An expedient route to 3-methoxy-2-furaldehyde

Article abstract of DOI:10.1055/s-0031-1290103

An expedient route to 3-methoxy-2-furaldehyde is presented. Georg Thieme Verlag Stuttgart. New York.

Full text of DOI:10.1055/s-0031-1290103

1
34
LETTER
An Expedient Route to 3-Methoxy-2-furaldehyde
A
M
n
Expedient Route to
a
3-Methoxy
g
-2-furaldehy
n
de us Ronn,* Ngiap-Kie Lim, Philip Hogan, Wu-Yan Zhang, Zhijian Zhu, Nicholas Dunwoody
Tetraphase Pharmaceuticals, Inc., Process Chemistry R&D, 480 Arsenal Street, Suite 110, Watertown, MA 02472, USA
Fax +1(617)9263557; E-mail: mronn@tphase.com
Received 4 October 2011
diate for tetracycline synthesis.2c Unfortunately, several
of the reported steps did not lend themselves to a robust
scale up. Several distillations to dryness resulted in poor
mixing, heat buildup and charred vessels. A sudden and
strong exotherm occurred in step 2 when the neat mixture
was heated under caustic condition to temperatures neces-
sary to induce the reaction. Finally the instability of 3-
methoxyfuran, presumably due to polymerization, as well
as the low yield in the final Vilsmeier–Haack formylation
prompted us to eventually abandon this route.
Abstract: An expedient route to 3-methoxy-2-furaldehyde is pre-
sented.
Key words: furans, aldehydes, heterocycles, nucleophilic substitu-
tion, Diels–Alder reaction
Furans are commonly used as substrates in Diels–Alder
1
chemistry. Despite the high reactivity of 3-alkoxyfurans
as dienes in Diels–Alder reactions, they have appeared
relatively infrequently in the literature, presumably due to
the cumbersome synthetic routes used to prepare them.
One advantage of using 3-alkoxyfurans in Diels–Alder
reactions is that the vinyl alkyl ethers products can be fur-
ther transformed into a number of different functional-
ities.
We believed that a significantly shorter and robust manu-
facturing route could be developed for manufacturing of
multi hundred gram amounts of the title compound with
the possibility of later targeting multi-kilogram manufac-
turing.
We investigated the use of 3-bromo-2-furaldehyde as sub-
strate for the displacement of bromide with methoxide
We have been exploiting an innovative synthetic platform
that provides access to a broad range of previously inac-
cessible tetracycline analogues where 3-methoxy-2-fural-
dehyde is a component of a convergent synthesis of one of
(
Scheme 2). This substrate could either be manufactured
using a lithiation–formylation approach from commer-
cially available 3-bromofuran or be purchased as a com-
mercial compound. We believed that the electron-
withdrawing carbonyl functionality would facilitate the
nucleophilic displacement of the bromide.
2
the key intermediates. Many substituted furans are com-
mercially available or easy to synthesize, but in this par-
ticular case a multistep route to 3-methoxy-2-furaldehyde
had to be used due to lack of expedient routes. Previously
published procedures for the manufacturing of the title Regioselective lithiation of 3-halofurans has previously
compound are impractical and not amendable to the syn- been described along with trapping of the resulting car-
4
thesis of the multi-kilogram quantities required.
banion with various electrophiles. Using 3-bromofuran
as substrate, we screened suitable bases for deprotonation
of the 2-position and found that a strong base such as lith-
ium diisopropylamide (LDA) was needed to get complete
deprotonation with high regioselectivity. We next investi-
gated the effect of solvents for the deprotonation–trapping
sequence. While THF and 2-methyltetrahydrofuran (2-
MeTHF) gave comparable results in terms of purity and
The initial attempts to develop a production route were
based on chemistry previously reported (Scheme 1). 3-
Methoxyfuran was prepared from commercially available
3
cheap starting materials and reagents. Vilsmeier–Haack
formylation of 3-methoxyfuran was reported by Myers et
al. in their disclosure of a rapid synthesis of a key interme-
OMe
CH(OMe)3
ZnCl2 (cat.)
OH
KOH/DBU
110 °C
O
O
~
2
Et O
O
O
O
O
HO
Cl
OMe
distillation
MeO
Cl
OMe
O
DMF
(COCl)2
CH Cl
TsOH (cat.)
MeOH
O
adipic acid (cat.)
200 °C
O
O
MeO
2
2
H
OMe
distillation
distillation
OMe
OMe
OMe
Scheme 1 Published route to 3-methoxy-2-furaldehyde
SYNLETT 2012, 23, 134–136
x
x.
x
x.
2
0
1
1
Advanced online publication: 09.12.2011
DOI: 10.1055/s-0031-1290103; Art ID: S51311ST
©
Georg Thieme Verlag Stuttgart · New York

LETTER
An Expedient Route to 3-Methoxy-2-furaldehyde
135
5
yields, the use of methyl tert-butyl ether (MTBE) resulted To minimize the side-reactions, we investigated slow ad-
in the formation of significant amounts of impurities as dition of the substrate to a refluxing solution of LiOMe in
well as lower yields. As the 2-substituted furaldehyde was MeOH. This resulted in a significant improvement with
our final target compound, we used N,N-dimethylforma- less than 15–20% impurities/starting material (LC–MS)
mide (DMF) for the trapping of the anion but other elec- using a five hour addition time. Despite these improve-
trophiles such as chlorocarbonates and CO were also ments, the isolated yield remained low (30–50%) prompt-
2
used successfully to form the corresponding furan deriva- ing investigations into how the workup and isolation was
tives.
performed.
The amount of LDA used in the lithiation step was impor- During the aqueous workup of our methoxy-substitution,
tant. Use of small excesses of LDA resulted in a side prod- insoluble dark material complicated the phase separation
uct tentatively assigned as a bis-formylated product. The between the two layers as well as purification via chroma-
lithiation–trapping sequence could be performed at tem- tography or reslurry/crystallization. This insoluble mate-
peratures as high as –20 °C without significantly affecting rial, presumably from polymerization, was believed to
the yield. However, as darker colored impurities appeared partly form during the workup and/or isolation stage. The
at higher temperatures, we chose to perform this reaction most convenient workup resulting in the best yields was
at –40 °C to –50 °C. We also opted to use 2-MeTHF for found to be a direct filtration of the crude reaction mixture
this reaction as it led to faster phase separations during through silica gel followed by elution of the silica using
workup, less color in the isolated product and a higher iso- isopropyl acetate. Concentration of the filtrate resulted in
lated yield. After these improvements, we were able to a residue that was submitted to a second filtration through
routinely produce multi hundred gram batches of 3-bro- silica gel to remove more polar polymeric materials. After
mo-2-furaldehyde in 85–95% yield.
concentration to dryness, the resultant mixture was reslur-
ried using MTBE and hexane to afford the desired product
in reasonable yield.
O
O
1) LDA, 2-Me-THF
) DMF
NaOMe/MeOH
reflux
O
O
O
2
After this successful development we went back to inves-
tigate the use of readily available and affordable NaOMe
instead of LiOMe. We found that the improvements de-
scribed above did indeed allow for use of NaOMe instead
of LiOMe without significantly changing the reaction pro-
file. We also found that closely monitoring the reaction
H
H
Br
Br
OMe
4% over two steps
5
Scheme 2 Improved route to 3-methoxy-2-furaldehyde
Having developed the first step and with sufficient mate- progress using HPLC was critical. If the reaction was al-
rial in our hands, we turned our attention to the displace- lowed to continue after completion, an increase of impu-
ment of the bromide using methoxide. Using NaOMe in rities resulting in lower purity and yield was observed. As
MeOH resulted in the conversion of the starting material mentioned before, a more dilute reaction led to a cleaner
into the corresponding hemiacetal (indicated by HPLC– reaction and in order to further improve the yield, we in-
MS) and potentially also the acetal followed by progres- vestigated the possibility of simultaneously adding both
sive formation of the product. The displacement step oc- the starting material and NaOMe into a refluxing solution
curred cleanly up to approximately 30% conversion at of MeOH. This did suppress the dimerization side reac-
which point a number of unidentified side products began tions described above resulting in higher purity and im-
to form resulting in a modest 50% isolated yield of the de- proved yield. In order to avoid the first direct silica gel
sired product. Modifications of this protocol such as slow filtration, we developed a workup that employed an aque-
addition of the substrate, higher reaction temperature in ous acetic acid neutralization of the basic reaction mixture
sealed tube (105 °C) with a shorter reaction time, the use followed by evaporation of most of the MeOH. Extraction
of various co-solvents (DMF, DMSO or DME) and water of the resulting aqueous layer using dichloromethane re-
did not provide any improvement in conversion and/or sulted in good mass recovery after concentration. At-
yield. We next investigated the use of other methoxide tempts to reslurry this crude material directly were not
sources [KOMe, LiOMe, NaOMe and Mg(OMe) ] as well successful due to the presence of insoluble polymeric ma-
2
as the effect of additives such as CuI (0, 0.5, 1 and 2 terials, therefore a filtration through silica gel was used to
equiv). We concluded that KOMe with 0.5 equivalent of remove these more polar materials.
CuI or LiOMe (with or without CuI) gave the best conver-
sion to the desired product. As the addition of CuI compli-
cated the workup we decided to further investigate the use
of LiOMe without the addition of CuI. The reactions were
cleaner when using a less concentrated solution of LiOMe
In conclusion, a convenient two-step process to 3-meth-
oxy-2-furaldehyde has been developed using commer-
cially available 3-bromofuran as starting material
resulting in an overall yield of 50–60%.
6
(
10, 20 or 40 volumes were investigated) while the use of
Acknowledgment
drying agents such as Na SO or molecular sieves did not
2
4
appear to improve the reaction profile.
We thank Asymchem Laboratories Inc. and WuXi Apptec for ma-
nufacturing on large scale.
©
Thieme Stuttgart · New York
Synlett 2012, 23, 134–136

1
36
M. Ronn et al.
LETTER
citric acid solution (1500 mL, w/v). The layers were
separated and the aqueous layer was extracted once with
EtOAc (1000 mL). The combined organics were subse-
References and Notes
(1) For a review, see: Kappe, C. O.; Murphree, S. S.; Padwa, A.
Tetrahedron 1997, 53, 14179.
quently washed with H O (500 mL) and then with sat. aq
2
(
2) (a) Charest, M. G.; Lerner, C. D.; Brubaker, J. D.; Siegel, D.
R.; Myers, A. G. Science 2005, 308, 395. (b) Charest, M.
G.; Siegel, D. R.; Myers, A. G. J. Am. Chem. Soc. 2005, 127,
NaCl (500 mL). The organic layer was concentrated under
reduced pressure at 30 °C to give a dark brown oil (178 g)
along with some visible solids. The residue was diluted with
MTBE (100 mL) and filtered through a medium fritted
funnel. An additional portion of MTBE (100 mL) was used
to wash the filter. The filtrate was concentrated to give 3-
bromo-2-furaldehyde (164 g, 92%) as a dark brown oil.
Analytical data were in accordance with those of a
commercial sample.
Preparation of 3-Methoxy-2-furaldehyde: To a 50-L
reactor equipped with an overhead stirrer, nitrogen inlet and
reflux condenser was added anhyd MeOH (20 L) followed
by NaOMe (25 mL, 25 w/w% solution in MeOH) to assure
a basic media at start of addition. The solution was heated to
Ti = 64 °C. 3-Bromo-2-furaldehyde (1034 g) was charged to
a 1-L addition funnel and NaOMe (5 L, 25 w/w% in MeOH,
8
3
292. (c) Brubaker, J. D.; Myers, A. G. Org. Lett. 2007, 9,
523. (d) Sun, C.; Wang, Q.; Brubaker, J. D.; Wright, P. M.;
Lerner, C. D.; Noson, K.; Charest, M.; Siegel, D. R.; Wang,
Y.-M.; Myers, A. G. J. Am. Chem. Soc. 2008, 130, 17913.
3) (a) Lyapkalo, I. M.; Webel, M.; Reißig, H.-U. Eur. J. Org.
Chem. 2001, 4189. (b) Mesiter, C.; Scharf, H.-D. Synthesis
(
(
1
981, 733. (c) Mesiter, C.; Scharf, H.-D. Synthesis 1981,
737.
4) (a) Ly, N. D.; Schlosser, M. Helv. Chim. Acta 1977, 60,
085. (b) Antonioletti, R.; D’Auria, M.; De Mico, A.;
Piancatelli, G.; Scettri, A. J. Chem. Soc., Perkin Trans. 1
985, 1285. (c) Sornay, R.; Meunier, J.-M.; Fournari, P.
2
1
Bull. Soc. Chim. Fr. 1971, 990.
(
(
5) Two of the major impurities were tentatively assigned using
3.7 equiv) was charged into a second funnel. The NaOMe
1
HPLC–MS and H NMR as a dimer and trimer likely
and 3-bromofurfural in the respective addition funnels were
simultaneously added dropwise over 5 h, and the mixture
was subsequently stirred at 64 °C for 13 h. After the reaction
was complete, the mixture was cooled to 20 °C and stirred at
that temperature for 3 h. The temperature was lowered
further to –30 °C and 12% aq AcOH (10 L) was added
resulting in a rise of the temperature to about 5 °C. Most of
the MeOH (about 23 L) was removed at reduced pressure at
35 °C and the remaining mixture (about 12 L) was extracted
with CH Cl (2 × 10 L). The combined organics were
resulting from side reaction of the starting material with the
product.
6) General Experimental: All reactions were performed under
1
a nitrogen atmosphere. H NMR spectra were recorded using
an Oxford ASR400 spectrometer operating at 400 MHz at a
probe temperature of 25 °C. Assays by HPLC–MS analyses
were performed on an Agilent 1200 using a Zorbax C18
column and H O–MeCN mobile phases that included 0.1%
2
formic acid. 3-Bromofuran was obtained from Penn
Specialty Chemicals and was separated from aqueous
materials included for stabilization before use. All
commercially available starting materials, reagents and
solvents were used as received.
Preparation of 3-Bromo-2-furaldehyde: To a 5-L, 4-neck
round-bottomed flask was charged 2-methyltetrahydrofuran
1000 mL) followed by i-Pr NH (173 mL, 1.3 equiv). The
mixture was cooled to T = –10 °C and n-BuLi (410 mL, 2.5
M in hexanes, 1.0 equiv) was charged into a 1-L addition
funnel. The n-BuLi was added dropwise to the reactor over
5 min while keeping the batch temperature below –10 °C.
2
2
washed with half-saturated aq NaHCO (10 L). The organics
3
were concentrated at reduced pressure at 25 °C to give a
black crystalline material. Reslurry in MTBE (4 L) and
hexane (4 L) gave about 250 g of product as filterable solids.
Additional material (260 g of black amorphous material
which was stuck to the flask) along with 120 g from
evaporation of the mother liquors was combined, dissolved
in CH Cl (1.5 L) and passed through silica gel (500 mL)
(
2
i
2
2
which was further eluted with 1.5 L CH Cl . The CH Cl
2
2
2
2
was evaporated and the resulting solids were combined with
the first solids (250 g) followed by suspending the solids in
MTBE (3 L). Hexanes (3 L) were then added over 5 h at r.t.
After stirring for an additional 17 h the slurry was cooled to
1
The resulting pale yellow solution was cooled to –45 °C and
kept for 10 min at that temperature. 3-Bromofuran (neat,
liquid, 150 g) was charged while maintaining batch tempera-
ture below –40 °C. The dark brown suspension was stirred at
0
°C. The solids were collected by filtration and provided
after drying 3-MeO-2-furaldehyde (440 g, 59%) as a brown
solid.
–
45 °C for 30 min and DMF (108 mL, 1.4 equiv) was added
dropwise over 10 min via addition funnel while observing an
exotherm. Temperature was maintained below –40 °C
during addition. The reaction was completed after 10 min at
Analytical data were in accordance with those described in
2c
the literature.
–40 °C and the batch was quenched by the addition of 25%
Synlett 2012, 23, 134–136
© Thieme Stuttgart · New York

Copyright of Synlett is the property of Georg Thieme Verlag Stuttgart and its content may not be copied or
emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.
However, users may print, download, or email articles for individual use.