Process development and pilot plant synthesis of methyl 2-bromo-6-chlorobenzoate

Article abstract of DOI:10.1021/op0501444

A scalable process for a pilot plant synthesis of methyl 2-bromo-6-chlorobenzoate (1) is described. The strategy employed for the synthesis hinged on the esterification of the sterically encumbered parent acid produced through an o-lithiation/ carboxylati

Full text of DOI:10.1021/op0501444

Organic Process Research & Development 2005, 9, 764−767
Process Development and Pilot Plant Synthesis of Methyl
-Bromo-6-chlorobenzoate
2
,
†
†
†
†
‡
‡
Matthew R. Hickey,* Shawn P. Allwein, Todd D. Nelson, Michael H. Kress, Osama S. Sudah, Aaron J. Moment,
‡
‡
‡
Stephen D. Rodgers, Mahmoud Kaba, and Paul Fernandez
Merck Research Laboratories, Department of Process Research, 466 DeVon Park DriVe,
Wayne, PennsylVania 19087, U.S.A., and Chemical Engineering Research and DeVelopment, 126 East Lincoln AVenue,
Rahway, New Jersey 07654, U.S.A.
Abstract:
Scheme 1. Reported syntheses of intermediate 2
A scalable process for a pilot plant synthesis of methyl 2-bromo-
6
-chlorobenzoate (1) is described. The strategy employed for
the synthesis hinged on the esterification of the sterically
encumbered parent acid produced through an o-lithiation/
carboxylation approach. Vigilant temperature control was
paramount for the success of this synthetic pathway. Initiation
of an exothermic benzyne decomposition pathway during the
o-lithiation step occurred at -70 °C. In the pilot plant, cooling
via the jacket service was suitable to provide batch temperatures
of -75 °C. However, control of reaction heat required liquid
nitrogen injection into the process vessel simultaneously with
the exothermic reagent charges. Esterification of the carboxylic
acid was accomplished with potassium carbonate/methyl iodide,
and subsequent crystallization resulted in 79% overall yield of
the title compound from inexpensive starting materials.
Scheme 2. Carboxymethylation of 3-bromochlorobenzene
using LDA and methyl chloroformate
Introduction
A recent program at Merck required kilogram quantities
of methyl 2-bromo-6-chlorobenzoate (1). To the best of our
knowledge, no reported synthesis of this deceptively simple
compound has been reported. Initial outsourcing efforts were
fruitless. Two vendors were approached for the large scale
delivery of 1; however, both independently defaulted on their
respective deliveries. Thus, rapid development of a robust
and scalable synthetic sequence was of utmost consequence.
We considered esterification of 2-bromo-6-chlorobenzoic
acid (2) the most direct approach for the synthesis of ester
Results and Discussion
Carboxymethylation of 3-Bromochlorobenzene. Treat-
ment of 4 with LDA at -78 °C, followed by quench of the
slurry with methyl chloroformate gave ester 1 in 79% yield.3
One of the major difficulties with this approach was
purification of the product mixture (Scheme 2). Formation
of ester 1 was accompanied by stoichiometric formation of
the N,N-diisopropylcarbamate (5). Attempts at direct crystal-
lization of ester 1 were unsuccessful. Only column chroma-
tography provided pure product, reinforcing the benefit of
having the acid function. The ability to purify acid 2 via an
acid base extraction sequence shifted focus to process
development of the o-lithiation/carboxylation chemistry
1
; our strategy hinged on the preparation and esterification
of acid 2.
The synthesis of acid 2 has been accomplished via two
1
basic strategies, (i) oxidation of 2-bromo-6-chlorotoluene (3)
using nitric acid and (ii) o-lithiation of 3-bromochlorobenzene
2
reported by Schlosser.
2
(4) with lithium 2,2,6,6-tetramethylpiperidide (LTMP),
Temperature Stability of 2-Bromo-6-chlorophenyl-
lithium (6). An initial perceived drawback to the o-lithiation
strategy is the use of cryogenics on a large scale. Our initial
approach was to determine the thermal stability of the anion
of 4 after LDA addition. In a typical experiment on a 5 g
scale, a 160 g/L solution of 4 in THF was treated with LDA
at -78 °C and aged for 2 h at that temperature. The dry ice
acetone bath was removed from the flask, and at an internal
temperature of -70 °C, the mixture began to darken. Within
followed by carboxylation with CO (Scheme 1).
We initially envisioned a third strategy, which was
founded on the latter o-lithiation chemistry. Instead of
carboxylation of the anion of 4 with CO to form the acid,
2
carboxymethylation through quench with methyl chlorofor-
mate or dimethyl carbonate was the focus (Scheme 2).
2
*
Corresponding author. E-mail: matthew_hickey@merck.com.
Department of Process Research.
Chemical Engineering Research and Development.
†
‡
3
0-45 s the batch temperature was 65 °C and refluxing.
(1) (a) Cohen, J. B.; Raper, H. S. J. Chem. Soc. 1904, 85, 1262. (b) Cohen, J.
B.; Smithells, C. J. J. Chem. Soc. 1914, 105, 1907.
(2) Mongin, F.; Schlosser, M. Tetrahedron Lett. 1997, 37, 1559.
(3) Based on analytically pure standard.
7
64
•
Vol. 9, No. 6, 2005 / Organic Process Research & Development
10.1021/op0501444 CCC: $30.25 © 2005 American Chemical Society
Published on Web 10/12/2005

Scheme 3. Benzyne type decomposition of
-bromo-6-chlorophenyllithium (6)
suggested a fast reaction rate. Unfortunately, precipitation
of various key intermediates clouded detailed kinetic mea-
surements.
2
Extractive workup of the reaction mixture removed
lithium salt 11 from the organic waste into 0.1 N NaOH,
with an aqueous back extraction to ensure complete removal
of the desired product from the organic waste phase (<3%
loss). The organic waste was discarded, and the aqueous pH
was adjusted to 1 with 6 N HCl. The resultant oil (bottom
phase) was collected, and the aqueous waste phase was salted
with solid NaCl and back extracted with THF, further
reducing the product loss below 3%. The combined organic
extracts contained product in typical assay yields of 88-
9
0% as a solution in THF.
Esterification. The initial discovery of the esterification
of acid 2 was not straightforward. Traditional Fisher type
esterification conditions using methanol failed to provide
suitable yield of the ester 1. The large size of the bromine
and chloride substituents flanking the acid prohibits passage
Assay of the product mixture revealed complete consumption
of 4 and generation of four major products. Structural
assignments by GC/MS and NMR (Scheme 3) corresponded
to products generated via benzyne type decomposition
analogous to observations from decomposition of 2,6-
dichloro- or 2,6-dibromophenyllithium.4
This insight into the mechanism of the decomposition
allowed for a rational solution. By diluting the reaction
concentration 4-fold to 40 g/L (based on 4), the increased
heat capacity of the system better absorbed heat generation
during benzyne formation. Additionally, the more dilute
conditions slowed the trapping of the benzyne intermediate
by possible nucleophiles (LDA, DIPA, or 6; Scheme 3). The
decomposition still occurred under the revised conditions;
however, the temperature did not rise above 0 °C (ice bath).
This eased our safety concerns for this exothermic pathway,
but temperature was still a parameter that was controlled and
monitored with vigilance.
In Situ Reaction Monitoring. A carboxylation reaction
using standard conditions was monitored by an in situ IR
probe. The results are plotted as Figure 1.
Bromochlorobenzene 4 was added to a solution of LDA
at -78 °C over 5 min. The formation of aryllithium 6 was
indicated by an increase in red signal, which coincided well
with the consumption of 4 (black signal). Decrease in the
red signal corresponded nicely with the visually observed
precipitation (equilibirum concentration ca. 0.11 M). The
internal temperature of the stirred slurry was maintained
6
through a tetrahedral intermediate. Attempts at activation
with DCC and 1-ethyl-3-[3-(dimethylamino)propyl] carbo-
diimide (EDC) also provided low levels of conversion.
Alternative methylation conditions using methyl iodide and
7
DBU gave 63% assay yield of ester 1. This reaction was
scalable and was carried out in similar yield to prepare 10
kg of ester 1 as a stream in toluene.
To minimize raw material cost for this process, a screen
of suitable bases was performed to replace the expensive
DBU in this transformation. This approach identified aqueous
potassium carbonate as a viable solution. Heating a mixture
of acid 2 with 1.3 equiv of methyl iodide and 1.5 equiv of
potassium carbonate (as a 50 wt % aqueous solution) to 65
°C under reflux routinely gave the ester in >95% assay yield.
After an extractive workup with MTBE and THF, the
solvents in the organic stream were switched to 2-propanol.
This solution was concentrated to a final ester concentration
of 350 g/L. Addition of two volumes of water (relative to
total batch volume) gave crystalline ester 1 in 93% isolated
yield (Scheme 4).
Pilot Plant Synthesis of 1. The foremost concern in
planning the pilot plant campaign was temperature control
of the o-lithiation/carboxylation procedure. The jacket service
was sufficient to reduce batch temperature to -75 °C.
However, jacket service alone could not control the heat of
reaction during the exothermic bromochlorobenzene 4 (37
between -70 and -76 °C for 5 h. Gaseous CO
2
(purple
2
kcal/mol) and CO (32 kcal/mol) charge. Temperatures below
signal) was added subsurface. Essentially no buildup of CO
2
-70 °C were ensured by equipping the vessel with a direct
in the solution was evident, until most of the aryllithium was
consumed, suggesting a fast reaction as anticipated. Pause
liquid nitrogen injection assembly. A flow meter would
monitor both flow rate and amount of liquid nitrogen used.
Care had to be exercised, however, since cooling the batch
below -100 °C exceeded the rating of the vessel. Thus, a
temperature window of -85 to -70 °C was established to
protect both the batch and the vessel.
in CO
both the red and blue lines. A sharp drop in the blue line
during the CO addition coincided with a visually observed
2
addition was visible in the plot as a slight plateau in
2
precipitation of the lithium carboxylate 11. A large exotherm
5
resulting from the carbon dioxide addition (32-34 kcal/mol)
(
6) (a) Thomas, J.; Canty, J. J. Pharm. Pharmacol. 1962, 14, 587. (b) A survey
of the literature suggests dominance of esterification of 2,6-dihalobenozic
acids by alkylative procedures. An instance of Fisher esterification of
difluoro-iodo benzoic acids has been reported in poor yields: Rausis, T.;
Schlosser, M. Eur. J. Org. Chem. 2002, 3351.
(
4) (a) 2,6-Dichlorophenyllithium: Kress, T. H.; Leanna, M. R. Synthesis 1988,
03. (b) 2,6-Dibromophenyllithium: Luli n˜ ski, S.; Serwatowski, J. J. Org.
8
Chem. 2003, 68, 5384. (c) Halophenyl triflates: Wickham, P. P.; Reuter,
K. H.; Senanayake, H. G.; Zalesky, M.; Scott, W. J. Tetrahedron Lett. 1993,
3
4, 7521.
(7) Ono, N.; Yamada, T.; Saito, T.; Tanaka, K.; Kaji, A. Bull. Chem. Soc. Jpn.
1978, 51, 2401.
(
5) Determined by reaction calorimetry.
Vol. 9, No. 6, 2005 / Organic Process Research & Development
•
765

Figure 1. FTIR data for a typical carboxylation experiment.
Scheme 4. Through process to synthesize 1
assay. Assay confirmed >95% assay yield. The aqueous
waste was separated, and the organics were diluted with
MTBE and washed with water to remove any residual
inorganic salts. Both the aqueous phase and the water wash
contained <1% ester 1. The organics from the first batch
were stored and then combined with the second batch10
during solvent switch to 2-propanol. It was imperative that
all THF was removed during the solvent switch, as the
solubility of ester 1 in THF was very high (.100 g/L) and
greatly affected solubility during the crystallization.
As above, LDA (0.2 M) was generated in THF at -50
°
C and cooled to -76 °C. 3-Bromochlorobenzene (4) was
added over 30 min, with concomitant liquid nitrogen charge
2 kg/min) to minimize potential heating of the batch. The
thin slurry was aged at -75 °C for 4 h. Control of the
(
Once the solvent switch was complete, two volumes of
water were added giving crystalline ester 1 in 93% isolated
yield. The single batch crystallization produced 120 kg of
7
subsurface addition of gaseous CO
a needle valve from manifolded cylinders on a scale. After
.2 equiv of CO were added, the exotherm subsided and
2
was ensured by using
1
1
ester 1 (>99 LCAP, >99 LCWP ). This represents a 79%
overall yield from 3-bromochlorobenzene and is directly in
accordance with the results obtained on a 2 kg scale.
In summary, we have developed a robust two-step process
to synthesize methyl 2-bromo-6-chlorobenzoate (1). Our
strategy relied on o-lithiation/carboxylation chemistry to
produce acid 2. Exploiting the acid function of this inter-
mediate allowed for consistent interbatch purity. Success of
this transformation was ensured through direct liquid nitrogen
injected cryogenic reaction conditions in the pilot plant. This
route has provided crystalline material consistent in yield
and purity on multiple 50 kg batches in 79-80% overall
isolated yield.
1
2
the batch was prepared for workup. The batch temperature
was slowly raised to 20 °C over 5.5 h to control offgassing
of dissolved CO
2
, which was then degassed for 15 min at
CO upon
600 mmHg. Degassing minimized formation of Na
2
3
transfer of the batch to aqueous NaOH and, hence, minimized
effervescence upon pH adjustment. After transfer of the batch
to the workup vessel, a similar workup protocol (vide supra)
was utilized to afford a final THF stream of the acid in 89%
assay yield.9
With the acid stream obtained above, 50 wt % aqueous
2 3
K CO and methyl iodide were charged to a vessel. The
vessel was sealed, and the batch was heated to 80 °C for 7
h and then cooled to room temperature for an end of reaction
Experimental Section
General. All reagents and solvents were obtained from
commercial sources and used without purification.
(
8) To control the exotherm on a 2 kg scale, the CO
headspace of the reaction at -70 °C. The addition rate was maintained
such that CO bubbling through oil was kept constant. The exotherm under
these conditions is much less pronounced in comparison to subsurface
addition. The yields are identical regardless of the mode of exposure of the
2
was added into the
2
batch to CO
2
.
(10) Batch 2 gave the acid in 90% assay yield followed by a 95% assay yield of
the esterification.
(11) LCAP represents LC area percent; LCWP represents LC weight percent.
(
9) The batch was drummed at this point to obtain an accurate assay yield, for
basis of the charges for the subsequent esterification.
766
•
Vol. 9, No. 6, 2005 / Organic Process Research & Development

HPLC Conditions: Zorbax Eclipse XBD C8 column, 5
micron, 4.6 mm × 150 mm; mobile phase gradient 0.1%
to a third glass lined 500 gal vessel. An additional 280 L of
0.1 N NaOH(aq) was charged into the organic waste phase
still in the quench vessel, and the mixture was agitated for
30 min and then settled for 1 h. The phases were separated,
and the second aqueous product stream was combined with
the first. To the aqueous product stream, 6 N HCl (143 L)
was added over 1 h to minimize offgassing and foaming.
An oily second phase formed, and the biphasic mixture at
pH 0.5 was allowed to settle for 1 h. The oily product phase
was drummed off (241.5 kg total), and the aqueous phase in
the vessel was treated with solid sodium chloride (48 kg)
and THF (286 L) for the back extraction. After agitation and
settling, the phases were separated and the aqueous phase
was sent to waste. The second organic phase (246.8 kg) was
drummed separately to get an accurate weight for assay
calculations. The combined assay yield was 63.9 kg of 1,
which represents an 88% yield.
The product streams were combined in a clean 500 gal
glass lined vessel, and 50% w/w solution of potassium
carbonate (90 kg) added, followed by addition of methyl
iodide (50 kg, 353 mol). After sealing the vessel, the biphasic
reaction mixture was heated to 80 °C for 7 h. Upon cooling
the reaction mixture, MTBE (86 kg) and water (115 kg) were
added and the phases separated. A 115 kg water wash of
the organic product phase was conducted to remove any
remaining inorganic salts, and the organic phase was
drummed.
3 4
aqueous H PO /acetonitrile 85/15 for 5 min, then to 5/95
over 15 min, then to 85/15 over 2 min; 1.5 mL/min; 40 °C;
detector 220 nm.
Retention times:
2
-Bromo-6-chlorobenzoic acid 9.20 min
Methyl 2-bromo-6-chlorobenzoate 14.28 min
-Bromochlorobenzene 15.74 min
3
Synthesis of 1 Using Methyl Chloroformate. A 50 mL
round-bottomed flask containing 5 mL of THF was cooled
to -70 °C in an ice-water bath. To the solvent n-
butyllithium (2.5 M in hexanes, 2.3 mL, 5.75 mmol) and
diisopropylamine (0.81 mL, 5.78 mmol) were added. The
mixture was cooled to -78 °C, and 3-bromochlorobenzene
(0.52 mL, 4.43 mmol) was added. After stirring the white
suspension for 3 h, a solution of methyl chloroformate (1.7
mL, 22.0 mmol) in THF (5 mL) was added over 15 min
such that the internal temperature did not exceed -70 °C.
The resulting yellow suspension was stirred at -78 °C for
1
h, and then the cooling bath was removed. Once at room
temperature, the reaction was treated with 10 mL of water
and the phases were separated. The aqueous was extracted
once with 10 mL of isopropyl acetate. Assay of the combined
organic extracts revealed a 79% assay yield of the product,
with little starting material remaining. Crude NMR of the
concentrated organic stream confirmed the major products
of the reaction to be 1 and 5 in 1:1 ratio. Other impurities
were present in lesser amounts and were uncharacterized.
Pilot Plant Process for the Synthesis of Methyl 2-Bromo-
A second batch of carboxylation and esterification was
conducted in the same manner on a similar scale in 90%
and 95% assay yields, respectively. The first batch product
stream was transferred to a 500 gal glass lined vessel, and
distillation started. During the distillation of the first batch
product stream, the second batch was fed in. Once both
batches were combined in a single vessel, the solvent switch
to 2-propanol was performed. When the target concentration
of ester 1 in 2-propanol (246 g/L) was achieved, and the
6-chlorobenzoate (1). Tetrahydrofuran (1000 L) was charged
to a 500 gal glass lined (cryo-rated) vessel, and cooling was
started. Once the solvent was at an internal temperature of
-
40 °C, 1.6 M n-butyllithium (169 kg, 398 mol) in hexanes
was charged over 25 min maintaining the batch temperature
below -34 °C. The batch was cooled further during the 15
min charge of diisopropylamine (43 kg, 429 mol) to a final
temperature of -77 °C, using liquid nitrogen injections into
the batch to supplement jacket cooling. 3-Bromochloroben-
zene (58.6 kg, 306 mol) was charged over 30 min, with a
constant liquid nitrogen flow of 2 kg/min into the batch. This
simultaneous liquid nitrogen charge ensured the batch to
remain below -75 °C. During the 4 h age, a quench vessel
was prepared by charging 560 L of 0.1 M NaOH(aq) to a
1
2
quantity of residual THF (<1% v/v ), 2 volumes (based on
the total batch volume) of water were added, resulting in
formation of crystalline ester 1. The slurry was filtered in
two drops to give a total physical yield of 120 kg (481 mol)
1
of 1 as large crystals: H NMR δ 7.49 (d, J ) 8.0 Hz, 1H),
7.37 (d, J ) 8.0 Hz, 1H), 7.21 (t, J ) 8.0 Hz, 1H), 3.98 (s,
1
3
3H); C NMR δ 166.0, 135.7, 131.8, 131.3, 131.0, 128.4,
119.9, 23.1.
1000 gal glass lined vessel. After a 4 h age at -75 °C, carbon
dioxide (35 kg, 795 mol) was charged subsurface, again with
a concomitant liquid nitrogen flow of 6 kg/min, maintaining
a batch temperature below -70 °C. To maintain control of
offgassing of dissolved carbon dioxide, the batch was slowly
warmed to 20 °C over 5.5 h and then degassed at 600 mmHg
for 10 min. The batch was transferred to the quench vessel,
agitated for 40 min, and then settled for 1 h. Phase separation
completed in 1 h, and the aqueous product was transferred
Acknowledgment
We thank Li Pan, Claire Lee, and Lili Zhou for analytical
support during the pilot plant campaign, Doug Frantz and
Damian Weaver for the Fisher esterification investigation,
and Peter Dormer for NMR structure analysis of the
decomposition products.
Received for review August 10, 2005.
OP0501444
(12) Determined by quantitative GC analysis.
Vol. 9, No. 6, 2005 / Organic Process Research & Development
•
767