Multikilogram-scale synthesis of a biphenyl carboxylic acid derivative using a Pd/C-mediated suzuki coupling approach

Article abstract of DOI:10.1021/op980079g

Reaction of 4-bromo-3-methylaniline with 4-chlorobutyryl chloride/TEA and subsequent treatment of the resulting secondary amide intermediate with KOt-Bu gives 1-(4-bromo-3methylphenyl)pyrrolidin-2-one in 65% yield. This procedure has been optimised (74-76% overall yield) and has been carried out on 41 molar scale. In a variation of this process, we have employed NaOH as the ring-closing base under phase-transfer conditions. NaOH is added to a mixture of 4-bromo-3-methylaniline, 4-chlorobutyryl chloride, and catalytic TBAC in THF/ H₂O. A further 2 equiv of aqueous NaOH is added, and the mixture is heated at 40-45 °C, providing access to cyclised product in an improved 86% yield. 1-(4-Bromo-3-methylphenyl)pyrrolidin-2-one is subsequently coupled with 4-carboxyphenylboronic acid under standard Suzuki coupling conditions [Pd(PPh₃)₄, Na₂CO₃, DME/H₂O] to give 2′-methyl-4′-(2-oxo-1-pyrrolidinyl)biphenyl-4-carboxylic acid in 64% yield, contaminated with 40-80 ppm of residual Pd. In a modification of this process, we have used Pd/C as the catalyst. Reaction in MeOH/ H₂O gives an improved yield of the biphenylcarboxylic acid with residual Pd levels of <6 ppm. This process has been carried out on 24 molar scale. The synthesis of the arylpyrrolidinone and subsequent Suzuki coupling have been combined into a one-pot procedure, providing access to 2′-methyl-4′-(2-oxo-1-pyrrolidinyl)biphenyl-4-carboxylic acid in 82% overall yield from 4-bromo-3-methylaniline.

Full text of DOI:10.1021/op980079g

Organic Process Research & Development 1999, 3, 248−252
Multikilogram-Scale Synthesis of a Biphenyl Carboxylic Acid Derivative Using a
Pd/C-Mediated Suzuki Coupling Approach
David S. Ennis,*,†,1 Julie McManus,† Wladyslaw Wood-Kaczmar,‡ John Richardson,‡ Gillian E. Smith,† and
Alexis Carstairs†
Synthetic Chemistry and Process Technologies, Chemical DeVelopment, SmithKline Beecham Pharmaceuticals,
New Frontiers Science Park, Third AVenue, Harlow, Essex, CM19 5AW, U.K.
Scheme 1
Abstract:
Reaction of 4-bromo-3-methylaniline with 4-chlorobutyryl
chloride/TEA and subsequent treatment of the resulting sec-
ondary amide intermediate with KOt-Bu gives 1-(4-bromo-3-
methylphenyl)pyrrolidin-2-one in 65% yield. This procedure
has been optimised (74-76% overall yield) and has been carried
out on 41 molar scale. In a variation of this process, we have
employed NaOH as the ring-closing base under phase-transfer
conditions. NaOH is added to a mixture of 4-bromo-3-methy-
laniline, 4-chlorobutyryl chloride, and catalytic TBAC in THF/
H2O. A further 2 equiv of aqueous NaOH is added, and the
mixture is heated at 40-45 °C, providing access to cyclised
product in an improved 86% yield. 1-(4-Bromo-3-methylphe-
nyl)pyrrolidin-2-one is subsequently coupled with 4-carbox-
yphenylboronic acid under standard Suzuki coupling conditions
[Pd(PPh3)4, Na2CO3, DME/H2O] to give 2′-methyl-4′-(2-oxo-1-
pyrrolidinyl)biphenyl-4-carboxylic acid in 64% yield, contami-
nated with 40-80 ppm of residual Pd. In a modification of this
process, we have used Pd/C as the catalyst. Reaction in MeOH/
H2O gives an improved yield of the biphenylcarboxylic acid with
residual Pd levels of <6 ppm. This process has been carried
out on 24 molar scale. The synthesis of the arylpyrrolidinone
and subsequent Suzuki coupling have been combined into a one-
pot procedure, providing access to 2′-methyl-4′-(2-oxo-1-pyr-
rolidinyl)biphenyl-4-carboxylic acid in 82% overall yield from
4-bromo-3-methylaniline.
Scheme 2
We now wish to describe the modification, optimisation,
and scale-up of this approach to 2 in order to make supplies
for Phases I and II clinical trials. Evaluation of alternative
syntheses of SB-249930 (6) will also be discussed.
SB-245570 (1), in development for the treatment of
depression, was prepared by the coupling of biphenylcar-
boxylic acid, SB-251475 (2), with indoline 3 (Scheme 1).
We sought an inexpensive and efficient synthesis of key
intermediate 2. The original route used to prepare 2 is
outlined in Scheme 2. 4-Bromo-3-methylaniline (4)2 was
treated with 4-chlorobutyryl chloride and TEA in THF at
15-20 °C to give secondary amide 5 that, on reaction with
KOt-Bu,3 underwent cyclisation to form SB-249930 (6) in
65% isolated yield after recrystallisation from MeOH. A Pd-
(0)-mediated Suzuki coupling reaction of 6 with 4-carbox-
yphenylboronic acid (7) provided access to biphenylcarbox-
ylic acid 2 in 64% yield after EtOH recrystallisation.
Multikilogram-Scale Synthesis of SB-249930 (6). As
described in Scheme 2, arylpyrrolidinone 6 was prepared by
a two-step process from aniline 4. There were two potential
problems associated with scaling-up the current procedure.
First, KOt-Bu was added in solid portions, providing poor
temperature control of the exothermic cyclisation; a more
controlled addition was achieved with a preformed solution
of base in THF. Second, crude product 6 was isolated after
workup by evaporation of EtOAc to dryness. Subsequent
recrystallisation from MeOH gave purified 6. The latter
purification procedure was achieved more efficiently by
azeotropic replacement of EtOAc with MeOH.
† Synthetic Chemistry.
With these problems addressed, we looked at improving
the yield of cyclisation. At 20 °C, KOt-Bu-mediated cycli-
sation provided 6, along with cyclopropyl impurity 8
[8-10% peak area ratio (PAR), HPLC], via a competitive
‡ Process Technologies.
(1) Current address: PR&D Loughborough, Astra Charnwood, Bakewell Road,
Loughborough, Leicestershire, LE11 5RH, UK.
(2) Commercially available from Wychem Ltd., Newmarket, Suffolk, UK.
(3) At least 2 equiv of KOt-Bu was required.
248
•
Vol. 3, No. 4, 1999 / Organic Process Research & Development
10.1021/op980079g CCC: $18.00 © 1999 American Chemical Society and The Royal Society of Chemistry
Published on Web 05/21/1999

Scheme 3
Scheme 4
cyclisation pathway.4 Reaction at 0-5 °C facilitated a more
Scheme 5
Scheme 6
efficient ring closure to give crude 6 contaminated with only
0.8-1.5% of impurity 8. The latter was removed in the
purification step, providing access to SB-249930 (6) (>99.5%
PAR, HPLC) in 74-76% isolated yield.
Scale-up of the modified reaction proceeded smoothly.
Crystallisation, however, was problematic; extremely rapid
precipitation of 6 occurred at 20-25 °C, leading to an
immobile suspension. By seeding the solution at 50-55 °C
with an authentic sample of 6, gradual crystallisation of
product took place reproducibly, and the suspension was
filtered without incident.
This method was successfully employed on the pilot plant
to prepare 2 × 7.5 kg batches of SB-249930 (6) as part of
the preparation of Phase I supplies of 1.
(2.5:1), and the further 2 equiv of NaOH was added in H2O
(appropriate quantity to give a THF/H2O ratio of 1:1). Only
1 mol % of TBAC was required in the reaction. A minor
modification to the workup was required; TBME was used
as the extraction solvent.6 After the usual acid and H2O
washes, TBME was azeotropically replaced by MeOH, from
which crystallisation was effected.
The new procedure was carried out on 100 g scale (with
respect to 4) to give 6 (PAR 100%, HPLC) in 86% yield.
Alternative Approaches to Arylpyrrolidinone 6. An
alternative strategy to prepare 6 was considered involving a
Pd(0)-mediated cross-coupling reaction (Scheme 5). Yama-
moto and Kurata7 have reported Ullmann type couplings of
2-pyrrolidinone with aryl halides. These reactions are,
however, limited in scope, requiring stoichiometric copper
and temperatures in excess of 240 °C. More recently, the
pioneering work of Wolfe and Buchwald8 has facilitated the
coupling of primary and secondary amines with aryl halides
or triflates using catalytic Pd(0) and lower temperatures
(<112 °C).
We have successfully prepared 6 in 58% yield by the
reaction of 5-iodo-2-bromotoluene (1 equiv) and 2-pyrroli-
dinone (1.2 equiv) in the presence of Pd2dba3 (5 mol %),
BINAP (16 mol %), and Cs2CO3 (1.4 equiv) in toluene at
reflux for 20 h (Scheme 6). Use of lower catalyst loadings
led to prolonged reaction times. With NaOt-Bu as the base,
no coupling was observed.
Although this was an efficient transformation, we sought
an alternative base to the expensive, air- and moisture-
sensitive KOt-Bu in THF. We have found that NaOH can
be employed to effect ring closure under phase-transfer type
conditions. The results are reported below.
Initial reaction of secondary amide 5 with NaOH in THF/
H2O at 25 °C gave SB-249930 (6) (96% PAR, HPLC), albeit
over a prolonged reaction time (24 h). In this example,
however, we did not detect the formation of cyclopropyl
impurity 8.
Subsequent use of the automated STEM BLOCK5 allowed
us to rapidly optimise the reaction: in a series of experiments,
we looked at the effect of (a) the proportion of H2O, (b)
temperature, (c) number of equivalents of NaOH, and (d)
the presence of a phase-transfer catalyst. The study culmi-
nated in selection of the conditions described in Scheme 3.
Reaction of 5 with NaOH (2 equiv) in THF/H2O (1:1) in
the presence of TBAC (1 mol %) at 40-45 °C for 1 h
provided access to cyclised product 6 (98% PAR, HPLC) in
96% isolated yield.
We have similarly shown that intermediate 5 can itself
be formed under phase-transfer conditions (Scheme 4).
Treatment of aniline 4 with NaOH (1.03 equiv) and 4-chlo-
robutyryl chloride (1.03 equiv) in the presence of TBAC (1
mol %) in THF/H2O (2.5:1) gave 5 in 97% isolated yield
(99% PAR, HPLC).
This procedure did not, however, compete in terms of
overall yield and efficiency with the current supply route.
Ohta and co-workers9 have reported the synthesis of
N-substituted pyrrolidinones by the reaction of the appropri-
(6) With EtOAc as the extraction solvent, some degree of saponification was
observed. The EtOH thus formed took some product into the aqueous layer,
leading to lower recoveries of 6. As TBME was successfully used as the
extraction solvent, we also explored its use as the reaction solvent in place
of THF. Although the preparation of intermediate 5 was possible, NaOH-
mediated cyclisation to form 6 proved problematic; significant hydrolysis
of 5 was observed under a variety of conditions.
In an extension of this work, we have combined the two
procedures into a “one-pot” preparation of 6 from aniline 4.
Thus, intermediate amide 5 was formed in THF/H2O
(7) Yamamoto, T.; Kurata, Y. Can. J. Chem. 1983, 61, 86.
(4) Cyclopropyl impurity 8 was shown in a control experiment not to be formed
from SB-249930 (6) in the presence of excess KOt-Bu.
(5) Anachem Synthesis Kit 233.
(8) Wolfe, J. P.; Buchwald, S. L. J. Org. Chem. 1997, 62, 6066 and references
therein.
(9) Ohta, S.; Sano, A.; Yamashita, M.; Okamoto, M. Heterocycles 1993, 743.
Vol. 3, No. 4, 1999 / Organic Process Research & Development
•
249

Scheme 7
Scheme 8
Figure 1. Particle size distribution of SB-251475 (2) after
acidification at 20 °C.
ate amine with γ-butyrolactone. In analogy to this work, we
have prepared 6 in 35% yield by heating γ-butyrolactone (1
equiv) and aniline 4 (1.1 equiv) at 180 °C in the presence of
catalytic concentrated HCl (0.3 equiv) for 4 h (Scheme 7).
The purity profile (97.7% PAR, HPLC) and moderate
isolated yield precluded further work in this area.
yield. The balance of material consisted of homocoupled
diacid 9 (0.5% PAR, HPLC) and biphenyl derivative 10
(0.3% PAR, HPLC), the latter being derived from residual
4-tolylboronic acid in substrate (7).16 Initially, purification
In summary, we have developed an inexpensive, efficient,
and robust process for the preparation of SB-249930 (6).
Pd-Mediated Coupling of 6 and 4-Carboxyphenylbo-
ronic Acid (7). Typical Suzuki coupling conditions10 were
originally employed to prepare biphenyl acid 2 (Scheme 2).
Although this was an efficient transformation on 4 molar
scale, we preferred to avoid the use of the expensive and
air-sensitive Pd(PPh3)4 catalyst that can vary in activity from
batch to batch. In addition, material prepared using this
catalyst was contaminated with residual Pd at levels of 40-
80 ppm.
Recently, Pd2dba3,11 Pd(OAc)2,12 and Pd/C13 have been
successfully employed as catalysts in Suzuki coupling
reactions, significantly in the absence of organophosphorus
ligands. Pd supported on charcoal is inexpensive and readily
available and can be removed by Celite filtration at the end
of the reaction. Its use, therefore, should overcome the Pd
contamination problem. To this end, we have prepared SB-
251475 (2) using modified and optimised Buchecker condi-
tions,13 as shown in Scheme 8. Arylpyrrolidinone 6 (1 equiv),
boronic acid 7 (1.06 equiv), Na2CO3 (1.92 equiv), and Pd/
C14 (1.18 mol %) in MeOH/H2O (1:1)15 were heated at reflux
for 5 h. The catalyst was removed by filtration through Celite,
and the resulting solution was acidified with concentrated
HCl, affording crude 2 (>98.5% PAR, HPLC) in 93-96%
of 2 was carried out by recrystallisation from EtOH. The
voluminous nature of this method (approximately 60 volumes
of EtOH was required to effect dissolution), however, led
us to explore a purification by slurry in EtOH. It was found
that heating a suspension of 2 in EtOH (25 volumes) at reflux
for 1 h, cooling to ambient conditions, and isolating gave a
90-93% recovery of product (PAR > 99.3%, HPLC),
identical in purity to that obtained by recrystallisation. EtOH
was subsequently replaced by IMS as the solvent of choice.
Both impurities 9 (0.3% PAR, HPLC) and 10 (0.2% PAR,
HPLC) present in product 2 were removed in the final step
(Scheme 1). In all cases, residual Pd levels in 2 were below
the limit of detection (<4-6 ppm, ICP analysis).
This Pd/C-mediated Suzuki coupling procedure was
successfully carried out in the pilot plant to prepare 2 × 6.3
kg batches of 2.
A processing problem encountered in the plant campaign
to make Phase I supplies had to be addressed. The acidifica-
tion step produced a thick suspension of fine material that
was extremely slow to filter, even with the use of a
centrifuge.
In the laboratory, with the aid of a Lasentec crystallisation
monitor,17 we were able to compare the particle size
distributions in suspensions of 2 formed under different
conditions and, therefore, obtain tangible evidence for
improved filtration. The results of this study are summarised
below.
(10) (a) Miyaura, N.; Yanagi, T.; Suzuki, A. Synth. Commun. 1981, 11, 513. (b)
Suzuki, A. Pure Appl. Chem. 1994, 66, 213.
(11) (a) Moreno-Manas, M.; Pajuelo, F.; Pleixats, R. J. Org. Chem. 1995, 60,
2396. (b) Wallow, T. I.; Novak, B. M. J. Org. Chem. 1994, 59, 5034.
(12) Campi, E. M.; Jackson, W. R.; Maruccio, S. M.; Naeslund, C. G. M. J.
Chem. Soc., Chem. Commun. 1994, 2395.
(13) (a) Marck, G.; Villiger, A.; Buchecker, R. Tetrahedron Lett. 1994, 35, 3277.
(b) Gala, D.; Stamford, J. J.; Kugelman, M. Org. Process Res. DeV. 1997,
1, 163.
(14) The type of Pd/C catalyst used was significant: intermediate catalyst type
487 (from Johnson Matthey), in which the Pd is distributed in the pores
and on the surface, proved much less efficient than type 58 catalyst (eggshell,
from Johnson Matthey), in which the Pd is entirely distributed on the carbon
surface. No other sources of catalyst were investigated at this stage. Lower
loadings (<1.18 molar %) of Pd/C catalyst were studied, but significant
reductions in the rate of reaction were observed.
Figure 1 shows the mean chord length (8.26 µm) of
particles obtained by acidification under standard conditions.
The fines constituted 66% of the suspension, and as a
consequence filtration was slow.
(15) The ratio of MeOH:H2O was critical. Optimum yields of 2 were obtained
with a 1:1 ratio. The reaction was less efficient in EtOH:H2O mixtures of
varying ratio, failing to proceed to completion. In DME:H2O (1:1), no
reaction took place.
(16) 4-Carboxyphenylboronic acid (7) (from MDA) was contaminated with
4-tolylboronic acid (0.9%). The former was prepared by KMnO4 oxidation
of 4-tolylboronic acid.
(17) Lasentec M400L, Lasentec, Redmond, WA.
250
•
Vol. 3, No. 4, 1999 / Organic Process Research & Development

Scheme 9
Figure 2. Particle size distribution of SB-251475 (2) after
acidification at 80 °C.
Experimental Section
General. Melting points were recorded on a capillary
melting point apparatus and are uncorrected. 1H NMR spectra
were recorded on a Bruker 250-MHz instrument. Chemical
shifts for 1H NMR are reported in ppm downfield (δ) relative
to TMS as an internal standard in CDCl3 or DMSO-d6. Mass
spectra were recorded on a Perkin-Elmer Sciex API-III
instrument. Residual heavy metal levels were measured by
inductively coupled plasma atomic emission spectroscopy
(ICP-AES) on a Jobin Yvon 24 spectrometer.
The rate of addition of concentrated HCl and the use of
different acids (concentrated H2SO4 and AcOH) had no
effect. Increasing the alcohol:H2O ratio caused a decrease
in the percentage of fines, and the filtration rate was
improved, but recovered yields of 2 dropped by 4-6%.
Decreasing the alcohol:H2O ratio did not improve the
filterability.
The temperature of acidification was shown to be crucial
(Figure 2). Acidification at 80 °C in MeOH/H2O (1:3) led
to particles with a mean chord length of 26.3 µm, and the
percentage of fines was significantly reduced to 34% (cf.
66% in control reaction). Subsequent filtration proceeded
rapidly, and recoveries were 91-93%. No esterification to
form 11 was observed under these conditions.
Reactions were monitored by HPLC analysis: RP Select
B column, 3:2 ratio of aqueous TFA (0.1%):MeCN mobile
phase at 1 mL/min; UV detection at 215 nm.
1-(4-Bromo-3-methylphenyl)pyrrolidin-2-one (6). (i)
KOt-Bu-Mediated Cyclisation Procedure. 4-Bromo-3-
methylaniline (4) (7.75 kg, 40.82 mol, 98%) in THF (116
L) was treated with TEA (4.25 kg, 42.04 mol) at 20-25 °C
under a nitrogen atmosphere. The solution was then cooled
to 15 °C, and 4-chlorobutyryl chloride (5.98 kg, 42.04 mol,
99.2%) was added slowly over 1 h, maintaining the reaction
temperature between 15 and 20 °C. The resulting suspension
was stirred at 15-20 °C for 1 h.
The reaction mixture was cooled to 0-5 °C, and a
solution of KOt-Bu (11.05 kg, 98.01 mol, 99.5%) in THF
(93 L) (prepared at ambient under nitrogen) was added over
1 h, maintaining the reaction temperature between 0 and 5
°C. The suspension was then stirred at this temperature for
a further 15 min.
On completion, the reaction mixture was diluted with H2O
(51 L) and EtOAc (67 L). The aqueous layer was separated,
and the organic layer was washed sequentially with 2 N HCl
(51 L) and brine (39 L). The THF/EtOAc was then
evaporated to low volume (approximately 230 L removed,
atmospheric distillation). The remaining EtOAc in the
concentrate was replaced by MeOH by a series of azeotropic
distillations. The methanolic solution was then cooled to 50-
55 °C, seeded with an authentic sample of arylpyrrolidinone
6, and cooled to 15-20 °C (crystallisation occurred at 44
°C). The suspension was filtered, and the solid cake was
washed with MeOH (9 L, 0-5 °C) and dried in vacuo (1
mmHg) at 40 °C to constant weight to give SB-249930 (6)
(7.5 kg, 74%) as a white solid: mp 107 °C; δ (CDCl3) 1.57
(s, 3H), 2.15 (m, 2H), 2.60 (t, 2H, J ) 7.9 Hz), 3.84 (t, 2H,
J ) 6.6 Hz), 7.30 (dd, 1H, J ) 1.7 and 6.6 Hz), 7.50 (d, 1H,
J ) 7.9 Hz), 7.52 (d, 1H, J ) 1.7 Hz); MS m/z (relative
The modified acidification procedure was successfully
scaled-up (2 molar scale) and safety tested in preparation
for the next plant campaign.
With regard to the coupling reaction, the level of Na2-
CO3 was reduced to 1.2 equiv (with respect to 6). Further
reduction in the catalytic amount of Pd/C catalyst, however,
proved unsuccessful, as reaction times increased significantly.
One-Pot Preparation of Biphenyl Acid 2 from Aniline
4. We have further optimised the preparation of 2 into a “one-
pot” procedure from bromoaniline 4, summarised in Scheme
9. The latter was transformed to SB-249930 (6). Usual
extractive workup into TBME followed by azeotropic
replacement with MeOH gave a methanolic solution of 6
(96% PAR, HPLC). To the solution were added phenylbo-
ronic acid 7, Pd/C (1.18 mol %), Na2CO3 (1.2 equiv), and
H2O (appropriate volume to give a 1:1 ratio of MeOH:H2O),
and the coupling reaction was carried out in the normal
manner. Purification (IMS slurry) gave 2 of typical purity
(99.3% PAR, HPLC) in 81% overall yield from 4, comparing
favourably with the stepwise process (72% overall yield
from 4).
In conclusion, we have developed an inexpensive, ef-
ficient, robust, and environmentally friendly synthesis of
biphenylcarboxylic acid 2 free of Pd contamination.
Vol. 3, No. 4, 1999 / Organic Process Research & Development
•
251

intensity) 255 (MH+, 100), 253 (MH+, 100); HPLC, 100%
peak area ratio (PAR).
was then dried (MgSO4) and the EtOAc removed in vacuo
to give crude product 6 (6.9 g). Purification by recrystalli-
sation from MeOH (30 mL) gave 6 (2.05 g, 35%) (97.7%
PAR, HPLC).
2′-Methyl-4′-(2-oxo-1-pyrrolidinyl)biphenyl-4-carboxy-
lic Acid (2). SB-249930 (6) (6.20 kg, 24.40 mol), 4-car-
boxyphenylboronic acid (7) (4.35 kg, 25.86 mol, 98.6%),
Na2CO3 (4.97 kg, 46.84 mol, 99.5%), and Pd/C (1.3 kg, 1.18
mol %) in MeOH (31 L) and H2O (31 L) were heated at
reflux (approximately 77.8 °C) for 5 h under a nitrogen
atmosphere.
On completion, the reaction mixture was filtered through
Celite (packed with H2O) at 60 °C, the bed was washed with
warm H2O (40 L, 50 °C), and the filtrate was cooled to 20-
25 °C. Acidification to pH 2 by addition of concentrated
HCl over 1 h gave a thick white precipitate that was stirred
at 20-25 °C for 0.5 h. The suspension was filtered (slow)
and the solid cake washed with H2O (140 L) to give the
crude product that was dried in vacuo (1 mmHg) at 40-45
°C to give 2 (6.89 kg, 96%). Biphenyl acid 2 was suspended
in IMS (172 L) and was heated at reflux for 1 h. The
suspension was then cooled to 20-25 °C and was stirred
for 0.5 h at this temperature. The suspension was filtered,
and the solid cake was washed with chilled IMS (21 L, 0-5
°C) and then dried in vacuo (1 mmHg) at 40-45 °C,
providing 2 (6.28 kg, 91%) as a white solid: mp 253 °C; δ
(DMSO-d6) 2.08 (m, 2H), 2.25 (s, 3H), 2.53 (t, 2H, 6.6 Hz),
3.86 (t, 2H, J ) 6.6 Hz), 7.23 (d, 1H, J ) 7.3 Hz), 7.46 (d,
2H, J ) 7.3 Hz), 7.58 (s, 1H), 7.60 (d, 1H, J ) 7.3 Hz),
7.98 (d, 2H, J ) 7.3 Hz), 12.98 (bs, 1H); MS m/z (relative
intensity) 296 (MH+, 100); HPLC, 99.3% PAR, 99.7% assay;
ICP, <6 ppm Pd,18 <4 ppm B.18
(ii) NaOH-Mediated Cyclisation Procedure. To a solu-
tion of 4-bromo-3-methylaniline (4) (100.0 g, 0.54 mol) in
THF (800 mL, 8 volumes) were added NaOH (22.14 g, 0.55
mol) in H2O (320 mL, 3.2 volumes) and tetrabutylammonium
chloride (TBAC, 1.60 g, 1 mol %) at 20-25 °C. 4-Chlo-
robutyryl chloride (62.53 mL, 0.55 mol, 99.2%) was added
to the biphasic mixture over 0.5 h, maintaining the reaction
temperature at 20-25 °C. After the mixture was stirred at
20-25 °C for a further 15 min, NaOH (43.0 g, 1.07 mol) in
H2O (480 mL, 4.8 volumes) was added over 10 min, and
the resulting reaction mixture was heated at 40-45 °C
for 2 h.
On completion, the reaction mixture was cooled to 20-
25 °C, TBME (900 mL, 9 volumes) was added, and the
solution was stirred for 15 min. The organic layer was
isolated and was washed sequentially with 2 N HCl (200
mL) and brine (400 mL). The organic layer was then
concentrated to 360 mL (3.6 volumes), and the residual
solvent (THF) was replaced with MeOH by a series of
azeotropic distillations (3 × 500 mL). The MeOH solution
was cooled to 20-25 °C, with seeding at 50 °C (to ensure
gradual crystallisation of product). After a further 1 h at 20-
25 °C, the suspension was filtered, and the solid cake was
washed with MeOH (5 mL, 1 volume) and dried in vacuo
(1 mmHg) at 40 °C to give 6 (117.4 g, 86%) as a white
solid: HPLC, 100% PAR.
(iii) Pd(0)-Mediated Coupling Procedure. 2-Bromo-5-
iodotoluene (500 mg, 1.68 mmol), 2-pyrrolidinone (172 mg,
2.02 mmol), Pd2dba3 (77 mg, 5 mol %), BINAP (168 mg,
16 mol %), and Cs2CO3 (770 mg, 2.36 mmol) in toluene (3
mL) were heated at 110 °C for 20 h under a nitrogen
atmosphere. The reaction mixture was cooled to ambient,
diluted with Et2O, and filtered. The filtrate was concentrated
under vacuum to give crude product as an orange solid.
Purification by column chromatography (silica gel, Et2O)
gave 6 (247 mg, 58%) as a beige solid: HPLC, 98.2% PAR.
Condensation of γ-Butyrolactone with Aniline 4. γ-Bu-
tyrolactone (2.5 g, 29 mmol), aniline 4 (5.94 g, 32 mmol),
and concentrated HCl (0.83 mL, 9.6 mmol, 11.6 M) were
heated at 200 °C for 4 h. On cooling, the product set solid.
The latter was extracted with EtOAc (3 × 80 mL), and the
organic fractions were combined and washed sequentially
with 1 N HCl (10 mL) and brine (20 mL). The organic layer
Acknowledgment
We are indebted to Martin Teasdale, Mark Blower, and
Adrian Bateman of Analytical Sciences for their help with
the measurement of residual Pd levels and the development
of HPLC systems. Special thanks also to Steve Buchwald
for a personal communication with regard to the intermo-
lecular coupling of 2-pyrrolidinone.
Received for review October 7, 1998.
OP980079G
(18) Below limit of detection.
252
•
Vol. 3, No. 4, 1999 / Organic Process Research & Development