condensed at y90 uC at the top of the rectification column. The
yield of methyl pentenoate collected over consecutive distillate
fractions (called ‘cumulative yield’ in Fig. 1) increased almost
linearly with time up to y80 mol%. The production rate remained
fairly high at 2.0 molproduct molcat21 h21 or 1.3 gproduct gcat21 h21
for most of the run. As the reaction approached completion, the
temperature of the reaction medium was raised to 230–250 uC and
the pentenoate content of the distillate (labelled ‘distillate
concentration’ in Fig. 1) dropped from y40 wt% to zero. At
that point, the yield of methyl pentenoates had reached 98%. The
esters consisted of a mixture of various isomers, namely the pent-4-
enoate (25–35%), the cis/trans pent-3-enoates (65–75%; mainly the
trans-isomer) and the cis/trans pent-2-enaotes (1–5%). The
distribution of these isomers changed slightly with time and
conversion. Traces of pentenoic acid isomers were the only side
products detected in the distillate. No dimethyl ether could be
detected in the N2 purge stream, even upon passing it through a
liquid N2 trap. The reaction residue consisted mainly of the methyl
ester of the pTSA catalyst, with small amounts of methyl
pentenoates and impurities that were present in the feed.
Scheme 3 Lactone feedstock.
with methyl pentenoate and pTSA in the presence or absence of
c-valerolactone.
Attempts to produce pentenoic acids by co-feeding water
instead of alcohols provided yields of 2–4 mol% after 7 h,
depending on the water feed rate. In this case, again, the product
has too high a boiling point (196–206 uC) to be easily distilled from
the reaction medium. Moreover, the reaction is thermodynamically
unfavorable, as mentioned above.
Substituting c-valerolactone (GVL) by other lactones encoun-
tered mixed success (Scheme 3). d-Hexanolactone (DHL) was
converted to the target methyl hexenoates with 70 mol% yield and
y80 wt% distillate concentration. Like c-valerolactone, this
lactone leads to a hydroxy ester intermediate with a secondary
hydroxyl group that is easily eliminated under acidic conditions. In
contrast, the c-butyrolactone (GBL), which leads to a primary
hydroxy-ester, formed the target methyl butenoates in trace
amounts and produced many oligomeric products.
Variations in reaction conditions affected the production rate
and concentration of methyl pentenoate in the distillate. For
instance, the production rate was reduced upon diluting the
c-valerolactone with sulfolane but remained unaffected by changes
in catalyst loading or methanol feed rate. The concentration of
methyl pentenoate in the distillate increased upon increasing the
catalyst loading and upon decreasing the methanol feed rate.
Substituting methanol for higher alcohols resulted in lower
yields of the corresponding pentenoate esters. The reaction
proceeded to lower conversion and was accompanied by the
formation of heavy products and some ethers (Fig. 2). This trend is
most likely due to the increased difficulty of withdrawing the
heavier alkyl pentenoates from the reaction medium by means of
distillation.
Several acidic and basic catalysts were also investigated under
these conditions. All acidic catalysts tested, whether homogeneous
or heterogeneous, showed some activity in this reaction. As
reported in Table 1, H2SO4 showed the highest initial activity,
followed by pTSA. Heterogeneous catalysts were also active, with
the initial activity (based on weight and H+) decreasing in the order
of their acid strength, i.e.
Nafion NR50 . zeolites . amorphous silica alumina (ASA)
It should be noticed that the Nafion and zeolites were more
active than the homogeneous acids when expressed per proton.
The Nafion NR50 is a strongly acidic resin. The zeolites include a
medium-pore H-ZSM-5 and two large-pore H-Beta zeolites,
having Si/Al atomic ratios of 15, 10 and 100, respectively.
Interestingly, the pore size and Si/Al ratio seem to have a limited
effect on the initial activity of the zeolite. All acidic catalysts
showed a comparable distribution of pentenoate isomers, a
comparably low coproduction of pentenoic acid and negligible
formation of diethyl ether or heavy products.
The benefit of distilling the ester was further confirmed by a
batch experiment, in which c-valerolactone, methanol and pTSA
were reacted for 22 h at 200 uC. The product contained no more
than 4 mol% methyl pentenoates, with traces of pentenoic acid and
dimethyl ether. This low yield could not be explained by a
degradation of the pentenoate esters during the batch experiment,
since it did not react during complementary batch experiments
Amphoteric or truly basic catalysts, such as c-alumina, La-
doped c-alumina and KOH, showed no significant activity in
producing methyl pentenoate or in converting c-valerolactone to
any significant level. Such catalysts are known to be active in
transesterification reactions. Their lack of activity here is most
likely due to their inability to dehydrate the hydroxy ester
intermediate under these mild conditions (Scheme 2).
In summary, the present work reports the possibility of con-
verting c-valerolactone and methanol to methyl pentenoate, under
mild conditions using strong acid catalysts, when performed under
catalytic distillation conditions. The reaction produces a distillate
that is rich in methyl pentenoates, alongside methanol, water and
minute amounts of pentenoic acid. The methyl pentenoates can be
further converted to nylon intermediates such as adipic acid,
dimethyl adipate or caprolactam. This provides an efficient route
for producing Nylon polymers from renewable feedstock.
Fig. 2 Composition of the bottom and top product of the transester-
ification of c-valerolactone with alcohols to alkyl pentenoates (after 25 h at
200 uC with 100 mol GVL, 3 wt% pTSA and alcohol feed rates of 25, 15
and 10 mol h21 for MeOH, EtOH and BuOH, respectively).
This journal is ß The Royal Society of Chemistry 2007
Chem. Commun., 2007, 3488–3490 | 3489
Lange, Jean-Paul
