Route development and bulk synthesis of CP-865,569

Article abstract of DOI:10.1021/op7000386

The synthesis of zwitterionic CP-865,569 by three different synthetic routes is described. The first two routes differ in the method of introducing the sulfonic acid at the penultimate step: by sulfite displacement of a benzylic chloride and by oxidation of a benzylic thioacetate. The third route is a convergent route to the drug candidate. The synthesis strategy was primarily driven by the need to introduce the sulfonic acid functionality at the final stage of the synthesis due to the high water solubility and low organic solubility of the desired product.

Full text of DOI:10.1021/op7000386

Organic Process Research & Development 2007, 11, 754−761
Route Development and Bulk Synthesis of CP-865,569
Katherine Belecki, Martin Berliner, Richard Todd Bibart, Cliff Meltz, Karl Ng, James Phillips,
David H. Brown Ripin,* and Michael Vetelino
Chemical Research and DeVelopment, Pfizer Global Research DiVision, Pfizer Inc., Eastern Point Road, Groton,
Connecticut 06340, U.S.A.
Abstract:
only crystalline intermediate in this sequence other than the
drug substance is salt 2, and chromatographic purification
of intermediates was required after two steps (5 and 8-Na).
First Bulk Campaign. For preparation of the first 500 g
of 8-Arg, an enabled discovery route was employed with
changes in the chemistry at the beginning and end of the
route. The reaction of trans-2,5-dimethylpiperzine with
p-fluorobenzyl chloride4 was uneventful and produced the
desired monoalkylated product (2) in 89% in situ yield (as
measured by HPLC) in addition to 10% of the dibenzylated
product and unreacted starting material. The mixture could
be separated cleanly by extraction. Unreacted piperazine was
primarily purged in the basic aqueous layer discarded after
the reaction. The desired product (2) could be separated from
bis-benzylated product and excess alkylating agent by
extracting 2 into water in the range of pH 4-5. Following
basification to above pH 12, the desired product could be
recovered cleanly, contaminated only with traces of unreacted
starting piperazine. Resolution with tartaric acid using 2 equiv
of acid (the salt is a 2:1 acid/product ratio) provides the
resolved amine in 41% yield and >99:1 er, provided the
filter-cake is washed thoroughly prior to drying and isolation.
On large scale, filtration on an unstirred Nutsche filter
provided variable results, whereas filtration on a filter/dryer
with a stirred wash provided consistent results between runs.
Following a salt break, resolved amine 2 was amidated with
chloroacetic anhydride, and the resulting chloride was
displaced with 5-chlorosalicylaldehyde. A solvent change
from DCM to toluene allowed us to avoid formation of an
undesired bis(salicylaldehyde) methylene acetal resulting
from salicylaldehyde anion and residual dichloromethane in
the second step, and obviated the need to isolate and handle
chloride 3, which had proven to be only moderately stable
on storage. After aqueous workup, the toluene solution of
chloride 3 was azeotropically dried and taken directly into
the alkylation, which proceeded uneventfully when 10% v/v
DMAc was added to the reaction. After a basic aqueous
workup to remove unreacted chlorosalicylaldehyde, aldehyde
5 was isolated as a low-melting amorphous solid by
precipitation from a DCM/IPE solution in 94% yield over
the two steps. Sodium borohydride reduction of the aldehyde
proceeded more efficiently in an EtOH/THF mixture than
in THF alone, allowing for a reduction of the NaBH4 charge
and a simplified workup. Conversion of the alcohol to
chloride 7 with thionyl chloride was followed by sulfonic
acid introduction, which was conducted with Na2SO3 in
The synthesis of zwitterionic CP-865,569 by three different
synthetic routes is described. The first two routes differ in the
method of introducing the sulfonic acid at the penultimate
step: by sulfite displacement of a benzylic chloride and by
oxidation of a benzylic thioacetate. The third route is a
convergent route to the drug candidate. The synthesis strategy
was primarily driven by the need to introduce the sulfonic acid
functionality at the final stage of the synthesis due to the high
water solubility and low organic solubility of the desired
product.
Introduction
CP-865,569 (8) is a CCR1 antagonist.1 The candidate
contains a tertiary amine and unusual benzylic sulfonic acid
functionalities and exists as a zwitterion as its neutral species.
The synthesis of this candidate involved a number of
synthetic and processing challenges due to the interesting
functionality present and the physical properties thereof.
Herein is described four iterations of bulk synthesis of the
candidate, utilizing successively improved and optimized
modifications of the discovery synthesis, the development
and scaling of a new synthetic process for the introduction
of the sulfonic acid functionality, and the development of a
convergent synthesis. Process safety studies and impurity
identification efforts are described in the following paper.2
Discovery Synthesis
The original route employed by discovery to identify the
candidate1 started with amidation of 2 with chloroacetyl
chloride followed by alkylation with 5-chlorosalicylaldehyde
to provide 5 as a low-melting solid (Scheme 1). Aldehyde 5
was reduced with NaBH4 in THF to provide alcohol 6 which
was chlorinated with excess SOCl2 and sulfonylated with
sodium sulfite in aqueous EtOH to give the sodium salt of
CP-865569 (8-Na) along with ∼10% ethyl ether and ∼10%
alcohol 6 from competitive solvolysis of the chloride during
the reaction. Purification of 8-Na away from the nonionic
impurities was accomplished by ion-exchange chromatog-
raphy,3 resulting in isolation of the zwitterion (8), which was
converted to the salt (8-Arg) by treatment with arginine. The
* To whom correspondence should be addressed. E-mail: David.B.Ripin@
pfizer.com.
(1) Hayward, M. M. U.S. Patent 2003/0083335 A1, 2003.
(2) Ripin, D. H. B.; Weisenburger, G. A.; am Ende, D. J.; Bill, D. R.; Clifford,
P. J.; Meltz, C. N.; Phillips, J. E. Org. Process Res. DeV. 2007, 11, 762-
765.
(3) Waters MCX resin, 10g/g loading, MeOH elution.
(4) Maw, G. N. WO 98/52929, 1998.
754
•
Vol. 11, No. 4, 2007 / Organic Process Research & Development
10.1021/op7000386 CCC: $37.00 © 2007 American Chemical Society
Published on Web 06/15/2007

Scheme 1. Discovery synthesis of CP-865,569 used in the preparation of the first bulka,b
a Reagents and conditions originally used: c) ClCH2COCl, Et3N, DCM; d) 5-chlorosalicylaldehyde, K2CO3, DMF, reflux; e) 2 equiv of NaBH4, THF; g) 2 equiv
of Socl2, Dcm; h) Na2SO3, EtOH/H2O, reflux 16 h; i) ion-exchange chromatography; j) arginine, MeOH; k) n-PrOH. bOptimized reagents and conditions: a) p-FPhCH2Cl,
NaOH, 5 mol % Bu4NCl, cyclohexane/H2O, 40 °C, 18 h; b) L-tartaric acid, MeOH; c) chloroacetyl chloride, Et3N, PhMe; d) 5-chlorosalicylaldehyde, K2CO3, PhMe,
DMAC; e) NaBH4, PhMe, EtOH; f) HCl; g) SOCl2, CH2Cl2; h) Na2SO3, MeCN; i) HCl, MeCN, iPrOH; j) L-arginine, EtOH; k) EtOH.
acetonitrile/water to eliminate formation of ethereal byprod-
ucts that consisted of 10-20% of the crude reaction mixture
under the original conditions. The resulting 9:1 mixture of
8-Na and 6 was separated by IPE extraction of the aceto-
nitrile solution of the mixture, and the purified sodium salt
was acidified with HCl in acetonitrile. The choice of ACN
as solvent for this neutralization was driven by the low
solubility of NaCl in acetonitrile across a wide range of
temperatures, but it turned out to be fortuitous as the presence
of acetonitrile was required to prevent gumming of the
sparingly soluble zwitterion during the neutralization reac-
tion. Zwitterion prepared in this manner contained ap-
proximately 2000 ppm sodium and was directly taken into
the salt formation. After a repulp (hot slurry) to reach full
crystallinity, low-sodium arginine salt 8-Arg was isolated
as a white, crystalline solid. Attempts to isolate crystalline
forms of 8-Na or the zwitterion (8) were unsuccessful; high-
melting amorphous solids were obtained under all conditions.
Interestingly, during preparation of the racemate for purposes
of establishing a chiral assay, it was discovered that racemic
8-Na is a highly crystalline solid.
Second Bulk Campaign. The next campaign resulted in
the preparation of 8 kg of 8-Arg and required modification
of the neutralization/salt formation step for 8 to accommodate
for limitations imposed by conducting the campaign in our
kilo-lab facility. Conversion of 1 to 8-Na proceeded as
described above with little change. After synthesizing 8-Na
and extractive removal of 6 with IPE washes, the acetonitrile
layer containing 8 was diluted with isopropanol and dried
by azeotropic distillation. Hydrogen chloride gas was intro-
duced to the ACN/IPA solution of 8-Na to pH 3, and the
resulting solution was added slowly to IPE to precipitate the
zwitterion (8), which was isolated by filtration and dried
under nitrogen sweep. Gumming of the zwitterion had been
observed during development of this process and was a
persistent concern during these operations. It was found
through empirical observations that the presence of aceto-
nitrile, the absence of water, and the addition rate were
critical to prevent gumming, but the failure limits of this
process are still not well understood. Solid zwitterion was
then taken back into a mixture of ACN and ethanol, filtered
through a 0.5 µm, and then a filtered water solution of
arginine was added. Removal of water and acetonitrile was
accomplished by repeated distillation and displacement with
ethanol. The resulting partially crystalline 8-Arg was then
rendered completely crystalline by a reslurry in hot ethanol.
In this way high-quality drug substance was delivered at the
expense of a very risky and laborious endgame process that
was impractical to repeat for any further large-scale cam-
paigns.
Development of a New Method for the Introduction
of the Sulfonic Acid. After encountering the difficulty in
removing the sodium ion from the API, a number of other
synthetic methods to install the sulfonic acid moiety were
investigated. Due to the high water solubility of the zwitter-
ionic intermediate and working under the assumption that
intermediates in the synthesis also containing the sulfonic
acid functionality would possess similar properties, the focus
of research was to introduce the acid as the final step of the
synthesis or to introduce it in a protected state. Ideas
investigated included other sulfite displacements of the
chloride in the absence of metal counterions, utilizing the
sodium sulfite displacement followed by reaction of the
sulfonic acid to form an organic soluble derivative which
could be hydrolyzed following ash removal, or introducing
the sulfonic acid via an entirely different method which
involves no metal ions.
Attempted displacement of chloride 7 with ammonium
sulfite resulted in a new material tentatively assigned as the
benzylic amine with no desired product observed. The
chloride could be very cleanly displaced by tritylmercaptan
or potassium thioacetate (Scheme 2). Either product could
Vol. 11, No. 4, 2007 / Organic Process Research & Development
•
755

Scheme 2. Installation of the sulfonic acid functionality of CP-865,569 via oxidation of thioacetate 9a
a Reagents and conditions: a) KSAc, PhMe; b) HCO2H, 30% H2O2 in H2O or 30% AcOOH in AcOH.
then be oxidized under a variety of conditions,5 including
some metal-ion-free conditions such as performic or peracetic
acid in formic or acetic acid. Under the acidic conditions,
no N-oxide formation was observed. The crystallinity of the
thioacetate derivative in conjunction with the higher atom
efficiency and lower cost of the acetate as compared with
that of the trityl protecting group led us to focus our attention
on the thioacetate route.
attempts to isolate the product by crystallization did not meet
with success. As the material produced by the two-step
process described above is high in purity (>94% by HPLC),
5 was carried directly into the next step.
The sodium borohydride reduction was carried out in
MTHF with 5.0 equiv of methanol added. When run entirely
in methanol, a borane-amine complex was isolated along
with the desired product in ∼10% yield. Further, the reaction
in methanol, ethanol, and/or THF required removal of the
reaction solvent prior to aqueous workup. In the MTHF case
the reaction was quenched with aqueous sodium citrate
followed by layer separation. Following workup, the alcohol
can be precipitated as the amorphous HCl salt, following
addition of anhydrous HCl. This procedure has the disad-
vantage that the benzylic alcohol, which is not very acid
stable, decomposes during the precipitation, and material of
only 86% purity is isolated from material that starts out at
>96% purity prior to introduction of the HCl. Once in the
solid state, the material does not appear to decompose, even
over a period of 6 months or more. Due to the decomposition
during salt formation and the high purity of the unisolated
material, the decision was made to telescope alcohol 6 into
the next reaction in lieu of isolation. This decision was
enabled by the discovery of the solid HCl salt of chloride 7
and crystalline thioacetate 9 (vide infra).
Chloride formation using thionyl chloride proceeds with-
out event in a number of solvents, and ethyl acetate was
chosen because the purity of the material out of the reaction
appeared highest. Following reaction completion, the un-
quenched ethyl acetate solution can be added to hexanes,
and the HCl salt of chloride 7 precipitates from the reaction
mixture. Little or no upgrade of purity is observed during
this isolation. As a consequence, the chloride was instead
telescoped one additional step prior to its isolation. In order
to telescope the material into the thioacetate displacement,
the reaction was quenched with base, and the free-base of
chloride 7 was carried forward in solution.
Third Bulk Campaign: Demonstration of the Thio-
acetate Oxidation To Introduce the Sulfonic Acid. In this
campaign amine 2 was prepared with less than a full
equivalent of p-fluorobenzyl chloride to reduce the quantity
of bis-alkylated byproduct formed in the reaction,6 and the
amount of tetrabutylammonium chloride was minimized from
5 mol % to 3.5 mol %. A biphasic Schotten-Baumann amide
formation using chloroacetyl chloride and aqueous sodium
hydroxide as base in place of triethylamine in 2-methyltet-
rahydrofuran (MTHF)7 proved superior to the original
conditions. The presence of chloroacetate anion (in the
chloroacetyl chloride), particularly under the single-phase
reaction conditions, resulted in formation of an alcoholic
impurity resulting when an acetate displaces the R-chloride
of the desired product and is hydrolyzed on workup. While
perhaps counterintuitive to add water to a reaction plagued
by hydrolysis, it seems that any acetate present in the reaction
mixture is sequestered in the water layer and is unavailable
to react with the product. The hydrolysis impurity is difficult
to purge, and its control using the modified reaction
conditions obviates significant processing needed to achieve
that purification. The R-chloroamide intermediate is not
indefinitely stable, and the elevated temperatures necessary
for azeotropic drying on scale result in decomposition. As a
result, a magnesium sulfate drying step was necessary.
The displacement of the chloride with 5-chlorosalicyl-
aldehyde proceeds cleanly in MTHF in the presence of
potassium carbonate and potassium iodide. The R-chloroa-
mide intermediate is not a solid, and as the reaction solvents
are the same, the dried reaction mixture from the previous
step is used directly in this reaction. Powdered potassium
carbonate is necessary to achieve reaction times of less than
12 h. Following reaction completion, the excess dark-yellow
phenolate can be washed down to <1% with two sodium
hydroxide washes.8 Following layer separation, the solution
can be azeotropically dried prior to the sodium borohydride
reduction. Aldehyde 5 is a low-melting solid, however
Chloride 7 is cleanly displaced in the presence of
potassium thioacetate in toluene. Following reaction comple-
tion and water washes, the toluene is distilled to a low
volume (∼2-4 L/kg), and hexanes are added to precipitate
the product. The product is isolated as a crystalline solid
with a good impurity purge. Typical purity of the material
isolated is >96%. The five step sequence (amide formation,
phenol displacement, borohydride reduction, chloride forma-
tion, and thioacetate displacement) proceeds in an average
of 66% overall yield on 50-kg scale, and all five of the
reactions proceed rapidly enough that the five steps can be
executed in one week in a pilot plant.
(5) Oxidation of thioacetates to sulfonic acids: Koenig, N. H.; Swern, D. J.
Am. Chem. Soc. 1957, 79, 4235-4237.
(6) It was later determined that unreacted starting material can carry through
the synthesis, reacting at both amine termini and forming a pseudodimeric
impurity that was observed at low levels in the finished product.
(7) Ripin, D. H. B.; Vetelino, M. Synlett 2003, 2353.
(8) The remainder of this material is reduced in the next step and washed away
in the aqueous workup of that step.
756
•
Vol. 11, No. 4, 2007 / Organic Process Research & Development

Scheme 3. Potential convergent route to CP-865,569
The success of the new route hinged upon the ability to
cleanly oxidize9 thioacetate 9 and quench the reaction in the
absence of any metal-ion-containing salts. This could be
accomplished using a performic acid or peracetic acid
oxidation followed by quenching of the excess peroxide with
activated carbon (Darco G-60). This quench, which generates
oxygen, is proposed to be catalyzed by the residual metals
in the activated carbon. No lot-to-lot variability was observed
in the effectiveness of the quench, either between different
lots of the same grade of carbon, or other types of carbon
(KBB was also tried). Each lot was use-tested prior to use
on scale. A significant amount of safety testing and engineer-
ing was undertaken in order to run this chemistry on a large
scale. The oxidation reaction was run at 15 °C with slow
addition of the oxidant to the substrate in order to control
heat generation and prevent a runaway reaction. To generate
the reagent in situ, 30% hydrogen peroxide was added slowly
to a solution of 9 in formic acid. The oxidant was subdivided
into nine separate charges to prevent accidental addition of
too much oxidant at once. The reaction was monitored by
in situ IR to determine reaction completion and halt oxidant
charging in order to minimize the amount of peroxide
requiring quenching. The completed reaction mixture was
added slowly to slurried activated carbon (5 wt % as
compared to the weight of 9 added) in formic or acetic acid
with generation of oxygen. A nitrogen purge was maintained
throughout the quench in order to keep the oxygen concen-
tration inside the reactor to a minimum. Increasing the
amount of carbon used in the quench led to more vigorous
gas generation and a significant decrease in isolated yield
of product. Quench completion was monitored with semi-
quantitative peroxide test strips, which indicated a final
peroxide concentration of less than 1 ppm. Following quench
and ARC10 testing, azeotropic drying with methanol and
precipitation with isopropanol or IPE provided solid zwit-
terion isolated by filtration.
temperatures to provide the ester and piperazine 2 in high
amounts (over 10% after 3 days at 95 °C). Initial observations
indicated that the cocrystal would not precipitate in wet
solvent, and when a solution of the mixture was azeotroped
dry, significant quantities of oil deposited on the sides of
the flask in addition to cocrystal precipitation. These issues
were resolved by dissolving the API and arginine in wet
1-propanol and treating the mixture with 7.5% w/w G-60
Darco followed by room-temperature filtration. The carbon
treatment effectively removed the oils that later precipitate
during the azeotropic drying and also served to remove a
significant amount of color from the solution. The wet
1-propanol solution could then be azeotropically displaced
with dry 1-propanol under vacuum with the pot temperature
remaining well below 50 °C, thus suppressing the solvolysis
reaction previously observed. The cocrystal thus obtained
could be recrystallized in methanol if necessary to upgrade
purity. The product precipitated from methanol as an
amorphous solid and had to be repulped in refluxing 2B
ethanol (toluene denatured ethanol, 99.5% EtOH, 0.5%
toluene) to re-establish crystallinity. For the oxidation,
cocrystal formation, and repulp, a 67-69% isolated yield
of clinically suitable material was obtained.
Convergent Route Development. An obvious retrosyn-
thetic disconnection to make to achieve a more convergent
synthesis would be the amide bond (Scheme 3). If amide
formation could be achieved on a fully elaborated carboxylic
acid precursor, the nine-step linear synthesis above could
be shortened to a synthesis with a longest linear route of
four to five steps. We made the decision not to pursue the
most convergent route possible, wherein the sulfonic acid
functionality would already be installed on the carboxylic
acid precursor prior to amide formation. This decision was
based on a number of reasons: the sulfonic acid would likely
participate in amide formation, there was difficulty in
handling the sulfonic acid functionality for more than one
isolated step, the final step of the synthesis would remain
the same as that for previous campaigns, and the number of
linear steps would remain the same. Additionally, the route
depicted in Scheme 3 has a reasonable regulatory synthesis
and starting materials associated with it: amide formation,
sulfonic acid formation, and salt formation.
We chose to investigate alternative conditions for the
formation of the arginine cocrystal (X-ray analysis indicates
a hydrogen-bonded complex) that did not involve the use of
acetonitrile or other solvents that are considered class 1 or 2
solvents by the FDA. Arginine is only soluble in water and
acetic acid, and therefore unless water was employed in the
cocrystallization, a solid-to-solid conversion would be neces-
sary. To complicate matters, alcoholic solvents proved to
react with the amide in the API at prolonged times at elevated
Alkylation of p-chlorophenol with methyl chloroacetate
followed by in situ hydrolysis resulted in a high yield of
acid 12 (Scheme 4). Potassium phosphate was preferred over
potassium carbonate due to a large increase in reaction
volume and gas output during acidification with potassium
carbonate. The actual isolated yield was not determined as
the solids contained a high amount of inorganic salts and
(9) For a review of large-scale oxidation reactions, see: Caron, C.; Dugger, R.
W.; Ruggeri, S. G.; Ragan, J. A.; Ripin, D. H. B. R. Chem. ReV. 2006, 106,
2943-2989. Dugger, R. W.; Ragan, J. A.; Ripin, D. H. B. Org. Process
Res. DeV. 2005, 9, 253-258.
(10) Accelerated Rate Calorimetry.
Vol. 11, No. 4, 2007 / Organic Process Research & Development
•
757

Scheme 4. Synthesis of functionalized acetic acid coupling partnersa
a Reagents and conditions: a) i) K3PO4, ClCH2CO2Me, acetone, 50 °C; ii) H2O, distill acetone to 85 °C; iii) concn HCl to pH 1. b) (HCOH)n, AcOH, HCl, H3PO4,
90 °C, 60%, 2 steps. c) KSAc, DMSO, 50 °C.
Scheme 5. Convergent synthesis of CP-865,569a
a Reagents and conditions: a) SOCl2 (or COCl)2, DMF, PhMe. b) i) PhMe, Et3N; ii) KSAc, EtOAc, 55 °C, 89%, 2 steps. c) i) HCO2H, H2O2; ii) G-60 Darco,
75-90%.
Table 1. Comparison of three routes to CP-865,569 from 2
was carried directly into the next reaction. This carboxylic
convergent
linear
discovery
route
acid could then be converted to benzylic chloride 13 directly
in a reaction wherein 12 and paraformaldehyde were
suspended in concn HCl, H3PO4, and HOAc and heated for
18 h. Despite our fears to the contrary, the deposition of
paraformaldehyde polymer on the reaction vessel setup was
not observed. After reaction completion, the product could
either be precipitated with the addition of water or taken
through an extractive workup with EtOAc. The extractive
workup was the preferred method as it purged some low-
level impurities that the precipitation did not.
At this stage in the synthesis, we were faced with a choice
of either taking acid 13 into an amide formation to recon-
verge with the linear synthesis at chloride 7 or displacing
the chloride with thioacetate prior to amide formation to
reconverge at thioacetate 9. Both sequences were investi-
gated, and the preferred method actually turned out to be a
blend of the two (Scheme 5). We found that chloride 13
was easier to handle than 14, and thus proceeded with the
coupling of 13 and 2. Following completion of the amide
formation, EtOAc and KSAc were added, and the thioacetate
was installed prior to isolation of the product. This procedure
provided 89% of the desired thioacetate 9 in high purity after
crystallization. Oxidation as previously described completed
the synthesis of CP-865,569. This route was demonstrated
on lab scale but has not been run on kilogram scale.
oxidation route oxidation route
% overall yield
(from 2)
no. of bulk
operations
safety
80
6
45
5
30
8
oxidation
reaction
123
oxidation
reaction
188
no issues
455
301
5/9
total waste/
L/kg
organic waste/
L/kg
solid
intermediates
118
6/6
144
5/9
was synthesized via a convergent route in 80% yield (from
2) and a longest linear route of four steps from p-
chlorophenol. A comparison of this route to the other routes
previously scaled is shown in Table 1.
Experimental Section
General. All materials were purchased from commercial
suppliers and used without further purification. All reactions
were conducted under an atmosphere of nitrogen unless noted
otherwise. All reactors were glass-lined steel vessels with
the exception of those used for catalytic hydrogenations,
which were Hastelloy. Reactions were monitored for comple-
tion by removing a small sample from the reaction mixture
and analyzing the sample by TLC, GC, or HPLC. HPLC
analyses were performed using one of the following sys-
tems: Chiralpak AD 4.6 mm × 250 mm Sn column
n-heptane/DEA (99.8/0.2 v/v) mobile phase or a Zorbax SB-
CN 4.6 mm × 150 mm column, and a mobile phase
consisting of acetonitrile and either 0.5% perchloric acid or
0.2% phosphoric acid. GC analyses were performed on a
DBWaxEtr (15 m × 0.25 mm i.d. × 0.25 µm film). Proton
and carbon NMR spectroscopy was performed on a Bruker-
Spectrospin Avance 400 MHz instrument. Karl Fischer
Conclusions
CP-865,569 has been synthesized on large-scale using
three different methods. Due to the water solubility of the
final product, the performic acid oxidation approach, which
produces the desired material in the absence of metal
counterions, proved to be superior to a sulfite displacement
approach to the sulfonic acid. The performic acid oxidation
proved safe and reliable to run on 30-kg scale. Using the
oxidation route, 96.2 kg of material was produced with an
overall campaign yield of 45% from amine 2. CP-865,569
758
•
Vol. 11, No. 4, 2007 / Organic Process Research & Development

measurements were obtained on a Mettler Toledo DL38
Volumetric Karl Fischer apparatus. Melting points were
determined on a Bu¨chi B-545 melting point apparatus.
Elemental analyses were performed by Quantitative Tech-
nologies Inc. of Whitehouse, NJ.
water (238 L), and 50% sodium hydroxide solution in water
(55 L). The reaction was stirred for 30 min for complete
dissolution. Chloroacetyl chloride (19.3 kg, 168 mol) was
added to the reaction mixture, while keeping the pot
temperature between 15 and 30 °C during the addition. The
reaction was held at 20-25 °C for 25 min. The reaction was
sampled and checked for completion. The reaction mixture
was settled for 20 min. The phases were separated, and
magnesium sulfate (19.9 kg) was added to the 2-methyltet-
rahydrofuran layer. The reaction was stirred for 20 min. The
reaction slurry was filtered and concentrated to 567 L using
vacuum distillation. Potassium carbonate (32.2 kg, 229 mol),
potassium iodide (27.8 kg, 168 mol), and 5-chlorosalicyl-
adehyde (25.0 kg, 160 mol) were added to the reaction
mixture. The reaction mixture was heated to 66-70 °C for
a minimum of 10 h. The reaction was cooled to 20-30 °C
and sampled for completion by HPLC. To the reaction was
added process water (159 L) and 50% sodium hydroxide
solution in water (6.35 L). The reaction was stirred for a
minimum of 15 min, settled for a minimum of 15 min, and
the phases were separated. This organic layer wash was
completed two more times. On the final wash the phases
were separated, and magnesium sulfate (19.9 kg) was added
to the 2-methyltetrahydrofuran layer. The reaction slurry was
filtered and concentrated to 114 L using vacuum distillation.
To the reaction mixture were added methanol (208 L) and
2-methyltetrahydrofuran (416 L). To the reaction mixture
was added sodium borohydride (1.12 kg, 28.9 mol) in four
equal portions while keeping the pot temperature 20-30 °C.
The reaction was stirred for 30 min at 20-30 °C. The
reaction was sampled for completion by HPLC. To the
reaction mixture was added ethyl acetate (1000 L), process
water (160 L), 50% sodium hydroxide solution in water (18
L), and citric acid (32.5 kg). The mixture was stirred for 20
min, settled for 15 min, and the phases were split. To the
organic layer was added magnesium sulfate (2.8 kg), and
the mixture was stirred for 20 min, and the insolubles were
filtered off. The reaction mixture was concentrated and used
as a solution in the next step. 1H NMR (400 MHz, CDCl3):
δ 7.28 (m, 3H), 7.18 (dd, 1H), 7.06 (t, 2H), 6.78 (m, 1H),
4.75 (m, 2H), 4.68 (s, 2H), 3.6 (d, 1H) 3.4 (d, 1H), 3.2 (s,-
1H) 3.04 (s, 1H) 2.70 (s, 1H) 2.2 (dd,1H) 1.75 (s,1H), 1.3
(dd,3H), 0.98 (dd,3H). 13C NMR (CDCl3): δ 128.65, 113.71,
77.58, 77.26, 76.95, 66.87, 61.54, 58.15.
(2S,5R)-1-(4-Fluorobenzyl)-2,5-dimethylpiperazine (2).
To a clean, dry, nitrogen-purged 1900-L reactor were charged
trans-2,5-dimethylpiperazine (1) (25.8 kg, 226 mol), tet-
rabutylammonium chloride (2.22 kg, 7.91 mol, cyclohexane
(258 L), process water (36.1 L), 50% sodium hydroxide
solution in water (23.6 L), and 4-fluorobenzyl chloride (31.3
kg, 215 mol). The reaction was heated to 35-45 °C for a
minimum of 6 h. The reaction was cooled 20-25 °C,
sampled, and checked for completion by GC. The mixture
was allowed to settle for at least 20 min. The phases were
separated, and process water (258 L) and 37% hydrochloric
acid (22.2 L) were added to the organic layer. The reaction
mixture was allowed to stir for at least 20 min after which
the pH of the lower aqueous layers was analyzed (pH range
is between pH 4 and 5). The reaction mixture was settled
for at least 25 min. The phases were separated, and to the
aqueous layer were added 2-methyltetrahydrofuran (387 L)
and 50% sodium hydroxide solution in water (23.6 L kg).
The mixture was stirred for a minimum of 20 min. The pH
of the lower aqueous layer was analyzed (pH range between
pH 12 and 14 recommended), and the contents were settled
for a minimum of 25 min. The phases were separated, and
the 2-methyltetrahydrofuran layer was concentrated atmo-
spherically until the pot temperature reached 82 °C and then
was vacuum stripped to a final volume of 75-90 L. The
solution was cooled to 20-30 °C, and methanol (129 L)
was added to the reaction. The reaction solution was
concentrated atmospherically to a final volume of 75-90
L. The previous step was repeated one more time. To the
concentrated mixture was added 1800 L of methanol and
L-tartaric acid (68.2 kg, 452 mol). The reaction mixture was
heated 60-70 °C and held for a minimum of an hour. The
solution was cooled to 10-20 °C until a white precipitate
came out of solution. The white slurry was heated to 45-
50° and held for a minimum of 12 h. The reaction was cooled
to 20-25 °C and then filtered on an adequately sized filter
and washed with methanol (52 L). After drying under
vacuum at 40-50 °C for at least 12 h with a slight nitrogen
bleed, 45 kg (86 mol) of (2S,5R)-1-(4-fluorobenzyl)-2,5-
dimethylpiperazine (2) was isolated in 38% yield (based on
2-(4-Chlorophenoxy)acetic Acid (12). To a clean, dry
2000-mL three-neck flask fitted with a reflux condenser,
temperature probe, and a glass stopper were added potassium
phosphate tribasic (413 g, 1.94 mol), p-chlorophenol (11)
(100 g, 0.778 mol), methyl chloroacetate (127 g, 1.17 mol),
and acetone (300 mL). The reaction mixture was heated to
40-60 °C for 10 h. The reaction mixture was analyzed by
HPLC, and process water (1000 mL) was added to the
reaction mixture. The reaction mixture was concentrated at
80-90 °C until all acetone distilled off. The reaction was
heated for 14 h at 80-90 °C. The reaction mixture was
cooled to 25-30 °C, and concentrated hydrochloric acid (210
mL) was added. The reaction was stirred for 20 min, cooled
to 10-15 °C, and the white product was filtered. The white
1
a maximum resolution yield of 50%) mp: 96-97 °C. H
NMR (400 MHz, methanol): δ 7.36 (m, 2H), 7.06 (m, 2H),
4.47 (s, 4H), 4.19 (d, 1H), 3.30 (m, 4H), 2.85 (m, 2H) 2.70
(m, 1H), 2.14 (m,1H) 1.26 (d, 3H) 1.20 (d, 3H). 13C NMR
(DMSO): δ 174.8, 72.6, 51.4, 49.2, 48.5, 48.3, 48.0, 47.8,
47.74, 47.72, 47.6, 47.4, 47.2, 15.9, 15.2. Anal. calcd for
C21H31FN2O12: C, 48.27; H, 5.98; N, 5.36. Found: C, 45.55;
H, 5.97; N, 4.78.
1-((2R,5S)-4-(4-Fluorobenzyl)-2,5-dimethylpiperazin-
1-yl)-2-(4-chloro-2-(hydroxymethyl)phenoxy)ethanone (6).
To a clean, dry, nitrogen-purged 1900-L reactor were charged
(2S,5R)-1-(4-fluorobenzyl)-2,5-dimethylpiperazine (2) (79.6
kg, 152 mol), 2-methyltetrahydrofuran (1651 L), process
Vol. 11, No. 4, 2007 / Organic Process Research & Development
•
759

product was washed with 20 mL of water and dried in a
none (9). To a nitrogen-purged ethyl acetate solution of
1-((2R,5S)-(4-(4-fluorobenzyl)-2,5-dimethylpiperazin-1-yl)-
2-(4-chloro-2-(chloromethyl)phenoxy)ethanone (7) (584 L/0.30
M) at 20 °C was added solid potassium thioacetate (30.1
kg/1.50 equiv.). The reaction was heated to 50 °C for 8 h,
and then cooled to 25 °C and washed with water (2 × 150
L). The organic layer was separated and concentrated
atmospherically to approximately 100 L (1.50 M, calculated
from starting benzyl alcohol). Toluene (215 L) was added
to the oil followed by reduced pressure distillation to
approximately 100 L (max. pot temperature not to exceed
40 °C). The toluene dilution/concentration was twice re-
peated. The ethanethioate oil was again diluted with toluene
(150 L) followed by hexane (230 L). Some precipitation was
evident at this point. The resulting mixture was heated to
60 °C and held until a complete solution was achieved. The
heated solution was cooled to 10 °C and stirred for 2 h
following initial precipitation. The precipitate was filtered,
rinsed with hexanes, and dried under vacuum at 50 °C.
vacuum oven for use in the following step. Mp: 152-153
1
°C. H NMR (400 MHz, CD3OD): δ 7.23 (dd, 2H), 6.88
(dd, 2H), 5.21 (s, 1H), 4.62 (s, 2H). 13C NMR (CD3OD): δ
171.26, 159.95, 129.21, 126.16, 116.00, 64.86, 47.92. Anal.
calcd for C8H7ClO3: C, 51.50; H, 3.78. Found: C, 50.33;
H, 3.54.
2-(4-Chloro-2-(chloromethyl)phenoxy)acetic Acid (13).
To a clean, dry 2-L vessel were added 2-(4-chlorophenoxy)-
acetic acid (12) (100 g, 0.536 mol), paraformaldehyde (72.4
g, 0.800 mol), acetic acid (400 mL, 6.98 mol), 37%
concentrated hydrochloric acid (368 mL, 4.29 mol), and
phosphoric acid (59.6 mL, 1.07 mol). The reaction was
heated to 85-95 °C for 18 h. The reaction was cooled to
20-30 °C and then analyzed for reaction completion by
HPLC. Ethyl acetate (500 mL) was added to the reaction
which was then stirred for 20 min. The phases were
separated, the organic layer was concentrated, and the white
product was filtered off and washed with ethyl acetate (20
mL). After drying under vacuum at 40-50 °C for 12 h with
a slight nitrogen bleed, 2-(4-chloro-2-(chloromethyl)phen-
1
Yield: 81% (two steps). H NMR (DMSO-d6): δ 0.92 (d,
3H), 1.22 (d, 3H), 2.22 (d, 1H), 2.31 (s, 3H), 2.70 (dd, 1H),
2.98 (m, 1H), 3.31(bs, 1H), 3.44 (d, 1H), 3.59 (d, 1H), 3.71-
(bs, 1H), 4.10 (s, 2H), 4.26 (bs, 1H), 4.76 (d, 1H), 4.86 (d,
1H), 6.95 (d, 1H), 7.09 (t, 2H), 7.22 (dd, 1H), 7.28 (d, 1H),
7.36 (t, 2H). 13C NMR (DMSO-d6): δ 7.0, 15.9, 20.1, 26.9,
29.7, 42.9, 45.9, 48.3, 51.0, 56.7, 66.9, 113.8, 114.4, 124.1,
126.4, 127.7, 128.2, 129.1, 129.6, 134.7, 154.6, 160.1, 161.7,
165.5.
1
oxy)acetic acid (13) was isolated. Mp: 112-113 °C. H
NMR (400 MHz, CD3OD): δ 7.4 (dd, 1H), 7.26 (qd, 1H),
6.9 (dd, 1H), 4.74 (s, 2H), 4.69 (s, 2H). 13C NMR (CD3-
OD): δ 170.87, 154.75, 130.09, 129.28, 128.72, 126.07,
113.61, 65.31, 39.98. Anal. calcd for C9H8Cl2O3: C, 45.99;
H, 3.43. Found: C, 45.33; H, 3.32.
1-((2R,5S)-4-(4-Fluorobenzyl)-2,5-dimethylpiperazin-
1-yl)-2-(4-chloro-2-(chloromethyl)phenoxy)ethanone (7).
To a concentrated solution of 1-((2R,5S)-4-(4-fluorobenzyl)-
2,5-dimethylpiperazin-1-yl)-2-(4-chloro-2-(hydroxymethyl)-
phenoxy)ethanone (6) in 2-methyltetrahydrofuran (74.0 kg/
75.9 L solution) was added ethyl acetate (214 L). The
resulting solution was concentrated under atmospheric pres-
sure to a minimum stir volume. To the concentrated solution
was added ethyl acetate (214L). Again, the solution was
concentrated to a minimum stir volume. The ethyl acetate
charge/concentration sequence was repeated until the water
content was reduced to 1.0% or less via KF analysis. Upon
completion of drying, ethyl acetate (874 L) was added to
the concentrated solution to create a final dried solution with
a concentration of 0.50 M. To a nitrogen-purged solution of
thionyl chloride (31.4 kg) in ethyl acetate (532 L) at 20 °C
was added the dried ethyl acetate solution of 1-((2R,5S)-4-
(4-fluorobenzyl)-2,5-dimethylpiperazin-1-yl)-2-(4-chloro-2-
(hydroxymethyl)phenoxy)ethanone (6) at a rate to ensure the
temperature did not exceed 35 °C. The reaction mixture was
stirred at 25 °C for 2 h followed by an aqueous quench (361
L). The organic layer was separated, washed with water, and
added to a 1.0 N NaOH solution (381 L) at a rate to ensure
the temperature did not exceed 30 °C. The resulting mixture
was stirred for 30 min at 20 °C, and the organic layer was
separated and dried with several atmospheric ethyl acetate
azeotropic distillations. The product (7) was delivered to the
next step as a dried ethyl acetate solution (0.30 M, calculated
as 100% yield from starting benzyl alcohol).
(2-(2-((2R,5S)-4-(4-Fluorobenzyl)-2,5-dimethylpiperazin-
1-yl)-2-oxoethoxy)-5-chlorophenyl)methanesulfonic Acid
(8). To a nitrogen-purged reactor at 20 °C were added solid
(2-(2-((2R,5S)-4-(4-fluorobenzyl)-2,5-dimethylpiperazin-1-
yl)-2-oxoethoxy)-5-chlorophenyl)methanesulfonic acid (9)
(34.2 kg/71.3 mol) and 96% formic acid (7.20 L/kg
ethanethioate). To the formic acid solution at 20 °C was
added 30% hydrogen peroxide (3.77 L/0.467 equiv). The
exotherm was noted, and the temperature rise was recorded.
The reaction was stirred at 20 °C for 2 h followed by the
addition of a second aliquot of 30% hydrogen peroxide (3.77
L/0.467 equiv). Again, the exotherm was noted, and the
temperature rise was recorded. The reaction was stirred at
20 °C for 2 h. The reaction mixture was sampled and
analyzed by HPLC to monitor starting material consumption
and sulfonic acid formation. (A lack of starting material
consumption and sulfonic acid formation implies a buildup
of hydrogen peroxide, and subsequent additions should be
discontinued.) Upon determination of safe hydrogen peroxide
consumption, the reaction mixture was treated with seven
further aliquots of 30% hydrogen peroxide (3.77 L/0.467
equiv) with 2-h reaction times at 20 °C. Samples for HPLC
analysis were taken following aliquots 5 and 8 to reaffirm
safe peroxide consumption. Upon complete addition of
hydrogen peroxide, the reaction mixture was added to a
suspension of 96% formic acid (39 L) and activated carbon
(Darco G-60) (1.70 kg) at 20 °C under a high-volume
nitrogen sweep over 2 h. A gentle gas evolution was noted
and confirmed a controlled consumption of excess hydrogen
peroxide. The resulting suspension was stirred at 20 °C for
1-((2R,5S)-(4-(4-Fluorobenzyl)-2,5-dimethylpiperazin-
1-yl)-2-(4-chloro-2-(acetoxythiomethyl)phenoxy)etha-
760
•
Vol. 11, No. 4, 2007 / Organic Process Research & Development

12 h, and a reaction sample was analyzed by accelerated
rate calorimetry to ensure complete peroxide consumption.
The quenched reaction was filtered through Celite, and the
filtrate was diluted with glacial acetic acid (316 L). The
resulting solution was concentrated under reduced pressure
to a volume of approximately 30 L with a maximum pot
temperature of 50 °C. The desired (2-(2-((2R,5S)-4-(4-
fluorobenzyl)-2,5-dimethylpiperazin-1-yl)-2-oxoethoxy)-5-
chlorophenyl)methanesulfonic acid (8) oil was diluted with
methanol (140 L) and added to isopropyl ether (1400 L) at
20 °C over1 h. The resulting slurry was stirred at 20 °C for
2 h followed by filtering, washing with isopropyl ether (35
L), and vacuum drying at 50 °C. Yield: 91%.
(2-(2-((2R,5S)-4-(4-Fluorobenzyl)-2,5-dimethylpiperazin-
1-yl)-2-oxoethoxy)-5-chlorophenyl)methanesulfonic Acid
L-Arginine-cocrystal (8-Arg). To a nitrogen-purged reactor
at 20 °C were added solid (2-(2-((2R,5S)-4-(4-fluorobenzyl)-
2,5-dimethylpiperazin-1-yl)-2-oxoethoxy)-5-chlorophenyl)-
methanesulfonic acid (8) (37.1 kg/69.6 mol), solid L-arginine
(12.4 kg/69.6 mol), 1-propanol (380 L), and water (76 L).
The resulting solution was stirred at 25 °C for 2 h followed
by addition of activated carbon (Darco KBB) (2.8 kg) and
Celite (1.9 kg). The new suspension was stirred at 25 °C for
2 h and then filtered through a pad of Celite (1.0 kg). The
filter pad was washed with 1-propanol, and the collected
filtrate was concentrated under reduced pressure to a volume
of approximately 110 L (maximum pot temperature of 55
°C). The concentrated solution was diluted with 1-propanol
(190 L) followed by distillation at reduced pressure to 110
L. The 1-propanol charge/concentration was repeated two
more times, a sample was removed, and water content was
determined by KF analysis. Upon completion of azeotropic
removal of water (water less than 1.0%), the slurry was
cooled to 20 °C and stirred for 6 h. The solids were filtered,
washed with 2B ethanol, and dried on the filter under
nitrogen flow. The collected solids were added to a reactor
followed by 2B ethanol (380 L). The slurry was stirred at
reflux for 12 h, cooled to 20 °C, and stirred for 1 h. The
solids were filtered, washed with 2B ethanol (75 L), and
dried under vacuum at 50 °C for 24 h. Yield: 67%. 1H NMR
(DMSO-d6): δ 0.91 (d, 3H), 1.21 (d, 3H), 2.21 (d, 1H), 2.70
(d, 1H), 2.96 (s, 1H), 3.29 (bs, 1H), 3.43 (d, 1H), 3.58 (d,
1H), 3.76 (bs, 1H), 3.84 (s, 2H), 4.29 (bs, 1H), 4.69 (d, 1H),
4.75 (d, 1H), 6.92 (d, 1H), 7.09 (t, 2H), 7.15 (dd, 1H), 7.36
(t, 2H), 7.57 (s, 1H). 13C NMR (DMSO-d6): δ 7.0, 15.9,
43.3, 45.8, 48.3, 49.4, 51.1, 56.8, 67.9, 114.3, 114.4, 124.2,
126.3, 126.7, 129.6, 130.0, 134.8, 154.8, 157.0, 160.1, 161.7,
166.0. Potency: 96.9%. HPLC purity: 97.5%. Arginine
content: 27.3% (26.4% theoretical). Water: 0.4%. Ethanol
0.24%.
Acknowledgment
We thank Professor Steven Ley and Drs. Robert Dugger,
Tamim Braish, Stephane Caron, and John Ragan for helpful
discussions. We also thank Dr. Gerald Weisenburger, Cliff
Meltz, David Bill, and Pamela Clifford for safety analysis
and Dr. Todd Bibart, Kyle Leeman, and Natalie Harner for
analytical assistance. We also thank Jeff Adler, Paul Allard,
Eric Allen, Joseph Bellavance, Thomas Brooks, Keith
Brouwer, Gary Bouchard, Todd Carden, Randy Church,
Baher Farag, Samuel Helme, Thomas Limani, Larry Lum-
bert, Brian Morgan, Alden Peckham, Stephen Stott, Ralph
Scheck, III, Scott Schelkly, Russell Shine, Walden Smolen,
and Charles Strutt for their execution of the large-scale
procedures.
Received for review February 14, 2007.
OP7000386
Vol. 11, No. 4, 2007 / Organic Process Research & Development
•
761