Large-scale preparation of 2-methyloxazole-4-carboxaldehyde

Article abstract of DOI:10.1021/op700198s

The large-scale preparation of 2-methyloxazole-4-carboxaldehyde presents a significant challenge due to the physical characteristics of the molecule. A method for the preparation of 10-kg batches of 2-methyloxazole-4-carboxaldehyde is described. The key reaction is the reduction of the corresponding N-methoxy-N-methyl amide using lithium aluminium hydride, followed by workup and isolation by crystallization.

Full text of DOI:10.1021/op700198s

Organic Process Research & Development 2008, 12, 88–95
Large-Scale Preparation of 2-Methyloxazole-4-carboxaldehyde
Georges-Emmanuel Benoit, John S. Carey,* Alan M. Chapman, Ranjit Chima, Nigel Hussain, Matthew E. Popkin,
Guillaume Roux, Bahareh Tavassoli, Carine Vaxelaire, Michael R. Webb, and David Whatrup
Synthetic Chemistry, GlaxoSmithKline Pharmaceuticals, Old Powder Mills, Leigh, nr. Tonbridge, Kent TN11 9AN, U.K.
Scheme 1
Abstract:
The large-scale preparation of 2-methyloxazole-4-carboxaldehyde
presents a significant challenge due to the physical characteristics
of the molecule. A method for the preparation of 10-kg batches
of 2-methyloxazole-4-carboxaldehyde is described. The key reac-
tion is the reduction of the corresponding N-methoxy-N-methyl
amide using lithium aluminium hydride, followed by workup and
isolation by crystallization.
To support the development of one of our drug candidate
molecules, we have had the requirement for many tens of
kilograms of 2-methyl-oxazole-4-carboxaldyde 1 as a starting
material. The supply base for electron-rich five-membered
heterocyclic starting materials is not as well developed as
electron-deficient six-membered ring compounds,1 and therefore
we had to devise our own synthesis that would be amenable to
scale-up to a multikilogram scale. The synthesis of oxazole
aldehyde 1 has been reported many times; however, the reported
scale of preparation is typically only a few grams.2,3 The
synthesis of oxazole aldehyde 1 on a large scale is a challenging
undertaking because the molecule has a low molecular weight,
is unstable4 to extremes of pH, is very soluble in almost all
solvents including water at levels >1g/mL, and has some
thermal instability issues; although it is a stable crystalline solid,
the melting point is low (58 °C) and it is prone to sublimation.2a,5
We report here our preliminary research into and evaluation of
synthetic routes6 to oxazole aldehyde 1 that would overcome
the constraints provided by the molecule and enable the rapid
initial preparation of multikilogram batches of high quality
product in multipurpose pilot plant equipment, without recourse
to chromatographic methods of purification.
We envisaged there could be many potential precursors for
the aldehyde 1 (Scheme 1). The alcohol 4, amides 5 and 6,
thioester 7, and acid chloride 9 would be available by manipula-
tion of the esters 2 or 3, whereas the nitrile 8 would require a
alternative oxazole synthesis.7
Our first synthetic method to methyl ester 2 was a modifica-
tion of a published procedure.8 The intermediate oxazoline 12
was readily prepared from ethyl acetimidate 10 and serine
methyl ester 11 and required an aqueous workup and then a
drying step prior to the oxidation reaction (Scheme 2). Drying
by azeotropic distillation was not applicable as partial hydrolysis
to form N-acetyl serine methyl ester was observed, so we took
recourse to drying the solution over MgSO4. The best conditions
we could develop for the oxidation were the addition of 2.0
equiv of DBU to a mixture of oxazoline and bromotrichlo-
romethane (1.5 equiv) over 5 h at 5 °C; however, during the
course of the reaction large amounts of dark, unidentified
polymeric materials were formed as well. Methyl ester 2 could
be isolated by crystallisation from heptane as a white solid once
a charcoal treatment had been performed. This procedure was
scaled up to 50 kg scale, and ester 2 was isolated as a solvent
wet cake containing 34% w/w heptane and in a corrected yield
of 56%. Subsequently we developed an alternative process that
obviated the need to dry the intermediate oxazoline, avoided
* To whom correspondence should be addressed. E-mail: john.s.carey@
gsk.com.
(1) Carey, J. S.; Laffan, D.; Thomson, C.; Williams, M. T. Org. Biomol.
Chem. 2006, 4, 2337–2347.
(2) References for the preparation of oxazole aldehyde 1 where experi-
mental details are provided: (a) Iso, Y.; Kozikowski, A. P. Synthesis
2006, 243–246. (b) Iso, Y.; Grajkowska, E.; Wroblewski, J. T.; Davis,
J.; Goeders, N. E.; Johnson, K. M.; Sanker, S.; Roth, B. L.;
Tueckmantel, W.; Kozikowski, A. P. J. Med. Chem. 2006, 49, 1080–
1100. (c) Lafontaine, J. A.; Provencal, D. P.; Gardelli, C.; Leahy, J. W.
J. Org. Chem. 2003, 68, 4215–4234. (d) Nicolaou, K. C.; He, Y.;
Ninkovic, S.; Pastor, J.; Roschangar, F.; Sarabia, F.; Vallberg, H.;
Vourloumis, D.; Winssinger, N.; Yang, Z.; King, N. P.; Finlay, M. R.
U.S. Patent 2002/6441186, 27/08/2002. (e) Klar, U.; Schwede, W.;
Skuballa, W.; Buchmann, B.; Hoffmann, J.; Lichtner, R. WO2000/
66589, 09/11/2000. (f) Chenard, B. L.; DeVries, K. M.; Welch, W. M.
WO1998/38187, 03/09/1998. (g) Welch, W. M.; DeVries, K. M.
WO1998/38173, 03/09/1998. (h) Boger, D. L.; Curran, T. T. J. Org.
Chem. 1992, 57, 2235–2244.
(3) Kim, H-S.; Kim, S. H.; Lee, H. K. Bull. Korean Chem. Soc. 1993,
14, 524–526.
(4) Rapid decomposition by hydrolysis to the 1,3-dialdehyde was observed
at pH < 3 and pH > 10.
(5) Some of these physical characteristics are shared by the precursor
compounds.
(7) Palmer, D. C.; Venkatraman, S. In Chemistry of Heterocy-
clic Compounds; Wiley and Sons: Hoboken, NJ, 2003; Vol. 60, pp
1-390.
(8) Williams, D. R.; Lowder, P. D.; Gu, Y.-G.; Brooks, D. A. Tetrahedron
Lett. 1997, 38, 331–334.
(6) Butters, M.; Catterick, D.; Craig, A.; Curzons, A.; Dale, D.; Gillmore,
A.; Green, S. P.; Marziano, I.; Sherlock, J-P.; White, W. Chem. ReV.
2006, 106, 3002–3027.
88
•
Vol. 12, No. 1, 2008 / Organic Process Research & Development
10.1021/op700198s CCC: $40.75
2008 American Chemical Society
Published on Web 11/20/2007

Scheme 2a
a Reagents and conditions: (i) NEt3, CH2Cl2; (ii) BrCCl3, DBU, CH2Cl2, 56%.
Scheme 3a
Scheme 5a
a Reagents and conditions: (i) LiAlH4, Et2O, -5 °C, 70%; (ii) (COCl)2,
DMSO, CH2Cl2, NEt3, -70 °C, 70%.
a Reagents and conditions: (i) NaOMe, MeOH, CH2Cl2; (ii) serine methyl
ester hydrochloride, CH2Cl2; (iii) DBU, CH2Cl2, 69%.
surrogate.13 The crude aldehyde obtained could then be recrys-
tallised to purity from diisopropylether in an overall yield of
69%. Although this was a promising lead, significant challenges
were perceived with respect to scaling up this chemistry; in
particular the reaction temperature of -70 °C was at the low
temperature extreme of our multipurpose pilot plant and we
were concerned that a long reagent addition time required to
control the reaction exotherm would reduce the chemoselec-
tivity. We were also concerned that the hydrochloric acid quench
would be incompatible with the molecule4 and that acetal
formation, from either the alcohol used in the quench or
liberated from the reaction, would occur during the workup or
subsequent solvent exchanges via distillation prior to isolation.
Because of the very short timelines in which we had to prepare
tens of kilograms of material, we did not pursue this approach;
however, with further development and greater process under-
standing it may be possible to use this short, direct route for
large-scale preparation.
Scheme 4a
a Reagents and conditions: (i) DIBALH, CH2Cl2, -70 °C, then iPr2O, 69%.
the use of bromotrichloromethane and therefore the formation
of CHCl3 as a byproduct, and avoided the need to charcoal treat
the product stream prior to crystallisation. An imidate was
formed from chloroacetonitrile 139 and MeOH followed by
addition of serine methyl ester to form the chloro-oxazoline 14
(Scheme 3). Treatment with DBU resulted in elimination of
HCl to form the oxazole ester 2. This was followed by an acidic
wash that furnished an organic stream of essentially pure ester
2; this could be isolated in 69% yield by crystallisation from
heptane if required.
Reduction of methyl ester 2 to the alcohol 4 was achieved
in 70% yield using LiAlH4 (Scheme 5). Oxidation of alcohol 4
to aldehyde 1 using the procedure of Swern has been
published.2c,14 This method was used to prepare some very early
supplies; however, due to the requirement for cryogenic
conditions, the generation of a large amount of gaseous
byproduct, and the stench of DMS produced, this method was
not considered for further scale-up. Additionally the aldehyde
produced using the Swern oxidation was contaminated with
some sulfurous residues that had a negative impact on the
downstream processing.
Replacement of the oxalyl chloride by SO3 ·pyridine com-
plex15 and running at ambient temperature was not practical; a
significant amount (10–15%) of the methyl thioether byproduct
15 was observed (Figure 1). A broad screen of alternative
oxidation reagents and conditions was performed.16 Several sets
of conditions that would affect a clean conversion were found,
Direct reduction of either ester 2 or 3 to aldehyde 1 has been
reported2a,b,d,e,10 using either DIBALH or LiAlH4. The reduction
of the closely related ethyl 4-oxazolecarboxylate using DIBALH
has been recently reported.11 In our hands, the low temperature
reduction of ester 312 using LiAlH42a,b resulted in significant
over-reduction to alcohol 4 with very little aldehyde 1 observed.
Preliminary investigations into the reduction of ester 3 using
DIBALH in dichloromethane yielded some promising leads in
that only 10% over-reduction to alcohol 4 was observed when
the reaction was performed at -70 °C (Scheme 4). It has been
suggested that over-reduction in such systems is suppressed
because the oxazole function may act as a Weinreb amide
(9) (a) Hermitage, S. A.; Cardwell, K. S.; Chapman, T.; Cooke, J. W. B.;
Newton, R. Org. Process. Res. DeV. 2001, 5, 37–44. (b) Cardwell,
K. S.; Hermitage, S. A.; Sjolin, A. Tetrahedron Lett. 2000, 41, 4239–
4242.
(10) (a) White, J. D.; Kranemann, C. L.; Kuntiyong, P. Org. Lett. 2001, 3,
4003–4006. (b) Nicolaou, K. C.; Vandelft, F.; Hosokawa, S.; Kim,
S.; Li, T.; Ohshima, T.; Pfefferkorn, J.; Vourloumis, D.; Xu, J.-Y.;
Winssinger, N. WO1999/21862, 06/05/1999. (c) Nicolaou, K. C.; He,
Y.; Ninkovic, S.; Pastor, J.; Roschangar, F.; Sarabia, F.; Vallberg,
H.; Vourloumis, D.; Winssinger, N.; Yang, Z.; King, N. P.; Finlay,
M. R. WO1998/25929, 18/06/1998. (d) Kende, A. S.; Blass, B. E.;
Henry, J. R. Tetrahedron Lett. 1995, 36, 4741–4744.
(13) Paterson, I.; Steven, A.; Luckhurst, C. A. Org. Biomol. Chem. 2004,
2, 3026–3038.
(14) (a) Provencal, D. P.; Gardelli, C.; Lafontaine, J. A.; Leahy, J. W.
Tetrahedron Lett. 1995, 36, 6033–6036. (b) Nakada, M.; Kobayashu,
S.; Shibaski, M.; Iwasaki, S.; Ohno, M. Tetrahedron Lett. 1993, 34,
1039–1042.
(11) Reeves, J. T.; Song, J. J.; Tan, T.; Lee, H.; Yee, N. K.; Senanayake,
C. H. Org. Lett. 2007, 9, 1875–1878.
(15) Parikh, J. R.; Doering, W. v. E. J. Am. Chem. Soc. 1967, 89, 5505–
5507.
(12) During the course of this work we were able to obtain supplies of the
ethyl ester 3 from third party suppliers. The methyl and ethyl esters
were used interchangeably.
(16) Caron, S.; Dugger, R. W.; Gut Ruggeri, S.; Ragan, J. A.; Brown Ripin,
D. H. Chem. ReV. 2006, 106, 2943–2989.
Vol. 12, No. 1, 2008 / Organic Process Research & Development
•
89

Scheme 7a
Figure 1
Scheme 6a
a Reagents and conditions: (i) NHMe(OMe) ·HCl, EDC, NEt3, CH2Cl2, 74%;
(ii) LiAlH4, THF, -35 °C, 50%.
to produce a stable homogeneous solution when we premixed
the wet carboxylic acid 16, dimethylhydroxylamine hydrochlo-
ride, and triethylamine in a small volume of the reaction solvent,
dichloromethane. This solution was then added to a cooled
slurry of EDC in dichloromethane to give clean conversion to
the amide. Once the reaction was complete, the solution was
washed with minimal amounts of citric acid solution and then
water prior to isolating the product by crystallization. In the
U.K., dichloromethane is a Class A VOC and its emissions to
the environment from our pilot plant facilities are carefully
monitored and controlled, so its use is not preferred. Unfortu-
nately dichloromethane was found to be the best solvent with
respect to minimizing product losses during the aqueous washes;
the use of 2-methyl tetrahydrofuran or tert-butylmethylether as
solvent lead to significantly decreased yields. Product 5 is very
soluble in dichloromethane, so to effect an efficient crystalliza-
tion it needed to be effectively removed. This was done by
several put and take atmospheric distillations with tert-butyl-
methylether until the level of dichloromethane was <30 mol
% with respect to amide 5. At this point a seed could be added
followed by the addition of heptane as an antisolvent. This
procedure was scaled up to a 75 kg input scale, and the product
was obtained in 74% yield.
The Weinreb amide 5 was cleanly reduced to the aldehyde
1 upon reaction with 0.34 equiv of lithium aluminium hydride
in THF at -35 °C, and no evidence of over-reduction to the
alcohol was observed, even with extended addition or reaction
times.19 Once the reduction was complete, the excess hydride
was quenched by the addition of acetic acid. The volume of
acetic acid used was optimized to ensure that the pH of the
solution post-addition of potassium/sodium tartrate was ap-
proximately pH 7. Use of ethanol to quench the excess hydride
resulted in the pH rising to pH 14 later in the workup, and
considerable decomposition occurred.20 To remove the alu-
minium salts, a concentrated solution of potassium/sodium
tartrate was added, at which point a wet THF layer containing
the product separated from the aqueous phase. The solution yield
of aldehyde 1 was 85%. A workup was then devised to remove
the byproduct dimethylhydroxylamine and then isolate the
product by crystallization. The THF solution was diluted with
tert-butylmethylether and then washed with a minimal volume
of 3.5 M aq sodium hydrogen sulfate. This acid was shown to
be strong enough to remove the dimethylhydroxylamine, unlike
an acetic acid wash, but not cause the same level of decomposi-
tion as hydrochloric acid. Failure to remove the dimethylhy-
droxylamine would lead to the partial formation of aminal 17
later in the process (Figure 2). Unfortunately a small but not
a Reagents and conditions: (i) NaOH, H2O, 92%.
such as either TEMPO/[bis(acetoxy)iodo]benzene17 or pyridi-
num dichromate (PDC). However, a suitable workup procedure
that would enable efficient separation of the product from the
reagents could not be found. Oxidation using tetra-n-propy-
lammonium perruthenate (TPAP) lacked sufficient chemose-
lectivity, and overoxidation to carboxylic acid 16 was observed
before all of alcohol 4 was consumed. The Dess-Martin
periodinane2f,g and 2-iodoxybenzoic acid (IBX) were successful,
but the stabilized formulation of IBX, more applicable for scale-
up,18 was not. Alternate TEMPO-based systems using either
sodium hypochlorite or Oxone failed to show any significant
oxidation, and a similar outcome was observed using MnO2.
Hydrolysis of ethyl ester 3 to carboxylic acid 16 was carried
out in water using aqueous NaOH followed by addition of
hydrochloric acid to precipitate the product (Scheme 6). The
volume of solvent was reduced to minimise losses to the mother
liquors. Acid 16 was shown to be thermally unstable by DSC
and ARC, with the onset of an exothermic gas-evolving
decomposition detected at 107 °C. Therefore to provide a
suitable safety margin we would not subject the material to
temperatures above 45 °C on a pilot plant scale. Small-scale
laboratory batches could be dried in a vacuum oven; however,
we felt that if the chemistry was scaled to multiple kilograms
in a pilot plant, vacuum drying of residual water would be very
slow. Due to solubility constraints no suitable water-miscible
organic solvent could be found that could be used to remove
the water by a displacement wash and so ease the drying
problem. Any subsequent large-scale chemistry that we devel-
oped would need to work using water wet acid 16. The
hydrolysis of ester 3 was scaled up to 100 kg scale, and the
acid 16 was obtained as a wet cake containing 19% w/w H2O
and a corrected yield of 92%.
From the survey of methods evaluated on a laboratory scale,
preparation and reduction of the Weinreb amide 5 appeared to
be the most attractive for rapid scale-up to the pilot plant scale.
The Weinreb amide 5 was prepared using the coupling reagent
1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride
(EDC) under conditions that were tolerant of significant levels
of water (Scheme 7). EDC is a known sensitizer, and we were
keen to develop a process that would reduce the exposure to
and manual handling of the reagent. To this end we were able
(17) (a) Camp, D.; Matthews, C. F.; Neville, S. T.; Rouns, M.; Scott, R. W.;
Truong, Y. Org. Process Res. DeV. 2006, 10, 814–821. (b) De Mico,
A.; Margarita, R.; Parlanti, L.; Vescovi, A.; Piancatelli, G. J. Org.
Chem. 1997, 62, 6974–6977.
(19) Nahm, S.; Weinreb, S. M. Tetrahedron Lett. 1981, 22, 3815–3818.
(20) Cornforth, J. W.; Fawaz, E.; Goldsworthy, L. J.; Robinson, R. J. Chem.
Soc. 1949, 1549–1553.
(18) Ozanne, A.; Pouysegu, L.; Depernet, D.; Francois, B.; Quideau, S.
Org. Lett. 2003, 5, 2903–2906.
90
•
Vol. 12, No. 1, 2008 / Organic Process Research & Development

Scheme 8a
Figure 2
a Reagents and conditions: (i) morpholine, NEt3, EDC, CH2Cl2, 78%.
insignificant quantity of aldehyde 1, approximately 10% as
determined by HPLC assay, was lost during the acidic wash.
To remove the residual sodium hydrogen sulfate and acetic acid
entrained within the organic layer a wash with a minimal
amount of sodium carbonate was installed. Again a small but
not insignificant quantity of aldehyde 1 was lost during this
wash. On a 50 g scale in the laboratory, the solution yield of
aldehyde 1 post-workup was 68–70% whereas on a 50 kg scale
in the pilot plant the solution yield had slipped to 60–63% due
to the increased contact times required for the washes. This
product stream was relatively free from any contaminants. To
effect an efficient crystallization it was essential to reduce the
level of water present to <0.5% w/w, this was achieved by
two put and take distillations at atmospheric pressure using tert-
butylmethylether. Upon cooling slightly, the product would
begin to crystallize, and then heptane was added as an
antisolvent to maximize recovery. Failure to remove sufficient
water would lead to product oiling along with the water when
the antisolvent was added. The isolated yield of product on a
50 g scale in the laboratory was 55–57%; however, in the pilot
plant the yield had diminished to 42–45%. The solubility of
aldehyde 1 in heptane/tert-butylmethylether was such that
approx 10–12% of yield was lost to the mother liquors,
irrespective of scale, and on a pilot plant scale the mother liquors
were reconcentrated by distillation under reduced pressure to
provide a further 5–7% of yield. Although the product filtered
readily, drying was not without difficulty as the product readily
sublimes. On a laboratory scale significant losses could be
observed when using a Buchner apparatus, and on a pilot scale
approximately 0.5–1 kg was lost from each batch during the
vacuum drying stage.
Schwartz’s reagent, Cp2Zr(H)Cl, has been reported as an
alternative reagent for the chemoselective reduction of Wein-
reb’s amides to aldehdyes.21 However, when amide 5 was
reacted with the reagent, extensive over-reduction to alcohol 4
was observed as well as the desired aldehyde 1.
The selective reduction of morpholine amides to aldehydes
has been reported to be a cost-effective alternative to the use
of the N,O-dimethylhydroxylamides.22 An additional attraction
to this method would be that an acid weaker than sodium
hydrogen sulfate should be able to wash out the morpholine
byproduct and therefore potentially reduce the levels of aldehyde
decomposition observed during the workup. The morpholine
amide 6 was readily prepared from the corresponding acid 16
(Scheme 8). Reaction of morpholine amide 6 with LiAlH4 in
THF at 0 °C proceeded within 1.5 h to produce aldehyde 1
contaminated with 10% of alcohol 4. The morpholine byproduct
Scheme 9a
a Reagents and conditions: (i) CH3(CH2)11SH, EDC, DMAP, CH2Cl2, 72%.
could be removed by washing with aqueous acetic acid.
However, even on a laboratory scale reaction, the isolated yield
of the crude product was no better than that achieved by
reduction of the Weinreb amide, and by the time a recystalli-
sation was performed to improve the quality, the yield had
diminished considerably.
The Fukuyama reduction of a thioester to an aldehyde
appeared as an attractive alternative procedure as it would
circumvent the use of active metal hydrides such as LiAlH4 or
DIBALH.23 We chose to prepare the dodecanethiol ester 7 as
the required dodecanethiol is odourless24 (Scheme 9). Various
solvents, palladium on charcoal catalysts and additives were
screened to try and find conditions that would give a chemose-
lective reduction.25 Under many of the reaction conditions
investigated, over-reduction to the alcohol 4 was observed. In
a series of control experiments it was shown that the aldehyde
1 could be readily reduced to the corresponding alcohol.
The best conditions we could find for the reduction of
thioester 7 to aldehyde 1 were 25% by weight of 5% palladium
on charcoal (Engelhard Escat 143), 1.65 equiv of Et3SiH, 2.0
equiv of MgSO4 in dichloromethane. After 1 h at ambient
temperature the reaction profile showed 5% unreacted thioester
7 and no alcohol 4 present; however, if the reaction time was
allowed to increase, then over-reduction occurred faster than
consumption of the unreacted thioester. The solids could be
removed by filtration to yield a 65% solution yield of aldehyde
1. Unfortunately an effective practical workup and isolation
procedure to remove the excess reagents and byproduct could
not be developed.26 This point coupled to the lack of chemose-
lectivity in the reaction and the sensitivity to reaction time meant
that this approach was not deemed suitable for large-scale
production.
(23) Fukuyama, T.; Lin, S.-C.; Li, L. J. Am. Chem. Soc. 1990, 112, 7050–
7051.
(21) (a) Spletstoser, J. T.; White, J. M.; Tunoori, A. R.; Georg, G. I. J. Am.
Chem. Soc. 2007, 129, 3408–3419. (b) White, J. M.; Tunoori, A. R.;
Georg, G. I. J. Am. Chem. Soc. 2000, 122, 11995–11996.
(22) Douat, C.; Heitz, A.; Martinez, J.; Fehrentz, J-A. Tetrahedron Lett.
2000, 41, 37–40.
(24) Miyazaki, T.; Han-ya, Y.; Tokuyama, H.; Fukuyama, T. Synlett 2004,
477–480.
(25) Kimura, M.; Seki, M. Tetrahedron Lett. 2004, 45, 3219–3223.
(26) During the course of the reaction a significant odour of thiol was
generated.
Vol. 12, No. 1, 2008 / Organic Process Research & Development
•
91

Scheme 10a
a Reagents and conditions: (i) NEt3, CH2Cl2; (ii) EtOCHO, KOtBu, CH2Cl2; (iii) TMSCl, CH2Cl2, ∆, 50%, iv) DIBALH, CH2Cl2, -20 °C 41%.
Scheme 11a
then evaluated in the key aldehyde-forming reactions. Reduction
of the Weinreb amide 5 was selected as the route to prepare
these early supplies as the chemistry required could readily be
accommodated in our multipurpose pilot plant equipment and
the reaction selectivity was not time-dependent. This chemistry
was scaled from 50 g scale in the laboratory to 50 kg scale in
the pilot plant. The isolated yield upon scale-up was lower than
that achieved in the laboratory, 45% yield versus 55%. The
longer contact times for the aqueous washes upon scale-up are
partially responsible for the loss in yield, along with some
sublimation during the extended distillation and drying stages
of the process. Several other shorter approaches, such as
reduction of ester 3 or nitrile 8, showed some promise and may
be suitable alternatives for large-scale aldehyde production after
further optimization, greater process understanding and use of
specific equipment.
a Reagents and conditions: (i) (COCl)2, DMF, CH2Cl2, 61%.
The known cyano-oxazole 827 was prepared by a modifica-
tion of the methods reported by Cornforth20,28 (Scheme 10).
Imidate 20 was prepared from ethyl acetimidate 18 and
aminoacetonitrile 19 and then acylated with ethyl formate in
the presence of potassium tert-butoxide, to give 21. Whilst this
intermediate could be isolated, it was more convenient to treat
it directly with chlorotrimethylsilane to effect cyclisation. This
modification of replacing the standard reagent,20,28 acetic acid,
with chlorotrimethylsilane increased the overall yield for the
three-step procedure from 8%27a to 50%.
Experimental
2-Methyloxazole-4-carboxylic Acid Methyl Ester (2).
Procedure 1. Serine methyl ester hydrochloride (50.0 kg, 321
mol) and ethyl acetimidate hydrochloride (47.7 kg, 386 mol)
were suspended in dichloromethane (500 L) at 20–25 °C, and
triethylamine (32.8 kg, 324 mol) was added. The resulting slurry
was stirred at 20–25 °C for 1 h. Water (250 L) was added, the
mixture was stirred for 10 min to ensure all solids had dissolved,
and then the aqueous and organic layers were separated. The
organic layer was washed with water (250 L) and then dried
over MgSO4 (50 kg). The solids were removed by filtration,
and the filtercake was washed with dichloromethane (250 L).
The filtrate and wash were combined, bromotrichloromethane
(95.6 kg, 482 mol) was added, and the solution was cooled to
3-7 °C. 1,8-Diazabicyclo[5.4.0]undec-7-ene (97.9 kg, 643 mol)
was added at 0-5 °C over 5 h. After a further 1 h at 3-7 °C
the solution was warmed to 20-25 °C, HCl (250 L; 2 M; aq)
was added, and then the organic layer was washed with water
(250 L). The organic layer was concentrated by distillation at
atmospheric pressure to leave a residual volume of 175 L.
Nuchar charcoal (5 kg) was added followed by heptane (500
L). The solution was concentrated by distillation at atmospheric
pressure until the distillation temperature reached 98 °C; approx
300 L of distillate was collected. Heptane (375 L) was added,
the content temperature was maintained within the range 70-90
°C while the charcoal was removed by filtration and the filterbed
was washed with hot heptane (100 L). The solution was cooled
to 45 °C and then a seed of methyl ester 2 was added to induce
crystallisation. The resulting slurry was cooled to 0-5 °C and
A brief screen of reduction conditions20 was performed using
Red-Al, PtO2/HCO2H, and DIBALH. It was found that the use
of DIBALH in CH2Cl2 at -20 °C afforded the aldehyde in
41% yield. Due to constraints it was not possible to further
explore this avenue; however, it may be possible to develop
this lead to establish a concise route for large-scale preparation.
The Rosemund reduction of acid chloride 9, readily prepared
from acid 16, was explored (Scheme 11). A screen of different
palladium hydrogenation catalysts (Pd/Ca2CO3 + Pb, Pd/
Ba2SO4, dry Pd/C) and solvents (EtOAc, toluene, NMP) in the
presence of a base (Na2CO3, K2CO3, NEt3) at various temper-
atures and pressures was undertaken. Unfortunately, either poor
conversion or over-reduction to alcohol 4 was observed, so this
avenue was not pursued further.
Conclusions
Tens of kilograms of oxazole aldehyde 1 were prepared to
support the early development of a drug candidate molecule.
To do this a series of potential precursors were prepared and
(27) (a) Khanna, I. K.; Yu, Y.; Huff, R. M.; Weier, R. M.; Xu, X.; Koszyk,
F. J.; Collins, P. W.; Cogburn, J. N.; Isakson, P. C.; Koboldt, C. M.;
Masferrer, J. L.; Perkins, W. E.; Seibert, K.; Veenhuizen, A. W.; Yuan,
J.; Yang, D.-C.; Zhang, Y. Y. J. Med. Chem. 2000, 43, 3168–3185.
(b) Ceulemans, E.; Dyall, L. K.; Dehaen, W. Tetrahedron 1999, 55,
1977–1988. (c) Sycheva, N. T.; Trupp, T. K.; Scehukina, M. N. J. Gen.
Chem. USSR (Engl.) 1962, 32, 1051–1055. (d) Rinderspacher, T.; Prijs,
B. HelV. Chim. Acta 1960, 43, 1522–1530.
(28) (a) Cornforth, J. W.; Huang, H. T. J. Chem. Soc. 1948, 1969–1971.
(b) Cornforth, J. W.; Cornforth, R. H. J. Chem Soc. 1947, 96–102.
92
•
Vol. 12, No. 1, 2008 / Organic Process Research & Development

aged for 6 h. The slurry was then filtered, and the filtercake
was washed with cold heptane (2 × 75 L). Methyl ester 2 was
isolated as a heptane wet cake, 38.4 kg, which was suitable for
onward processing. Loss on drying ) 34% w/w. Theoretical
dry weight ) 25.3 kg (56%).
dimethylhydroxylamine hydrochloride (57 kg, 584 mol) were
suspended in dichloromethane (225 L) and water (20 L).
Triethylamine (60 kg, 593 mol) was added and then stirred at
ambient temperature for 1 h to achieve a clear solution that
was then cooled to 0-5 °C. 1-(3-dimethylaminopropyl)-3-
ethylcarbodiimide hydrochloride (137 kg, 715 mol) was sus-
pended in dichloromethane (375 L) and then cooled to 0-5
°C. The solution of carboxylic acid was added to the slurry of
diimide while maintaining the temperature <10 °C. Once the
addition was complete the solution was allowed to warm to
ambient temperature and then stirred for 1 h. A solution of citric
acid (45 kg) in water (180 L) was added, and the layers were
separated. The aqueous solution was re-extracted with dichlo-
romethane (225 L). The organic extracts were combined and
then washed with water (112 L). The solution was concentrated
by distillation at atmospheric pressure to leave a residual volume
) 225 L. tert-Butylmethylether (375 L) was added, and the
solution was concentrated by distillation at atmospheric pressure
to leave a residual volume of 188 L. tert-Butylmethylether (375
L) was added, and the solution was concentrated by distillation
at atmospheric pressure to leave a residual volume of 188 L.
The solution was cooled to 35 °C, and then tert-butylmethyl-
ether (188 L) was added. The solution was seeded with amide
5 at 35 °C, and heptane (375 L) was added at 35 °C. Once the
addition was complete, the slurry was aged at 35 °C prior to
being cooled to ambient temperature and aged for a further 2 h.
The slurry was cooled to 0-5 °C and aged for 1 h, and the
product isolated by filtration. The filtercake was washed with
heptane (75 L) and then dried under vacuum to afford the amide
5 as a white solid (74.3 kg, 74%, 68% yield over 2 stages). Mp
) 53 °C; 1H NMR (400 MHz, CDCl3) δ 8.10 (s, 1 H), 3.77 (s,
3 H), 3.37 (s, 3 H), 2.51 (s, 3 H); 13C NMR (100 MHz, CDCl3)
δ 161.1, 161.0, 142.3, 133.1, 61.2, 33.0, 13.5; IR (neat cm-1)
1639, 1595; LRMS (CI +ve) m/z 171 (M+ + H).
2-Methyloxazole-4-carboxaldehyde (1). Procedure 1. Wein-
reb amide 5 (50 kg, 294 mol) was dissolved in tetrahydrofuran
(300 L) and cooled to -35 °C. Lithium aluminium hydride
solution (1 M in tetrahydrofuran, 100 L, 100 mol) was added
while maintaining the temperature at -30 to -35 °C. After 30
min a solution of acetic acid (15 L) in tetrahydrofuran (50 L)
was added while maintaining the temperature at -30 to -35
°C. The resulting solution was warmed to -15 °C, and a
solution of potassium/sodium tartrate (125 kg of K/Na tartrate
tetrahydrate in 250 L H2O) was added while maintaining the
temperature <10 °C. Upon completion of the addition the
mixture was warmed to 20 °C, and the layers separated. The
organic layer was diluted with tert-butylmethylether (500 L)
and was then washed with sodium hydrogen sulfate (150 L;
3.5 M; aq). The organic layer was then washed with sodium
carbonate (125 L; 8% w/v; aq) and then concentrated by
distillation at atmospheric pressure to leave a residual volume
of 150 L. tert-Butylmethylether (250 L) was added and then
concentrated by distillation at atmospheric pressure to leave a
residual volume of 150 L. tert-Butylmethylether (250 L) was
added and then concentrated by distillation at atmospheric
pressure to leave a residual volume of 150 L. The solution was
then sampled for water content (<0.5% w/w by Karl Fischer
titrimetry). The resulting solution was cooled to room temper-
Procedure 2. Sodium methoxide (25% w/w solution in
methanol, 1.65 mL, 7.20 mmol) was added to a solution of
methanol (50 mL) in dichloromethane (450 mL) at 0-5 °C.
After 5 min, chloroacetonitrile (50 mL, 790 mmol) was added,
and the solution was stirred at 0-5 °C for 1.5 h. Serine methyl
ester hydrochloride (99.0 g, 636 mmol) was added at 0-5 °C,
and the resulting slurry was allowed to warm to room temper-
ature and stirred for 18 h. Water (200 mL) was added, the
mixture stirred for 10 min to ensure all solids had dissolved,
and then the aqueous and organic layers were separated. The
organic layer was washed with water (200 mL) and then
concentrated by distillation at atmospheric pressure to leave a
residual volume of 150 mL. Fresh dichloromethane (500 mL)
was added, and the solution was warmed to 30 °C. 1,8-
Diazabicyclo[5.4.0]undec-7-ene (95 mL, 635 mmol) was added
over 2 h while maintaining the internal temperature of the
reaction at 30 °C. Upon completion of the addition the reaction
mixture was cooled to room temperature, and HCl (200 mL; 2
M; aq) was added. Then the organic layer was washed with
water (200 mL). The organic layer was concentrated by
distillation at atmospheric pressure to leave a residual volume
of 300 mL. Heptane (400 mL) was added, and the solution was
concentrated by distillation at atmospheric pressure to leave a
residual volume of 300 mL. Heptane (100 mL) was added, the
solution was cooled to 45 °C, and then a seed of methyl ester
2 was added to induce crystallisation. The resulting slurry was
cooled to 0-5 °C and aged for 2 h. The slurry was filtered,
and the filtercake was washed with cold heptane and then dried
to give methyl ester 2 (62.11 g, 69%). Mp ) 48 °C; 1H NMR
(400 MHz, CDCl3) δ 8.16 (s, 1 H), 3.91 (s, 3 H), 2.52 (s, 3 H);
13C NMR (100 MHz, CDCl3) δ 162.2, 161.6, 143.7, 133.1, 51.9,
13.7; IR (neat cm-1) 1732, 1585; LRMS (CI +ve) m/z 142
(M+ + H).
2-Methyloxazole-4-carboxylic Acid (16). Ethyl ester 3 (100
kg, 645 mol) was dissolved in water (250 L) at 20–25 °C, and
NaOH (96 kg; 32% w/w; aq, 768 mol) added over 30 min,
while maintaining the temperature in the range 20-25 °C. After
1 h HCl (78 kg; 37% w/w; aq, 770 mol) was added while
maintaining temperature in the range 20-25 °C. Upon comple-
tion of the addition the resulting slurry was cooled to 0 °C and
aged for 1 h. The product was isolated by filtration, and the
filtercake was washed with cold water (100 kg). Wet weight
of acid ) 93 kg, loss on drying ) 19% w/w, theoretical weight
of dry acid ) 75 kg (92%). A small sample was dried for
analytical purposes. Mp ) 77 °C; 1H NMR (400 MHz, DMSO-
d6) δ 12.90 (s, 1 H), 8.57 (s, 1 H), 2.44 (s, 3 H); 13C NMR
(100 MHz, DMSO-d6) δ 167.3, 167.0, 150.0, 138.4, 18.5; IR
(neat cm-1) 1715, 1585; LRMS (CI - ve) m/z 126 (M+ - H);
water content <0.1% w/w by Karl Fischer titrimetry; residue
on ignition ) 0.2% w/w.
N-Methoxy-N-methyl-2-methyloxazole-4-carboxamide (5).
Carboxylic acid 16 (93 kg, loss on drying ) 19% w/w,
theoretical weight of dry acid ) 75 kg, 591 mol) and N,O-
Vol. 12, No. 1, 2008 / Organic Process Research & Development
•
93

ature, during which time crystallization occurred. Heptane (400
L) was added over 2 h, and then the slurry was cooled to 0 °C.
The slurry was filtered, and the filtercake washed with cold
heptane (50 L) and then dried under vacuum at ambient
temperature to afford aldehyde 1 as an off-white solid (14.8
kg, 45%). The mother liquors and wash liquors were combined
and then concentrated by distillation under reduced pressure to
leave a residual volume of 75 L. The resulting slurry was cooled
to -5 to -10 °C and filtered, and the filtercake was washed
with cold heptane (7.5 L) and then dried under vacuum at
ambient temperature to afford a second crop of aldehyde 1 as
an off-white solid (1.78 kg, 5%). Mp ) 49 °C; 1H NMR (400
with HCl (50 mL; 2 M; aq), water (50 mL), and then brine (50
mL). The organic solution was dried over MgSO4, filtered, and
then concentrated to dryness to leave the product as a white
solid (11.45 g, 93%). This essentially pure product could be
recrystallised from methanol (35 mL) to afford thioester 7 as a
white solid (8.87 g, 72%). 1H NMR (400 MHz, CDCl3) δ 8.08
(s, 1 H), 3.03 (t, J ) 7 Hz, 2 H), 2.52 (s, 3 H), 1.65 (quin, J )
7 Hz, 2 H), 1.44–1.22 (m, 18 H), 0.88 (t, J ) 7 Hz, 3 H); 13
C
NMR (100 MHz, CDCl3) δ 185.6, 162.0, 139.8, 139.5, 31.9,
29.6, 29.6, 29.5, 29.5, 29.3, 29.1, 28.8, 22.7, 14.1, 13.8; IR (neat
cm-1) 2915, 1651, 869; LRMS (CI +ve) m/z 312 (M+ + H).
2-Methyloxazole-4-carbonitrile (8). Ethyl acetimidate hy-
drochloride (25.0 g, 202 mmol) and aminoacetonitrile hydro-
chloride (23.3 g, 252 mmol) were suspended in dichloromethane
(500 mL) at room temperature. Triethylamine (28.5 mL, 204
mmol) was added over 1.5 h, and once the addition was
complete, the slurry was cooled to 5 °C. Water (125 mL) was
added, the mixture was stirred to ensure all solids had dissolved,
and then the aqueous and organic layers were separated. The
aqueous layer was re-extracted with dichloromethane (50 mL).
The organic extracts were combined and then washed with
water (125 mL), dried over MgSO4 (15 g), filtered, and then
concentrated by distillation to leave a residual volume of 50
mL. tert-Butylmethylether (400 mL) was added, and the solution
was cooled to 0 °C. Ethyl formate (14.8 mL, 184 mmol) was
added followed by a portionwise addition of potassium tert-
butoxide (20.6 g, 184 mmol) while maintaining the temper-
ature <10 °C. Once the addition was complete the
resulting slurry was stirred at 0 °C for 1.5 h prior to being
heated to reflux. Chlorotrimethylsilane (53 mL, 415 mmol)
was added at reflux over 1.5 h. Once the addition was
complete the mixture was cooled to room temperature,
and water (100 mL) was added. Once all of the solids
had dissolved, the aqueous and organic layers were
separated. The aqueous layer was re-extracted with
dichloromethane (100 mL). The organic extracts were
combined and then washed with water (100 mL), and the
solvent was removed by distillation to leave an oil, which
could be purified by short path distillation (oven temper-
ature ) 170 °C, 20 mmHg) to give cyano-oxazole 8 as
MHz, CDCl3) δ 9.92 (s, 1 H), 8.20 (s, 1 H), 2.54 (s, 3 H); 13
C
NMR (100 MHz, CDCl3) δ 183.8, 163.0, 144.6, 140.9, 13.7;
IR (neat cm-1) 1674, 1595; LRMS (CI +ve) m/z 112 (M+ +
H).
2-Methyloxazole-4-methanol (4). Methyl ester 2 (500 g,
3.55 mol) was suspended in diethyl ether (7.7 L) and cooled to
-5 °C. Lithium aluminium hydride solution (1.0 M in tetrahy-
drofuran, 2.13 L, 2.13 mol) was added while maintaining the
temperature at -5 °C. After 1 h, water (81 mL) was added,
followed by NaOH (81 mL; 15% w/v; aq) and further water
(242 mL). Sodium sulfate (1.02 kg) was added, and the resulting
suspension was allowed to warm to ambient room temperature.
The slurry was filtered and the filtercake was washed with
dichloromethane (3 × 2.5 L). The filtrate and washes were
combined and then concentrated to dryness to afford alcohol 4
as an off-white solid (280 g, 70%). The alcohol could be
1
recrystallised from heptane if required. H NMR (400 MHz,
CDCl3) δ 7.48 (s, 1 H), 5.00 (br s, 1 H), 4.53 (s, 2 H), 2.43 (s,
3 H); 13C NMR (100 MHz, CDCl3) δ 162.2, 140.3, 134.9, 55.8,
13.7; IR (neat cm-1) 3212, 1578; LRMS (CI +ve) m/z 114
(M+ + H).
N-Morpholine-2-methyloxazole-4-carboxamide (6). Car-
boxylic acid 16 (51.39 g, 405 mmol) was suspended in
dichloromethane (515 mL). Triethylamine (59.3 mL, 425
mmol), then morpholine (37.2 mL, 425 mmol) and then 1-(3-
dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (101
g, 527 mmol) were added, and the resulting mixture was stirred
at ambient temperature for 70 h. A solution of citric acid (40
g) in water (150 mL) was added, and the layers were separated.
The organic layer was washed with water (150 mL), dried over
MgSO4, filtered, and then concentrated to dryness to leave the
product 6 as a white solid (62.31 g, 78%). 1H NMR (400 MHz,
CDCl3) δ 8.07 (s, 1 H), 4.15 (br s, 2 H), 3.74 (br s, 6 H), 2.48
(s, 3 H); 13C NMR (100 MHz, CDCl3) δ 160.8, 160.6, 143.4,
136.5, 67.0, 46.9, 42.8, 13.8; IR (neat cm-1) 1738, 1611, 1586;
LRMS (CI +ve) m/z 197 (M+ + H).
2-Methyloxazole-4-carboxylic Acid Dodecanthioester (7).
Carboxylic acid 16 (5.00 g, 39.4 mmol) was suspended in
dichloromethane (50 mL) and cooled to 0-5 °C. 1-(3-
Dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (9.33
g, 48.7 mmol), N,N-dimethylaminopyridine (240 mg, 1.96
mmol), and dodecanthiol (9.42 mL, 39.3 mmol) were added at
0-5 °C, and the resulting mixture was allowed to warm to
ambient temperature. After 1 h the reaction was complete, and
the dichloromethane was removed on a rotary evaporator. Ethyl
acetate (100 mL) was added, and the organic layer was washed
1
an oil (10.9 g, 50%). H NMR (400 MHz, CDCl3) δ 8.08
(s, 1 H), 2.55 (s, 3 H); 13C NMR (100 MHz, CDCl3) δ
163.2, 146.2, 114.8, 111.8, 13.7; IR (neat cm-1) 2247,
1589; LRMS (CI +ve) m/z 109 (M+ + H).
2-Methyloxazole-4-carboxylic Acid Chloride (9). Car-
boxylic acid 16 (5.00 g, 39.4 mmol) was suspended in
dichloromethane (20 mL) and cooled to 0-5 °C. Oxalyl
chloride (6.75 mL, 77.4 mmol) and then DMF (50 µL, 0.65
mmol) were added. Once the additions were complete, the
mixture was allowed to warm to ambient temperature and then
aged for 3 h. The resulting slurry was cooled to 0-5 °C and
aged for 1 h, and then heptane (50 mL) was added. The slurry
was aged at 0-5 °C for 1.5 h, and then the product was isolated
by filtration and dried under vacuum to afford acid chloride 9
as a white solid, which colourised slightly upon standing (3.47
g, 61%). 1H NMR (400 MHz, DMSO-d6) δ 8.60 (s, 1 H), 2.46
(s, 3 H); 13C NMR (100 MHz, DMSO-d6) δ 161.2, 161.1, 144.8,
133.1, 133.0; IR (neat cm-1) 1755, 820.
94
•
Vol. 12, No. 1, 2008 / Organic Process Research & Development

2-Methyloxazole-4-carboxaldehyde (1). Procedure 2. Ethyl
ester 3 (24.27 g, 156 mmol) was dissolved in dichloromethane
(300 mL) and then cooled to -70 °C. DIBALH (56 mL, 314
mmol) was added over 45 min while maintaining temperature
at -65 to -70 °C. After 45 min, methanol (50 mL) was added,
and the solution was allowed to warm to ambient temperature,
during which period some solids precipitated from solution and
stirring became difficult. HCl (150 mL; 2 M; aq) was added,
all the solids dissolved, and the layers were separated. The
aqueous layer was extracted with dichloromethane (200 mL),
and the organic extracts were combined, dried over MgSO4,
and concentrated to dryness to afford the crude product 1 (15.7
g). Recrystallisation from iPr2O (60 mL) gave aldehyde 1 as
an off-white solid (12.0 g, 69%).
Procedure 3. Dimethylsulphoxide (391 mL, 5.50 mol) in
dichloromethane (1.0 L) was added to a solution of oxalyl
chloride (236 mL, 2.71 mol) in dichloromethane (3.75 L) while
maintaining the temperature at -60 to -55 °C. After 5 min a
solution of alcohol 4 (289 g, 2.11 mol) in dichloromethane (1.35
L) was added while maintaining the temperature at -60 to -55
°C. After 20 min, triethylamine (1.78 L, 12.77 mol) was added
while maintaining the temperature at -60 to -55 °C. After 45
min the resulting slurry was allowed to warm to ambient room
temperature, and water (7.23 L) was added. The layers were
separated, and the aqueous layer was re-extracted with dichlo-
romethane (4.34 L). The dichloromethane extracts were com-
bined and dried over MgSO4 (290 g). The solvent was removed,
and the residue purified by column chromatography (Biotage
system, eluent petroleum ether/ethyl acetate 3:1) to afford
aldehyde 1 as an off-white solid (199 g, 70%).
Procedure 4. Cyano-oxazole 8 (1.45 g, 13.43 mmol) was
dissolved in dichloromethane (6 mL) and then cooled to -20
°C. DIBALH (1.0 M solution in dichloromethane, 20 mL, 20
mmol) was added over 1 h while maintaining the temperature
at -20 °C. After 15 h, methanol (5 mL) was added, and the
solution was allowed to warm to ambient temperature, Further
dichloromethane (20 mL) was added followed by HCl (7.5 mL;
2 M; aq). The layers were separated, and the aqueous layer
was re-extracted with dichloromethane (7.5 mL). The organic
extracts were combined, dried over MgSO4, and concentrated
to dryness to afford the product 1 as a light tan solid (612 mg,
41%).
Acknowledgment
We thank Mark Buswell for hydrogenation support, Erica
Vit for process safety, and Charles Dexter and David Thompson
for pilot plant support.
Received for review September 3, 2007.
OP700198S
Vol. 12, No. 1, 2008 / Organic Process Research & Development
•
95