A new approach to the rapid parallel development of four neurokinin antagonists. Part 3. Assembly of neurokinin antagonists

Article abstract of DOI:10.1021/op020066+

Four neurokinin antagonists were assembled using a rapid parallel development approach. Research Department processes were scaled up if process safety and robustness were not compromised. Using this approach, 1 kg of each compound was rapidly delivered fo

Full text of DOI:10.1021/op020066+

Organic Process Research & Development 2003, 7, 67−73
A New Approach to the Rapid Parallel Development of Four Neurokinin
Antagonists. Part 3. Assembly of Neurokinin Antagonists
Jeremy S. Parker,* Sharon A. Bowden, Catherine R. Firkin, Jonathan D. Moseley, Paul M. Murray, Matthew J. Welham,
Richard Wisedale, Maureen J. Young, and William O. Moss
AstraZeneca, Process Research and DeVelopment, AVlon Works, SeVern Road, Hallen, Bristol BS10 7ZE, UK
Abstract:
Four neurokinin antagonists were assembled using a rapid
parallel development approach. Research Department processes
were scaled up if process safety and robustness were not
compromised. Using this approach, 1 kg of each compound was
rapidly delivered for clinical trials.
Introduction
Zeneca Pharmaceuticals devised a rapid parallel develop-
ment approach for the delivery of 1 kg of a compound for
preliminary toxicity and clinical studies. This approach was
applied to the preparation of a series of four neurokinin
antagonists: ZD6021 1, ZD2249 2, ZD4979 3, and ZM374979
4.1
Figure 1.
Results and Discussion
The first transformation required in this sequence was the
coupling of the cyano acids 5-7 to ZD6021 N-methylamine
fumarate 8, forming the corresponding amide alcohols 11-
13. ZD6021 N-methylamine fumarate 8 was converted to
ZD6021 N-methylamine 14 by washing with aqueous sodium
bicarbonate solution. The coupling was then achieved by
activating the cyano acids 5-7 and reacting these with
ZD6021 N-methylamine 14 (Scheme 1).
The assembly of these neurokinin antagonists was un-
dertaken from a number of advanced intermediates.1 For this
For ZD6021 amide alcohol 11, the Discovery conditions
involving the preparation of the ZD6021 acid chloride 15
were used. Thus, ZD6021 cyano acid 5 was reacted with
oxalyl chloride, catalysed by DMF, affording the intermediate
ZD6021 acid chloride 15 (Scheme 2), which was then further
reacted with ZD6021 N-methylamine 14 to afford ZD6021
amide alcohol 11. Concerns about gas evolution from the
acid chloride formation led to a two-batch campaign, the
first being run on a smaller scale (0.86 mol) to investigate
the control of gas evolution during the reaction. In the event,
control of gas evolution was not problematic, and the size
of the reaction was increased to 2.43 mol for the second
batch. The combined batches of product afforded 1.54 kg
of crude ZD6021 amide alcohol 11 which was used without
further purification in the subsequent stage.
work, kilogram quantities of ZD6021 cyano acid 5,2,3
ZD4974 cyano acid 6, and ZM374979 cyano acid 74 were
prepared, together with ZD6021 N-methylamine fumarate 8,
ZD7944 pip sulfoxide 9, and ZD2249 methoxy sulfoxide 105
(Figure 1). The preparation of a number of these molecules
will be discussed in subsequent papers in this series.
This contribution describes the assembly of the four
neurokinin antagonists from these advanced intermediates,
using the rapid parallel development approach.
(1) Moseley, J. D.; Moss, W. O. Org. Process Res. DeV. 2002, 6, 53-57.
(2) Moseley, J. D.; Moss, W. O.; Welham, M. J.; Ancell, C. L.; Banister, J.;
Bowden, S. A.; Norton, G.; Young, M. J. Org. Process Res. DeV. 2002, 6,
58-66.
(3) Ashworth, I.; Bowden, M. C.; Dembofsky, B.; Levin, D.; Moss, W. O.;
Robinson, E.; Szczur, N.; Virica, J. Org. Process Res. DeV. 2002, 6, 74-
81.
(4) Parker, J. S.; Smith, N. A.; Welham, M. J.; Moss, W. O. Org. Process Res.
DeV. Manuscript in preparation.
(5) Bowden, S. A.; Burke, J. N.; Gray, F.; McKown, S.; Moseley, J. D.; Moss,
W. O.; Murray, P. M.; Welham, M. J. Org. Process Res. DeV. Manuscript
in preparation.
This procedure, although effective at preparing ZD6021
amide alcohol 11, raised concerns due to the potential
production of carcinogenic dimethyl carbamyl chloride
10.1021/op020066+ CCC: $25.00 © 2003 American Chemical Society
Published on Web 12/17/2002
Vol. 7, No. 1, 2003 / Organic Process Research & Development
•
67

Scheme 1
Table 1. Values of parameters for selection of acid chloride
formation
temperature
stoichiometry
(mol equiv)
catalyst
(°C)
DMF
NMP
DMAP
0
20
40
0.05
0.075
0.01
However, on careful analysis of the crude product, a new
impurity 18 was observed, which was slow-running on HPLC
and thus had not been identified previously. This impurity,
typically at a level of 10%, was shown to inhibit the
crystallisation of the desired ZD6021 amide alcohol 11; thus,
work was undertaken to reduce the level of the impurity 18
in the crude product. Tituration of the crude, the use of
cleaner input materials, and changing the order of addition
and the amounts of the reagents were all investigated without
significant effect. Eventually it was decided to hydrolyse the
ester bond of the impurity, converting this material to
ZD6021 amide alcohol 11. Thus, treatment of an MTBE
solution of the crude product with sodium hydroxide solution
at 45 °C reduced the impurity level to 0.3% in 3 h. The
crude material obtained after extraction of the hydrolysis
reaction mixture was found to contain ZD6021 amide alcohol
11 at 88% strength. This material was successfully recys-
tallised from acetonitrile/water to afford 996 g of ZD6021
amide alcohol 11 at 90% strength. This represents a yield
of 62%.
Scheme 2
(DMCC).6 This concern led to investigations into the use of
an EDCI coupling to effect this transformation for subsequent
batches of ZD6021 amide alcohol 11, which were being
prepared for the second compound in the series, ZD2249 2.
Initial investigations involved the addition of ZD6021
N-methylamine 14 in dichloromethane to a stirred slurry of
ZD6021 cyano acid 5, EDCI, and triethylamine in dichloro-
methane.7 This resulted in the formation of ZD6021 amide
alcohol 11 (56%) but with the accompaniment of two
impurities (27% and 7%). These were identified by LC-
MS as the two possible N-acyl ureas, 16 and 17, generated
by rearrangement of the initially formed O-acyl urea.
For delivery of the subsequent compounds in this series,
and considering the problems identified during the previous
preparations, it was decided to return to the preparation of
acid chlorides to effect the coupling reaction to prepare amide
alcohols 12 and 13, for the preparation of ZD4974 3 and
ZM374979 4. Alternative reaction conditions were sought
for this transformation, particularly focusing on changing the
reaction catalyst from DMF to avoid the issue of DMCC.
Factorial experimental design (FED) work was carried out
to investigate this change, evaluating NMP and DMAP as
alternative catalysts for the reaction. The parameters evalu-
ated are shown in Table 1.
The amounts of these impurities could be reduced by the
slow addition of a solution of EDCI to a solution of ZD6021
cyano acid 5 and ZD6021 N-methylamine 14, with the
addition of HOBt to the reaction mixture. This procedure
was operated on a large scale, affording the required ZD6021
amide alcohol 11 with the impurities at low levels (1 and
0.5%, respectively).
These parameters were evaluated in a nine-experiment
Latin square factorial design as shown in Table 2.
As expected, DMF gave the best-quality product in the
shortest reaction time. However, NMP at 40 °C gave
comparable quality and yield, after a slightly longer reaction
time, and this was selected as the catalyst for further
reactions. The reaction stoichiometry and temperature were
then further optimised using FED. The parameters evaluated
are shown in Table 3.
(6) Levin, D. Chem. Ind. 1997, 2; Levin, D. Chem. Br. 1997, 33, 20.
(7) Estenne, G.; Leclerc, G. J. Heterocycl. Chem. 1994, 31, 1121.
68
•
Vol. 7, No. 1, 2003 / Organic Process Research & Development

Table 2. Results of Latin square factorial design for
selection of acid chloride formation
Table 6. Results of fractional factorial design for
optimisation of amide formation
temp stoichiometry corrected reaction
N-methyl amine
addition
time
addition
temp
corrected
yield (%)
experiment catalyst (°C)
(mol equiv) yield (%) time (h)
expt
charge
1
2
3
4
5
6
7
8
9
DMF
NMP
DMAP
DMF
NMP
DMAP 20
DMF
NMP
0
0
0
20
20
0.05
0.075
0.10
0.075
0.10
0.05
0.10
0.05
0.075
97.4
24.4
4.3
97.5
91.4
40.7
98.4
96.2
96.2
18
19
18
1
19
22
0.16
0.66
6.5
1
2
3
4
-1
-1
1
-1
1
-1
1
-1
1
1
96.6
95.6
95.0
96.5
1
-1
Scheme 3
40
40
DMAP 40
Table 3. Values of parameters for optimisation of acid
chloride formation
parameter
low level
high level
NMP charge
oxalyl chloride charge
temperature
0.10 mol equiv
1.0 mol equiv
30 °C
0.15 mol equiv
1.1 mol equiv
reflux
Scheme 4
Table 4. Results of fractional factorial design for
optimisation of acid chloride formation
NMP oxalyl chloride
corrected
reaction
expt charge
charge
temp yield (%) time (min)
1
2
3
4
-1
-1
1
-1
1
-1
1
-1
1
1
98.4
97.6
96.9
98.3
300
80
80
1
-1
300
facture. The temperature effect was minimal and was set at
ambient for convenience.
Table 5. Values of parameters for optimisation of amide
formation
These optimised conditions were utilised for the prepara-
tion of both ZD4974 amide alcohol 12 and ZD374979 amide
alcohol 13 (Scheme 3). The crude ZD4974 amide alcohol
12 was purified by recrystallisation from tetrahydrofuran:
isohexane, affording 3.67 kg of material with a typical
strength of 96%, representing a 95% yield. Crude ZM374979
amide alcohol 13 (954 g) was manufactured and used without
further purification in the subsequent stage.
The produced amide alcohols 11-13 now required
coupling to the appropriate piperidine (9 or 10) to afford
the crude products 21-24. This was achieved by oxidation
of the amide alcohols 11-13 to aldehydes 25-27, followed
by a reductive amination with appropriate piperidine (9 or
10) (Scheme 4).
The Discovery procedures for these transformations were
investigated for the preparation of ZD6021 crude 21.
Two initial concerns with the Discovery methodology
were the need to conduct the Swern oxidation at low
temperatures and containment of the dimethyl sulfide gener-
ated during the reaction. However, as appropriate equipment
was available in the Macclesfield Large Scale Laboratory
to address these concerns, it was decided to attempt scale-
up of the process. The process utilised by Discovery involved
the addition of DMSO to a solution of oxalyl chloride in
dichloromethane at -78 °C. In practice, this led to the
DMSO freezing on the side of the vessel; consequently, the
procedure was modified to involve addition of DMSO as a
parameter
low level
high level
N-methyl amine charge
addition time
addition temperature
1.0 mol equiv
30 min
10 °C
1.1 mol equiv
60 min
30 °C
These parameters were evaluated in a four-experiment
fractional factorial design as shown in Table 4.
Temperature was shown to be the dominant parameter,
with a slower reaction occurring at the lower temperature.
However, the reaction was restricted to a maximum of 30
°C for manufacture, given concern for the hazard of gas
evolution during charging of oxalyl chloride to the batch at
reflux. The effect of the other factors was negligible; thus,
the stoichiometry was fixed at 1.2 mol equiv oxalyl chloride
and 0.1 mol equiv of NMP on the basis of trends.
As an extension to this work, it was decided to optimise
the coupling reaction between the produced acid chlorides
19 and 20 and ZD6021 N-methylamine 14. This also was
achieved using FED studies, the parameters of which are
shown in Table 5.
These parameters were evaluated in a four-experiment
fractional factorial design as shown in Table 6.
With minimal difference between the results, the predicted
best combination of 1.1 mol equiv ZD6021 N-methylamine
14 added over 30 min was selected for large-scale manu-
Vol. 7, No. 1, 2003 / Organic Process Research & Development
•
69

Table 7. Values of parameters for optimisation of Swern
oxidation
Table 9. Values of parameters for optimisation of reductive
amination
parameter
low level
high level
parameter
low level
0.9 equiv
high level
1.1 equiv
oxalyl chloride/DMSO charge 1.5/3 mol equiv 3/6 mol equiv
amide alcohol addition time
hold time before Et3N addition 30 min
Et3N charge
pip sulfoxide charge
CH2Cl2 charge for pip
sulfoxide solution
water charge for pip
sulfoxide solution
order of addition
10 min
60 min
60 min
7.5 mol equiv
30 min
0 vols
0.7 vols
0 vols
0.7 vols
4.5 mol equiv
5 min
Et3N addition time
aldehyde to
pip sulfoxide
(normal)
22 °C
pip sulfoxide to
aldehyde
(inverse)
32 °C
Table 8. Results of fractional factorial design for
optimisation of Swern oxidation
reaction temperature
hold time before
borane-pyridine addition
borane-pyridine charge
0 h
1 h
oxalyl
amide hold time
chloride/ alcohol
DMSO addition
before
Et3N
Et3N corrected
0.8 equiv
1.2 equiv
Et3N addition
addition charge time
yield
(%)
expt charge
time
12 and triethylamine were also beneficial to a lesser extent,
but were more difficult to achieve on scale-up for a reaction
requiring -60 °C.
The reductive amination was evaluated by a similar
approach. The parameters evaluated for the optimisation of
reductive amination stage are shown in Table 9. The addition
of the water or dichloromethane was investigated to improve
the solubility of the ZD7944 pip sulfoxide 9.
1
2
3
4
5
6
7
8
-1
-1
-1
-1
1
1
1
-1
-1
1
-1
1
-1
1
1
-1
1
-1
1
-1
1
-1
1
-1
1
-1
1
1
-1
1
-1
-1
1
69.7
68.1
64.0
69.7
73.0
77.0
71.4
70.0
1
-1
-1
1
1
1
-1
These parameters were evaluated in a 16-experiment
fractional factorial design (resolution III) using a Benchmate
robot, as shown in Table 10
These results showed that ZD7944 pip sulfoxide 9 and
the water charges were the most important parameters. A
higher ZD7944 pip sulfoxide 9 charge gave improved
conversion at the end of reaction. In contrast the presence
of water was detrimental to the reaction.
The optimised conditions identified in these experiments
were applied to the preparation of 3.67 kg of ZD4974 crude
23, with a yield of 79% achieved over the two steps from
ZD4974 amide alcohol 12, and product strength of 85%.
These conditions were then applied, without modification,
to the preparation of ZM374979 crude 24, successfully
producing 1.38 kg of product, with a yield of 87% achieved
over the two steps from ZM374979 amine alcohol 13 and
product strength of 82%.
The final step required for delivery of the final active
pharmaceutical ingredients was a crystallisation of crude
materials as amine salts. The purity target for these final
stages was to produce material with strength greater than
95%.
For the crystallisation of ZD6021 hydrogen fumarate pure
1, the Discovery recrystallisation conditions were used,
involving dissolution in hot ethanol and the addition of
fumaric acid. The purity of the material obtained from this
stage was 90%, whereas Discovery had produced ZD6021
hydrogen fumarate pure 1 with a strength >98% using the
same procedure.
The Discovery synthesis had made extensive use of
chromatography and had produced ZD6021 crude 21 with
strength of 98.7%. In contrast, the material produced by
Development varied in strength from 65 to 75%, the material
having been carried though five synthetic steps without
crystallisation or chromatography.
solution in dichloromethane. The reaction was shown to
operate well between -60 and -45 °C, but incomplete
reaction was observed at -20 °C.
Due to concerns about the stability of the produced
aldehyde, it was decided not to purify this material, using
the crude product in the reductive amination stage. The
reducing reagent was successfully changed from sodium
cyanoborohydride used by Discovery, to the less toxic
borane-pyridine. The reaction was first investigated in
methanol but a competing side reaction occurred, thought
to be the formation of the dimethyl acetal. To avoid this
problem the reaction solvent was changed to dichloro-
methane.
With both stages for the formation of the ZD6021 crude
21 being conducted in dichloromethane, it was decided to
investigate telescoping the two stages together. This proce-
dure proved to be effective and was utilized for the delivery
of 2.06 kg of ZD6021 crude 21, with a yield of 76% achieved
over the two steps from ZD6021 amide alcohol 11 and
product strength of 68% (product contained 25% residual
dichloromethane).
An analogous procedure was used for the delivery of 1.95
kg of ZD2249 crude 22, with a yield of 79% achieved over
the two steps from ZD6021 amide alcohol 11 and product
strength of 75%.
For the preparation of ZD4974 crude 23, extensive FED
work was carried out to optimise and demonstrate the
robustness of the oxidation. The parameters evaluated for
the optimisation of oxidation stage are show in Table 7.
These parameters were evaluated in an eight-experiment
fractional factorial design (resolution III) as shown in Table
8.
This work showed that high charges of oxalyl chloride,
DMSO, and triethylamine provided the most effective
reaction. Short addition times for the ZD4974 amide alcohol
70
•
Vol. 7, No. 1, 2003 / Organic Process Research & Development

Table 10. Results of fractional factorial design for optimisation of reductive amination
pip
sulfoxide
CH2Cl2
charge
water
charge
order of
addition
hold
time
borane-
pyridine
solution
conversion (%)
expt
temp
1
2
3
4
5
6
7
8
-1
-1
-1
-1
-1
-1
-1
-1
1
1
1
1
1
-1
-1
-1
-1
1
1
1
-1
-1
1
-1
-1
1
1
1
-1
1
-1
1
-1
1
-1
1
-1
1
-1
1
-1
1
-1
1
-1
1
-1
1
1
-1
1
-1
1
-1
1
-1
-1
1
-1
1
-1
1
1
-1
-1
1
1
-1
1
77.8
64.0
35.3
26.0
56.9
64.4
23.6
52.2
80.9
85.1
22.1
67.0
83.6
80.1
23.9
31.1
1
-1
-1
1
1
-1
-1
1
1
1
9
-1
-1
-1
-1
1
1
1
-1
-1
1
10
11
12
13
14
15
16
1
-1
-1
1
-1
-1
-1
-1
1
1
-1
-1
1
1
1
1
1
-1
-1
1
1
1
1
Thus, a further purification step was required, and a
retreatment of semi-pure material was developed involving
recrystallisation from an ethanol:water mixture. This proce-
dure increased the strength of the material to 94.8%, which
was judged to be of sufficient purity to use in toxicity trials.
The retreatment yielded 980 g of ZD6021 hydrogen fumarate
pure 1, which represents a yield of 59%.
ZM374979 crude 24 was also purified by dry flash
column chromatography, increasing the purity to 95.5%
with a 75% material recovery (1.03 kg, mixture of four
atropisomers). The semi-pure material was then crystallised
as the maleate salt from hot methanol, affording 237 g of
ZM374979 maleate pure 4 (single atropisomer), with a
strength of 94.7%, which was judged to be of sufficient purity
for toxicity trials. The maleate salt formation and crystalli-
sation yield was 23%.
The learning from the preparation of ZD6021 hydrogen
fumarate pure 1 was applied to the development of a process
for ZD2249 hydrogen fumarate pure 2. Extensive screening
of crystallisation solvents showed that an ethanol:water
mixture remained the most effective solvent system for this
transformation, which would increase purity to ca. 92%.
Retreatment of this material was considered, but due to low
recoveries from the screening experiments it was decided to
investigate the use of slurry washes. The use of hot aqueous
ethanol washes were shown to give an increase in purity
with low losses to liquors, as was the use of ethyl acetate
washes. These techniques were combined, and the ZD2249
hydrogen fumarate pure 2 was crystallised from ethanol:water
and then slurry-washed with hot ethanol and ethyl acetate,
affording 950 g of product with a strength of 94.8%. As with
ZD6021 hydrogen fumarate pure 1, this material was
considered to be of sufficient purity to use in toxicity trials.
The yield for this stage was 75%.
The crystallisations of both ZD4974 citrate pure 3 and
ZM374979 maleate pure 4 were complicated by the existence
of atropisomerism in these molecules. The issue raised by
the atropisomerism and the development work carried out
on selective atropisomer crystallisation will be discussed in
a subsequent paper.4
With the problems previously encountered with purifica-
tion of both ZD6021 and ZD2249 hydrogen fumarate pures
1 and 2, it was decided to purify ZD4974 and ZM374979
crudes 23 and 24 using flash column chromatography. Thus,
ZD4974 crude 23 was purified by dry flash column chro-
matography, improving the purity to 97.3% with a 72%
material recovery. This semi-pure material was crystallised
as the citrate salt from hot ethanol, affording 1.22 kg of
ZD4974 citrate pure 3, with a strength of 97.3%. The citrate
salt formation and crystallisation yield was 66%.
Conclusions
The use of the rapid parallel development approach
allowed delivery of 1 kg of each of the neurokinin antagonists
(ZM374979 as a mixture of atropisomers) providing a further
example of the success of this approach.1
The key factors of robustness and safety remained the
highest priorities during this work, robustness being dem-
onstrated by extensive factorial experimental design work
and the safety by adequate hazard evaluation.
As previously noted, the focus of this approach is on the
short-term delivery of these compounds. For longer-term
manufacture the use of a Swern oxidation would not be
viable. However, for the manufacture of 1-kg quantities, the
use of Research procedures, or a modification of these, has
been successful in affording the necessary products within
the scope of the rapid parallel development approach.
Experimental Section
General Procedures. Melting points were determined
using a Griffin melting point apparatus (aluminium heating
block) and are uncorrected. 1H NMR spectra were recorded
on a Varian Inova 400 MHz spectrometer with chemical
shifts given in ppm relative to TMS at δ ) 0. Electrospray
(ES) mass spectra were determined on a Micromass Platform
LC. The reaction mixtures and products were analysed by
reverse phase HPLC on a Hewlett-Packard 1100 according
to the following conditions: column, Waters Spherisorb
S5ODS1, 250 mm × 4.6 mm i.d.; eluent, 560:440 acetoni-
trile:water with 0.1% v/v trifluoroacetic acid; flow rate, 1.0
mL/min; wavelength, 235 nm; injection volume, 5 µL,
column temperature, 25 °C. HPLC purities were %w/w
Vol. 7, No. 1, 2003 / Organic Process Research & Development
•
71

1
against a standard of known strength as determined by H
86.1%, tR 12.7 min (single atropisomer); mp 131-132 °C;
1H NMR (399.895 MHz, d6-DMSO) 1.25 (3H, t, J ) 7.4 Hz,
CH2CH3), 1.77-1.98 (2H, m, Cl2PhCHCH2CH2), 2.54-2.59
(3H, s, NCH3), 2.63-2.76 (1H, m, CHH′CH3), 2.80-2.93
(1H, m, CHH′CH3), 3.23-3.53 (4H, m, Cl2PhCHCH2CH2,
Cl2PhCH and CH3NCHH′), 4.51-4.65 (2H, m, CH3NCHH′
and OH), 6.39 (1H, d, J ) 8.5 Hz, ArH), 7.40 (1H, t, J )
7.4 Hz, ArH), 7.52 (1H, dd, J ) 8.5, 1.8 Hz, ArH), 7.62-
7.70 (1H, m, ArH), 7.72-7.81 (2H, m, ArH), 8.03-8.11
(1H, m, ArH), 8.63 (1H, s, ArH); MS (ES+) 459 (8, MH+,
37Cl2), 457 (41, MH+, 35Cl37Cl), 455 (100, MH+, 35Cl2).
Preparation of ZM374979 Maleate Pure 4. A 1-L
round-bottomed flask (previously oven-dried and cooled to
ambient temperature under a nitrogen purge) was equipped
with an overhead stirrer, a pressure-equalising additions
funnel, and a nitrogen inlet/outlet; it was then purged with
nitrogen for 10 min. Dichloromethane (125 mL) and oxalyl
chloride (14.35 mL, 165 mmol, 3.00 equiv) were then
charged to the flask. This solution was stirred and cooled to
-70 °C, at which temperature it was held for 15 min. A
solution of dimethyl sulfoxide (23.35 mL, 329 mmol, 6.00
equiv) in dichloromethane (45 mL) was charged to the
additions funnel and then added dropwise at a rate such that
the batch temperature was maintained below -65 °C. On
completion of addition, the mixture was stirred at -70 °C
for 30 min. A solution of ZM374979 amide alcohol 13 (25.00
g, 54.9 mmol, 1.00 equiv) in dimethyl sulfoxide (100 mL)
and dichloromethane (137 mL) was charged to the additions
funnel and then added dropwise at a rate such that the batch
temperature was maintained below -60 °C. On completion
of addition, the mixture was stirred at -70 °C for 30 min.
Triethylamine (57.10 mL, 411.7 mmol, 7.50 equiv) was
charged to the additions funnel and then added over a 5-min
period, maintaining the reaction temperature below -60 °C.
The reaction was stirred at -70 °C for 4 h and then warmed
slowly to ambient temperature. Hydrochloric acid (1 M,
aqueous, 275 mL) was added; the mixture was stirred for
15 min and then left for 15 min for the layers to separate.
The lower, organic phase was removed and retained, the
aqueous phase being discarded. The organic solution was
recharged to the flask, and hydrochloric acid (1 M, aqueous,
275 mL) was added. The mixture was stirred for 15 min
and then left for 15 min for the layers to separate. The lower,
organic phase was removed and retained, the aqueous phase
being discarded. The organic solution was recharged to the
flask, and sodium hydrogen carbonate solution (saturated,
aqueous, 275 mL) was added. The mixture was stirred for
15 min and then left for 15 min for the layers to separate.
The lower, organic phase was removed and retained, the
aqueous phase being discarded. The organic solution was
recharged to the flask, and sodium chloride solution (satu-
rated, aqueous, 137 mL) and water (137 mL) were added.
The mixture was stirred for 15 min and then left for 15 min
for the layers to separate. The lower organic phase was
removed and retained, the aqueous phase being discarded.
A 1-L round-bottomed flask was equipped with an overhead
stirrer, a pressure-equalising additions funnel, and a nitrogen
inlet/outlet. ZD7944 pip sulfoxide 9 (15.91 g, 71.4 mmol,
NMR. Analytical TLC was carried out on commercially
prepared plates coated with 0.25 mm of self-indicating Merck
Kieselgel 60 F254 and visualised by UV light at 254 nm.
Preparative scale silica gel flash chromatography was carried
out by standard procedures using Merck Kieselgel 60 (230-
400 mesh).
Preparation of ZM374979 Amide Alcohol 13. A 250-
mL round-bottomed flask was equipped with overhead stirrer,
condenser, thermometer, and nitrogen inlet/outlet and then
was purged with nitrogen for 10 min. ZM374979 cyano acid
7 (12.00 g, 53.3 mmol, 1.00 equiv) was then charged,
followed by dichloromethane (120 mL). The resulting brown
slurry was stirred, and NMP (0.51 mL, 5.33 mmol, 0.10
equiv) was charged in a single portion. The mixture was then
heated to 30 °C and stirred at this temperature for 30 min.
Oxalyl chloride (5.70 mL, 64.0 mmol, 1.20 equiv) was added
over a 30-min period, during which time effervescence was
observed. On completion of addition, the batch was held
overnight at 30 °C, affording a dark brown solution; the
reaction mixture was then cooled to ambient temperature.
A 1000-mL round-bottomed flask was equipped with con-
denser, overhead stirrer, additions funnel, and thermometer.
ZD6021 amine fumarate (17.95 g, 29.33 mmol, 0.55 equiv)
was charged to this flask with sodium carbonate (16.96 g,
160.0 mmol, 3.00 equiv), dichloromethane (240 mL), and
water (240 mL). This mixture was stirred at ambient
temperature for 30 min, affording a biphasic system, with a
white, cloudy, lower organic phase. The acid chloride
solution was added dropwise over a 30-minute period and
then the additions funnel was washed with dichloromethane
(12 mL). The reaction mixture was stirred at ambient
temperature for 2 h. Agitation was then stopped, and the
layers were allowed to settle for 15 min. The lower organic
phase was removed, dichloromethane (72 mL) was added,
and the mixture was stirred for 15 min. Agitation was then
stopped, and the layers were allowed to settle for 15 min.
The lower organic phase was removed and combined with
the other organic portions, the aqueous portion being
discarded. The organic solution was recharged to the flask,
and hydrochloric acid (1 M, aqueous, 96 mL) was added.
The mixture was stirred for 15 min and then left for 15 min
for the layers to separate. The lower organic phase was
removed and retained, the aqueous phase being discarded.
The organic solution was recharged to the flask, and sodium
hydrogen carbonate solution (saturated, aqueous, 96 mL) was
added. The mixture was stirred for 15 min and then left for
15 min for the layers to separate. The lower organic phase
was removed and retained, the aqueous phase being dis-
carded. The organic solution was recharged to the flask, and
sodium chloride solution (saturated, aqueous, 96 mL) was
added. The mixture was stirred for 15 min and then left for
15 min for the layers to separate. The lower organic phase
was removed and the solvent removed using a rotary
evaporator, affording a brown foam. This material was trans-
ferred to a vacuum oven at 40 °C and dried to a constant
weight, affording ZM374979 amide alcohol 13 as a brown
solid (25.00 g, 89% corrected for strength). HPLC purity
72
•
Vol. 7, No. 1, 2003 / Organic Process Research & Development

1.30 equiv) was charged to this vessel with methanol (100
mL). The mixture was stirred, and the solution of ZM374979
aldehyde 27 was added to the flask. The mixture was stirred
for 15 min, and then borane-pyridine (5.55 mL, 54.9 mmol,
1.00 equiv) was added dropwise over a 30 min period. The
reaction was stirred for 20 h, and then hydrochloric acid (1
M, aqueous, 275 mL) was added cautiously to the mixture.
The mixture was stirred for 15 min a nd then left for 15 min
for the layers to separate. The lower, organic phase was
removed and retained, the aqueous phase being discarded.
The organic solution was recharged to the flask, and
hydrochloric acid (1M, aqueous, 275 mL) was added. The
mixture was stirred for 15 min and then left for 15 min for
the layers to separate. The lower, organic phase was removed
and retained, the aqueous phase being discarded. The organic
solution was recharged to the flask, and sodium hydrogen
carbonate solution (saturated, aqueous, 275 mL) was added.
The mixture was stirred for 15 min and then left for 15 min
for the layers to separate. The lower organic phase was
removed and retained, the aqueous phase being discarded.
The organic solution was recharged to the flask, and sodium
chloride solution (saturated, aqueous, 137 mL) and water
(137 mL) were added. The mixture was stirred for 15 min
and then left for 15 min for the layers to separate. The lower,
organic phase was removed and the solvent removed using
a rotary evaporator, affording ZM374979 crude 24 as a
brown foam (36.90 g, 93% corrected for strength). HPLC
purity 91.7%, tR 26.2 min (single atropisomer).
A 10-cm diameter sinter funnel (grade 3) was packed with
silica (244.3 g, 12 times crude weight loading, depth 68 mm)
slurried in 5% methanol in ethyl acetate. The solvent level
was run down to the top of the silica. ZM374979 crude 24
(20.36 g, 30.85 mmol) was dissolved in 5% methanol in ethyl
acetate (25 mL), and this solution was carefully applied onto
the column by pipet. Quartz sand (20.4 g, 1 times crude
weight loading) was carefully applied to the column, giving
a level surface. Methanol (5%) in ethyl acetate (20 mL) was
run down the side of the column to wet the sand, and then
the column was filled with the same solvent, taking care not
to disturb the surface of the sand. The column was eluted,
collecting 100-mL fractions, ensuring that the solvent flow
through that column was not too fast. When collected
fractions contained only product spots by thin layer chro-
matography analysis (fraction 23), the eluent was changed
to 5% methanol in dichloromethane. Elution was continued
until no further product was detected in the collected fractions
(fraction 40). The product fractions were combined, and the
solvent was removed using a rotary evaporator, affording
ZM374979 semi-pure as a yellow foam (14.11 g, 66%
corrected for strength). HPLC purity 95.6%, tR 26.2 min
(single atropisomer).
semi-pure (5.00 g, 7.58 mmol, 1.00 equiv) was charged with
methanol (20 mL). The mixture was heated to 60 °C and
then held at this temperature for 48 h. A 100-mL round-
bottomed flask was equipped with overhead stirrer, con-
denser, and thermometer. A solution of maleic acid (0.89 g,
7.58 mmol, 1.00 equiv) in methanol (5 mL) was charged
to this second flask, and this solution was heated to 40 °C.
The solution of ZM374979 semi-pure was cooled to 40 °C
and then transferred to the second flask via an in-line filter.
The first flask was washed with methanol (15 mL), and
this wash was transferred to the second flask via the in-line
filter. The resulting solution was stirred at 40 °C for 1 h and
then cooled to 23 °C over a 1-h period. The solution was
stirred at 23 °C for 1 h, and then seed crystals of ZM374979
maleate pure 4 (0.5 mg, 6.0 × 10-4 mmol, 8.5 × 10-5 equiv)
were added. The solution was then held for a further 24 h
at 23 °C, after which time the mixture was a thick, white
slurry. The product was then isolated by filtration, and
the flask and cake were washed with methanol (5 mL). The
solid was pulled dry using vacuum and then washed with
methanol:isohexane (1:1, 5 mL). The solid was pulled dry
using vacuum and then washed with isohexane (5 mL). The
solid was pulled dry using vacuum and then transferred to a
vacuum oven at 30 °C and dried to constant weight, affording
ZM374979 maleate pure 4 as a white, crystalline solid (1.41
g, 23% corrected for strength). HPLC purity 94.7%, tR 26.2
min (single atropisomer); mp 170-172 °C; 1H NMR
(399.895 MHz, d6-DMSO, 80 °C) 1.23 (3H, t, J ) 7.2 Hz,
CH2CH3), 1.86-2.18 (6H, m, MeSOPhCH(CH2)2 and Cl2-
PhCHCH2CH2), 2.54 (3H, s, NCH3), 2.63-2.88 (6H, m,
PhSOCH3, CH2CH3 and MeSOPhCH), 2.99-3.16 (4H, m,
Cl2PhCHCH2CH2 and MeSOPhCH(CH2CHH′)2), 3.34-3.36
(1H, m, Cl2PhCH), 3.44-3.51 (2H, m, MeSOPhCH(CH2-
CHH′)2), 3.58-3.63 (1H, m, CH3NCHH′), 4.66 (1H, t, J )
12.8 Hz, CH3NCHH′), 5.67-6.58 (1H, m, ArH), 7.36-7.44
(2H, m, ArH), 7.50-7.62 (4H, m, ArH), 7.71 (1H, d, J )
8.4 Hz, ArH), 7.77 (1H, s, ArH), 7.88 (1H, d, J ) 7.6 Hz,
ArH), 8.01 (1H, d, J ) 8.0 Hz, ArH), 8.53 (1H, s, ArH);
MS (ES+) 664 (30, MH+, 37Cl2), 662 (65, MH+, 35Cl37Cl),
660 (100, MH+, 35Cl2).
Acknowledgment
Manufacture was performed in the Macclesfield LSL by
John Banister, Nigel Burke, Steve Knight, Steve McKown,
Glenn Norton, and Kevin Vare. Analytical support was
provided by Steve Baldwin, Nathalie Roberts, and Matt
Young. Hazard Studies were performed by Steve Hallam and
Paul Gillespie. We thank Bob Osborne (PR&D Avlon) for
the use of the Benchmate robot and both him and Rob Shaw
(Macclesfield) for help with the FEDs and statistical analysis.
Received for review July 25, 2002.
OP020066+
A 50-mL round-bottomed flask was equipped with
overhead stirrer, condenser, and thermometer. ZM374979
Vol. 7, No. 1, 2003 / Organic Process Research & Development
•
73