Synthesis of the NK1 receptor antagonist GW597599. Part 1: development of a scalable route to a key chirally pure arylpiperazine

Article abstract of DOI:10.1021/op800146d

GW5975991 is a novel NK-1 antagonist currently under investigation for the treatment of CNS disorders and emesis. The initial synthetic route devised from the medicinal chemistry one, used several hazardous reagents, gave low yields, and produced high lev

Full text of DOI:10.1021/op800146d

Organic Process Research & Development 2008, 12, 1188–1194
Synthesis of the NK1 Receptor Antagonist GW597599. Part 1: Development of a
Scalable Route to a Key Chirally Pure Arylpiperazine
Giuseppe Guercio,* Sergio Bacchi,* Michael Goodyear, Antonella Carangio, Francesco Tinazzi, and Stefano Curti
Department of Chemical DeVelopment, GlaxoSmithKline, Via Fleming 4, 37135 Verona, Italy
Scheme 1. Initial synthetic route
Abstract:
GW597599 1 is a novel NK-1 antagonist currently under investiga-
tion for the treatment of CNS disorders and emesis. The initial
synthetic route devised from the medicinal chemistry one, used
several hazardous reagents, gave low yields, and produced high
levels of wastes. By targeted process of research and development,
application of novel techniques, and extensive route scouting, a
novel synthetic route for GW597599 has been developed.This
paper reports the optimisation work of the first stage in the
chemical synthesis of GW597599: the development of a pilot-plant
suitable process for the synthesis of the arylpiperazine derivative
7 in an optically pure fashion. In particular, the process definition
allowed eliminating the initial need for cryogenic conditions and
copper catalysis in Grignard chemistry. It also allowed replacing
a classical resolution step with a more efficient dynamic kinetic
resolution, substantially enhancing the overall yield and throughput.
1. Introduction
Substance P (SP) is a member of the neurokinin family
that exerts its pleiotropic role by preferentially binding
to the neurokinin receptor classified as NK1. The role of
SP in mood disorders may also partly depend on its
interaction with serotonin (5-HT) receptors. Preclinical
studies suggest that NK1 antagonists and selective sero-
tonin reuptake inhibitors (SSRIs) may synergise to produce
a final common effect on serotonergic neurotransmission.
In fact GW597599 is under development to tackle central
nervous system (CNS) diseases. Furthermore, substance
P and the NK1-receptors that mediate its activity are
present in the brain stem centres that elicit the emetic
reflex. Nausea and vomiting have been reported as the
most distressing side effects of chemotherapy, and the
disruptive effects of these symptoms on patients’ daily
lives have been well documented. In light of the need for
continued routine use of emetogenic chemotherapy, ef-
fective prevention of chemotherapy-induced nausea and
vomiting (CINV) is a central goal for physicians admin-
istering cancer chemotherapy. GW597599 1 is a novel
NK-1 antagonist currently being developed to relieve also
those symptoms. In this paper we describe the successful
efforts to define a synthetic route for large-scale manu-
facturing.
2.1. Experimental Details. The synthesis started with the
in situ preparation of a Grignard reagent of 2-bromo-5-
fluorotoluene 2 which was then added to a solution of methyl
oxalyl chloride 3 in THF. The temperature was maintained
between -55 to -65 °C in order to minimize the formation of
the tertiary alcohol 13a.
Furthermore, the heat evolution was controlled by diluting
the reaction to such an extent that the conversion of even gram
quantities of 2-bromo-5-fluorotoluene might have required
unacceptably high volumes of solvents.1
The keto-ester derivative 4 was reacted with ethylenediamine
52 in the presence of anhydrous sodium sulfate to give imine
6, which was reduced Via catalytic hydrogenation, though
2. Initial synthetic Route
The synthetic route employed to synthesise the first few
grams of GW597599 is shown in Scheme 1
* To whom correspondence should be addressed. E-mail: giuseppe.2.guercio@
gsk.com; sergio.k.bacchi@gsk.com.
1188
•
Vol. 12, No. 6, 2008 / Organic Process Research & Development
10.1021/op800146d CCC: $40.75
2008 American Chemical Society
Published on Web 10/16/2008

requiring long reaction time and a high hydrogen pressure, to
give racemic ketopiperazine 7.
required for a successful initiation step. The use of TMSCl in
the activation of magnesium for Grignard reagents has been
reported to be effective for large-scale production.7 When
applying this procedure to our chemistry on the laboratory scale,
we found that by addition of 5 mol % of TMSCl, the Grignard
formation initiated smoothly and quickly at room temperature,
and the amount of catalyst could be reduced as low as 0.5 mol
% without affecting the yield and the initiation time. After
initiation, the reaction was kept at room temperature reaching
completion within 2 h with no need for higher temperature.
During the following scale-up and optimisation it was seen that,
with a temperature of 25 °C and 0.5 mol % TMSCl, the
initiation time might be temperamental upon scale-up. Less
borderline conditions (1 mol % TMSCl and 35 °C) were
therefore tested on several grams scale and were subsequently
chosen as working conditions for the 200 kg pilot-plant
campaign.
A further improvement in the process was achieved by
decreasing the amount of magnesium used in the synthesis (from
1.2 to 1 equiv with respect to 2). This last change in the process,
whilst having no effect on the yield and the quality of the final
intermediate 14, was of pivotal importance from an operational
point of view in the pilot plant, as it significantly decreased the
filtration time of the residual metal from several hours to 30
min.
The carbamoyl chloride formed by the treatment of the
racemic ketopiperazine 7 with the highly toxic triphosgene,3
was reacted with the benzylamine 12 giving the urea 8 as a
mixture of four stereoisomers. The amide group in the pipera-
zine ring was then reduced by treatment with a solution of
borane in THF at reflux. The resulting crude material was
purified by chromatography to isolate a single diastereoisomer
of 1 which underwent further classical resolution procedures
to obtain the optically pure GW597599. The active moiety was
finally crystallized as its acetate salt.
2.2. Evaluation of the Route. The overall yield obtained
from the initial route was only about 2% theoretical; thus, a
more efficient alternative was sought to provide larger quantities
of GW597599. In addition to the low yield, when a purified
single diastereoisomer isolated from mixture 8 was reduced with
borane, partial epimerization was observed at the chiral centre
of the piperazine ring.4 It was therefore not possible to improve
the efficiency of the process by isolating the ketopiperazine 7
in enantiomerically pure form by maintaining the borane
reduction as the final step. Also, the use of BH3 solution in
THF was not recommended as pilot-plant first choice. This is
because the borane exists in solution as a dimeric species with
a problematically low auto-ignition temperature (38 °C). Finally,
the acetate salt of 1 proved to be unstable at above 60 °C due
to loss of acetic acid, which would cause difficulties both
preparing and drying the final compound.5,6
A quick win was the replacement of methyl oxalyl chloride
3 with diethyl oxalate 17 that increased the yield from 50-60%
to almost 80% theoretical. Additionally the less reactive diethyl
oxalate did not require smoothing the hardness of the Grignard
reagent Via copper transmetalation, resulting in a simplified
workup when getting rid of relatively low soluble copper salts.
However, the low-temperature chemistry was still required for
this reaction to minimise formation of tertiary alcohol 13b,
which despite a temperature as low as -50 °C, accounted for
∼17% of the final output weight.
3. Development of a Practical Convergent Synthesis
3.1. Synthetic Strategy. The disconnection of 1 into the
two chiral fragments 15 and 16, shown in Scheme 2, provided
the basis of a more efficient route, avoiding the need for the
last-stage resolution and the issue of the racemisation in the
final chemical step caused by borane. The synthesis of the chiral
ketopiperazine 14 followed by reduction and final coupling,
formed the basis of our potential new route.
The low yield was made even worse by the cyclisation of
the residual diethyl oxalate 17 with the ethylenediamine 5
employed in the subsequent step (Scheme 3).
Scheme 2. New potential route
Scheme 3. Side-product identified during cyclisation step 2
The resulting cyclic amide 18 was able to undermine the
resolution of racemic ketopiperazine to optically pure 14, in a
(3) (a) Rosiak, A.; Hoenke, C.; Christoffers, J. Eur. J. Org. Chem. 2007,
26, 4376–4382. (b) Jacobsen, E. J.; Stelzer, L. S.; Belonga, K. L.; Carter,
D. B.; Im, W. B.; Sethy, V. H.; Tang, A. H.; VonVoigtlander, P. F.;
Petke, J. D. J. Med. Chem. 1996, 39 (19), 3820–3836.
(4) Donohue, A. C.; Jackson, W. R. Aust. J. Chem. 1995, 48 (10), 1741–
1746.
(5) Alvaro, G.; Di Fabio, R.; Giovannini, R.; Guercio, G.; St. Denis, Y.;
Ursini, A. Preparation of 2-phenylpiperazine-1-carboxylic acid benzy-
lamides as tachykinin antagonists. PCT Int. Appl. 2001, WO/0125219,
2001.
(6) Alvaro G.; Di Fabio, R., Maragni, P.; Tampieri, M.; Tranquillini, M. E.
PCT Int. Appl. WO/0232867 A1 20020425, 2002.
3.2. Grignard. The activation of magnesium in the synthesis
of the Grignard reagent 19 was performed with the Gillman
catalyst;1 under these conditions a temperature of 50 °C was
(1) Gilman, H.; Heck, L. L. Bull. Soc. Chim. Fr. 1929, 45, 250.
(2) (a) Fokin, A. S.; Burgart Ya, V.; Saloutin, V. I. Russ. J. Org. Chem.
2006, 42 (6), 887–896. (b) Ogawa and Co, Jpn. Kokai Tokkyo Koho,
57062269, 1982.
(7) (a) Bellingham, R.; Ace, K.; Armitage, M.; Blackler, P.; Ennis, D.;
Hussain, N.; Lathbury, D.; Morgan, D.; O’Connor, N.; Oakes, G.;
Passey, S.; Powling, L. Org. Process Res. DeV. 2001, 5, 479. (b) See
Supporting Information for the Process Safety Evaluation of 19.
Vol. 12, No. 6, 2008 / Organic Process Research & Development
•
1189

Table 2. Additives screening for cyclisation in refluxing
way that a silica pad or a multiple crops precipitation, time and
yield demanding, was necessary to achieve the target 96% ee.
A plethora of possible route modifications was tested, but
the real breakthrough was found when substituting the ketoester
4 with the Weinreb amide derivative 21 (Scheme 4).8
methanol
additive
none
yield (%)
24 (%)
78
77
79
84
83
79
78
81
78
79
82
81
2.8
2.7
2.6
1.2
2.1
2.7
2.7
2.4
2.5
2.7
2.3
2.8
MgSO4
KF
CH3COOH
BF3 ·Et2O
HCl·Et2O
CeCl3
Scheme 4. Weinreb amide approach
AlCl3
ZnCl2
CSA
CF3COOH
citric acid
Scheme 5. New side products identified during cyclisation
step 2
simplified workup procedure and a consistent lower level
of the impurity 24 and consequently of 25.
3.4. Reduction. The reduction step was somehow prob-
lematic to be performed at plant scale due to very long
reaction time, high hydrogen pressure, and high catalyst
loading.
The optimisation of the reduction of the imine 6 implied
a screening of catalysts which allowed the identification
of wet platinum on charcoal as a faster and safer
hydrogenation catalyst rather than the dry palladium on
charcoal (Figure 1) initially employed in the synthesis. A
further refinement of the reaction conditions via a half-
factorial statistical design of experiment (DoE), four
The R-keto amide derivative 21 was easily prepared from
the same Grignard 19 by reaction with N,N′-dimethoxy-N,N′-
dimethylethanediamide 20, in a quantitative yield at 0 °C,
avoiding the tertiary alcohol contamination.9 Moreover, 21
underwent the transformations showed in the Scheme 4 achiev-
ing the targeted 96% ee for 14 in a single crop and with a
considerably higher yield.
3.3. Cyclisation. A solvent screening (Table 1) identified
methanol as the best solvent to use to perform the cyclisation
step. The use of molar quantities of additives (Table 2) allowed
gaining some additional yield. In particular acetic acid aided
the minimisation of the level of the impurity 24 formed by the
ethylendiamine nucleophilic substitution of the aromatic fluorine
atom (Scheme 5).
Table 1. Solvent screening for cyclisation (24 h up to
90-95°C)
solvent
yield (%)
24 (%)
toluene
2-propanol
ethanol
IMS
MeOH
1,4-dioxane
MTBE
19
50
63
68
78
28
24
25
63
45
39
24
<3
<3
<3
<3
<3
<3
<3
<3
19
ethylene glycol dimethyl ether
N,N-dimethyl-acetamide
1-methyl-2-pyrrolidinone
acetonitrile
41
<3
<3
ethyl acetate
A further step of fine optimisation was performed to
minimise volumes of methanol and amounts of both
ethylenediamine 5 and acetic acid. The result was a
Figure 1. Screen of hydrogenation catalysts.
1190
•
Vol. 12, No. 6, 2008 / Organic Process Research & Development

Scheme 6. DKR
reactions only, permitted identifying the optimal process
operating region to maximise the throughput (Table 3).
Table 3. Hydrogenation DoE ranges and best conditions
low
high
best conditions
hydrogen pressure (bar)
temperature (°C)
catalyst charge (% wt/wt)
0.3
20
2
4.3
40
10
1.5
40
4
3.5. Preparation of the Optically Pure Ketopiperazine
via Dynamic Kinetic Resolution (DKR). The resolution of
the racemic ketopiperazine into the pure enantiomer 14
was initially achieved by classical fractional crystallization
using the S-(+)-mandelic acid in ethyl acetate in a 38%
theoretical yield.
To avoid wasting the unwanted enantiomer we con-
sidered implementing a DKR (Scheme 6).10 The piperazine
secondary amine could form an iminium derivative with
an appropriate aldehyde increasing the acidity of the
benzylic proton. The resulting derivative could easily
racemise, allowing the wanted enantiomer to precipitate
out as its mandelate salt leaving the unwanted enantiomer
in solution ready to be further racemised. After a short
screen, the dichlorosalicylaldehyde 26 was selected,
leading in the end to ∼90% yield of a pure enantiomer of
the ketopiperazine. Unfortunately, this process required
purification on silica gel because it was easily undermined
by the amine impurities such as 5, 24, and 25.
While, as mentioned before, optimisation of the cycli-
sation step greatly decreased the amount of 24 and 25,
impurity 5 can only by removed by introducing an aqueous
workup before the DKR step. This meant that the water
contamination of the organic layer containing the raceme
ketopiperazine had to be eliminated before the aldehyde
addition. On a larger scale, it was very difficult to
azeotropically remove the water to the target level, thus
minimizing the formation of nonreactive hydrate of the
aldehyde. In fact while the yield achieved on a 5 L scale
was about 70%, the overall yield in plant was only about
50%. In addition, the slurry was difficult to stir, leaving
a considerable amount of product adhering to the reactor
walls. Hence, it was evident that the DKR needed further
optimization. This was accomplished by means of full
factorial DoE study, implying 12 reactions and two center
points to evaluate the curvature, which looked into reaction
conditions and stirring rates (Figure 2). As a result of the
DoE study, stirring rate was the most significant factor
influencing the primary yield of the reaction. Moreover,
a larger amount of water was buffered by a larger amount
of aldehyde. Finally, computer modeling enabled us to
design an appropriate reactor configuration to ensure
efficient mixing. The initially employed round bottomed
vessel with baffles led to the tendency to form dead mixing
zones (see Figure 3, yellow/green/blue areas) which
subsequently turned into encrustation. The computer
modelling suggested using a conical vessel without baffles,
as this would be ideal for thick suspensions and would
avoid the dead mixing zones issue. When the process was
repeated on pilot-plant scale with the new conditions (more
aldehyde and the suggested vessel configuration) it was
possible to reproduce the yield and purity achieved on 5
L scale.
(8) (a) Nahm, S.; Weinreb, S. Tetrahedron Lett. 1981, 22, 3815. For a
review see: (b) Sibi, M. P. Organic Prep. Proced. Int. 1993, 25 (1),
15–40. (c) Sibi, M. P.; Marvin, M.; Sharma, R. J. Org. Chem. 1995,
60, 5016–5023. (d) De Luca, L.; Giacomelli, G.; Taddei, M. J. Org.
Chem. 2001, 66, 2534–2537.
(9) Mukund, P. S.; Rajiv, S. J. Org. Chem. 1995, 60, 5016–5023.
(10) (a) Kubo, A.; Kubota, H.; Takahashi, M.; Nunami, K. J. Org. Chem.
1997, 62, 5830–5837. (b) Noyori, R.; Tokunaga, M.; Kitamura, M.
Bull. Chem. Soc. Jpn. 1995, 68, 36. (c) Ward, R. S. Tetrahedron:
Asymmetry 1995, 6, 1475. (d) Noyori, R.; Ikeda, T.; Ohkuma, T.;
Widhalm, M.; Kitamura, M.; Takaya, H.; Akutagawa, S.; Sayo, N.;
Saito, T.; Taketomi, T.; Kumobayashi, H. J. Am. Chem. Soc. 1989,
111, 9134. (e) Kitamura, M.; Ohkuma, T.; Tokunaga, M.; Noyori, R.
Tetrahedron: Asymmetry 1990, 1, 1. (f) Kitamura, M.; Tokunaga, M.;
Noyori, R. J. Am. Chem. Soc. 1993, 115, 144. (g) Kitamura, M.;
Tokunaga, M.; Noyori, R. Tetrahedron 1993, 49, 1853. (h) Seebach,
D. Angew. Chem., Int. Ed. Engl. 1990, 29, 1320.
4. Conclusions
In summary, we have developed a safe, robust and
scalable process for the first stage of the synthesis of
GW597599 1. When the sequence of Grignard addition,
cyclisation, reduction, and DKR (Scheme 7) was pro-
gressed into the rest of the GW597599 possible manu-
Vol. 12, No. 6, 2008 / Organic Process Research & Development
•
1191

environmentally friendly conditions such as lower amounts
of solvents and reagents, better reproducibility and plant
operability compared to those of the original route.
The new process was successfully scaled up in pilot plant
and delivered almost 400 kg of intermediate 14 in high yield
(average 75% over four steps) and excellent purity.
Scheme 7. Final manufacturing route
Experimental Section
N-N′-Dimethoxy-N-N′-dimethyl oxaldiamide 20. To N,O-
Dimethylhydroxylamine hydrochloride 23 (231 g, 2.37 mol)
in dichloromethane (2300 mL) was added oxalyl chloride 22
(100 mL, 1.18 mol) all at once. The resulting mixture was
cooled down to 0 °C. A solution of triethylamine (666 mL,
4.78 mol) in dichloromethane (1800 mL) was added dropwise
over 30 min. The suspension was stirred at 0 °C and then
washed twice with water (2 × 1500 mL).
The aqueous layer was extracted two additional times
with dichloromethane (2 × 1300 mL). The combined
organic phases were concentrated to ∼380 mL, and
cyclohexane (3100 mL) was added dropwise. Crystalli-
sation occurred almost immediately. The resulting white
solid was stirred for 30 min at room temperature and for
1.5 h at 0 °C, filtered, washed with cyclohexane (1000
mL), and finally dried under high vacuum to obtain 188 g
of 20 as a white solid (yield ) 90%).
1H NMR (400 MHz, DMSO-d6): δ 3.13-3.18 (bs, 6H);
3.65-3.70 (bs, 6H); MS (ES+) m/z 176.17 (MH+); IR (nujol):
1679-1655 (str. CdO) (cm-1); mp 89-90 °C.
HPLC Zorbax Eclipse XDB C8; Mobile phase A: water and
B: acetonitrile; Gradient: 0 min 55% A to 18 min 20% A. Flow
1 mL/min; Detector UV DAD @215 nm. Retention time 20:
2.00 min. Purity >99.8% a/a.
Figure 2. DoE on parameters affecting DKR.
Bromo-(4-fluoro-2-methylphenyl)magnesium 19. To a
suspension of magnesium turnings (16.5 g, 0.68 mol) in
anhydrous THF (450 mL), TMSCl (0.75 mL, 5.91 mmol) was
added at 25 °C under nitrogen. A portion (∼5 mL) of 2-bromo-
5-fluorotoluene 2 (134 g, 0.71 mol) was added to the magne-
sium in THF. The suspension was stirred at 35 °C (jacket
temperature) until initiation of reaction was seen (heat evolution,
∼10 °C internal temperature increase and mixture discoloration).
The remaining 2-bromo-5-fluorotoluene 2 was added, keeping
the internal temperature in the range of 35-45 °C. The progress
Figure 3. computer modelling.
facturing route, it contributed considerably to the overall
scoring of the new process in terms of safer and more
1192
•
Vol. 12, No. 6, 2008 / Organic Process Research & Development

of the reaction was followed by continuously monitoring the
heat released during addition of 2. The reaction mixture was
stirred for 3 h and then cooled to 25 °C and used as such in the
synthesis of 21.
HPLC Phenomenex LUNA C18; Mobile phase A: 0.05%
TFA/water and B: 0.05% TFA/acetonitrile; Gradient: 0 min 0%
B to 8 min 95% B. Flow 1 mL/min; Detector UV DAD @210
nm. Retention time 7: 2.47 min. Purity >80% a/a.
2-{4-[(2-Aminoethyl)amino]-2-methylphenyl}-N-methyl-
N-(methyloxy)-2-oxoacetamide 24: isolated by chromato-
graphic column (EtOAc/MeOH 9/1) from the reaction mixture
containing 7 as yellowish oil.
2-(4-Fluoro-2-methylphenyl)-N-methyl-N-(methoxy)-2-
oxoacetamide 21. 20 (100 g, 0.56 mol) was suspended in
anhydrous THF (450 mL) and cooled to 0 °C. The
previously prepared solution of Grignard reagent 19 (0.71
mol) was added through an online filter to the suspension
of 20 over at least 40 min, maintaining a temperature of
0 °C. The resulting pale-green solution was stirred at 0-5
°C for 2-2.5 h. An aqueous solution of NH4Cl 13% w/w
(550 mL) was added, keeping the temperature below 10
°C. After heat ceased to evolve, the temperature was
increased to 25 °C, and the mixture was stirred for 15-30
min at 25 °C, dissolving any solid present. The two phases
were allowed to separate, and the aqueous phase was
discarded. Methanol (1200 mL) was added and the mixture
concentrated under reduced pressure to 600 mL, cooled
to 25 °C, and used as such in the subsequent step.
1H NMR (400 MHz, DMSO-d6): δ 2.57 (s, 3H); 3.28 (s,
3H); 3.58 (s, 3H); 7.25 (dt, J ) 8.38, 2.73 Hz, 1H); 7.31 (dd,
J ) 9.94, 2.53 Hz, 1H); 7.76 (dd, J ) 8.58, 5.85 Hz, 1H); MS
(ES+) m/z 226.1 (MH+).
1H NMR (600 MHz, DMSO-d6): δ 7.30 (d, 1H), 6.84 (t,
1H), 6.47 (d, 1H), 6.45 (s, 1H), 3.55 (s, 3H), 3.21 (s, 3H), 3.09
(m, 2H), 2.69 (t, 2H), 2.45 (s, 3H), 1.85 (d, 2H).
13C NMR (150 MHz, DMSO-d6 derived from the 2D
experiments): δ 170.1, 154.3, 144.2, 137.1, 120.5, 116.3, 110.2,
47.1, 42.4, 32.2, 23.4.
(3S)-3-(4-Fluoro-2-methylphenyl)-2-piperazinone (S)-
mandelate Salt (1:1) 14. The above solution of 7 (∼0.65
mol) was diluted with ethyl acetate (1400 mL), treated
with 60 mL of a 15% w/w Na2CO3 solution, and stirred
at 35 °C for 30 min. The phases were allowed to separate
at this temperature. A further wash with water (40 mL)
was performed at 25 °C. The organic phase was concen-
trated under vacuum to 400 mL. Ethyl acetate (1000 mL)
was added, and the mixture was concentrated again to 400
mL to azeotropically remove water.
HPLC Phenomenex LUNA C18; Mobile phase A: 0.05%
TFA/water and B: 0.05% TFA/acetonitrile; Gradient: 0 min 0%
B to 8 min 95% B. Flow 1 mL/min; Detector UV DAD @210
nm. Retention time 21: 2.47 min. Purity >75% a/a.
3-(4-Fluoro-2-methylphenyl)-5,6-dihydro-2(1H)-pyrazi-
none 6. Methanol (900 mL) was added to the previous solution
of 21 (∼0.69 mol), followed by ethylenediamine 5 (113 mL,
1.69 mol) and glacial acetic acid (119 mL, 2.08 mol), keeping
the temperature at 25 °C. The resulting pale-yellow solution
was heated to reflux for 16-20 h. The solution was concentrated
under reduced pressure to 800 mL and used as such in the
subsequent step.
The resulting suspension was cooled to 25 °C, (L)-
mandelic acid (86 g, 0.56 mol) and 3,5-dichlorosalicyla-
ldehyde 26 (11 g, 0.058 mol) were added, and the mixture
was heated to 43 °C. Initial, almost complete, dissolution
was observed, and crystallisation occurred within 10-30
min. The suspension was aged at 43 °C for 16 h, cooled
to 20 °C, and stirred for a further 2 h.
The mixture was filtered, washed first with ethyl acetate (100
mL) and then with a mixture ethyl acetate/isooctane, 2:1 (2 ×
250 mL), and dried under vacuum at not above 45 °C to obtain
an overall yield of 185.4 g of 14 as ethyl acetate solvate (80.4%
w/w purity, theoretical yield ) 73% from 20).
1H NMR (400 MHz, DMSO-d6): δ 2.21 (s, 3H); 3.38 (t, J
) 5.9 Hz, 2H), 3.79 (t, J ) 6.2 Hz, 2H), 6.99-7.09 (m, 2H),
7.26-7.29 (m, 1H), 8.49 (s, 1H), MS (ES+) m/z 207.1 (MH+).
HPLC Phenomenex LUNA C18; Mobile phase A: 0.05%
TFA/water and B: 0.05% TFA/acetonitrile; Gradient: 0 min 0%
B to 8 min 95% B. Flow 1 mL/min; Detector UV DAD @210
nm. Retention time 6: 2.64 min. Purity >75% a/a.
3-(4-Fluoro-2-methylphenyl)-2-piperazinone 7. Platinum
on charcoal (5% loading, 50% water wet, 4 g, 0.5 mmol) was
added to the previously prepared solution of 6 (∼0.65 mol),
and the resulting suspension was stirred at 40 °C under 2.5 bar
hydrogen pressure for ∼1 h. The mixture was filtered and the
solution concentrated under reduced pressure to 400 mL. Ethyl
acetate (1200 mL) was added and the solution concentrated
under reduced pressure to 400 mL and used as such in the
subsequent step.
1H NMR (400 MHz, DMSO-d6): δ 1.18 (t, J ) 7.10 Hz,
3H); 1.99 (s, 3H); 2.35 (s, 3H); 2.85-2.91 (m, 2H); 2.95-3.00
(m, 1H); 3.14-3.20 (m, 1H); 3.31-3.38 (m, 1H); 4.03 (q, J )
7.10 Hz, 2H); 4.49 (s, 1H); 5.00 (s, 1H); 6.94 (td, J ) 8.56,
2.85 Hz, 1H); 6.98 (dd, J ) 10.10, 2.64 Hz, 1H); 7.24-7.28
(m, 1H); 7.28-7.30 (m, 1H); 7.31-7.36 (m, 2H), 7.40-7.43
(m, 2H), 7.82-7.84 (bm, 1H).
13C NMR (100 MHz, DMSO-d6): δ 174.18, 169.24, 161.09
(d), 140.44, 139.95 (d), 135.26, 130.90 (d), 128.06, 127.54,
126.61, 116.52 (d), 111.68 (d), 72.46, 60.75, 41.97, 41.38, 19.14,
14.40.
IR (nujol): 3604 (OH), 3472, 3182-2000 (NH), 1724
(OdC-NH), 1680 (acid CdO) (cm-1).
HPLC Phenomenex LUNA C18; Mobile phase A: 0.05%
TFA/water and B: 0.05% TFA/acetonitrile; Gradient: 0 min 0%
B to 30 min 95% B. Flow 1 mL/min; Detector UV DAD @210
nm. Retention time 14: 5.78 min. Purity >99% a/a.
HPLC Chiralcel OJ (4.6 mm × 250 mm) from Daicel;
Mobile phase n-hexane/ethanol 80/20 v/v; Flow 1 mL/
min; Detector UV @265nm. Retention times 14: 7.84 min
and opposite enantiomer 7.07 min, ee 99%.
1H NMR (400 MHz, DMSO-d6): δ 2.36 (s, 3H);
2.82-2.89 (m, 1H); 2.94-2.99 (m, 1H); 3.15-3.19 (m,
1H); 3.31-3.37 (m, 1H); 4.46 (s, 1H); 6.94 (dt, J ) 8.56,
2.85 Hz, 1H); 7.00 (dd, J ) 10.10, 2.64 Hz, 1H); 7.26
(dd, J ) 8.34, 2.20 Hz, 1H); 7.77-7.84 (bm, 1H); MS
(ES+) m/z 209.1 (MH+).
[R]2D0j ) +17.9 (1.17, CHCl3).
Vol. 12, No. 6, 2008 / Organic Process Research & Development
•
1193

Acknowledgment
Supporting Information Available
Process safety evaluation of bromo(4-fluoro-2-methylphe-
nyl)magnesium, 19, and chiral HPLC of (3S)-3-(4-fluoro-2-
methylphenyl)-2-piperazinone, 14. This material is available free
of charge via the Internet at http://pubs.acs.org.
We thank Stefano Provera and Lucilla Turco for
spectroscopic characterisation of the compounds and for
the valuable discussion. We thank Giulio Camurri and
Luca Martini for the HPLC methods development, Fran-
cois Ricard for the computational modelling studies, and
Angelo Manzo for his support in defining the pro-
cess.
Received for review June 20, 2008.
OP800146D
1194
•
Vol. 12, No. 6, 2008 / Organic Process Research & Development