The process development of RG 12525 (2-{[4-(Tetrazol-5-ylmethylphenyl)-methoxy]phenoxymethyl}quinoline)

Article abstract of DOI:10.1021/op0000677

This contribution describes process improvements to provide a practical and cost-effective synthesis for the manufacture of RG 12525 which resulted in a 3-fold increase in overall yield. Improved solvent systems for chlorination and azidation reactions are described. Adjustments to the tetrazole-forming step eliminated azide sublimation and minimised this risk on scale-up. A robust solvent system was found to control the polymorphic form during crystallisation, which had hitherto been difficult due to the near-equivalence of melting points (154 and 157 °C) of the two known forms.

Full text of DOI:10.1021/op0000677

Organic Process Research & Development 2001, 5, 9−15
Articles
The Process Development of RG 12525
(2-{[4-(Tetrazol-5-ylmethylphenyl)-methoxy]phenoxymethyl}quinoline)
Andrew W. Bridge,† Ronald H. Jones, Humayun Kabir, Alex A. Kee,* David J. Lythgoe, Mustafa Nakach,‡
Clive Pemberton, and John A. Wrightman
Process DeVelopment, AVentis Pharma Ltd.1, Dagenham Research Centre, Rainham Road South,
Dagenham, Essex RM10 7XS, UK and Centre-Researche Vitry-AlfortVille, 13, Quai Jules Guesde,
BP 14, Vitry-Sur-Seine, 94403 Cedex, France
Abstract:
of this nonisolated intermediate with excess hydroquinone
This contribution describes process improvements to provide
a practical and cost-effective synthesis for the manufacture of
RG 12525 which resulted in a 3-fold increase in overall yield.
Improved solvent systems for chlorination and azidation reac-
tions are described. Adjustments to the tetrazole-forming step
eliminated azide sublimation and minimised this risk on scale-
up. A robust solvent system was found to control the polymor-
phic form during crystallisation, which had hitherto been
difficult due to the near-equivalence of melting points (154 and
157 °C) of the two known forms.
gave the phenol 4. A large excess of R,R′-dichloroxylene
ensured monoalkylation, and then cyanide displacement of
the intermediate chloromethylbenzyl ether 5 gave the pen-
ultimate nitrile 6. This was converted by azide treatment
(NaN3/NH4Cl, DMF) to the tetrazole 1. Although this
1
Introduction
synthesis was short, the following selectivity and industrial
hygiene/safety issues needed to be addressed:
(a) Significant amounts of the dimeric ether 7 are formed
in the hydroquinone treatment of 3.
RG 12525 1 was originally identified as a potent leuko-
triene D4 receptor antagonist2 but was later found to be a
potent PPARγ (peroxisome proliferator-activated receptor)
agonist with an additional indication for late onset diabetes.3
The original synthetic route (Scheme 1) was used to prepare
initial multikilogram quantities for clinical development, but
process improvements were required to make further scale-
up a viable proposition. Alternative routes have also been
described4 but were not deemed suitable, and therefore it
was decided to identify potential improvements to the
existing route. Quinaldine 2 was converted by in situ
chlorination to 2-(chloromethyl)quinoline 3, and treatment
7
(b) An excess of the lachrymatory R,R′-dichloroxylene
was required to minimise dimer formation in the alkylation
of 4.
(c) Sublimation of ammonium azide was observed in the
tetrazole step.
(d) Polymorphism of the final drug substance was difficult
to control.5
(e) Colour and assay quality of the final drug substance
was often variable.
† Please send correspondence to: Andrew Bridge, Aventis Pharmaceuticals,
Inc., Rt. 202-206, Bridgewater, NJ 08807.
‡ Centre-Researche Vitry-Alfortville.
(1) Formerly known as Rhoˆne-Poulenc Rorer Ltd.
(2) Huang, Fu-Chih; Galemmo, Robert A.; Johnson, William H., Jr.; Poli,
Gregory B., Jr.; Morrisette, Matthew M.; Mencel, James J.; Warus, James
D.; Campbell, Henry F.; Nuss, George W.; Carnathan, Gilbert W.; Van
Inwegen, Richard G. J. Med. Chem. 1990, 33, 3, 1194-1200
(3) Jayyosi, Zaid; McGeehan, Gerard M.; Kelley, Michael F. PPAR-g-binding
quinoline derivatives, their preparation, and their therapeutic use. PCT Int.
Appl. 1999, 125 pp. CODEN: PIXXD2 WO 9920275 A1 19990429 CAN
130:320864 AN 1999: 282096.
Stage 1: Chlorination and Subsequent Alkylation of
Quinaldine. Initial synthetic investigations utilising pure
2-(chloromethyl)quinoline hydrochloride (>99.8% norma-
lised peak area by HPLC) established that the colour and
assay problems inherent in the bulk API (active pharmaceuti-
cal ingredient) were associated with the use of crude
(4) O’Brien, M. K.; Sledeski, A. W.; Truesdale, L. K. Tetrahedron Lett. 1997,
38, 509-512; Sledeski, A. W.; O’Brien, M. K.; Truesdale, L. K. Tetrahedron
Lett. 1997, 38, 1129-1132.
(5) (a) Thompson, M. D.; Authelin, J.-R. Chemical Development of the Drug
Substance Solid Form. In Process Chemistry in the Pharmaceutical Industry;
Gadamasetti, K. G., Ed.; Marcel-Dekker: New York, 1999; p 371-388;
ISBN 0-8247-1981-6. (b) Carlton, R. A.; Difeo, T. J.; Powner, T. H.; Santos,
I.; Thompson, M. D. J. Pharm. Sci. 1996, 85, 461-467.
10.1021/op0000677 CCC: $20.00 © 2001 American Chemical Society and The Royal Society of Chemistry
Published on Web 12/30/2000
Vol. 5, No. 1, 2001 / Organic Process Research & Development
•
9

Scheme 1a
Table 1: Extraction composition by HPLC (% npa)
quinaldine monochloro dichlorinated
extract
2
salt 3
8
reaction mixture
1st aqueous extract
organic layer
2nd aqueous extract
organic layer
1.8
3.6
0
1.9
0
91.4
96.0
75.5
97.7
61.3
6.1
0
24.5
0
38.6
equiv) was also found to be important for safety reasons,
since this promotes the formation of oxygen dichloride OCl2
as opposed to the more explosive chlorine dioxide ClO2.8
This water could be conveniently provided by using the
requisite amount of the sodium carbonate monohydrate.
Alternatively the use of sodium bicarbonate precluded the
need for any added water as it is probable that under the
reaction conditions sodium carbonate is initially converted
to sodium bicarbonate. The use of dichloromethane resulted
in chlorination of the solvent to give some chloroform and
carbon tetrachloride. 1,2,4-Trichlorobenzene handled well,
but its high melting point (16 °C) and inclusion on the UK
Red List9 of chemicals led us to consider alternative solvents.
Benzotrifluoride and 4-chlorobenzotrifluoride proved viable
alternatives, but the former was chosen on the grounds of
ease of solvent recovery and cost. Under the established
reaction conditions no chlorination of this solvent was
observed. The optimum temperature was found to be 30-
35 °C; lower temperatures gave a sluggish reaction, whilst
higher temperatures gave higher levels of the dichlorinated
impurity 8. Small-scale experiments suffered from blockage
of the dip tube, but this was alleviated at larger scale by the
higher chlorine gas flow rates.
8
Typically, on bubbling 1.4 equiv of chlorine over 4 h,
the reaction was shown to be complete by HPLC (quinaldine
< 2%) with the monochloro 3-to-dichloro 8 ratio being 92:
5. Any attempt to continue the chlorination beyond this point
would result in lower yield as the product 3 is then
chlorinated faster than the residual quinaldine 2. After
degassing, the inorganics were dissolved by addition of water.
Following charcoal treatment and filtration, extraction of the
organic layer with dilute hydrochloric acid was discovered
to preferentially extract the desired product 3, leaving the
dichloro impurity 8 in the organic phase (Table 1).
The combined aqueous phase was washed with TBME
(to remove any benzotrifluoride, the presence of which causes
the product in the next step to form large agglomerates). This
aqueous phase can then be used directly in the next step and
can be stored at ambient temperature for 48 h without
significant decomposition of 3 (<1%).
a Original conditions: (a) chlorine, Na2CO3, water, 1,2,4-trichlorobenzene;
(b) hydroquinone, methanol, water, NaOH; (c) R,R′-dichloroxylene, NaOH,
tetrabutylammonium hydrogen sulphate, xylene, water; (d) NaCN, xylene, water;
(e) sodium azide, DMF, ammonium chloride, 130 °C; (f) recrystallisation from
methanol.
2-(chloromethyl)quinoline 3. However, this material is a
severe skin irritant, and supply difficulties and the cost of
the material meant that an in situ preparation and purification
of the (chloromethyl)quinoline free base 3 was necessary.
Chlorination of the cheap and readily available quinaldine 2
with chlorine and anhydrous sodium carbonate as base in
either carbon tetrachloride6 or chloroform7 has been reported,
but for health and environmental reasons we chose to
examine alternative solvents. The addition of water (0.5
(8) Hain, J.; McGee, H. A. Ind. Eng. Chem. Prod. Res. DeV. 1984, 23, 490.
(9) Department of the Environment, Transport and the Regions, Greening
Government Operations available from http://www.environment.detr.gov.uk/
greening/greenpro/gguide/gg2annc.htm.
(6) Mathes, W.; Schu¨ly, H. Angew. Chem., Int. Ed. Engl. 1963, 2, 144.
(7) Nakanishi, S.; Saito, T. (Sumika Fine Chemicals Co. Ltd.). Jpn. Kokai
Tokkyo, Koho JP 92-360753, 1994.
10
•
Vol. 5, No. 1, 2001 / Organic Process Research & Development

Table 2: Effect of DCX level
product assay (% w/w)
This purification procedure was found to be essential to
eliminate colour that would otherwise be carried through the
process until the final step. Also polymeric material which
is not detected by HPLC but which is carried through to the
final product is removed in this protocol. As a demonstration
of the validity of the purification regime, commercial
2-(chloromethyl)quinoline hydrochloride (>99.8% norma-
lised peak area by HPLC) was taken through to RG 12525
1 to give product quality comparable to that obtained by the
new process.
equiv of DCX
product 5
dimer 9
yield (%) of 5
2.5
2.0
86
85
11
12
81
71
75
63.6
1.5
1.05
87
75.7
10
19.3
The model favoured a high water content (detrimental to
solubility) and lower temperatures which might cause the
DCX to be present as a solid (mp ) 55 °C). Further
experiments to refine the model produced a product: dimer
ratio of 87.5/12.5 using 2.5 equiv of DCX, 1.05 equiv of
sodium hydroxide, 10 vol of IPA and 7 vol of water. To
minimise raw material, disposal and environmental costs we
were keen to recycle the excess DCX but were unable to
obtain satisfactory recovery. This approach was abandoned
on the grounds of minimising exposure to this material and
the intermediate 5, both being strong irritants. Consequently
it was decided to carry out the next step (reaction with
cyanide) without isolation of the product and to destroy the
excess DCX with cyanide to give R,R′-dicyano xylene 10.
Alkylation of the (chloromethyl)quinoline 3 solution by
hydroquinone (3 equiv) in the presence of excess sodium
hydroxide solution in aqueous methanol was straightforward
and gave a 9:1 ratio10 of the desired phenol 4 to the dimer
7.
The crude product could be purified by dissolution in hot
sodium hydroxide solution (70 °C) and filtration to remove
the insoluble dimer 7. The quality of this product is high
(assay > 99.5%, <0.1% dimer) with a yield of 59% over 2
steps.
Stage 2: Alkylation with R,R′-Dichloroxylene and
Subsequent Cyanation. Initially, alkylation of 4 was carried
out in a xylene and water mixture with tetrabutylammonium
hydrogen sulphate as phase transfer catalyst, sodium hy-
droxide as base, and an excess of R,R′-dichloro-o-xylene
(DCX) (1.7 equiv) to favour monoalkylation. The cyanation
was carried out without isolation to avoid handling the
intermediate 5. This process gave highly coloured product,
low yield and assay necessitating improved conditions. The
main impurity in the alkylation with DCX was the dialkylated
product 9.
This pathway rendered recycling obsolete and increased
the pressure to reduce DCX usage to minimise cost. It was
found that limiting the DCX level to 1.5 equiv gave the best
compromise (Table 2).
Once the reaction was complete, the crude reaction
mixture was reacted with sodium cyanide (2.5 equiv, 3 h at
reflux). The reaction was clean and facile. The product was
isolated by cooling to ambient temperature and filtration. The
crude product which contained the dimer 9 was used without
purification in the next step.
9
Stage 3: Formation of Tetrazole. The tetrazole 1 was
originally obtained using ammonium chloride and sodium
azide in DMF heating at 130 °C for 5 h, but a copious
sublimate of ammonium azide was formed in the condensor.
The explosive potential of this material (shock-sensitive)12
made it prudent to find alternative ways to effect this
transformation. In addition, this solvent led to a protracted
work-up involving dilution with water and multiple extrac-
tions.
The replacement of ammonium chloride by triethylamine
hydrochloride reduced the sublimate, and this held true using
NMP as solvent. The most successful procedure involved
in situ generated triethylammonium acetate (by addition of
triethylamine and acetic acid) which gave an equivalent
reaction rate in NMP. The reaction could be applied in xylene
but the limited solubility of the tetrazole as sodium salt made
the work-up more difficult.
The first modification increased the DCX to 2 equiv and
utilised a 2-propanol/water mix (3:1) at 60 °C to give product
5/dimer 9 ratio of 70/30. Attempts to reduce the formation
of dimer 9 by slow addition of 4 to DCX failed.
A statistical model (using MODDE)11 was set up to
examine the important parameters with the goal of maximiz-
ing product/dimer ratio. The model used 19 experiments
taking DCX, temperature, excess of sodium hydroxide, water,
and 2-propanol as factors. A reasonable fit was initially
obtained, but the prediction fit was poor. Subsequent
elimination of insignificant factors and factor interactions
led to a good fit for predictions. This showed that the
favoured conditions were:
(a) large excess DCX,
(b) minimum NaOH excess,
(c) maximum water content, and
(d) low temperature.
(12) The relationship between the kinetic data of the low-temperature thermolysis
and the heats of explosion of inorganic azides. Zeman, Svatopluk; Dimun,
Milan; Truchlik, Stefan; Kabatova, Viera. Thermochim. Acta 1984, 80 (1),
137-41.
(10) Based on isolated yields.
(11) MODDE is a registered trademark of UMETRI AB, Box 7960, S-907 19
Umeå, Sweden.
Vol. 5, No. 1, 2001 / Organic Process Research & Development
•
11

Figure 1. Solubility and supersaturation vs temperature in methanol.
Table 3: Effect of solvent on tetrazole formation
(>99.0% w/w by HPLC) after only a single recrystallisation.
However polymorphism was a major issue. It has been shown
that RG 12525 exists as two polymorphs.5 Form I (needles,
mp 154 °C) had been selected as the form required for
development over Form II (rhombs, mp 157 °C) on the basis
of stability.
Recrystallisation from methanol, the original choice, was
associated with several drawbacks:
(a) The solubility of RG 12525 in methanol is poor
needing 30 volumes of solvent to achieve solution.
(b) Attempts to recover a second crop from methanol were
unsuccessful.
LC-MS analysis of the liquors showed a high content
of a compound corresponding to the addition of methanol
to the product 1. Efforts to produce this compound by boiling
RG 12525 in methanol alone proved fruitless, indicating that
the process is probably catalysed by charcoal used in the
procedure.
solvent
yield (%)
assay (% w/w)
DMF
NMP
xylene
cyclohexanone
acetophenone
anisole
72.5
92.4
89.0
89.6
93.2
95.4
89.9
99.0
98.5
92.8
88.8
98.6
99.1
99.0
hexan-1-ol
Cyclohexanone, acetophenone, anisole, and hexan-1-ol
were found to facilitate complete reaction and also avoid
formation of sublimate (See Table 3).
Anisole was chosen as optimal solvent because it provides
a number of advantages over the other alternatives:
(a) a reaction temperature of 130 °C and a 6 h reaction
period,
(b) sublimation was avoided,
(c) it is a recoverable solvent,
Pilot plant batches achieved recrystallisation returns of
only 80% and these contained up to 4% of Form II. Solubility
studies in methanol (Figure 1) showed that at low super-
saturation (i.e., the region above the supersaturation curve)
both Form I (F1) and Form II (F2) are present during primary
nucleation (i.e., in the absence of seeding) so that there will
be no control of polymorphic form on crystallisation.
Solubility studies identified 10% aqueous acetone (i.e., 1
vol of water/9 vol of acetone) as a better solvent with this
combination giving the maximum solubility at 50 °C. Only
Form I (F1) is produced during primary nucleation in the
region of supersaturation so that control of polymorph is
possible by selection of solvent. The return on material was
85% (See Figures 2 and 3). In addition, a much higher
throughput could be achieved as only 10 vol of solvent were
required leading to adoption as the solvent of choice. A
comparison between methanol and acetone/water is shown
in the Van’t Hoff diagram (Figure 4).
(d) immisibility with water facilitates removal of inor-
ganics leaving the product in anisole,
(e) the stability of the solvent to base allows the organic
solvent to be added directly to hot alkali for extraction of
product as the sodium salt into the aqueous layer, leaving
the dimer 9 in the anisole layer,
(f) the solvent is unreactive towards azide.
This combination provided the product in the best yield
and quality.
Stage 4: Recrystallisation. With the original overall
synthetic procedure, which did not involve an extraction into
acid of the (chloromethyl)quinoline at stage 1, it was
necessary to employ at least two recrystallisations of crude
product 1 to achieve the requisite specification both in
colour13 and assay.
The improvements described in this paper gave product
that was now of a good colour specification and assay
(13) Colour as determined by solution colour; Y6 or higher was acceptable.
12
•
Vol. 5, No. 1, 2001 / Organic Process Research & Development

Figure 2. Solubility vs % of water in acetone at 50 °C.
Figure 3. Solubility vs temperature in acetone-water (90-10 v/v).
As a final comparison the yields before and after the
described process improvements are given in Table 4.
NMR spectra were recorded on a Varian Unity Inova 400
spectrometer in CDCl3 or d6-DMSO solutions and TMS as
internal standard for calibration, mass spectra on a Bruker
Daltonics BioApex 30esICR spectrometer, infrared on a
Nicolet 20SXB spectrometer, the sample being dispersed in
a KBr disk. Elemental analysis were conducted using a Carlo
Erba 1108 elemental analyser.
Conclusions
A route suitable for scale-up of RG 12525 has been
described with a 3-fold improvement in yield. Benzotrifluo-
ride was demonstrated to be a solvent suitable for chlorina-
tion reactions. Extraction of the chlorination product into
aqueous acid was found to be the key to downstream
improvements in colour and assay. The use of statistical
methods and the need to minimise the usage of dichloroxy-
lene gave the optimal conditions for step 2. Anisole in
combination with acetic acid/triethylamine/sodium azide was
established as the best conditions for tetrazole formation in
this system. Finally, a single recrystallisation with 10%
aqueous acetone provided a high return and complete control
over the polymorphic form.
HPLC used a Hewlett Packard HP-1100 using a Hichrom
NC100-5C18-2874 column. The mobile phase was acetoni-
trile/water/phosphoric acid (70/30/0.1) with detection at 230
nm, and flow rate at 1 mL/min.
DSC was run on a Mettler DSC20 in aluminium crucibles
with a heating rate of 6 °C min-1.
4-(Quinolin-2-yl)methoxyphenol, 4. Quinaldine (25.0 g,
Fluka > 90.0%, 175 mmol), benzotrifluoride (150 mL) and
sodium carbonate (10.2 g, 96.3 mmol) and sodium carbonate
monohydrate (11.9 g, 96.3 mmol) were charged to a 250
mL round-bottom flask and the mixture stirred (200 rpm)
Experimental Section
General. Reactions were carried out under a nitrogen
atmosphere except in the case of the chlorination.
Vol. 5, No. 1, 2001 / Organic Process Research & Development
•
13

Figure 4. Vant’Hoff diagram.
Table 4: Summary of effect of process changes on yield
shows <1% of 2-(chloromethyl)quinoline) and then allowed
to cool to room temperature. The reaction mixture was
filtered and washed with water (2 × 150 mL). The damp
solid (24.7 g) was returned to a 250 mL flask containing
sodium hydroxide (5.2 g) in water (158 mL) and heated to
70 °C for 15 min. The suspension was then filtered (yield
of dimer (7) ) 2.2 g, 6.3%) and washed with water (30 mL).
The combined aqueous filtrate was returned to the reaction
vessel and cooled to <30 °C. Concentrated hydrochloric acid
(20.2 g, 37%, 205 mmol) was added over 1 h, keeping the
temperature <30 °C (pH ) 5). The product was filtered,
washed with water (40 mL), and dried in an oven at 55 °C.
Yield ) 26.0 g (59.2%).
original conditions
new
stage 1a
stage 2b
stage 3c
stage 4d
total
54.6% (91.0% w/w)
45.9% (59.6% w/w)
68.8% (99.0% w/w)
75.5% (99.5%)
9.7%
57. 0% (99.0% w/w)
75.0% (87.0% w/w)
94.9% (99.5% w/w)
85.0% (99.9% w/w)
34.9%
a Summary of new conditions. Stage 1: (a) Benzotrifluoride, 2, sodium
carbonate monohydrate (1.1 equiv) and chlorine (1.4 equiv), 35 °C- 40 °C.
(b)1 M HCl (1.4 equiv). (c) Hydroquinone (3 equiv), methanol, water, aq sodium
hydroxide (4.4 equiv), 3. b Stage 2: (a) R,R′-Dichloro-o-xylene (1.5 equiv), 4,
sodium hydroxide, water, 60 °C. (b) Sodium cyanide (2.5 equiv), 2-propanol,
reflux. c Stage 3: Anisole, 6, triethylamine (2.2 equiv), acetic acid (2.2 equiv),
sodium azide (1.6 equiv), 130 °C, 7 h. d Stage 4: 1, Acetone:water (9:1),
acticarbone 3SA.
1H NMR (CDCl3) 5.3 (2 H, s), 6.7-6.9 (4 H, AB quartet),
7.6 (1 H, t), 7.7-7.8 (2 H, 2t), 7.85 (1 H, dd), 8.05 (1 H, d),
8.2 (1 H, d), 8.25 (1 H, s, OH).
and heated to 35 °C. Chlorine (20.3 g, 286 mmol)14 was
bubbled into the reaction mixture via a subsurface dip-tube
over a period of 180 min, ensuring that the temperature
remained between 35- 40 °C. After the reaction was
complete (by HPLC quinaldine < 2.0%), nitrogen was
bubbled through the mixture for 1 h. To the reaction mixture
was added water (130 mL) followed by charcoal (acticarbone
ENO, 1.54 g), and the mixture stirred for 15 min. The
mixture was filtered through a pad of Hyflo (12.0 g) and
washed with benzotrifluoride (2 × 25 mL). The organic layer
was separated and extracted with 1 M hydrochloric acid (175
mL then 75 mL), and the combined aqueous layer was
washed with TBME (15 mL) to give a solution of 2-(chlo-
romethyl)quinoline hydrochloride (267.8 g).
Hydroquinone (57.7 g, 524 mmol) was charged to a 1 L
reaction flask containing methanol (174 mL) and water (125
mL) under nitrogen atmosphere and the resulting solution
cooled to 10 °C. To the reaction mixture was added a solution
of sodium hydroxide (30.9 g, 773 mmol) in water (31 mL)
over 10 min, followed by the solution of the 2-(chloro-
methyl)quinoline hydrochloride salt (267.8 g) over 5 min.
The reaction mixture was heated to 50 °C for 3 h (HPLC
13C NMR (CDCl3) 71.5, 115.9, 116.0, 119.5, 126.5, 127.3,
128.6, 129.8, 136.9, 147.1, 151.1, 151.8, 158.2; IR (cm-1)
3600-2500, 1515, 1460, 1231, 1200, 1064; MS 252 (M +
H, 100%).
C16H13NO2 requires C 76.48, H 5.21, N 5.57. Found C
76.08, H 5.12, N 5.69%.
Dimer 7: 1H NMR (DMSO) 5.30 (4 H, s), 7.01 (4 H, s),
7.62 (2 H, m), 7.66 (2 H, d), 7.78 (2 H, m), 7.99 (2 H, d),
8.01(2 H, d), 8.41 (2 H, d); IR (cm-1) 3100-2900 (C-H),
1505 (CdC), 1429, 1241(C-O), 1225 (C-O); MS (ES+)
393 (M + H, 55%), 250 (100%), 143 (25%).
2-[4-(Quinolin-2-yl-methoxy)phenoxy]methylphenyl Ac-
etonitrile, 6. A 2 L flask was charged with R,R′-dichloro-
o-xylene (78.4 g, 448 mmol) and 2-propanol (225 mL) and
the mixture heated to 50 °C to give a solution. 4-(Quinolin-
2-yl)methoxyphenol 4 (75.0 g, 299 mmol) was charged
followed by a solution of sodium hydroxide (13.8 g, 345
mmol) in water (510 mL) and the mixture heated to 60 °C
for 2.5 h. (HPLC analysis shows 4-(quinolin-2-yl)methox-
yphenol 4 < 1.0%). Sodium cyanide (36.6 g, 746 mmol)
was added to the reaction mixture followed by 2-propanol
(450 mL) and the mixture heated to reflux (84 °C) for 4 h.
(14) Of this quantity 3.03 g was collected in a sodium hypochlorite trap so that
only 17.27 g of chlorine (243 mmol, 1.39 equiv) is required for reaction.
14
•
Vol. 5, No. 1, 2001 / Organic Process Research & Development

(HPLC shows stage 2A < 1.0%). The batch was cooled to
70 °C and seeded with 2-[4-(quinolin-2-yl-methoxy)phenoxy]-
methylphenyl acetonitrile 6 (0.56 g). Then it was cooled to
25 °C over 5 h and stirred at this temperature for 16 h. The
material was filtered and washed with 2-propanol (450 mL)
and hot water (70 °C, 750 mL) and then dried in an oven at
60 °C to constant weight.
phase was added to a hot (80 °C) solution of sodium
hydroxide (6.45 g, 157 mmol) in water (360 mL) and stirred
for 15 min. The aqueous phase was separated, keeping the
temperature at 60-65 °C. The solution was washed with
anisole (3 × 100 mL). The aqueous solution was cooled to
52 °C and then acidified to pH 5 by the addition of water/
hydrochloric acid (3:1, 60 mL) over 40 min. The product
was filtered, washed with water (2 × 100 mL), and dried to
constant weight at 55 °C.
Yield ) 98.2 g, 75.3% corrected yield.15
1H NMR (DMSO) 4.09 (2 H, s), 5.09 (2 H, s), 5.32 (2 H,
s), 6.97-7.05 (4 H, m), 7.36-7.41 (2 H, m), 7.46-7.50 (2
H, m), 7.62 (1 H, t), 7.67 (1 H, d), 7.79 (1 H, t), 8.00 (2 H,
t), 8.41 (1 H, d); IR (cm-1) 3100-2900 (CH), 2220 (CN),
1507 (CdC), 1217 (C-O), 1070 (C-O), MS: 381 (M+ +
H, 100%), 272 (15%).
Dimer 9: 1H NMR (DMSO) 5.14 (4 H, s), 5.30 (4 H, s),
6.95-7.00 (8 H, m), 7.34 (2 H, m), 7.49 (2 H, m), 7.61 (2
H, dt), 7.66 (2 H, d), 7.78 (2 H, m), 7.99 (2 H, d), 8.02 (2
H, d), 8.40 (2 H, d); IR (cm-1) 3100-2900 (CH), 1510 (Cd
C), 1250 (C-O), 1075 (C-O), MS 627 (M + Na+, 100%),
605 (M + H, 58%).
Yield ) 26.2 g (90.2%, corrected for assay). Assay )
99.5% w/w.
Recrystallisation. Crude RG 12525 1 (10.0 g), acetone
(90 mL), and water (10 mL) were heated to 58 °C to ensure
complete dissolution. Then charcoal (acticarbone 3SA, 0.5
g) was added and the mixture heated to reflux for 20 min.
The mixture was cooled to 58 °C and then filtered, rinsing
with 10% acetone/water (5 mL), and the filtrate was heated
to 55 °C. RG12525 Form 1 (0.1 g) was added as seeds and
the mixture cooled to 20 °C over 1 h. The product was
isolated on a filter, washed with 10% acetone/water (2 ×
10 mL), and dried to constant weight in an oven at 55 °C.
Yield ) 8.4 g (84.3%). Assay ) 99.9%. Mp ) 153.5 °C
(Form 1).
2-[4-(2-chloromethylphenylmethoxy)phenoxymethyl]-
1
quinoline, 5: H NMR (DMSO) 4.88 (2 H, s), 5.18 (2 H,
s), 5.32 (2 H, s), 6.99-7.05 (4 H, m), 7.35-7.40 (2 H, m),
7.49 (2 H, m), 7.62 (1 H, m), 7.67 (1 H, d), 7.79 (1 H, dt),
8.00 (2 H, t), 8.41 (1 H, d); IR (cm-1) 3100-2900 (C-H),
1500 (CdC), 1250 (C-O); MS 392(M + H, 55%), 390 (M
+ H, 100%).
1H NMR (DMSO) 4.40 (2 H, s), 5.05 (2 H, s), 5.30 (2 H,
s), 6.9-7.0 (4 H, AB quartet, ArH), 7.22 (1 H, q), 7.30 (2
H, m), 7.5 (1 H, d), 7.63 (1 H, t), 7.70 (1 H, d), 7.77 (1 H,
t), 8.0 (2 H, dd), 8.4 (1 H, d).
2-{[4-(Tetrazol-5-ylmethylphenyl)-methoxy]phenoxy-
methyl}quinoline (RG 12525) 1. Anisole (85 mL) and 2-[4-
(quinolin-2-yl-methoxy)phenol]methylphenyl acetonitrile 6
(30.0 g) (assay ) 87% w/w) were charged to a 500 mL flask
and heated to 60 °C under nitrogen. Triethylamine (16.1 g,
160 mmol) was added, followed by acetic acid (9.6 g, 160
mmol), then sodium azide (7.62 g, 116 mmol) and the
mixture heated to 130 °C for 7 h. (HPLC analysis indicates
(6) < 4.0%). The reaction mixture was cooled to 90 °C, and
then water (100 mL) was added, stirring for 5 min at 60-
70 °C. An additional charge of anisole (25 mL) was added
and the aqueous phase separated. The organic phase was
washed with water (100 mL) and separated. The organic
13C NMR 26.0, 67.9, 71.2, 115.6 (x2), 119.5, 126.5, 127.1,
127.3, 127.9, 128.4, 128.5, 129.8 (×2), 134.5, 135.3, 136.9,
146.9, 152.4, 155.0, 157.8; IR (cm-1) 3140 (NH), 3060 (CH),
2900 (CH), 1507 (CdC), 1217 (C-O), 1070 (C-O); MS
(ES) 424 (M + H, 100%); C25H21N5O2 requires C 70.91, H
5.00, N 16.54. Found C 70.83, H 4.78, N 16.60%.
Acknowledgment
We thank Ron Cook, Kiran Virji, Peter Shade, Dave
Thomas, and Mike Podmore for analytical support and Drs.
Michel Mulhauser and Jean-Re´ne Authelin for stimulating
discussions.
Received for review June 15, 2000.
OP0000677
(15) The isolated yield represents 86.5%; the assay of the material is 87.6%;
hence, corrected yield ) 86.5 × 0.876 ) 75.7%
Vol. 5, No. 1, 2001 / Organic Process Research & Development
•
15