D. Sieben et al. / Tetrahedron Letters 57 (2016) 808–810
809
O
O
O
O
N
catalyst
+
R1
OH
+
R
O
R1OH
R
O
O
O
N
BF3
1 solvolytic
N
O
2a-g
3a-g
1
O
O
yield: 8% - 94%
f g
O
4a/b
2a
2/3
a
b
c
d
e
Scheme 2. Supposed enol ether prepared under Lewis acid conditions.
R
C3H7 C5H11 C7H15 C9H19 C11H23 C13H27 C15H31
R1 = Et, Me
toluenesulfonic acid showed a rising amount of the desired pro-
duct over time without significant amounts of by-products. After
24 h of reflux the reaction seemed to have reached an equilibrium
at about 80% of conversion to the nicotinyl ester. In order to shift
the equilibrium, we removed the resultant ethanol from the solu-
tion by using an apparatus equipped with a soxhlet device filled
with molecular sieve 4 Å. Refluxing the mixture overnight gave
quantitative conversion and a total yield of 82% of the desired pro-
duct after workup and distillation. We were able to apply the reac-
tion conditions to the other even 3-oxo esters, which also resulted
in good purified yields as shown in Table 1.
The fact that removal of the ethanol was crucial for completing
the reaction was in accordance with Bader et al. who transesteri-
fied ethyl esters by continuous distilling off the ethanol formed
during the reaction.4a However, contrary to the good results
reported with simple alcohols, we obtained only poor
conversions when heating a mixture of 1 and 2a up to 100 °C in
a short way distillation apparatus to expel the ethanol. Again,
Scheme 1. Preparation of 3-oxo fatty acid nicotinyl esters from the respective alkyl
esters using alkaline and acidic catalysts. Yields depend on the catalyst used and the
rest R of the starting compound. See Table 1.
neatly. Yields were improved considerably by increasing the
reaction temperature to 100 °C. After 1 h, 70% of the starting
material of an equimolar reaction mixture was converted to the
desired nicotinyl ester, the rest to undefined matter. Due to these
contaminations purification of the product could only be
accomplished by repeated chromatography to give an overall
yield of 23% of the nicotinyl ester. Nevertheless, the conditions
could be transferred to the homologous even 3-oxo fatty acid
esters up to the octadecanoic acid ester with an average yield of
25% (Table 1).
Contrary to the simple fatty acid esters, the 3-oxo species have
highly acidic
a-hydrogens and the high amount of by-products
might have been due to reactions from the enolate form of the
ester. To avoid these adverse reactions and to improve the yield,
we tested a weaker base. Indeed, when using zinc(II)oxide instead
of the alcoholate and a ten-fold excess of 1, yields were more than
doubled (Table 1).
According to the published reports transesterification of the
oxo-esters should theoretically be possible either under basic or
acidic catalysis or even under neutral conditions.4 Since basic
catalysis showed poor to moderate yields it might be interesting
to find out how acidic catalysis would improve the yields.
Similar to the preparation of fatty acid methyl esters from
neither
a ten-fold excess of nicotinyl alcohol nor of 3-oxo
hexanoic acid ethyl ester improved the results. Except for an
increasing amount of side and decomposition products only up
to 10% conversion to the nicotinyl ester could be detected and
only traces of ethanol distilled off using
a bulb to bulb
distillation apparatus. We hypothesized that according to
Raoult’s law the vapor pressure of the formed ethanol was very
low at 100 °C when solved in excess nicotinyl alcohol. Therefore,
solvolytic conditions did not seem to be the right choice. The
previous experiments had shown that azeotropic distillation with
toluene did expel the alcohol so that it could be absorbed by the
molecular sieve. The question was whether an acidic catalysis
was necessary. So the reaction was repeated without the sulfonic
acid catalyst. Refluxing 1 and 2a in toluene 24 h over molecular
sieve in fact led to quantitative conversion to the desired
nicotinyl ester. As no elaborate workup was needed the yields of
the even 3-oxo fatty acid nicotinyl esters from C6 to C18
improved even more (Table 1). Interestingly, the respective 3-
hydroxy fatty acid ethyl or methyl esters gave no conversion
under the same reaction conditions.
triglycerides 50 ll of 2a was solved in 5 ml of a 10% solution of
BF3 in 1.9 The mixture was stirred first at room temperature for
4 days and then for 96 h at 60 °C but none of the desired product
was detected by GC/MS and HPLC. Except for lots of adverse reac-
tion products the chromatograms and spectra only showed two
isomeric substances. Their m/z values were according to the iso-
meric Z/E enol ethers 4a and 4b but the substances were not yet
isolated or characterized (Scheme 2).
Again, when the temperature was increased to 100 °C we found
about 50% conversion to the desired nicotinyl ester after 24 h,
however, plenty of by-products as well. Due to the excess of nico-
tinyl alcohol, purification of the desired product was not
successful.
Increasing the temperature may be preferable over an excess of
one reactant as a mean to improve yields. Indeed, an equimolar
reaction mixture refluxed in toluene with a catalytic amount of
Acknowledgement
We gratefully acknowledge the financial support from the
Hochschule Darmstadt.
Supplementary data
Supplementary data (full experimental details, chromatograms
and copies of 1H NMR, 13C NMR, IR and HRMS spectra) associated
with this article can be found, in the online version, at http://
Table 1
Isolated yields (%) of 3-oxo fatty acid nicotinyl esters after transesterification of the
respective ethyl or methyl esters with and without different catalysts
Catalyst
Mol sieve
3a
3b
3c
3d
3e
3f
3g
tert-Butylate
—
—
4 Å
4 Å
23
73
82
93
22
60
75
92
25
50
75
94
18
35
69
91
31
44
66
92
27
46
72
92
8
34
58
82
References and notes
Zinc oxide
H+
—
Mol sieve 4 Å was used to absorb formed ethanol and methanol.
Sieben, Daniela
