A highly regioselective preparation of 4-chloromethyl-5-methyl-2-aryl-1,3- oxazoles

Article abstract of DOI:10.1002/adsc.200404135

A facile highly regioselective process is described for the formation of 4-chloromethyl-1,3-oxazoles from 1,3-oxazole N-oxide/HCl salts. An explanation is presented for the high regioselectivity in deoxygenation-chlorination using POCl₃ with HC

Full text of DOI:10.1002/adsc.200404135

FULL PAPERS
A Highly Regioselective Preparation of
4-Chloromethyl-5-methyl-2-aryl-1,3-oxazoles
a
a
b
a,
ˇa
George T. Lee, Xinglong Jiang, T. R. Vedananda, Kapa Prasad, * Oljan Repic
a
Process R&D, Chemical and Analytical Development, Novartis Institute for Biomedical Research, One Health Plaza,
East Hanover, NJ 07936, USA
Fax: (þ1)-973-781-2188, Prasad.kapa@pharma.novartis.com
b
Novartis Institutes for Biomedical Research, Inc., 400 Technology Square, Cambridge, MA 02139, USA
Received: May 12, 2004·Accepted: August 23, 2004
Abstract: A facile highly regioselective process is de- method is quite general and the products are isolated
scribed for the formation of 4-chloromethyl-1,3-oxa- by direct precipitation in all cases studied.
zoles from 1,3-oxazole N-oxide/HCl salts. An explan-
ation is presented for the high regioselectivity in de-
oxygenation-chlorination using POCl3 withHCl salts
Keywords: 4-chloromethyl-1,3-oxazoles; deoxygena-
compared to the corresponding free N-oxides. The tion-chlorination; N-oxide/HCl salts; POCl3
Introduction
and we needed multi-kilo quantities of 4a for making
different active pharmaceutical ingredients. In spite of
Peroxisome proliferator activated receptors (PPAR) the low yield, we embarked on optimizing this approach
are a highly conserved set of ligand activated transcrip- as we believed that this is the most direct way to the tar-
tion factors in the nuclear hormone receptor superfam- geted compounds. These investigations led to a highly
ily. Three distinct PPAR subtypes, PPARa, PPARg and regioselective and practical preparation of the title com-
PPARd, have been identified in most mammalian spe- pounds, and these results are reported in this publica-
cies. These receptors function as sensors of metabolic tion.
state, responding to metabolites and dietary compo-
nents, and triggering the required metabolic adaptation
at the level of the cell, the organ and the organism. We Results and Discussion
embarked on a program[1] to develop agonists of PPARs
that necessitated the development of a short, versatile At the outset, compound 3 was prepared by the conden-
and regioselective synthesis of suitably substituted 4- sation of 1 with 2 in acetic acid/HCl as reported in the lit-
chloromethyloxazoles as alkylating agent. In the litera- erature.[4] The free N-oxide 3 was isolated after quench-
ture[2] they are typically made from condensation of al- ing withammonia and extractive work-up withdichloro-
dehyde 1a with2,3-butanedione monoxime 2 withthe
methane. Compound 3 on deoxygenation-chlorination
intermediacy of compound 3. The second step in this withPOCl in refluxing chloroform yielded a mixture
3
process is a deoxygenation-chlorination step as shown of 4a and its regioisomer 5 in the ratio of 65:35 (Table,
in Scheme 1.
entry 1).[5] The desired 4 was isolated by chromatograph-
The reported yields[3] of 65% for 3 and 30% (after ic purification in 43% yield. The structures of 4 and 5
chromatographic purification) for 4a are rather low, were confirmed by NMR and by a single-crystal X-ray
Scheme 1. Preparation of 4-chloromethyloxazole.
Adv. Synth. Catal. 2004, 346, 1461–1464
DOI: 10.1002/adsc.200404135
ꢀ 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1461

George T. Lee et al.
FULL PAPERS
of the drug substance produced from 4a. As the selectiv-
The dramatic change in the regioselectivity in the
ity obtained is not practical, we investigated the effect of presence of excess chloride ion was a pleasant surprise,
solvent and other additives in the deoxygenation-chlori- but understandable, if one considers the mechanism
nation step. By switching to acetonitrile as the solvent, for the deoxygenation-chlorination process as shown
the selectivity remained unchanged (Table 1, entry 2). in the Scheme 2. In principle, there are two possible
Interestingly by adding tetrabutylammonium chloride pathways. The desired compound 4 can be formed either
to the reaction medium, the selectivity dramatically by an intramolecular pathway (a) or an intermolecular
shifted to>99% (Table 1, entry 3). Assuming the shift pathway (b, not shown in Scheme 2), while the unde-
in selectivity was attributable to the presence of excess sired 5 can form only via pathway (b). In the presence
chloride ion, we studied the use of the N-oxide HCl of excess chloride ion (entries 3 and 4) the pathway (a)
salt 6a, which is the initial intermediate in the condensa- incorporating an intramolecular step predominates
tion of 1a and 2 under these deoxygenation-chlorination thus explaining the increase in regioselectivity.
conditions.
Conclusion
A facile process for a highly regioselective formation of
4-chloromethyl-1,3-oxazoles is described starting from
readily available raw materials. An explanation for ob-
served high regioselectivity in the presence of excess
chloride ion is proposed. The method is quite general
and the products are isolated by direct precipitation in
Subjecting salt 6a to the deoxygenation-chlorination all cases studied.
step withPOCl 3 in acetonitrile, we obtained 4a in higher
than 99% regioselectivity. With this high selectivity in
hand the desired 4a was isolated by precipitation with Experimental Section
water, followed by filtration, in 85% yield. This simpli-
fied the isolation and in turn avoided chromatography.
General Methods
Application of this method to other substituted ben-
zaldehydes gave the 4-chloromethyloxazoles 4b–d in a
highly regioselective manner, and these results are sum- All chemical were obtained from commercial suppliers
marized in the Table 1.
and used without further purification. 1H and
Table 1. Preparation of 4-chloromethyloxazoles.
Entry
Compound
used
Deoxygenation-chlorinating agent/solvent/additive
Ratio[a]
of 4: 5
Product isolated (yield)
1
2
3
4
5
6
7
3
3
3
6a
6b
6c
6d
POCl3 (1.1 equivs.)/CHCl3
65: 35
65: 35
>99/1
>99/1
>99/1
>99/1
>99/1
4a (43% after chromatography)
Not isolated
Not isolated
4a (85%)
4b (86%)
4c (83%)
POCl3 (1.86 equivs.)/CH3CN
POCl3 (1.86 equivs.), CH3CN/Bu4NþClÀ (2 equivs.)
POCl3 (1.86 equivs.)/CH3CN
POCl3 (1.86 equivs.)/CH3CN
POCl3 (1.86 equivs.)/CH3CN
POCl3 (1.86 equivs.)/CH3CN
4d (90%)
[a]
1
Ratio of 4:5 is based on H NMR of crude product mixture.
1462
ꢀ 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2004, 346, 1461–1464

Regioselective Preparation of 4-Chloromethyl-5-methyl-2-aryl-1,3-oxazoles
FULL PAPERS
Scheme 2. Mechanism for the deoxygenation-chlorination process.
145.59, 143.30, 132.25, 131.82, 131.38, 130.95, 130.22, 129.54,
126.60, 126.13, 126.08, 125.98, 125.93, 125.21, 124.96, 122.32,
118.71, 11.47, 6.61; MS (ESI): m/z¼258 (Mþ); anal. calcd. for
C12H11ClF3NO2: C 49.08, H 3.78, N 4.77, Cl 12.07; found: C
49.04, H 3.72, N 4.66, Cl 12.17.
Compounds 6b, 6c and 6d were prepared as above starting
from the corresponding benzaldehydes, and their physical
properties and spectral data are as follows.
13C NMR spectra were recorded at 500 or 300 and at 125
and 75 MHz respectively, in CDCl3 unless otherwise
mentioned. Proton and carbon chemical shifts are ex-
pressed in ppm relative to internal tetramethylsilane;
coupling constants (J) are expressed in Hertz. Melting
points were measured on a Bꢁchi 535 melting point ap-
paratus.
4,5-Dimethyl-2-phenyloxazole 3-oxide hydrochloride
(6b): Yield: 94%; white solid; mp 179–1818C; 1H NMR
(300 MHz, CDCl3): d¼8.32–8.35 (2H, d, J¼0.03 Hz), 7.67–
7.70 (1H, t, J¼0.03 Hz), 7.58–7.63 (2H, t, J¼0.03 Hz), 2.49
(3H, s), 2.47 (3H, s); 13C NMR (300 MHz, CDCl3): d¼154.32,
145.13, 135.02, 129.98, 128.71, 128.50, 119.77, 11.39, 7.56; MS
(ESI): m/z¼190 (Mþ); anal. calcd. for C11H12ClNO2: C 58.54,
H 5.36, N 6.21;, Cl 15.71; found: C 58.46, H 5.54, N 6.26, Cl
15.95.
4,5-Dimethyl-2-(4-fluorophenyl)-oxazole 3-oxide hydro-
chloride (6c): Yield: 91%; white solid; mp 181–1828C;
1H NMR (300 MHz, CDCl3): d¼8.37–8.42 (2H, q, J¼
0.01 Hz), 7.72–7.33 (2H, t, J¼0.03 Hz), 2.49 (3H, s), 2.46
(3H, s); 13C NMR (300 MHz, CDCl3): d¼168.25, 164.81,
153.47, 145.12, 131.61, 131.48, 128.45, 117.79, 117.49, 116.22,
116.18, 11.38, 7.51; MS (ESI): m/z¼208 (Mþ); anal. calcd. for
C11H11ClFNO2: C 54.22, H 4.55, N 5.75, Cl 14.55; found: C
54.08, H 4.22, N 5.67, Cl 15.16.
4,5-Dimethyl-2-(4-trifluoromethylphenyl)-oxazole
3-Oxide Hydrochloride (6a)
Into a solution of 4-trifluoromethylbenzaldehyde (1, 400 g,
2.273 mol) and 2,3-butanedione monoxime (2, 212 g,
2.055 mol) in 1 L of acetic acid at 2–58C, hydrogen chloride
gas (250 g, 6.781 mol) was slowly bubbled at 2–58C over 1 h.
The pale-yellow homogeneous mixture was stirred at 2–58C
for 1 h. Methyl t-butyl ether (3.75 L) was slowly added into
the mixture at 5!258C over 30 min. The resulting white sus-
pension was stirred at 228C for 30 min. The contents were
cooled to 108C and filtered. The filter cake was washed with
2Â250 mL methyl t-butyl ether. The filter cake was dried un-
der vacuum at 558C overnight to give 3d as a white solid; yield:
550 g (91%); mp 182–1848C; 1H NMR (300 MHz, CDCl3): d¼
8.48–8.51 (2H, d, J¼0.03 Hz), 7.85–7.88 (2H, d, J¼0.03 Hz),
2.53 (3H, s), 2.48 (3H, s); 13C NMR (300 MHz, CDCl3): d¼
4,5-Dimethyl-2-(4-chlorophenyl)-oxazole 3-oxide hydro-
chloride (6d): Yield: 84%; white solid; mp 188–1908C;
Adv. Synth. Catal. 2004, 346, 1461–1464
asc.wiley-vch.de
ꢀ 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
1463

George T. Lee et al.
FULL PAPERS
1H NMR (300 MHz, CDCl3): d¼8.28–8.31(2H, d, J¼
0.03 Hz), 7.57–7.60 (2H, d, J¼0.03 Hz), 2.50 (3H, s), 2.45
(3H, s); 13C NMR (300 MHz, CDCl3): d¼153.40, 145.51,
141.62, 130.42, 129.88, 128.63, 118.19, 11.42, 7.48; MS (ESI):
m/z¼224 (Mþ); anal. calcd. for C11H11Cl2NO2: C 50.79, H
4.26, N 5.38, Cl 27.26; found: C 51.06, H 3.75, N 5.32, Cl 27.59.
(3H, s); 13C NMR (300 MHz, CDCl3): d¼160.45, 146.97,
132.28, 130.67, 129.10, 127.59, 126.57, 37.69, 10.78; MS (ESI):
m/z¼208 (MHþ); anal. calcd. for C11H10ClNO: C 63.62, H
4.85, N 6.75, Cl 17.07; found: C 63.39, H 4.63, N 6.39, Cl 17.17.
4-Chloromethyl-5-methyl-2-(4-fluorophenyl)-oxazole
(4c): Yield: 83%; white solid; mp 77–788C; 1H NMR
(300 MHz, CDCl3): d¼7.96–8.01 (2H, m), 7.09–7.15 (2H,
m), 4.54 (2H, s), 2.41 (3H, s); 13C NMR (300 MHz, CDCl3):
d¼166.02, 162.69, 159.61, 146.98, 133.30, 128.72, 128.61,
123.98, 123.93, 116.42, 116.13, 37.60, 10.71; MS (ESI): m/z¼
226 (MHþ); anal. calcd. for C11H9ClFNO: C 58.55, H 4.02, N
6.21, Cl 15.71; found: C 58.66, H 3.94, N 6.24, Cl 15.66.
4-Chloromethyl-5-methyl-2-(4-chlorophenyl)-oxazole
(5c): Yield: 90%; white solid; mp 97.5–98.58C; 1H NMR
(300 MHz, CDCl3): d¼7.90–7.95 (2H, m), 7.27–7.42 (2H,
m), 4.54 (2H, s), 2.41 (3H, s); 13C NMR (300 MHz, CDCl3):
d¼159.50, 147.22, 136.72, 133.49, 129.81, 129.42, 129.10,
128.72, 128.10, 127.83, 126.06, 37.54, 10.77; MS (ESI): m/z¼
242 (MHþ); anal. calcd. for C11H9Cl2NO: C 54.57, H 3.75, N
5.79, Cl 29.29; found: C 54.43, H 3.60, N 5.56, Cl 29.28.
4,5-Dimethyl-2-(4-trifluoromethylphenyl)-oxazole
3-Oxide (3)
Into a suspension of 6a (30 g, 0.102 mol) in 270 mL of water and
300 mL of methylene chloride, concentrated ammonium hy-
droxide (30 mL) was slowly added and the mixture stirred at
228C for 30 min. The bottom organic extract was dried over an-
hydrous magnesium sulfate, filtered, and concentrated under
vacuum to give 4d as a white solid; yield: 25.5 g (97%); mp
147–147.58C; 1H NMR (300 MHz, CDCl3): d¼8.56–8.59
(2H, d, J¼0.03 Hz), 7.72–7.7.75 (2H, d, J¼0.03 Hz), 2.39
(3H, s), 2.2.22 (3H, s).
4-Chloromethyl-5-methyl-2-(4-trifluoromethylphenyl)-
oxazole (4a)
References and Notes
[1] a) A. T. Bach, P. K. Kapa, G. T. Lee, E. M. Loeser, M. L.
Sabio, J. L. Stanton, T. R. Vedananda, Patent WO 03/
043985 A1, 2003; Chem. Abstr. 2003, 139, 6767; b) M.
Boehringer, D. Hunziker, H. Kuehne, B. M. Loeffler, R.
Sarabu, H. P. Wessel, Patent WO 03/037327 A1, 2003;
Chem. Abstr. 2003, 138, 368754; c) Y. Momose, T. Maeka-
wa, T. Yamano, M. Kawada, H. Odaka, H. Ikeda, T. Soh-
da, J. Med. Chem. 2002, 45, 1518–1534 and references cit-
ed therein.
[2] M. S. Malamas, J. Sredy, I. Gunawan, B. Mihan, D. R. Sa-
wicki, L. Seestaller, D. Sullivan, B. R. Flam, J. Med. Chem.
2000, 43, 995–1010 and references cited therein.
[3] Our work was started with the study of 4-trifluoromethyl-
benzaldehyde as the starting material mentioned in the
above reference. The importance of the effect of substitu-
ents on the aromatic ring on the regioselectivity of deoxy-
genation-chlorination of free N-oxides was not addressed
by us. However, the corresponding HCl salts are highly se-
lective.
A suspension of 4,5-dimethyl-2-(4-trifluoromethylphenyl)-ox-
azole 3-oxide hydrochloride (3d, 500 g, 1.702 mol) in 4.06 L of
acetonitrile was stirred at 228C for 30 min and cooled to 108C.
Phosphorus oxychloride (491 g, 3.17 mol) was added at 15–
188C over 15 min. The reaction mixture was warmed up slowly
to 288C by its own exotherm over 1 h and stirred at 28 ! 228C
for 16 h. TLC analysis (5% CH3OH/CH2Cl2) of the reaction
mixture showed no starting material. Water (6.0 L) was slowly
added into the mixture at 10!258C over 1 h. The resulting
white suspension was stirred at 22–288C for 16 h. The contents
were filtered at 228C, and the filter cake was washed with 4Â
500 mL water. The solid was dried under vacuum at 608C over-
night to give 4a as a white solid; yield: 400 g (85%); mp 97–
1
988C; H NMR (300 MHz, CDCl3): d¼8.09–8.12 (2H, d, J¼
0.03 Hz), 7.68–7.71 (2H, d, J¼0.03 Hz), 4.56 (2H, s), 2.45
(3H, s); 13C NMR (300 MHz, CDCl3): d¼159.01, 147.88,
133.88, 132.44, 130.68, 126.77, 126.17, 126.12, 126.07, 37.37,
10.80; MS (ESI). m/z¼276 (MHþ); anal. calcd. for C12H9ClF3
NO: C 52.29, H 3.29, N 5.08, Cl 12.86; found: C 52.13, H 3.29,
N 4.97, Cl 12.85.
[4] P. M. Weintraub, J. Med. Chem. 1972, 15, 419–420 and ref-
erences cited therein.
[5] To our knowledge, no one has addressed the regioselectiv-
ity issue in this type of deoxygenation-chlorinations using
oxazole N-oxides.
Compounds 4b, 4c, and 4d were prepared following the pro-
cedure described for 4a, and their physical and spectroscopic
data are as follows.
4-Chloromethyl-5-methyl-2-phenyloxazole (4b): Yield:
86%; white solid; mp 84–858C; 1H NMR (300 MHz, CDCl3):
d¼7.99–8.02 (2H, m), 7.42–7.45(3H, m), 4.55 (2H, s), 2.42
1464
ꢀ 2004 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
Adv. Synth. Catal. 2004, 346, 1461–1464