Process development of an inherently safer oxidation: Synthesis of 2-chloro-6-methylbenzoic acid in the R411 manufacturing process

Article abstract of DOI:10.1021/op050083+

Many oxidation reactions can be hazardous when run on large scale. The manufacturing process for the production of R411, a developmental compound indicated for the treatment of asthma, includes the oxidation of 2-chloro-6-methylbenzaldehyde to the corresponding carboxylic acid. The use of sodium chlorite in this transformation was efficient and economical, but there were safety concerns regarding the use of hydrogen peroxide to scavenge unwanted hypochlorite, which was generated as a byproduct of the reaction. During the development of the R411 manufacturing process, an inherently safer oxidation system was discovered using a stoichiometric quantity of dimethyl sulfoxide (DMSO) as scavenger. The new process provided equivalent yields and purities to the hydrogen peroxide procedure, thus maintaining the economic viability of the process. The developed process was demonstrated in fixed equipment on a 300 gal scale.

Full text of DOI:10.1021/op050083+

Organic Process Research & Development 2005, 9, 1003−1008
Technical Notes
Process Development of an Inherently Safer Oxidation: Synthesis of
2-Chloro-6-methylbenzoic Acid in the R411 Manufacturing Process
Flavio Chavez Lopez,* Ashish Shankar, Maethonia Thompson, Brandon Shealy, Dobbert Locklear, Thimma Rawalpally,
Thomas Cleary, and Charles Gagliardi†
Pharmaceutical Process DeVelopment, Roche Carolina, Inc., Florence, South Carolina 29506, U.S.A.
Abstract:
has been found to be a potent and specific inhibitor of T-cell
and eosinophil activation in vitro.2 In animal models of
asthma, R411 shows potential as a novel nonsteroidal
approach for the treatment of asthmatic patients.
Many oxidation reactions can be hazardous when run on large
scale. The manufacturing process for the production of R411,
a developmental compound indicated for the treatment of
asthma, includes the oxidation of 2-chloro-6-methylbenzalde-
hyde to the corresponding carboxylic acid. The use of sodium
chlorite in this transformation was efficient and economical,
but there were safety concerns regarding the use of hydrogen
peroxide to scavenge unwanted hypochlorite, which was gener-
ated as a byproduct of the reaction. During the development
of the R411 manufacturing process, an inherently safer oxida-
tion system was discovered using a stoichiometric quantity of
dimethyl sulfoxide (DMSO) as scavenger. The new process
provided equivalent yields and purities to the hydrogen peroxide
procedure, thus maintaining the economic viability of the
process. The developed process was demonstrated in fixed
equipment on a 300 gal scale.
2-Chloro-6-methylbenzoic acid (2) is a key raw material
in the synthesis of 1, but despite its simple structure there
are currently no economical sources for commercial quanti-
ties of this material and the compound has been difficult to
prepare in high yields and purities. Several synthetic routes
to 1 were evaluated in Process Research, and the most
commercially viable route was determined to be from
2-chloro-6-fluorobenzaldehyde (3). Protection of the alde-
hyde as the imine 4, followed by introduction of a methyl
group by a Grignard reaction, formed the 2-chloro-6-
methylbenzaldehyde 5. Oxidation of 5 to the corresponding
benzoic acid was then carried out using sodium chlorite
(Scheme 1).3
While this method provided good yields and was adequate
for clinical supplies of multi-kilo quantities, there were safety
concerns on scaling the reaction to a commercial process
due to the use of hydrogen peroxide in the reaction.
Hypochlorite, which forms as a byproduct of the oxidation,
can cause side reactions as its concentration in the system
increases (eq 1).4 Hydrogen peroxide is added to the reaction
to scavenge hypochlorite; however the use of this reagent
leads to the loss of an inert atmosphere in the reactor due to
the generation of oxygen during consumption of hypochlorite
(eq 2).4
Introduction
The integrin VLA-4 (R4â1) is expressed on a variety of
leukocytes including T-cells, eosinophils, and mast cells.
Recruitment, activation, and retention of these types of cells
are implicated in the inflammatory cascade associated with
asthma by releasing a variety of substances that cause both
acute and chronic damage to the airways.1 R411 (N-(2-
chloro-6-methylbenzoyl)-4-[(2,6-dichlorobenzoyl)amino]-L-
phenylalanine-2-(diethylamino)ethyl ester hydrochloride, 1)
RCHO + HClO2 f RCO2H + HOCl
HOCl + H2O2 f HCl + O2 + H2O
(1)
(2)
(1) (a) Abraham, W. M.; Gill, A.; Ahmed, A.; Sielczak, M. W.; Lauredo, L.
T.; Botinnikova, Y.; Lin, K. C.; Pepinsky, B.; Leone, D. R.; Lobb, R. R.;
Adams, S. P. Am. J. Respir. Crit. Care Med. 2000, 162, 603. (b) Lin, K.
C.; Ateeq, H. S.; Hsiung, S. H.; Chong, L. T.; Zimmerman, C. N.; Castro,
A.; Lee, W. C.; Hammond, C. E.; Kalkunte, S.; Chen, L. L.; Pepinsky, R.
B.; Leone, D. R.; Sprague, A. G.; Abraham, W. M.; Gill, A.; Lobb, R. R.;
Adams, S. P. J. Med. Chem. 1999, 42, 920. (c) Boer, J.; Gottschling, D.;
Schuster, A.; Semmrich, M.; Holzmann, B.; Kessler, H. J. Med. Chem.
2001, 44, 2586.
(2) (a) Sidduri, A.; Tilley, J. W.; Lou, J. P.; Chen, L.; Kaplan, G.; Mennona,
F.; Campbell, R.; Guthrie, R.; Huang, T. N.; Rowan, K.; Schwinge, V.;
Renzetti, L. M. Bioorg. Med. Chem. Lett. 2002, 12, 2479. (b) Hull, K. G.;
Achytharao, S.; Tilley, J. W. Int. Patent Appl. WO200048994. (c) Chen,
L.; Guthrie, R. W.; Huang, T. N.; Hull, K. G.; Achytharao, S.; Tilley, J.
W. Int. Patent Appl. WO9910312.
* Author for correspondence. E-mail: flavio.chavez_lopez@roche.com.
† Present address: Process Engineering, Roche Colorado Corporation, 2075
North 55th Street, Boulder, CO 80301.
10.1021/op050083+ CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/14/2005
Vol. 9, No. 6, 2005 / Organic Process Research & Development
•
1003

Scheme 1. Synthesis of 2-chloro-6-methylbenzoic acid from 2-chloro-6-fluorobenzaldehyde
Another concern was the potential accumulation of
hydrogen peroxide in the reactor. The decomposition of
hydrogen peroxide is known to be highly exothermic, and
while the decomposition rate depends strongly on both
composition and temperature, it can also be catalyzed by
the materials of construction used in the reactor or by various
trace contaminants.5,6 Although some hydrogen peroxide
chemistry is routinely performed in properly passivated
stainless steel or Hastelloy C reactors, exposure to unpas-
sivated metal surfaces in the reactor train may result in highly
accelerated decomposition even at ambient temperatures.
Hydrogen peroxide is also known to form explosive mixtures
with various organic compounds (including aldehydes and
ketones) in certain concentration ranges.5,6 For these reasons,
the hazards of using hydrogen peroxide on this step along
with potential alternates to hydrogen peroxide were inves-
tigated to improve the safety profile of this reaction for future
commercial operation.
Hastelloy C bomb, showed an exotherm onset temperature
of approximately 50 °C (æ ) 1.6; Cp ) 2.175 kJ/kg K),
with a calculated adiabatic temperature rise of roughly 155
°C. This compared rather poorly with the ∼122 °C that
would be indicated exclusively from the decomposition of
the peroxide if it had partitioned fully with the organic
components. The excess energy indicated the occurrence of
some undesired reaction in the system. A second exotherm
was then observed, with an onset temperature of about 175
°C, which was tracked to ∼250 °C before the experiment
was terminated. Since this second exotherm was terminated
while the self-heat rate was still increasing, we can only
roughly estimate the associated energy release as g600 kJ/
kg (g ∼260 °C adiabatic temperature rise). The experimental
temperature and pressure profiles for both exotherms are
shown in Figure 1, and the associated self-heat and pressure
rates are shown in Figure 2. Taken together, the two
exotherms showed the potential for a very significant energy
release.
After considering the potential for developing explosive
compositions in the vapor space of the reactor, the magnitude
of the potential energy release at elevated temperatures, and
the potential for accidental catalysis of undesired chemistry,
an investigation was carried out to find an inherently safer
oxidation system. DMSO has been shown to be an effective
HOCl scavenger when used as solvent since it is oxidized
to dimethyl sulfone;4 however the high boiling point of the
solvent would present downstream processing issues if used
in substantial quantities. To avoid the presence of excess
DMSO, the compound was added as a stoichiometric charge,
in the absence of hydrogen peroxide, and these conditions
were found to be adequate to drive the reaction to completion.
Under these reaction conditions, the oxidation of the aldehyde
is performed as described before (eq 1) and the hypochlorite
byproduct is consumed by oxidizing DMSO to dimethyl
sulfone (eq 3).
Results and Discussion
The energy release associated with the decomposition of
hydrogen peroxide is well defined.5 It was calculated that if
the reaction was run on a 300 gal scale, 43 kg of aldehyde
3 and 37.6 kg of 35% H2O2 would be required per batch.
Full decomposition of the H2O2 would produce 6.19 kg (or
193 mol or ∼1250 gal @ 1 atm and 25 °C) of O2 gas,
accompanied by an adiabatic temperature rise of ∼43 °C
for a well-mixed system (281.5 kg total mass; ∼3.15 kJ/kg
K heat capacity). In this case, the oxygen release to the
headspace and the associated risks of deflagration would pose
a much greater threat than the temperature rise of the batch.
The organic and aqueous layers quickly separate in the
absence of agitation. If the peroxide partitioned entirely into
the aqueous phase, the expected adiabatic temperature rise
would be 66 °C (137.5 kg; ∼4.18 kJ/kg K). If the peroxide
partitioned entirely into the organic phase (worst case
scenario), the anticipated adiabatic temperature rise would
be ∼122 °C (144 kg; ∼2.175 kJ/kg K) without phase-change
corrections. Using a value of 808.8 kJ/kg for the heat of
vaporization, approximately 22 kg (∼22%) of the acetonitrile
in the system would be vaporized at the boiling point of the
organic phase (∼82 °C) in a vented system.
RCHO + HClO2 f RCO2H + HOCl
DMSO + HClO f DMSO2 + HCl
(1)
(3)
As an additional processing benefit, the process capacity
was improved by a factor of 2 and waste was minimized
using the DMSO procedure since buffer was no longer
required in the reaction. The dimethyl sulfone byproduct of
the oxidation was removed in the aqueous extractions during
workup. The tandem base-acid purification workup was no
longer required, and the products were isolated by direct
crystallization from the reaction mixture, resulting in reduced
process cycle time and additional reductions in waste.
Overall, the new process provided better yields and similar
purities to the hydrogen peroxide procedure, thus maintaining
the economic viability of the process. To show the general
The effect of the organic components in the system was
evaluated using adiabatic calorimetry. An accelerating rate
calorimetry (ARC) test on the organic layer, run in a
(3) Daniewski, A. R.; Liu, W.; Puntener, K.; Scalone, M. Org. Process Res.
DeV. 2002, 6, 220-224.
(4) Dalcanale, E.; Montanari, F. J. Org. Chem. 1986, 51, 567-569.
(5) Bretherick’s Handbook of ReactiVe Chemical Hazards, 5th ed.; Urben, P.
G., Ed.; Butterworth-Heinemann: 1995; Vol. 1, pp 1519-1538, Vol. 2,
pp 291-296.
(6) Jeff, M. Hydrogen Peroxide - The Safe Supply and Handling of HTP;
Solvay Interox, Inc.: 3333 Richmond Ave., Houston, Texas 77098.
1004
•
Vol. 9, No. 6, 2005 / Organic Process Research & Development

Figure 1. Temperature and pressure profiles for two exotherms observed during the ARC screening of the organic phase, sampled
after the hydrogen peroxide addition.
Figure 2. Self-heat and pressure rates for two exotherms observed during the ARC screening of the organic phase, sampled after
the hydrogen peroxide addition.
applicability of the process, some simple aldehyde substrates
were oxidized to their carboxylic acids using the DMSO
procedure, and the results are shown in Table 1.
Prior to scale-up, the new process was tested for thermal
stability. The first reaction mixture tested in the ARC (æ )
1.45) after the addition of sulfuric acid to a mixture of
2-chloro-6-methylbenzaldehyde, acetonitrile, water, and DMSO
showed an onset at 84 °C with a calculated adiabatic
temperature rise of approximately 8.4 °C. A second exotherm
was observed with an onset of 134 °C and a calculated
adiabatic temperature rise of >109 °C. (The ARC experiment
was terminated at 250 °C.)
Vol. 9, No. 6, 2005 / Organic Process Research & Development
•
1005

Table 1. Oxidation of aldehydes using HClO2 and DMSO as HOCl scavenger
A second ARC investigation was carried out on the
reaction mixture after the addition of sodium chlorite. The
ARC run (æ ) 1.34) showed an initial minor exotherm with
an onset of 139 °C and a calculated adiabatic temperature
rise of 2.6 °C. On continued heating a second minor
exotherm was detected with an onset of 154 °C and a
calculated adiabatic temperature rise of 3.1 °C. On further
heating and screening, a third exotherm with an onset of 164
°C and a calculated adiabatic temperature rise of 78 °C was
measured giving a final calculated adiabatic temperature of
242 °C. Raw temperature and pressure tracking data of this
ARC run and corresponding self-heat and pressure rise rates
are shown in Figures 3 and 4. This process compares
favorably with the first process using hydrogen peroxide
where the exotherm onset temperature was approximately
50 °C and the adiabatic temperature rise > 155 °C.
There are three additional exothermic events in the new
process, which include the exothermic addition of sulfuric
acid, the exothermic addition of sodium chlorite (aq), and
the exothermic quench with sodium sulfite (aq). The addition
of sulfuric acid is moderately exothermic with a measured
reaction enthalpy of -56.9 kJ/kg (reaction mixture at addition
end), and the adiabatic temperature rise is calculated at 19.3
°C. The addition is performed at 10 °C, and therefore this
gives a maximum reaction temperature of 29.3 °C in the
absence of cooling. The addition was performed over 4 min,
and >95% of measured energy release was complete at the
end of the addition of sulfuric acid. ARC screening reported
above shows a mild self-heat exotherm with an onset
temperature of 84 °C which is >60 °C over the maximum
predicted adiabatic reaction temperature.
Figure 4. Self-heat and pressure rate exotherms observed
during the ARC screening of the reaction mixture, sampled
after the sodium chlorite addition.
Figure 3. Temperature and pressure profiles for exotherms
observed during the ARC screening of the reaction mixture,
sampled after the sodium chlorite addition.
1006
•
Vol. 9, No. 6, 2005 / Organic Process Research & Development

The addition of sodium chlorite is also exothermic with
a measured reaction enthalpy of -458.7 kJ/kg (of reaction
mixture at addition end), and the adiabatic temperature rise
is calculated at 141.6 °C. The addition is performed at 10
°C, and this gives a maximum reaction temperature of 151.6
°C under complete containment. There exists the potential
of uncontrolled boiling and release of reactor contents from
the reactor if all sodium chlorite were to be added at once.
In the laboratory the addition was performed over 30 min,
and energy release was found to be approximately linear with
the addition mass with >95% of total energy released at the
end of addition. No accumulation of reaction enthalpy was
measured. In the plant because of potential limits in cooling
capacity, the rate of addition is adjusted to ensure that reactor
contents are maintained at 10 °C. Additionally, in ARC
testing reported above, the first major self-heat exotherm was
measured at an onset of 164 °C.
The addition of sodium sulfite is mildly exothermic, and
the reaction enthalpy was measured at -14.4 kJ/kg (of
reaction mixture at addition end). The predicted adiabatic
temperature rise is 4.4 °C. Approximately 75% of reaction
enthalpy is released over the addition time of 10 min. The
residual accumulated reaction enthalpy is predicted to cause
an adiabatic temperature rise of ∼1 °C, which was considered
insignificant.
mixture was agitated and brought to a temperature of 15 °C
( 5 °C. Water (125.1 mL) was added followed by concen-
trated sulfuric acid (33.92 g) while continuing to maintain
the temperature at 15 °C ( 5 °C (slightly exothermic).
Hydrogen peroxide (121.8 g of 35% solution) was then
charged, while maintaining 15 °C ( 5 °C (slightly exother-
mic). A solution of sodium chlorite (141.6 g) in water (368
mL) was then carefully charged while maintaining the batch
below 20 °C (highly exothermic). When less than 0.5%
starting material remained, the reaction was carefully quenched
with a solution of sodium sulfite (112.7 g) in water (448
mL) while maintaining the batch below 20 °C (highly
exothermic). After a 3 h hold, the batch was distilled under
vacuum at 40 °C and diluted with toluene (500 mL), and
then the phases were separated. The toluene solution of crude
product was extracted with 9% NaOH (780 mL), and the
organic phase was discarded. The aqueous extract was then
acidified with 37% HCl (186.7 mL) and diluted with water
(500 mL), and the product crystallized by cooling from 35
°C to 6 °C. After filtering, the batch was washed with water
and dried under vacuum at <40 °C for 24 h to provide 2 in
85% yield and 99.9 area% purity.
1H NMR (400 MHz, CDCl3): δ 3.87 (s, 3H), 6.56 (d,
1H), 7.33 (d, 1H), 7.61 (d, 1H), 7.97(d, 1H), 8.27 (d, 1H).
13C NMR (400 MHz, CDCl3): δ 33.14, 102.03, 106.24,
114.68, 120.58, 133.16, 134.54, 135.14, 142.78.
Based on this favorable thermal data, the DMSO process
was run on a 300 gal scale. Yields and purities were similar
to those in the lab.
Oxidation of 2-Chloro-6-methylbenzaldehyde (5a) with
DMSO Scavenger: To a 1 L reactor was added 2-chloro-
6-methylbenzaldehyde (40.0 g, 257 mmol), acetonitrile (100
mL), DMSO (24.28 g, 311 mmol), and water (33 mL). The
mixture was agitated and brought to a temperature of 10 °C
( 5 °C. Concentrated sulfuric acid (14 g) was added while
continuing to maintain the temperature at 10 °C ( 5 °C
(slightly exothermic). A solution of sodium chlorite (34.85
g) in water (170 mL) was then carefully added while
maintaining the batch below 10 °C (highly exothermic).
When less than 0.5% starting material remained, the reaction
was carefully quenched with a solution of sodium sulfite (6.7
g) in water (40 mL) while maintaining the batch below 20
°C (highly exothermic). After a 30 min hold, the batch was
distilled under vacuum at 40 °C until a slurry was observed.
The mixture was then diluted with toluene (300 mL), and
the phases were separated. The organic phase was washed
with water (2 × 60 mL), and then the volume was reduced
by removing ∼200 mL of toluene by vacuum distillation.
The resulting solution was assayed for both product and water
content; usually 0.2% g of 5a/mL and water <0.01%.
Alternatively, the carboxylic acid 2 can be isolated from
toluene-heptane crystallization to provide ∼90-95% yield
and assay >99 area%.
Conclusion
Hydrogen peroxide was successfully removed from the
R411 process by substituting DMSO as a HOCl scavenger
in the oxidation of 2-chloro-6-methylbenzaldehyde to its
carboxylic acid. In addition to the improved safety profile
for the reaction step, the new process had improved process
cycle times, reduced waste, and better yields with similar
purities to those of the hydrogen peroxide procedure, thus
maintaining the economic viability of the process. The
developed process was demonstrated in fixed equipment on
a 300 gal scale.
Experimental Section
Solvents and reagents were obtained from commercial
1
sources and used without further purification. H and 13C
NMR were performed on a Varian 400 MHz spectrometer.
HPLC analysis was performed on a Hewlett-Packard 1100
HPLC using a Zorbax SB-CN (4.6 mm × 250 mm, 5 µ)
column; mobile phase was a gradient of A (water/TFA
1000:1 (v/v)) and B (acetonitrile/TFA 1000:1 (v/v)) in ratios
(%A) of 95% (25 min), 20% (10 min), and 95% (5 min);
ambient temperature; flow rate ) 1.0 mL/min; UV detection
at 220 nm.
Oxidation of 2-Chloro-6-methylbenzaldehyde (5a) with
Hydrogen Peroxide Scavenger: To a 2 L reactor was added
2-chloro-6-methylbenzaldehyde (139.1 g, 0.89 mol), aceto-
nitrile (417 mL), and a solution of sodium dihydrogenphos-
phate monohydrate (36.8 g) in water (126.54 mL). The
2-Chlorobenzoic Acid (2b). Prepared as described above
and isolated from acetonitrile-water.
1H NMR (500 MHz, DMSO): δ 7.41 (m, 1H), 7.50-
7.55 (m, 2H), 7.78 (m, 1H), 13.36 (bs, 1H). 13C NMR (500
MHz, DMSO): δ 127.1, 130.5, 130.7, 131.4, 131.6, 132.5,
166.6.
2-Methylbenzoic Acid (2c). Prepared as described above
and isolated from methylene chloride.
Vol. 9, No. 6, 2005 / Organic Process Research & Development
•
1007

1H NMR (500 MHz, DMSO): δ 7.27 (t, 1H, ovlp), 7.28
(d, 1H, ovlp), 7.42 (td, 1H), 7.82 (d, 1H), 12.78 (bs, 1H).
13C NMR (500 MHz, DMSO): δ 21.1, 125.7, 130.1, 130.4,
131.4, 131.6, 138.9, 168.6.
4-Methoxybenzoic Acid (2d). Prepared as described
above and isolated from methylene chloride.
1H NMR (500 MHz, DMSO): δ 7.00 (d, 1H), 7.91 (d,
1H), 12.6 (bs, 1H). 13C NMR (500 MHz, DMSO): δ 55.3,
113.7, 122.9, 131.3, 162.8, 166.9.
1H NMR (500 MHz, DMSO): δ 7.60-7.55 (m, 2H), 7.64
(td, 1H), 8.00 (d, 1H), 8.14 (d, 1H), 8.17 (dd, 1H), 8.89 (d,
1H), 13.15 (bs, 1H). 13C NMR (500 MHz, DMSO): δ 124.8,
125.4, 126.1, 127.5, 127.6, 128.5, 129.8, 130.6, 132.9, 133.4,
168.9.
Acknowledgment
Our thanks to Michael Hall, Thomas Williamson, and
Eileen Zhao for NMR and HPLC data.
Note Added after ASAP Publication: There was an
error in Scheme 1 in the version published on the Web
October 14, 2005. The version published October 19, 2005
and the print version are correct.
4-Fluorobenzoic Acid (2e). Prepared as described above
and isolated from acetonitrile-water.
1H NMR (500 MHz, DMSO): δ 7.29 (t, 2H), 8.00 (dd,
2H), 13.02 (bs, 1H). 13C NMR (500 MHz, DMSO): δ 115.5,
127.3, 132.1, 164.8, 166.3.
Received for review May 27, 2005.
OP050083+
1-Naphthoic Acid (2f). Prepared as described above and
isolated from methylene chloride-toluene.
1008
•
Vol. 9, No. 6, 2005 / Organic Process Research & Development