-1H-ben-
zotriazole, is developed and described as a lecture* transcript.
Cheap and safe reagents have been used. A typical procedure
for oxidative destruction of aqueous cyanide waste, and stability
data on N-acetyl hydrazine are provided. Special focus is on
cost analysis as an important tool in developing performant
synthetic methods. The method consists of preparing a chiral
monosubstituted hydrazine which is ring closed to the chiral
1-alkylated 1,2,4-triazole title compound.
Scheme 2
I. Racemic R 83842
At first, racemic R 83842 ((()-(1) was selected as a highly
selective aromatase inhibitor.1 An important element in that
decision was that nor by synthesis nor by chromatography,
both enantiomers could be obtained in sufficient quantity to
investigate and differentiate them. During the development
however, we were forced to a racemic switch. Racemic R
83842 is discussed first. The discovery route is summarized
in schemes 1-5. Suitable 4-nitro benzoic acid derivatives
11 were not commercially available in larger quantities.
Hence, it was prepared as in Scheme 3.
As shown in schemes 3-5, the discovery route is fully
linear, comprises 10 synthetic steps and a chromatographic
step to remove the isomeric by-product 15. The overall yield
is about 6% which is very high. The average yield per step
is 72%. However, some steps raise a concern for the plant
like the usage of MnO2 (abrasive and jeopardising cleaning),
KMnO4 (same as MnO2), LiAlH4, NaH, dimethylsulphate,
and a 4:1 ratio of (()-1 and 15 in the final step. The
chromatographic yield to isolate 1 is very high, but a
crystallisation to achieve the removal of 15 will significantly
reduce the overall yield.
We managed to solve most of the problems by synthe-
sising (()-1 as indicated in Schemes 6 and 7.
The starting materials are abundantly available and very
cheap. Key steps are the regioselective amination with
methylamine, the phase-transfer catalysed arylation of 4-chloro-
phenyl acetonitrile (18), and subsequent air oxidation giving
20.
The benzhydrylic alcohol 4 is obtained in six steps,
avoiding the inconvenient reagents mentioned earlier. The
regioselective alkylation of 1,2,4-triazole (3) is achieved by
† Lecture given at the 14th SCI Process Development Symposium in
Manchester, and at the 1st International Conference on Organic Process Research
and Development in San Francisco.
(1) Raeymaekers, A.; Freyne, E.; Van Gelder, J.; Venet, M. (Janssen Phar-
maceutica). Eur. Pat. B-O 293,978, 1987.
162
•
Vol. 4, No. 3, 2000 / Organic Process Research & Development
10.1021/op990081n CCC: $19.00 © 2000 American Chemical Society and The Royal Society of Chemistry
Published on Web 03/02/2000

Scheme 3
Scheme 6
Scheme 4
Scheme 7
Scheme 5
The air oxidation of benzhydrylcyanides to benzophe-
nones was studied by A. Donetti.3 For diphenylacetonitrile
in DMSO, the oxidation appears to almost exclusively follow
pathway A, producing mainly cyanate as a by-product
(Scheme 8).
We found the air oxidation to proceed smoothly in DMA
using N-benzyl triethylammonium chloride as a phase-
transfer catalyst and by bubbling air through the reaction
mixture.
Surprisingly for nitrile 19, in DMA and using a phase-
transfer catalyst, we found pathway B dominating. Next to
only 9.5% of cyanate, 90.5% of cyanide was formed.
As a consequence, the aqueous cyanide waste needed
further processing, that is, oxidation, to reduce the cyanide
content below 3 ppm before sending it to the waste water
plant.
alkylating 4-aminotriazole (21) and by deamination, a proven
methodology2 (Scheme 7).
In the meantime some enantiomeric R 83842 was obtained
in a very laborious way, intensively using chromatography.
The (S)-enantiomer proved superior to the (R)-enantiomer,
and thus we decided for a racemic switch. Before dealing
with the racemic switch, a few interesting notes on the
chemistry are outlined in Schemes 6 and 7.
The oxidation is complete within 1 h at room temperature,
but the decomposition of the cyanate formed is much
(2) Scriven, E.; Astleford, B. A.; Goe, G. L.; Keay, J. G. J. Org. Chem. 1989,
54 (3), 731.
(3) Donetti, A.; Boniardi, O. Synthesis 1980, 1009.
Vol. 4, No. 3, 2000 / Organic Process Research & Development
•
163

Scheme 8
Scheme 10
Scheme 11
Scheme 9
acids. Chiral chromatography is a practical way for kilogram
quantities, less practical for larger scale. 1,2,4-Triazoles can
be manufactured from hydrazine derivatives by ring closure
with, for example, formamide,6 formamidine,7 and 1,3,5-
triazine.8 Preliminary experiments showed little loss in optical
purity if chiral hydrazines were used in the ring closure. Fine-
tuning of the reaction conditions gave adequate control of
the optical purity. Enantioselective reduction of hydrazones
is a known route for chiral hydrazines,9 but unfortunately it
is unsuccessful for hydrazones of benzophenones (Scheme
10).
slower, as we found out incidently. The aqueous solution
was stored in a closed drum. We should have controlled the
decomposition of cyanate before charging in the drums and
should have had pressure relief in place, as a drum ruptured
overnight.
Therefore, we set off screening for resolutions of hydra-
zine derivatives suitable for ring closure to the desired 1,2,4-
triazole 1 (Scheme 11).
Another example showing the importance of addressing
the total reaction equation is the deamination of 22. One
might expect N2 to be formed as a by-product, but to our
surprise, N2O is also formed. We should have known, as
Fischer already published this in 1877.4 We confirmed it
qualitatively, but M. De Rosa5 quantified the amount of
nitrous oxide to be 35% of theory for 1,1-diphenylhydrazine
(Scheme 9). This is more than just nice to know, especially
in the plant.
We found the monosubstituted hydrazine 23 (RdH) the
most suitable for both the resolution and ring closure.10
Hydrazine derivative 23 can be obtained either by direct
alkylation of 14 on aqueous hydrazine, or on mono-
acetylhydrazine and subsequent hydrolysis (Scheme 12).
For safety reasons, we do not want to handle hydrazine
in our plants. As mono-acetyl hydrazine is a very cheap solid,
we accepted the extra step going over 24. Note, however,
that mono-acetyl hydrazine is hygroscopic and should be
stored cold for not too long a period of time, as it rearranges
slowly on standing, giving di-acetylhydrazine and hydrazine.
II. Enantiomeric (S)-R 83842 ((-)-1)
Racemic R 83842 ((()-(1) is a very weak base. The pKa
of triazole is 2.3 and for benzimidazole the pKa is 1.6, which
made it hard to establish a resolution method using chiral
(6) Petree, H. (Ciba-Geigy). U.S. Patent 79-92 257,791,107, 1981.
(7) Huang, Y. Zhongguo Yiyao Gongye Zazhi 1990, 21 (10), 466.
(8) Grundmann, C.; Ra¨tz, R. J. Org. Chem. 1956, 21, 1037.
(9) Burk, M.; Feaster, J. E. J. Am. Chem. Soc. 1992, 114, 6266.
(10) Willemsens, B.; Verstappen, W.; Copmans, A.; Vandendriessche, A.; De
Knaep, A.; Venet, M. (Janssen Pharmaceutica). Eur. Pat. A-O 668,861,
1992.
(4) Fischer, E. J. Liebigs Ann. Chem. 1877, 170, 158.
(5) De Rosa, M.; Haberfield, P. J. Org. Chem. 1981, 46, 2639.
164
•
Vol. 4, No. 3, 2000 / Organic Process Research & Development

Scheme 12
Table 1
full cost
%
1. labour
2. equipment
3. THF
4. DMA
65.5
8.9
3.6
2.8
2.5
2.3
2.3
2.3
2.2.
5. NRSa low Cl
6. NRS high Cl
7. Pt 5% on carbon
8. HBr acetic acid solution
9. formamidine acetate
a NRS ) non recoverable solvent.
Table 2
raw material cost
%
solvent + solvent waste
first building block
50.8
8.6
total building blocks
miscellaneous
26.6
14.0
Scheme 13
(additives, cats, packaging, inorganics, ...)
reagents, but also solvents used and waste produced are
considered, as well as operations.
Table 1 shows that indeed we succeeded in using very
cheap reagents, as the first building block (formamidine) is
only in position 9 of the full cost breakdown, contributing
only 2.2%. The major cost is labour accounting for two/
thirds of the costs. It is clear where our focus for further
cost reduction should be.
Solvents often surprise many beginning development
chemists, and solvent waste, even more.
Isn’t it surprising that simple solvents such as THF and
DMA take position 3 and 4?
Low chlorine-containing solvent waste (<5% chlorine)
and higher are in position 5 and 6!
The raw material cost breakdown is even more striking
(Table 2). Solvents and solvent waste make up half of it,
almost double the total cost of all building blocks. Note also
that aids (miscellaneous) account for half of the cost of
building blocks. Clearly, cost analysis during development
is critical.
And there is more to a good synthetic method than the
number of steps and the simplicity of the reagents. Process
engineering, solvent recycling and recovery, mastering waste,
and safety assessment of reactions, reagents, and intermedi-
ates are equally important elements for chemical develop-
ment.
Experimental Section
1. Oxidative Destruction of Cyanide in Aqueous Waste.
Typical Procedure. The pH of 500 mL of aqueous cyanide
solution (8100 ppm or 0.156 mol of cyanide) is adjusted to
10.5 by adding sulphuric acid. At 50 °C is added dropwise
over 1 h, 1 equiv of H2O2 (15 g, 35.4 wt %), and the pH is
kept between 9.5 and 10.5. At higher pH (>11) the reaction
slows down, and some H2O2 may decompose. Typically, the
oxidation is done within 1 h. Cyanide concentration should
Scheme 13 gives a summary of the optimised synthesis
of (S)-1. It is an 11-step linear synthesis using cheap and
safe reagents that was scaled up and validated successfully.
During development, cost analysis is a standard tool used
to achieve optimal economics. Not only raw materials and
Vol. 4, No. 3, 2000 / Organic Process Research & Development
•
165

have dropped <3 ppm; if not, some additional H2O2 is added.
DO NOT store the resulting solution in a closed system,
as pressure may build up as a result of cyanate decomposi-
tion. Always check the compatibility of all constituents of
the aqueous cyanide solution with H2O2.
2. Stability of N-Acetyl Hydrazine. N-acetylhydrazine
was found to be very hygroscopic and prone to decomposi-
tion giving N,N′-di-acetylhydrazine along with hydrazine.
The decomposition is relatively fast at room temperature.
After a few months already, the solid material has turned
into a semi-liquid mass, with product concentrations going
<80%. At 4 °C, however, in a triple polyethylene bag with
a desiccant present, the product is reasonably stable (no
decomposition after 3 months). N-acetylhydrazine HCl was
found to be stable at room temperature (no change after 4
months).
Received for review October 1, 1999.
OP990081N
166
•
Vol. 4, No. 3, 2000 / Organic Process Research & Development