Synthesis of biobased N-methylpyrrolidone by one-pot cyclization and methylation of γ-aminobutyric acid

Article abstract of DOI:10.1039/c0gc00061b

N-Methylpyrrolidone (NMP) is an industrial solvent that is currently based on fossil resources. In order to prepare it in a biobased way, the possibility to synthesize NMP from γ-aminobutyric acid (GABA) was investigated, since GABA can be obtained from glutamic acid, an amino acid that is present in many plant proteins. Cyclization of GABA to 2-pyrrolidone and subsequent methylation of 2-pyrrolidone to NMP was achieved in a one-pot procedure, using methanol as the methylating agent and a halogen salt (i.e. ammonium bromide) as a catalyst. A selectivity above 90% was achieved, as well as a high conversion. Methylation of 2-pyrrolidone could also be done with dimethyl carbonate, but then the selectivity for NMP was less (67%).

Full text of DOI:10.1039/c0gc00061b

View Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/greenchem | Green Chemistry
Synthesis of biobased N-methylpyrrolidone by one-pot cyclization and
methylation of c-aminobutyric acid
Tijs M. Lammens,a Maurice C. R. Franssen,b Elinor L. Scott*a and Johan P. M. Sandersa
Received 29th April 2010, Accepted 3rd June 2010
DOI: 10.1039/c0gc00061b
N-Methylpyrrolidone (NMP) is an industrial solvent that is currently based on fossil resources. In
order to prepare it in a biobased way, the possibility to synthesize NMP from g-aminobutyric acid
(GABA) was investigated, since GABA can be obtained from glutamic acid, an amino acid that is
present in many plant proteins. Cyclization of GABA to 2-pyrrolidone and subsequent
methylation of 2-pyrrolidone to NMP was achieved in a one-pot procedure, using methanol as the
methylating agent and a halogen salt (i.e. ammonium bromide) as a catalyst. A selectivity above
90% was achieved, as well as a high conversion. Methylation of 2-pyrrolidone could also be done
with dimethyl carbonate, but then the selectivity for NMP was less (67%).
of 36 kton. If this would all be converted into NMP, it would
Introduction
result in an annual NMP production of 23 kton, a significant
In a world where fossil fuels will become scarce and therefore
part of the world market. Once glutamic acid becomes available
more and more expensive, it is important to look for alternative
in large volumes from byproduct streams such as DDGS as
sources for chemical products. That is not only the case for trans-
opposed to the current production method of fermentation,8
portation fuels, but also for industrial chemicals such as solvents
it will become less expensive. Then the production of NMP
and polymer precursors. One example of an industrial solvent
that is currently based on fossil resources is N-methylpyrrolidone
or NMP. NMP has high chemical and thermal stability making
from glutamic acid could become an interesting process for
chemical companies, if the process can be competitive with
current production technology.
it suitable for a range of applications including the use as a
Scheme 1 shows the current and the proposed processes for
solvent for plastics and as an ingredient in paint removers.1
the production of NMP. Currently NMP is made by reacting
It has been reported that NMP is safe and without any acute
butyrolactone with methylamine, at high temperature (200–
harmful effects,2 although there have been other reports of
350 ◦C) and pressure (100 bar). Methylamine is a corrosive and
toxicity developing upon inhalation or oral exposure.3
The global annual production of NMP is estimated to be 100–
highly flammable gas.10 So if NMP could be made by methylation
of 2-pyrrolidone, which is readily formed by the cyclization of
150 kton.3,4 From a commercial point of view, manufacturing
GABA,11 the NMP would become largely biomass based instead
this product from a biorefinery instead of from conventional
of fossil based, and its production process could become more
petrochemical products is potentially worthwhile, because the
energy-efficient, safer and dependent on fewer reagents. If the
production volume is large enough to consider investing in a new
methylating agent would be biobased as well, NMP production
process. In a previous paper we showed that glutamic acid, which
would become fully biobased.
can be derived from waste streams from biofuel production, can
Traditional methylating agents, such as dimethyl sulfate and
be enzymatically decarboxylated to form g-aminobutyric acid
(GABA), in a process that we expect to be both technically and
economically feasible.5 GABA could then be an intermediate
methyl iodide, are toxic and not environmentally friendly,
because in their use stoichiometric amounts of waste salts are
produced. In search of a better methylating agent, dimethyl
for the synthesis of a variety of nitrogen containing industrial
carbonate (DMC, produced from methanol, oxygen and carbon
chemicals, such as NMP.
monoxide) was shown to be capable of methylating vari-
Taking the large bioethanol production plant that the com-
ous molecules and forming only methanol and CO2 as side
pany Abengoa is currently building in the Port of Rotterdam in
products.12,13 In a reaction catalyzed by zeolites, DMC can
the Netherlands as an example, this plant will annually produce
methylate different phenols and amines.14 Ben Taleb et al.
around 500 million litres of grain-based ethanol, resulting in
showed that it is possible to use DMC to methylate amides, such
360 kton of the byproduct dried distiller’s grains with solubles
as acetamide and several lactams (e.g. 2-pyrrolidone), with the
(DDGS).6 Wheat DDGS can contain as much as 10 wt%
quaternary ammonium salt cetyl trimethylammonium bromide
glutamic acid,7 resulting in a potential glutamic acid stream
(CTAB) as a catalyst.15 Unfortunately, methanol did not show
any methylating activity under these conditions (220 ◦C in an au-
toclave). To perform a methylation reaction with just methanol
aValorisation of Plant Production Chains, Wageningen University and
would be highly advantageous as only water is generated as a
Research Centre, PO Box 17, 6700 AA, Wageningen, The Netherlands.
co-product and also the atom efficiency would be better than
using DMC. Oku et al. showed that N-methylation of several
amines with methanol is possible under supercritical conditions
E-mail: elinor.scott@wur.nl; Fax: +31 317 483011; Tel: +31 317 481172
bLaboratory of Organic Chemistry, Wageningen University and Research
Centre, Wageningen, The Netherlands.
1430 | Green Chem., 2010, 12, 1430–1436
This journal is
The Royal Society of Chemistry 2010
©

View Online
Scheme 1 Simplified overview of all process steps towards NMP. A: Fossil based;9 B: Biomass based.
with a bifunctional acid–base catalyst.16 A prerequisite for good
High resolution MS spectra were recorded on an Exactive
conversion with their catalyst is an ‘anchoring group’ on the
amine, such as an alcohol or another amine that can form a bond
with the surface of the catalyst. Therefore the catalyst worked
well for 2-aminoethanol, but resulted in a poor conversion with
molecules such as aniline and 2-pyrrolidone. Other methylation
reactions of lactams such as 2-pyrrolidone with methanol that
are reported in literature were performed in ◦the gas phase, at
temperatures that are typically as high as 400 C.17
The goal of this paper is to investigate the possibility to syn-
thesize NMP from GABA, in order to obtain biobased NMP. We
will show that it is possible to perform the cyclization of GABA
and the subsequent catalytic methylation of 2-pyrrolidone with
methanol in a one-pot procedure, under conditions that are
milder than those of the current process for the production
of NMP from butyrolactone, with a catalytic amount of an
ammonium salt such as CTAB or ammonium bromide. We
also investigated the possibility to use DMC in combination
with a NaY zeolite catalyst to perform the methylation of 2-
pyrrolidone. This will be compared with the use of methanol as
a methylating agent.
apparatus from Thermo Scientific, equipped with an ESI probe.
Spectra were recorded both in positive and negative mode. m/z
ratios were detected from 50 to 500.
GC-MS was performed with a Finnigan GC8000top appa-
ratus connected to a Finnigan Automass II quadrupole EI-
MS system. The used column was a BPX5 from SGE, 30 m x
0.25 mm x 0.25 mm. Helium carrier gas was applied at 100 kPa,
the temperature program 50–300 ◦C at 10 ◦C min-1. m/z ratios
were detected from 35 to 500.
Methylation procedure with methanol
In a typical experiment, a glass liner was charged with ammo-
nium bromide (59 mg, 0.60 mmol), 2-pyrrolidone (0.90 mL,
12 mmol) and methanol (3.0 mL, 74 mmol), and placed in an
autoclave. Before reaction, the atmosphere in the autoclave was
replaced with nitrogen by applying five vacuum-nitrogen cycles.
Each reactor was heated to 250 ◦C in 20 min and left at 250 ◦C
for five hours (the pressure in the reactor is then approximately
5 bar), after which heating was stopped and the reactor cooled to
100 ◦C in one hour and down to room temperature in five hours
at which point the reactor was opened and the contents removed.
Methanol was evaporated under reduced pressure and a crude
sample was taken for determining the conversion and selectivity
by 1H-NMR (400 MHz). GC-MS and high resolution ESI-MS
were further used to identify the formed products. No sampling
was performed during the reaction, in order to avoid interference
with the experiment. For purification, column chromatography
was used with a 4 : 1 (vol.) mixture of chloroform and ether
as the mobile phase. On TLC, the Rf values of NMP and
2-pyrrolidone with the same mobile phase were 0.2 and 0.1,
respectively.
Experimental
Materials and equipment
g-Aminobutyric acid (Fisher, >99%), 2-pyrrolidone (Sigma,
>99%), anhydrous methanol (Sigma, >99.8%), ammo-
nium bromide (Sigma, >99%), ammonium chloride (Sigma,
>99.5%), ammonium iodide (Fisher, >99%), cetyltrimethy-
lammonium bromide (CTAB, Fisher, >99%), cesium bromide
(Sigma, >99.5%), sodium bromide (Sigma, >99%), 1-butyl-
3-methylimidazolium bromide (BMIM, Sigma, >98.5%), and
methyltriphenylphosphonium bromide (MTPB, Sigma, >98%)
were all used as received. Dimethyl carbonate (DMC, Sigma,
>99%) was dried by refluxing over sodium sulfate and distilled
under N2 prior to use. NaY (CBV100, Zeolyst) was activated
overnight at 70 ◦C under vacuum prior to use.
Methylation procedure with DMC
In a typical experiment, a glass liner was charged with NaY
zeolite (0.50 g), 2-pyrrolidone (0.50 mL, 6.6 mmol), DMC (8 mL,
0.1 mol) and diglyme (0.20 ml, 1.4 mmol, internal standard)
and placed in an autoclave. Before reaction, the atmosphere
in the autoclave was replaced with nitrogen by applying five
vacuum-nitrogen cycles. The reactors were heated to the required
All reactions were performed in a Parr Series 5000 Multiple
Reactor System, with six stainless steel autoclaves of 75 mL
internal volume used in parallel.
This journal is
The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 1430–1436 | 1431
©

View Online
temperature and kept at this temperature for the set amount of
time, after which heating was stopped and the reactor cooled
down to room temperature in about five hours at which point
the reactor was opened and the contents removed. DMC was
then evaporated under reduced pressure and a crude sample
was taken for determining the conversion and selectivity by
1H-NMR (400 MHz). Yield, conversion and selectivity were
proposed reaction sequence, shown in Scheme 2, is analogous
to the one proposed for the methylation with DMC or methyl
formate.15
1
determined by comparison of the H-NMR signal of diglyme
with those of 2-pyrrolidone and NMP. GC-MS was further used
to identify the formed products. No sampling was performed
during the reaction, in order to avoid interference with the
experiment.
Scheme 2 Proposed reaction sequence for the methylation of pyrroli-
done. In this scheme, NH4 can be substituted by any of the bromide
salts.
Results and discussion
As the selectivity is not 100%, the formation of side products
was investigated. Five side products were identified by high
resolution MS (Table 1): g-hydroxybutyric acid (GHB), g-
hydroxybutyric acid methyl ester (MGHB), g-butyrolactone
(GBL), trimethylamine and tetramethylamine. The formation of
these compounds was unexpected, as it means that the nitrogen
atom from the lactam, which is a poor leaving group, has
been replaced by oxygen. Scheme 3 shows a suggested reaction
mechanism for the formation of these side products. Methylation
of NMP would turn the nitrogen atom into a good leaving group,
making the amide susceptible to a nucleophilic ring opening.
Further methylation then yields a quaternary nitrogen that is
eventually substituted by water, giving GHB and trimethylamine
as products. GHB can cyclize to GBL, or react with methanol
to form MGHB. Trimethylamine can be methylated to form a
tetramethylammonium salt.
Catalyst screening
Ben Taleb et al. showed that CTAB can be used for the
methylation of 2-pyrrolidone with DMC.15 Here the use of
different halogen salts as catalysts for the methylation of 2-
pyrrolidone with methanol was investigated. This was done in
order to determine the possibility to use CTAB as a catalyst, to
investigate the possibility of using other halogen salts, and to
obtain mechanistic information.
As can be seen in Fig. 1, bromide containing salts
such as cetyl trimethylammonium bromide (CTAB), 1-butyl-
3-methylimidazolium bromide ([Bmim]Br), methyltriphenyl-
phosphonium bromide (MTPB) and ammonium bromide show
an excellent ability to catalyze the methylation with methanol
◦
at 250 C. Similar results were also achieved with ammonium
bromide and ammonium iodide. Ammonium bromide was
chosen as the preferred catalyst for further studies. Although it
does not provide the highest conversion, it was chosen because
it has the same high selectivity and is a readily available and
inexpensive salt.
Strengthening the postulated reaction mechanism is that both
the presence of GHB, MGHB and GBL were detected (by H-
NMR and ESI-MS), and also the presence of trimethylamine
and the tetramethylammonium salt (by ESI-MS). The latter two
molecules could have been formed by methylation of ammonium
bromide, but they were also detected when MTPB was used
as catalyst. As MTPB contains no nitrogen, the methylated
amines must originate from 2-pyrrolidone and are likely to be
side products from the formation of GHB. Table 1 shows the
exact masses and the anticipated products, from high resolution
ESI-MS in both positive and negative mode.
Ammonium bromide catalyzed methylation
The influence of pressure and temperature on the conversion and
selectivity of the reaction were studied with ammonium bromide
as a catalyst, to determine the optimal reaction conditions. Our
results showed that an increased pressure does not lead to a
significant change in conversion, but does lead to a decrease
in NMP selectivity (from >90% to 70%), due to the formation
of more of the side products GHB, MGHB and GBL. The
best pressure to perform this reaction was found to be 5 bar
(obtained by starting with atmospheric pressure and heating the
closed autoclave).
Fig. 1 Conversion and selectivity determined for the methylation of
2-pyrrolidone with methanol in the presence of different catalysts.
Experiments were carried out with 0.6 mmol catalyst, 1.0 g (12 mmol)
2-pyrrolidone and 3.0 mL (74 mmol) methanol, at 250 ◦C, for 5 h.
Reaction mechanism
The fact that ammonium chloride is less active than ammonium
bromide which in turn is less active than ammonium iodide
(Fig. 1), leads us to believe that a halomethane is involved
in the reaction sequence. In such a mechanism, the hydroxy
group of methanol is replaced by a halide ion in an SN2
reaction, which is easiest for a good nucleophile like iodide.18 The
Fig. 2 shows that there is a temperature threshold value
between 200 and 225 ◦C, from which the reaction rate increased
dramatically. At very high temperatures the selectivity of the
reaction decreases due to the formation of a black tar.
A study of conversion and selectivity in time (Fig. 3) shows
that next to a high selectivity it is also possible to achieve high
1432 | Green Chem., 2010, 12, 1430–1436
This journal is
The Royal Society of Chemistry 2010
©

View Online
Table 1 Measured m/z ratios with the molecular formula and the anticipated products
m/z
Molecular formula (theoretical mass)
Derived compound
Positive mode
Negative mode
58.0656
60.0812
74.0967
86.0602
100.0759
78.9178,
80.9157
84.0444
85.0284
103.0390
117.0574
126.9040
C3H8N (58.0651)
C3H10N (60.0808)
C4H12N (74.0964)
C4H8ON (86.0600)
C5H10ON (100.0757)
Br (78.9178)
Trimethylamine (-H)
Trimethylamine (+H)
Tetramethylammonium
2-Pyrrolidone (+H)
N-Methylpyrrolidone (+H)
Bromide (2 isotopes)
C4H6ON (84.0444)
C4H5O2 (85.0284)
C4H7O3 (103.0390)
C5H9O3 (117.0546)
I (126.9039)
2-Pyrrolidone (-H)
g-Butyrolactone (-H)
g-Hydroxybutyric acid (-H)
g-Hydroxybutyric acid methyl ester (-H)
Iodide
Scheme 3 Suggested reaction mechanism for the formation of the side-product GHB from methylated 2-pyrrolidone. GHB may then react with
methanol to form MGHB or cyclize to GBL.
Fig. 2 Conversion and selectivity as a function of the reaction
temperature. Experiments were carried out with 59 mg NH4Br, 1.0 g
2-pyrrolidone and 3.0 mL methanol, for 5 h. Error margins indicate
the differences between two duplicate experiments. Selectivity at low
temperature could not be accurately determined because of the low
conversion.
Fig. 3 Conversion and selectivity as a function of the reaction time.
Experiments were carried out with 59 mg NH4Br, 1.0 g 2-pyrrolidone
and 3.0 mL methanol, at 250 ◦C. Error margins indicate the standard
deviation between two or three identical experiments. Fitting the data
points from 1 to 9 h to a linear function gives an R2 value of 0.99.
Scope of the alkylation
conversions, by allowing the reaction to proceed for a longer
period of time. After nine hours, the conversion achieved is
88 6% and the corresponding selectivity 87 4%. While it
is possible to obtain high conversions with this reaction, the
employed reaction time in our studies is five hours, as it offers
a high selectivity in combination with a medium conversion,
making it possible to study how other factors such as the reaction
temperature can influence the conversion.
To investigate the scope of the alkylation reaction, it was
also performed with other, more sterically hindered alcohols.
Chosen for this were: ethanol, n-propanol, n-butanol, 2-butanol,
t-butanol and benzyl alcohol. The results are shown in Table 2.
For the sake of comparison, the ratios (determined by 1H-NMR
without evaporating the alcohol) between 2-pyrrolidone and the
N-alkylated pyrrolidone (NAP) are shown, which are a good
indication of the reaction progress after five hours.
An isolated yield of 57 mol% NMP was obtained after
purification of the product of a five-hour experiment at 250 ◦C
by column chromatography, which is consistent with the data
shown in Fig. 3.
These results show that methanol is more reactive than
ethanol, n-propanol and n-butanol. This is probably related
to the initial substitution of the alcohol with bromide, which
proceeds more readily when it is less sterically hindered. In
This journal is
The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 1430–1436 | 1433
©

View Online
Table 2 N-alkylation of 2-pyrrolidone, with different alcohols. Shown
here is the ratio between 2-pyrrolidone and N-alkylated pyrrolidone
(NAP) in the crude reaction mixture, after 5 h reaction time. Experiments
were carried out with 59 mg NH4Br, 1.0 g 2-pyrrolidone and 72 mmol
alcohol, at 250 ◦C. Error margins indicate the variation between two
identical experiments
under the applied conditions. However, when the reaction was
performed at an increased pressure, NMP became unstable with
ammonium bromide present and formed 17 mol% of the GHB-
related side products (data not shown), which is in accordance
with the postulated mechanism in Scheme 3.
NAP/Pyrrolidone (mol/mol)
Process design
Methanol
Ethanol
n-Propanol
n-Butanol
2-Butanol
t-Butanol
Benzyl alcohol
1.2 0.3
0.15 0.03
0.07 0.01
0.16 0.02
0
In an industrial process design, the removal of water from the
reactor could be incorporated in a continuous process, together
with continuous product removal by vacuum distillation. 2-
Pyrrolidone and NMP are easily separable by distillation (their
boiling points at 1 atm are 245 and 202 ◦C, respectively), so
one could think of a product stream continuously leaving the
reactor, passing through a water absorber (for example a zeolite
drier) into a first distillation column for the removal of the excess
of methanol and then into a second distillation column, where
the top fraction would be NMP and the bottom fraction, 2-
pyrrolidone, could be either isolated as a co-product, or could
be recycled to the reactor. Schematically this process is shown
in Fig. 5. What should be considered is the fate of the catalyst.
As the salt is a homogeneous catalyst, part of it will be removed
from the reactor with the product removal and will end up in
the rest of the process. A part will probably stay in the zeolite
dryer where it may be recovered and the rest will end up in the
2-pyrrolidone stream and can directly be recycled to the reactor.
When warm 2-pyrrolidone is recycled, ammonium bromide will
stay in solution, but it could also be cooled to room temperature
at which point the ammonium bromide is sparingly soluble in
2-pyrrolidone and could be filtered off to a large extent.
0
1.4 0.1
between ethanol, n-propanol and n-butanol, no trend could be
observed. 2- and t-butanol yielded no alkylated product. In their
case, elimination of the alcohol probably took place instead of
bromination, leading to the formation of gaseous (iso)butene,
as an increase of pressure (from 3 to 7 bar) was observed
during the reaction. Another indication for the formation of
butenes was that after the reaction, 2- and t-butanol had been
partially consumed in the reaction mixture. In the case of benzyl
alcohol, N-benzylpyrrolidone was formed at a similar rate as
N-methylpyrrolidone. In this case the mechanism can be SN2
or SN1, because it is well known that adjacent p-bonds enhance
both mechanisms.
One-pot cyclization and methylation of GABA
From literature, it is known that GABA can cyclize rather easily,
for example in boiling toluene, catalyzed by silica using a soxhlet
set-up for water removal.11 Therefore our hypothesis was that it
should be possible to cyclize and subsequently methylate GABA
in a simple one-pot procedure, in order to obtain biobased
NMP. Fig. 4 shows that upon heating GABA in methanol
without added catalyst, 2-pyrrolidone is indeed obtained in
a high yield, even without water removal. When GABA was
heated in methanol with ammonium bromide present, there
was only 2 mol% GABA left after five hours at 250 ◦C and
the subsequent conversion of 2-pyrrolidone was 42 mol%, with
a selectivity for NMP of 92%. This means that it is actually
quite straightforward to convert GABA to NMP in a one-pot
procedure, with methanol and a catalytic amount (5 mol%) of
ammonium bromide. Fig. 4 also shows that when starting with
NMP, 99% of it was recovered, indicating that NMP is stable
Fig. 5 Process scheme for the continuous production of NMP from
GABA and methanol.
Comparison with DMC as methylating agent
Although the atom efficiency of the above described process
is relatively high (73%), the use of a homogeneous halogen-
based catalyst may be perceived as a drawback in a commercial
application, for there is the risk of product contamination with
traces of bromide. Oku et al. reported an acid–base bifunctional
zeolite that catalyzes methylation reactions with supercritical
methanol, but for 2-pyrrolidone the conversion was quite poor.16
As was discussed in the introduction, DMC is a versatile
methylating agent, also in combination with heterogeneous
catalysts such as NaY zeolite.12–14 However, to our knowledge
there is no report of the use of DMC for the methylation of
a (cyclic) amide, in combination with a heterogeneous catalyst.
Fig. 4 Molar ratio of the products present in the reaction mixture after
5 h at 250 ◦C, when starting with different materials, in the presence and
absence of NH4Br catalyst. Determined by 1H-NMR.
1434 | Green Chem., 2010, 12, 1430–1436
This journal is
The Royal Society of Chemistry 2010
©

View Online
Table 3 Results of N-methylation procedure with DMC, in the presence and absence of NaY
Temp/◦C
Catalyst
Time/min
2-Pyrrolidone Conversion (mol%)
NMP Selectivity (mol%)
NMP Yield (mol%)
130
130
130
180
180
180
250
250
250
—
60
8
7
5
76
81
84
90
86
85
0
0
0
4
17
23
48
66
67
0
0
0
3
14
19
43
57
57
NaY
NaY
—
NaY
NaY
—
120
240
60
60
240
60
NaY
NaY
30
60
Therefore we attempted the methylation of 2-pyrrolidone with
this reagent, in the presence of NaY zeolite. The results, shown
in Table 3, indicate that it is also possible to methylate 2-
pyrrolidone with DMC. At a temperature of 250 ◦C, a maximum
NMP yield of 57 mol% was obtained after 30–60 min of
reaction. The results show that a high temperature is needed
for the N-methylation of 2-pyrrolidone with DMC, because
at 180 ◦C the NMP yield goes down while the conversion of
2-pyrrolidone remains the same, indicating that the selectivity
towards NMP goes down, and at 130 ◦C 2-pyrrolidone is hardly
although we found that at an increased pressure the selectivity
towards NMP goes down.
The methylation of 2-pyrrolidone can also be done with DMC
in the presence of NaY zeolite. However, the selectivity that was
found for this methylation procedure (67%) was significantly less
than that of the methylation with methanol and a halogen salt
catalyst. Combining this with the lower atom efficiency in the
use of DMC, the preferred method of methylation would be the
procedure with methanol.
This paper shows that there is now a straight-forward route
to synthesize biobased NMP, based on glutamic acid. This
can be done by combining the enzymatic decarboxylation of
glutamic acid to form GABA with the one-pot cyclization and
methylation of GABA to form NMP.
◦
converted. At lower temperatures (130 and 180 C), we found
that 2-pyrrolidone is partly N-methoxycarbonylated instead of
N-methylated. This is in line with previous reports by Selva
et al., who indicated that at a lower temperature DMC acts as
a methoxycarbonylating agent and at higher temperatures as a
methylating agent, although there is not always a clear cut-off
at which temperature each reactions occurs.12
Acknowledgements
Both at 180 and 250 ◦C, the results indicate that NaY has
a directive effect towards the methylation reaction, because the
selectivity towards NMP formation increases with the zeolite
present, although this effect is not large.
We wish to thank Barend van Lagen for performing NMR
measurements and Frank Claassen for the MS measurements.
Furthermore, we are grateful to NWO-Aspect for funding of
this work.
When comparing these results with the results that were
obtained with the methylation procedure with methanol, they
seem rather poor. Although the DMC reaction is quicker, both
the atom efficiency (51%) and the achieved selectivity (67%)
with DMC are less than with methanol, which leads to the
conclusion that the methanol procedure is preferable over the
DMC procedure.
Notes and references
1 http://www2.basf.us/diols/bcdiolsnmp.html Official website of
BASF chemical company, Germany (accessed Mar 30, 2010).
2 http://worldaccount.basf.com/wa/PublicMSDS/PDF/
09007af88017cc55.pdf Official website of BASF chemical company,
Germany (accessed Mar 30, 2010).
3 B. Flick, C. E. Talsness, R. Ja¨ckh, R. Buesen and S. Klug, Toxicol.
Appl. Pharmacol., 2009, 237, 154–167.
4 M. Reich, Chem. Eng. News, 2008, 86, 32.
Conclusion
5 T. M. Lammens, D. De Biase, M. C. R. Franssen, E. L. Scott and
J. P. M. Sanders, Green Chem., 2009, 11, 1562–1567.
6 http://www.abengoabioenergy.com Official website of Abengoa
Bioenergy, Spain (accessed Mar 30, 2010).
The goal of this paper was to investigate the possibility to
synthesize NMP from GABA, that can be obtained by the
a-decarboxylation of glutamic acid.5 In this way NMP could
be made largely biobased, and with the use of a biobased
methylating agent (for example biomethanol19) NMP would
become fully biobased.
The synthesis of NMP from GABA was done in two steps,
the first being the cyclization of GABA to form 2-pyrrolidone,
and the second the N-methylation of 2-pyrrolidone to form
NMP. We found that this is possible in a one-pot procedure,
where the methylation reaction can be performed with methanol
as the methylating agent, catalyzed by a halogen salt such
as ammonium bromide, ammonium iodide or CTAB. The
selectivity that was achieved for this reaction is greater than 90%,
7 V. Deike, J. De Mol, Wheat distillers grains and solubles, in: Raw
Material Compendium, 2nd edition reprint, 1996, p. 117, Novus
International Inc. Brussels.
8 T. Kawakita1, C. Sano, S. Shioya, M. Takehara and S. Yamaguchi,
Monosodium Glutamate, in: Ullmann’s Encyclopedia of Industrial
Chemistry, Online edn, 2000, Wiley InterScience.
9 W. Schwarz, J. Schossig, R. Rossbacher, and H. Ho¨ke, Butyrolactone,
in: Ullmann’s Encyclopedia of Industrial Chemistry, Online edn, 2000,
Wiley InterScience.
10 http://www.praxair.com/praxair.nsf/AllContent/25C6766DFFF-
78DC785256A860081A221/$File/p4626d.pdf, Official website of
Praxair (accessed Mar 30, 2010).
11 A. Blade-Font, Tetrahedron Lett., 1980, 21, 2443.
12 P. Tundo and M. Selva, Acc. Chem. Res., 2002, 35, 706–716.
13 M. Selva and A. Perosa, Green Chem., 2008, 10, 457–464.
This journal is
The Royal Society of Chemistry 2010
Green Chem., 2010, 12, 1430–1436 | 1435
©

View Online
14 (a) M. Selva, E. Militello and M. Fabris, Green Chem., 2008, 10,
73–79; (b) M. Selva, A. Perosa and M. Fabris, Green Chem., 2008,
10, 1068–1077.
17 (a) T. Setoyama and T Ogoshi, JP2001288164A, 2001; (b) T.
Setoyama and T Ogoshi, JP2001288165A, 2001.
18 P. Y. Bruice, Substitution reactions of alkyl halides, in: Organic
Chemistry, 2nd edition, 1998, Prentice-Hall, Inc. New Jersey.
19 (a) http://www.biomcn.eu Official website of BioMCN, The Nether-
lands (accessed Mar 30, 2010); (b) A. Demirbas, Energy Sources, Part
A: Recovery, Util. Environ. Eff., 2008, 30, 565–572.
15 A. Ben Taleb and G. Jenner, J. Mol. Catal., 1993, 84, L131–
L136.
16 T. Oku, Y. Arita, H. Tsuneki and T. Ikariya, J. Am. Chem. Soc., 2004,
126, 7368–7377.
1436 | Green Chem., 2010, 12, 1430–1436
This journal is
The Royal Society of Chemistry 2010
©