One-Pot Selective Catalytic Synthesis of Pyrrolidone Derivatives from Ethyl Levulinate and Nitro Compounds

Article abstract of DOI:10.1002/cssc.201601333

N-substituted 5-methyl-2-pyrrolidones were prepared in a one-pot process starting from ethyl levulinate and nitro compounds in the presence of a nanosized Pt-based catalyst. Pt supported on TiO₂nanotubes (Pt/TiO₂-NT) catalyzed the synthesis of N-substituted 5-methyl-2-pyrrolidones through a cascade process involving the reduction of nitro compounds, formation of the intermediary imine, hydrogenation, and subsequent cyclization. A bifunctional metal–acid system was a suitable catalyst for the process. Pt supported on TiO₂showed lower catalytic activity than Pt/TiO₂-NT owing to the strong adsorption of nitro compounds during the first reaction step that poisoned the acidic sites and strongly decreased the rate of amination and cyclization. However, Pt/TiO₂-NT with milder acid sites was less affected by the adsorption of nitro compounds and the full cascade process could proceed. The results indicate that the prepared Pt/TiO₂-NT is a chemoselective and reusable catalyst that can be applied to the synthesis of a variety of N-substituted 5-methyl-2-pyrrolidones starting from nitro compounds with excellent yields in absence of an additional organic solvent under mild reaction conditions.

Full text of DOI:10.1002/cssc.201601333

Accepted Article
Title: One-pot selective catalytic synthesis of pyrrolidone derivatives
from ethyl levulinate and nitro compounds
Authors: Avelino Corma, JUAN DIEGO VIDAL, MARIA JOSE
CLIMENT, PATRICIA CONCEPCION, and SARA IBORRA
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemSusChem 10.1002/cssc.201601333
Link to VoR: http://dx.doi.org/10.1002/cssc.201601333
A Journal of
www.chemsuschem.org

ChemSusChem
10.1002/cssc.201601333
1
One-pot selective catalytic synthesis of pyrrolidone derivatives from
ethyl levulinate and nitro compounds
Juan D. Vidal, Maria J. Climent, Avelino Corma*, Patricia Concepcion, and Sara Iborra*
Instituto de Tecnología Química (UPV-CSIC). Avda dels Tarongers s/n, Universitat
Politècnica de València, 46022, Valencia, Spain
*
To whom correspondence should be addressed
E-mail
acorma@itq.upv.es
siborra@itq.upv.es
Fax:
(+34) 963877809
(+34) 963877800
Phone:
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
2
Abstract
N-substituted-5-methyl-2-pyrrolidones have been prepared in one pot process
starting from ethyl levulinate and nitro compounds and in the presence of nanosized
platinum based catalyst.
Pt supported on TiO nanotubes (Pt/TiO -NT) catalyzes the synthesis of N-substituted-
2
2
5
-methyl-2-pyrrolidones through a cascade process which involves the reduction of
nitro compounds, formation of the intermediary imine, hydrogenation and subsequent
cyclization. A bifunctional metal-acid system works for the process. Pt on TiO gives
2
lower catalytic activity than Pt/TiO -NT due to the strong adsorption of nitro
2
compounds during the first reaction step that poisons the acidic sites and strongly
decreases the rate of amination and cyclization steps. On the other hand, Pt/TiO -NT
2
with milder acid sites is less affected by the adsorption of nitro compounds and the full
cascade process can take place. The results indicate that the prepared Pt/TiO -NT is a
2
chemoselective and reusable catalyst that can be applied to the synthesis of a variety
of N-substituted-5-methyl-2-pyrrolidones starting from nitro compounds with
excellent yields in absence of an additional organic solvent under mild reaction
conditions .
Keywords: Ethyl levulinate, nitro compounds, reductive amination, platinum nanotube
titanate catalyst, pyrrolidones.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
3
Introduction
The current dependence of fossil feedstoks and associated environmental problems
have promoted an important research in the transformation of biomass as a
1
2
renewable source of carbon for the production of fuels and chemicals. The
transformation of platform molecules, derived from lignocellulosic biomass is
3
recognized as a key approach to convert biomass into fuels and chemicals. In this way,
levulinic acid (LA) and its esters which can be produced trough a low-cost route by
2
c
acidic dehydration of carbohydrates (it has been estimated that it can be produced at
4
less than 0.50 euro/L ) are among the most promising chemicals from biomass
5
3b
feedstock for the production of a variety of interesting derivatives. Among them, N-
alkyl-5-methyl-2-pyrrolidones, which can be obtained by reductive amination of
levulinic acid (or esters), are value-added products widely used as surfactants,
solvents, intermediates for pharmaceuticals, dispersants in fuel additives
6
compositions, and in the manufacture of agrochemicals, etc.
N-substituted-5-methyl-2-piyrrolidones can be produced by reductive amination of
levulinic acid (LA) or its esters with amines, using molecular hydrogen in the presence
of a hydrogenation catalyst. The process involves several steps that starts with the
formation of an intermediate imine by the nucleophilic attack of the amino group to
the ketonic group of LA (or ester). Then, the hydrogenation of imine to amino group
followed by the subsequent cyclization of the amino ester formed gives the
corresponding N-substituted- 5-methyl-2-pyrrolidone (Scheme 1).
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
4
Since formic acid is produced in equimolecular amounts with levulinic acid during the
transformation of carbohydrates under acidic aqueous conditions, a variety of recent
studies report the use of formic acid as a source of hydrogen for the reductive
7
8
amination of LA in the presence of homogeneous and heterogeneous catalysts. More
recently, efficient catalyst-free transformations of levulinic acid into pyrrolidones using
formic acid have been reported, however a stoichiometric amount of triethylamine in
9
10
DMSO as a solvent or high pressure and reaction temperature are required.
From an economical and ecological point of view the use of molecular hydrogen in the
presence of a heterogeneous catalyst to transform LA, or its esters, into N-alkyl-5-
methyl-2-pyrrolidones is an attractive approach, although the number of studies
reported in literature are scarce. In the liquid phase, it has been described the
6
,11
reductive amination of LA or its ester with amines , using different metals (Ni, Pd, Pt,
Ru, Rh, and Ir) supported on carbon or metal oxides such as SiO and Al O and over
2
2 3
1
2
carbon-supported FeNi nanoparticles. However, 55-85 bar hydrogen pressures,
o
organic solvents such as dioxane, ethanol or 2-methyltetrahydrofuran and 150-180 C
reaction temperature were required. Under those reaction conditions the selectivity to
pyrrolidones strongly depends on the nature of the amine involved, and particularly
when amines bearing aromatic rings or other reducible groups are used, over reduced
products are often formed decreasing the selectivity to the pyrrolidone. Recently,
1
3
Touchy et al. have reported that Pt and MoO co-loaded TiO (Pt-MoO /TiO ) is an
x
2
x
2
efficient catalyst for the reductive amination of LA with primary amines under solvent-
o
free and mild (100 C, 3bar of H ) conditions. However long reaction times (20 h) were
2
required, while the catalyst did not show chemoselectivity when amines bearing
hydrogen-accepting groups such as C-C double and triple bonds are used. More
1
4
o
15
recently, Pt/TiO2 activated at 450 C and Pd/ZrO2 have showed excellent activity for
the reductive amination of ethyl levulinate and levulinic acid respectively with primary
amines giving the corresponding N-substituted-5-methyl-2-pyrrolidones in high
1
4
conversion and selectivity. Particularly, we showed in previous work that Pt/TiO
2
with Pt crystal faces decorated with TiO was an excellent and selective catalyst to
2
,
obtain N-substituted-5-methyl-2-pyrrolidones from EL and primary amines under
solvent free conditions, even when other groups susceptible of hydrogenation were
present in the amine moiety. We showed that the controlling step of the reaction is
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
5
the formation of the imine intermediate catalyzed by the acid sites of the catalyst, and
that the preferential adsorption of the imine against other reducible groups can be
controlled by controlling the crystal size of the metal. Thus, the Pt/TiO sample with
2
the lowest crystal size, corresponding to a sample with a 0.2 wt % Pt loading presented
the highest activity and selectivity to 2-pyrrolidones.
On the other hand, one pot procedures involving multiple catalytic events have
become an important area of research in organic chemistry since they improve energy
and raw materials consumption, atom economy and lower E factor (Kg subproduct/Kg
1
6
product) . Then, since Pt/TiO has showed excellent chemoselectivity for reducing
2
1
7
nitro compounds , we envisaged that this catalyst could be effective for the
production of N-substituted-5-methyl-2-pyrrolydones in one pot process starting from
nitro compounds and alkyl levulinates. As far as we know, only one report in the
1
8
patent literature describes the synthesis of pyrrolidones by one pot starting from
nitro compounds and ethyl levulinate (EL). There, Manzer, reported that N-
substituted- 5-methyl-2-pyrrolidones can be produced by the treatment of EL and nitro
compounds using solvents under molecular H and in the presence of a hydrogenation
2
catalyst (palladium, ruthenium, rhenium, rhodium, iridium, platinum, nickel, cobalt,
copper, iron and osmium supported on different oxides). However, the main
drawbacks in the example showed are the use of organic solvents, the requirement of
o
a high excess of H pressures in the range of 48-69 bar, 150-200 C reaction
2
temperature and high substrate to catalyst ratio. Under those conditions selectivity to
the desired product was limited. For instance, starting from nitrobenzene and working
o
at 200 C under 69 bar of hydrogen, 97% conversion of ethyl levulinate with 32%
selectivity to N-phenyl-5-methyl-2-pyrrolidone can be achieved after 24h reaction
time.
The catalytic activity of metal supported catalyst can be modified by tuning not only
the size, morphology and composition of the metal particles but also the nature of the
support. The discovery of titanate nanotubes opens a possibility to use it as support for
metal catalysts, since they posses large surface area and can effectively disperse the
1
9
loading metal due to strong metal-support interaction. Particularly, platinum
supported on TiO -Nanotubes (Pt/TiO -NT) is a catalytic system recently studied for
2
2
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
6
2
0
the hydrogenation of cinnamaldehyde, while other applications are related with their
2
1
photocatalytic properties.
In the present work it will be shown that Pt supported on nanotubes of titania
Pt/TiO -NT) is an excellent chemoselective catalyst for the production of N-
(
2
substituted-5-methyl-2-pyrrolidones starting from nitro compounds and ethyl
levulinate in a one pot process. We will show that on the contrary Pt/TiO gives much
2
lower yields due to a strong adsorption of the nitro compound that strongly deplete
the available active sites on the surface of the catalyst. These results show the
relevance of adequately coupling adsorption-reaction for achieving chemoselective
hydrogenations.
Experimental section
Commercially available organic and inorganic compounds (from Sigma-Aldrich and
Degussa) were used without further purification. GC (Agilent GC-7980A) and GC-MS
(Agilent GC 6890N and MS 5973N) analyses were carried out with HP-5 capillary
column (Agilent) using N and He as the carrier gas.
2
Catalysts Preparation.
0
.2w% Pt/TiO catalyst was prepared by a modified deposition-precipitation (DP)
2
method by mixing TiO (P25) powders (5g) with appropriate amounts of aqueous
2
solutions of hexachloroplatinic acid (10mL). After over-night stirring at room
o
temperature under inert atmosphere, the catalyst was dried at 100 C in a forced air
o
22
oven for 5h, followed by reduction with a stream of H at 450 C for 3h. The BET
2
2
-1
surface area of the resultant Pt/TiO catalyst was 50m g . The concentration of
2
platinum was 0.2% Pt by weight determined by ICP-AES analysis.
2
3
TiO -Nanotubes (TiO -N) were prepared following the reference using commercial
2
2
TiO powder as starting material. The TiO powder (1g) was dispersed in 15 mL of an
2
2
aqueous solution of NaOH (10M) and charged into a teflon-lined autoclave, which was
o
heated in an oven at 150 C for 20h. The precipitate was filtered; the solid was
suspended with deionized water, filtered and washed repeatedly until pH of the
filtrate was near 7–8. Then, the solid was treated with a 100mL of HNO 0.1M solution,
3
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
7
filtered again and washed until pH of the filtrate was between 5–7. The material was
o
dried overnight in an air-circulating oven at 120 C.
0
.2wt% Pt/TiO -NT catalyst was prepared by impregnation of TiO -N with an
2
2
aqueous solution of hexachloroplatinic acid (H PtCl ) at room temperature under
2
6
o
stirring. The solvent was then evaporated and then dried at 100 C in a forced air oven
o
for 5h, followed by reduction with a stream of H at 450 C for 3h. The BET surface area
2
2
–1
of the resultant Pt/TiO -NT catalyst was 260m g . The concentration of platinum was
2
0
.2wt% Pt by weight (ICP-AES analysis). Metal nanoparticle size was measured by TEM
and the physicochemical characteristic of the catalysts are presented in Table S1.
Catalytic Tests
Catalytic experiments were performed in absence of solvent under H atmosphere
2
in a 2mL glass reactor. For each reaction, a mixture of EL (3 mmol), nitro compounds (1
mmol), supported metal catalysts (metal 0.05mol %), and dodecane (0.1mmol,) as
internal standard were charged into the reactor. After sealing the reactor, air was
purged by flushing twice with 10bar of hydrogen. Then, the mixture of substrates and
catalyst was heated to the desired temperature in less than 2 minutes and it was
pressurized with H at the targeted reaction pressure.
2
During the experiment, the pressure was maintained constant and the stirring rate
was fixed at 1000 rpm. Aliquots were taken from the reactor at different reaction
times and the product composition was analyzed on Agilent GC-7980A gas
chromatograph equipped with a capillary column HP-5 (30m × 0.25mm x 0.25m) and
FID detector. The identification of the products was performed by using a GC-MS
0
-1
spectrometer. The initial reaction rates (r (mmol h )) were calculated in all cases for a
conversion of EL below 10%
Catalyst characterization
TEM analysis: Transmission electron microscopy (TEM) images for supported platinum
catalysts were taken with a JEOL 2100 electron microscope operating at 200 kV. Before
being transferred into the TEM chamber, the samples dispersed with dichloromethane
were deposited onto a carbon-coated copper grid and then quickly moved into the
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
8
vacuum evaporator. The size distribution of the metal nanoclusters was determined by
measuring about 200 random particles on the images.
Additionally, to reveal subtle details of the surface composition of the Pt
nanoparticles in the catalysts, nanoanalysis by STEM-EELS have been performed. The
Pt loading of the catalysts was measured by inductively coupled plasma atomic
emission spectroscopy (ICP-AES) using a Varian 715-ES spectrometer.
IR spectroscopic experiments
IR spectroscopy of the catalysts and adsorbed CO and nitrobenzene was carried out
with a Bruker Vertex 70 spectrophotometer with detector DTGS-TEC, and resolution 4
-
1
cm and 32 scans.
2
Briefly, the samples 10 mg/cm self-supporting pellets of the catalyst powder were
o
pre-activated in the IR cell at 200 C for 2h in hydrogen, followed by 1h outgassing at
o
-6
2
50 C and 10 mbar dynamic vacuum. For the FTIR-CO experiments, the samples, after
o
pre-activation, were cooled down to -175 C and CO was adsorbed at increasing dosing
(0.2mbar-2mbar). Spectra were collected at each dosing. For the nitrobenzene FTIR
o
studies, the samples, after pre-activation, were cooled down to 25 C and nitrobenzene
was adsorbed until saturation (0.2-0.4mbar). Afterwards the sample has been
-6
evacuated at room temperature under dynamic vacuum (10 mbar) followed by raising
o
o
the temperature to 70 C and 120 C. Spectra were recorded at each temperature. For
spectra treatment ORIGIN software was used and all spectra have been normalized to
sample weight.
X-ray diffraction (XRD) Characterization
XRD patterns analysis were obtained by X-ray powder diffraction (XRD) with a
Panalytical CUBIX diffractometer using monochromatic Cu Kα radiation (λ=
0
.15417nm) at 45kV and 40mA. The angle (2θ) was measured in a scan range between
2
.00° and 90.03° in steps of 0.0401° with a counting time of 34.92s.
Hydrogen chemisorption
The metal dispersion of the catalysts was measured by hydrogen chemisorptions using
a Quantachrome Instruments (Autosorb-1) under a hydrogen pressure from 0 mm Hg
o
to 800 mm Hg and isotherm of 40 C.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
9
Results and discussion
Catalytic activity of Pt/TiO -NT and Pt/TiO in the reductive amination of EL with
2
2
amines
To check the influence of the morphology of TiO support on the catalytic activity for
2
the synthesis of N-aryl pyrrolidones, titania nanotubes was prepared and Pt was
supported by wetness impregnation (Pt loading 0.2wt%, see experimental section) and
decorating the exposed (111) and (110) Pt crystal faces with TiO from the support by
2
o
14
activation under hydrogen flow at 450 C for 3 hours. TEM images of the different
prepared materials as well as the most important physicochemical characteristic of the
samples used in this work are presented in Table S1 . As expected, BET surface area of
TiO -NT based materials are considerably larger than those prepared with regular TiO ,
2
2
while the size of Pt nanoparticles are similar (in the range of 1.4-1.8 nm) for all the Pt
supported catalysts. Moreover, hydrogen chemisorption experiments showed that the
metal dispersion on both catalysts was similar (see Table S2). For comparison, in Table
S2 is included the dispersion of Pt over non-decorated catalysts (the catalysts were
o
activated at 250 C under hydrogen flow). It can be observed around 80% dispersion
for the two non-decorated samples, while the metal dispersion is reduced to 5% for
the Pt/TiO and Pt/TiO -NT catalysts after decoration.
2
2
The reductive amination of EL was carried out first no with the nitro derivative but
with two different amines (aniline and 2-phenylethylamine) under 10 bar of hydrogen,
o
in absence of solvent at 120 C, using 0.2wt% Pt/TiO -NT and 0.2wt% Pt/TiO (both
2
2
o
activated under H atmosphere at 450 C) as catalysts. As can be observed in Table 1,
2
excellent conversions and selectivities to pyrrolidones were achieved. However, when
turnover frequency (TOF) calculated on the bases on initial reaction rate were
compared for the two catalysts, it can be observed that the catalytic activity of the
Pt/TiO -NT was lower and longer reaction times were required to achieve high
2
conversions (Table 1).
To check if the lower activity of the Pt/TiO -NT could be due to changes in the
2
structure of the nanotubular Pt/TiO -NT caused by the reduction process, the XRD
2
o
analysis of a series of the Pt/TiO -NT samples reduced between 150 and 650 C was
2
carried out. XRD results (Figure S1) demonstrate that the nanotubular structure of the
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
1
0
o
Pt/TiO -NT is preserved after reduction with H at 450 C, while when the reduction is
2
2
o
performed at temperatures higher than 500 C the nanotubular structure completely
disappears giving more crystalline samples with a dominant anatase phase.
At this point, and if one considers that both catalysts possess similar Pt nanoparticle
size, the different catalytic activity should be attributed to the different microstructure
of the support. We have demonstrated previously that in the reductive amination of
ethyl levulinate to pyrrolidones using Pt/TiO as catalyst, the controlling reaction step
2
1
4
was the imine formation that could be catalyzed by acid sites of the support. Then, it
could be that the difference in activity of Pt/TiO and Pt/TiO -NT is related to
2
2
differences in acidity of the support. To check that, the reaction between aniline and
EL was performed in absence of H using Pt/TiO and Pt/TiO -NT as catalysts. Reaction
2
2
2
results presented in Figures 1 and 2 show that the imine intermediate was formed at
higher initial reaction rate on Pt/TiO than on Pt/TiO -NT. Moreover the acidity of
2
2
both catalysts was measured by IR spectroscopy of CO as probe molecule showing a
higher amount (c.a twofold) and a higher strength of Lewis acid sites in the
0
.2wt%Pt/TiO sample versus the 0.2wt%Pt/TiO -NT sample (see figure S2A).
2
2
-1
CO adsorption on 0.2wt% Pt/TiO sample shows two IR bands at 2178 and 2172 cm
2
4
+
24
associated to Ti Lewis acid sites of the TiO support, and two IR bands at 2164 and
2
-
1
2
155 cm corresponding to TiO hydroxyl groups, evidenced by the corresponding shift
2
in the OH region. Similar peak positions have been observed on the pure TiO support.
2
2
+
Contribution in the aforementioned IR bands due to CO interacting with Pt cannot be
2
5
discarded, though if present, it should be low due to H pre-reduction of the samples
2
-1
prior to the FTIR-CO studies. In addition, an IR band at 2097 cm related to CO
26
interacting with Pt metal sites is clearly observed (Figure S2B). In opposite, in the
-
1
0
.2wt% Pt/TiO -NT sample, two IR bands at lower frequencies (2168 and 2161 cm )
2
4
+
are observed which could be associated to less acid Ti -Lewis acid sites, while
2
+
contribution of Pt can again not be discarded. Since hydroxyl groups have not been
-1
-1
observed in the 4000-3000 cm IR region, the low frequency IR band at 2150 cm has
+
24
-1
to be related to less oxidised Pt surface sites. In addition, the IR band at 2085 cm
(Figure S2B) corresponds to CO interacting with more unsaturated Pt metal sites than
2
7
in the 0.2wt% Pt/TiO sample.
2
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
1
1
From these results we can conclude that the lower activity of Pt/TiO -NT in the
2
reductive amination of EL with respect to Pt/TiO can be attributed to their lower
2
acidity.
One-pot synthesis of N-substituted-5-methyl-2-piyrrolidones starting from nitro
compounds
Encouraged by our previous results on the synthesis of N-substituted-5-methyl-2-
piyrrolidones by reductive amination of EL with amines, and considering the good
performance and chemoselectivity of Pt/TiO for reducing nitro groups into amino
2
22
groups, we thought that this material would be an excellent candidate to perform
the synthesis of pyrrolidones starting from nitro compounds and ethyl levulinate (EL)
in one-pot process (Scheme 2).
First, we performed the one pot reaction starting from nitrobenzene and EL, under 10
o
bar of hydrogen at 120 C in absence of solvent and using 0.2wt% Pt/TiO as catalyst.
2
As expected, nitrobenzene was converted almost quantitatively into aniline after 2h
reaction (Figure 3), however low yield of 5-methyl-N-phenyl-2-pyrrolidone (9%) at 19%
EL conversion was produced after 6h of reaction time. Other products such as N,N-
dicyclohexyl amine (3) and N-cyclohexyl-N-phenyl amine (4) (see Scheme 3), which can
1
4,17
be formed by over reduction of nitrobenzene followed by coupling reactions,
also detected in minor amounts (Scheme 3).
were
The low catalytic activity showed by Pt/TiO for the one-pot process, suggests that
2
during the first step, i.e. reduction of the nitro group, the catalyst becomes deactivated
for performing the reductive amination of EL. To study the deactivation of the catalyst,
the used solid was thoroughly extracted with methanol using a Soxhlet apparatus and
then analyzed by Temperature Programmed Desorption Mass Spectrometry (TPD-MS).
The results indicate the presence of nitrobenzene. In fact, TG analysis of the catalyst,
after Soxhlet extraction, showed that 4 wt % of organic material was retained on the
catalyst surface. Moreover, an IR spectroscopy analysis of the catalyst after reaction
-
1
showed absorption bands at 1350 and 1525 cm corresponding to nitrobenzene (see
Figure 4). These results may suggest that the remaining nitrobenzene adsorbed on the
catalyst during the first step of the reaction could be the cause of catalyst deactivation
for the second step.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
1
2
In fact, when the catalyst used in the first step, i.e. hydrogenation of nitrobenzene,
was used in the reductive amination of EL with aniline, the yield of pyrrolidone was
very low (see Figure 5) showing that the catalyst became strongly deactivated during
the first step.
We performed an additional experiment starting from aniline and EL, in where
different percentages of nitrobenzene (NB) (<1 wt%) were added to the reaction
mixture. Figure 6 shows a considerable decrease in the yield of pyrrolidone when
nitrobenzene was added in amounts between 0.4 and 1 wt% respect to EL.
At this point, and in order to elucidate if deactivation of the catalysts affects the
formation of imine and/or to the imine hydrogenation during the second step (see
Scheme 2), both process were studied separately. Thus, the imine intermediate was
o
obtained from EL and aniline in absence of hydrogen at 120 C in the presence of a
catalyst previously used in the first step, i.e. for nitrobenzene hydrogenation. Results
are depicted in Figure 7 and were compared with the ones obtained when the reaction
was carried out using fresh catalysts in the same reaction conditions (Figure 1). As can
be observed in Figure 7, the deactivation of the catalyst strongly affect the imine
formation. Taking into account that formation of imine should be promoted by Ti(IV)
Lewis acid sites of the support, these results suggest that Lewis acid sites become
strongly poisoned by the organic adsorbed on the catalyst.
On the other hand, the catalytic activity of the catalyst used in the first step (the
reduction of nitrobenzene) was tested in the hydrogenation of imine formed by
reaction between EL and aniline and compared with the catalytic activity of the fresh
catalyst. Figure 8 shows the yield of pyrrolidone as a function of time in the imine
hydrogenation with fresh and used catalyst. An important decrease in the rate of
formation of pyrrolidone when the reaction was carried out in the presence of used
catalyst was observed. In fact the initial rate for imine hydrogenation to give
pyrrolidone with the fresh catalyst was 0.57 mmol/h, while in the presence of used
catalyst the initial hydrogenation rate was reduced to one half (0.25 mmol/h).
The results presented above for the formation of 5-methyl-N-phenyl-2-pyrrolidone
from nitrobenzene and EL in one pot allows to conclude that during the first step
(nitrobenzene hydrogenation) a strong deactivation of the catalyst occurs owing to the
strong adsorption of the remaining nitrobenzene on the acid sites of the support. It
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
1
3
appears that adsorption has a negative effect on both processes, the imine formation
and imine hydrogenation, being the former the most strongly affected. Then,
1
4
considering that the controlling step of the reaction is the imine formation, the
nitrobenzene adsorption is the major cause of the low performance observed for the
one pot process with Pt/TiO2.
One pot process on Pt supported on Titania nanotubes
At this point, the Pt/TiO -NT catalyst was tested in the synthesis of 5-methyl-N-phenyl-
2
2
-pyrrolidone in one pot, starting from nitrobenzene and ethyl levulinate. Results in
Figure 9 show that the initial rate of reduction of the nitro group using the Pt/TiO -NT
2
-
1
-1
(
0.27 molh ) was slower that when using the regular Pt/TiO (0.40 molh ) (see Figure
2
3
).
However, unlike Pt/TiO , Pt/TiO -NT was able to promote efficiently the subsequent
2
2
reductive amination of ethyl levulinate with the aniline formed, achieving high
conversion and selectivity to 5-methyl-N-phenyl-2-pyrrolidone. These results appear to
indicate that in this case the deactivation of the catalyst by the strong absorption of
nitrobenzene on the acid sites was negligible. In fact, when the used Pt/TiO -NT
2
catalyst was analysed by IR spectroscopy no absorption bands of nitrobenzene at 1350
-
1
and 1525 cm were detected (see Figure 10), while the total amount of organic
retained on the catalyst was lower than 2 wt%.
To explain these results we studied the interaction strength of nitrobenzene on
both catalysts by means of FTIR spectroscopy. As can be seen in Figure S3,
nitrobenzene is adsorbed on two sites on both catalysts, i.e on the TiO support, with
2
-
1
-1

(NO)s at 1528cm , and on Pt-Ti interface sites, with (NO)s at 1518 cm . Similar
adsorption sites has already been observed on other systems and associated to nitro
2
8
group adsorbed on the support and on metal-support interface sites. For analysing
the stability of the adsorbed nitrobenzene IR spectra were recorded at increasing
o
o
temperature (70 C and 120 C), and different interaction strength could be observed
for the different adsorption sites depending on the catalysts. In detail, on the 0.2wt%
o
-1
Pt/TiO -NT sample the IR peak maxima at 120 C appears at 1518cm (associated to
2
o
interface sites), while on the 0.2wt% Pt/TiO sample the IR peak maxima at 120 C
2
-
1
appears at 1528cm (due to the support). Thus, on 0.2wt% Pt/TiO sample a higher
2
interaction of the nitro group with support sites is observed, while the opposite trend
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
1
4
is observed on 0.2wt% Pt/TiO -NT sample, where higher interaction strength with
2
interface sites versus support sites is observed. This result agrees with the higher acid
strength of Lewis acid sites observed in the 0.2wt%Pt/TiO sample (Figure S2A) being
2
responsible of a fast deactivation of the catalyst by a strong adsorption of
nitrobenzene by blocking Ti(IV) Lewis sites. On the other hand, the higher interaction
strength of the nitro group on Pt-Ti interface sites in the 0.2wt% Pt/TiO -NT sample is
2
related to the higher unsaturation degree of metal Pt surface sites, as observed in
0
Figure S2B, where a lower CO stretching frequency due to Pt -CO is observed. This
higher interaction lead to a lower hydrogenation activity, being Pt-Ti interface sites
2
9
active sites in the hydrogenation step.
Stability and reusability of Pt/TiO -NT catalyst
2
The stability and reusability of the Pt/TiO -NT catalyst was tested in the one pot
2
o
process starting from nitrobenzene and EL, at 120 C and in absence of solvent. After
the first cycle, the catalyst was removed by filtering and washed thoroughly with
methanol before each reuse (see experimental section). As can be observed in Figure
1
1, the catalyst could be reused in three consecutive cycles observing a small decrease
th
in catalytic activity. After the 4 cycle the catalyst was regenerated by calcination at
o
4
50 C and the initial activity and selectivity was recovered.
Scope of the reaction
To show the scope of the reaction, the one-pot process for the production of 5-methyl-
N-aryl-pyrrolidones was performed with EL and different nitro compounds. Since,
some of the nitro compounds selected are solid, in this case the EL was used as
reactant and as a solvent, and reactions were performed using an EL:nitro compound
o
molar ratio of 3:1, at 120 C under 10 bar of hydrogen. The results obtained for the
different substrates are presented in Table 2. Excellent selectivities to the
corresponding 5-methyl-2-pyrrolidones were achieved in all cases, also when using
nitro compounds bearing reducible groups such as nitrile, aldehyde, ketone and
alkenes (entries 4, 5,7, 9 and 12). In these cases products corresponding to the
reduced form of the different functions were not obtained, and the only byproduct
detected in all cases was the imine intermediate. The high chemoselectivity exhibited
by Pt/TiO -NT for reducing nitro groups in the presence of other reducible groups is in
2
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
1
5
good agreement with previously reported results on the reduction of nitroaromatics
1
4, 22
using Pt/TiO2 catalyst.
On the other hand, it can be observed that the highest conversion is achieved with
nitrobenzene (entry 1), whereas when another group is present in the aromatic ring
(regardless of their electron withdrawing o donor character) the conversion achieved is
lower. This could be related with steric effects for adsorption of reagents on the active
site of the catalyst. Notice, that in the case of nitroaromatics with a group in ortho
position the conversion is sensibly affected (compare for instance entry 8 and 13). For
comparison purposes in Table 2 (Entry 2) the reported results obtained in the synthesis
of 5-methyl-N-phenyl-2-pyrrolidone in one pot process using the 5wt%Pt-Al O at 150
2
3
o
18a
C, under 51.7 bar of hydrogen are presented.
Conclusions
It has been shown that it is possible to synthesize pyrrolidone derivatives directly from
EL and nitro compounds in a one pot reaction in a chemoselective way without an
additional organic solvent and under mild conditions with Pt supported on titania
nanotubes. The reaction requires a fine tuning of the hydrogenating and acid function
of the catalyst. The hydrogenating function works for the chemoselective nitro and
imine reduction while the acidic function catalyzes the imine formation. Too strong
acidities become deactivated by the starting nitro compounds and the cascade process
can not continue after formation of the amine. Pt/TiO shows higher rate for
2
hydrogenation of the nitro function but the catalyst becomes deactivate for
performing the subsequent amination and cyclization reactions. Pt supported on
titania nanotubes can perform the cascade process without observing deactivation
after the nitro compound hydrogenation step.
Acknowledgments
Financial support by Consolider –Ingenio 2010 (Project Multicat), Spanish MICINN
Project (CTQ-2015-67592-P), Generalitat Valenciana (Prometeo Program) and Program
Severo Ochoa are gratefully acknowledged.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
1
6
References
1) (a T. P. Vispute, H. Zhang, A. Sanna, R. Xiao, G.W. Huber, Science 2010, 330, 1222-
227; b) M. J. Climent, A. Corma, S. Iborra, Green Chem. 2014, 16, 516-547.
2) a) A. Corma, S. Iborra, A. Velty, Chem. Rev. 2007, 107, 2411-2502; b) M. J. Climent,
A. Corma, S. Iborra, Green Chem. 2011, 13, 3, 520-540; c) P. Gallezot, Chem. Soc. Rev.
012, 41, 1538-1558.
3) a) T. Werpy, G. R. Petersen, Top Value Added Chemicals from Biomass. Volume I.
(
1
(
2
(
Results of Screening for Potential Candidates from Sugars and Synthesis Gas, U.S.D.
Energy, 2004; b) J. J. Bozell, G. R. Petersen, Green Chem. 2010, 12, 539-554.
(
4) L. Manzer, “Biomass Derivatives: A Sustainable Source of Chemicals” Chapter 4 (41-
1) - Feedstocks for the Future. ACS Symposium Series; American Chemical Society:
Washington, DC 2006.
5) E. I. Gürbüz, D. M. Alonso, J. Q. Bond, J. A. Dumesic, ChemSusChem 2011, 4, 357.
DOI: 10.1002/cssc.201000396.
6) a) L. E. Manzer, US Patent 6743819 B1, 2004; b) L. E. Manzer, US Patent 0247443
A1, 2006.
7) a) Y. Wei, C. Wang, X. Jiang, D. Xue, J. Li, J. Xiao, Chem. Comm. 2013, 49, 5408-5410;
b) Y. B. Huang, J-J. Dai, X-J. Deng, Y-C. Qu, Q-X. Guo, Y. Fu, ChemSusChem 2011, 4,
5
(
(
(
1
578-1581; c) C. Ortiz-Cervantes, M. Flores-Alamo, J. J. García, Tetrahedron Letters, 57,
016, 766–771.
2
(
8) X-L. Du, L. He, S. Zhao, Y-M. Liu, Y. Cao, H-Y. He, K-N. Fan, Angew. Chem. Int. Ed.
011, 50, 7815-7819.
9) Y. Wei, C. Wang, X. Jiang, D. Xue, Z. T. Liu, J. Xiao, Green Chem.. 2014, 16, 1093-
096.
10) A. Ledoux, L. Sandjong Kuigwa, E. Framery, B. Andrioletti, Green Chem. 2015, 17,
251-3254.
11) a) L. E. Manzer US Patent 7129362 B2, 2006; b) L. E. Manzer US Patent US
743819 B1, 2004; c) L. E. Manzer Patent US 0247444 A1, 2006.
2
(
1
(
3
(
6
(12) G. Chieffi, M. Braun, D. Esposito, ChemSusChem 2015, 8, 3590-3594.
(13) A. S. Touchy, S.M.A.H. Siddiki, K. Kon, K. Shimizu, ACS Catal. 2014, 4, 3045-3050.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
1
7
(
14) J. D. Vidal, M. J. Climent, P. Concepcion, A. Corma, S. Iborra, M. J. Sabater, ACS
Catal. 2015, 5, 5812–5821, DOI: 10.1021/acscatal.5b01113.
15) J. Zhang, B. Xie, L. Wang, X. Yi, C. Wang, G. Wang, Z. Daí, A. Zheng, F.S. Xiao,
ChemCatChem 2016, 8, 1-8.
16) a) M. J. Climent, A. Corma, S. Iborra, Chem. Rev. 2011, 111, 1072, DOI:
(
(
1
5
0.1021/cr1002084; b) M. J. Climent, A. Corma, S. Iborra, RSC Advances, 2012, 2, 16-
8.
(
17) H. Yoshida, N. Igarashi, S. Fujita, J. Panpranot, M. Arai, Catal. Lett. 2015, 145, 606–
11. DOI: 10.1007/s10562-014-1404-4.
18) a) L. E. Manzer, US Patent 6855731 B2, 2005; b) L. E. Manzer, US Patent 6818593
B2, 2004.
19) X. Yang, X. Yu, L. Long, T. Wang, L. Ma, L. Wu, Y. Bai, X. Li, S. Liao, Chem. Commun.
014, 50, 2794-2796. DOI: 10.1039/C3CC49331H.
20) a) Chin-Yao Hsu, Tsai-Chin Chiu, Meng-Hung Shih, Wei-Je Tsai, Wei-Yi Chen, Chiu-
6
(
(
2
(
Hsun Lin, J. Phys. Chem. C, 2010, 114 (10), 4502–4510, DOI: 10.1021/jp9095198; b) T.
C. Chiu, H. Y. Lee, P. H. Li, J. H. Chao, O. H. Lin Nanotechnology 24, 2013, 11, 115601,
DOI: 10.1088/0957-4484/24/11/115601
(21) Baolin Zhu, Kairong Li, Shurong Wang, Shoumin Zhang, Shihua Wu, Weiping
Huang, Journal of Dispersion Science and Technology, 29(10), 2008, 1408-1411,
DOI: 10.1080/01932690802313311; b) Hu Xiao-Jing, J. Chil. Chem. Soc. 57, 1, 2012,
1
008-1011, DOI: 10.4067/S0717-97072012000100012: c) T. Kubo, H. Nagata, M.
Takeuchi, M. Matsuoka, M. Anpo, A. Nakahira, Res. Chem. Intermed. Vol. 34, No. 4,
008, 339–346, DOI: 10.1163/156856708784040605.
22) a) P. Serna, M. Lopez-Haro, J. J. Calvino, A. Corma, J. Catal. 2009, 263, 328; b) M. J.
Climent, A. Corma, S. Iborra, L. Santos Chemistry-A European Journal 2009, 15, 8834-
841. M.J. Climent, A, Corma, S. Iborra, L. Marti ACS Catalysis 2015, 5, 157-166.
23) T. Kasuga, M. Hiramatsu, A. Hoson, T. Sekino, K. Niihara, Langmuir 1998, 14,
160-3163, DOI: 10.1021/la9713816.
2
(
8
(
3
(
(
(
24) K.I. Hadjiivanov, G.N. Vayssilov, Adv. Catal. 2002, 47, 307-511.
25) K. Hadjiivanov, J. Chem. Soc., Faraday Trans 1998, 94, 1901-1904.
26) S. Shen, X. Wang, Q. Ding, S. Jin, Z. Feng, C. Li, Chin. J. Catal. 2014, 35, 1900-1906.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
1
8
(27) a) R. G. Greenler, K. D. Burch, K. Kretzschmar, R. Klauser, A. M. Bradshaw, B. E.
Hayden, Surf. Sci. 1985, 152-153, 338-345; b) F. Jiang, L. Zeng, S. Li, G. Liu, S. Wang, J.
Gong, ACS Catal, 2015, 5, 438-447.
(
28) A. Corma, P. Concepcion, P. Serna, Angew. Chem. Int. Ed, 46(38), 2007, 7266-
269.
29) A. Corma, P. Serna, P. Concepcion, J. J. Calvino, J. Am. Chem. Soc. 2008, 130, 8748-
753.
7
(
8
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
1
9
Table 1. Results of reductive amination of EL and different amines using 0.2%wt
Pt/TiO and 0.2wt% Pt/TiO -NT as catalyst.
2
2
a
TOF
Time Conv. Yield
Sel.
(%)
o
-1
Catalyst
Reactant
r (mmol h )
-
1
(h )
(h)
(%)
(%)
7
33
49
87
90
0.376
6
98
91
93
97
98
97
0
0
.2wt%Pt/TiO2
7
0.384
0.096
0.200
2
48
8
97
85
94
83
97
1
3
.2wt%Pt/TiO2
-
NT
100
Reaction conditions: ethyl levulinate (2 mmol), amine (2 mmol), hydrogen (10 bar),
o
catalyst (0.05 mol%), at 120 C. In all cases the only byproduct detected was the imine
o
intermediate. r (Initial rate of appearance of pyrrolidone), TOF (initial rate of
a
appearance of pyrrolidone per mmol of Pt). GC yield
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
2
0
Table 2. Synthesis of different pyrrolidones by reductive amination of EL and different
nitro compounds using 0.2%wt Pt/TiO -NTs as catalyst.
2
c
Entry
Reactant
Conv. (%)
94
Yield (%)
Sel. (%)
98
a
1
92
b
2
17
78
90
9.6
77
75
56.5
98
3
4
5
6
84
80
66
63
61
80
92
7
8
9
62
68
63
59
55
47
95
81
74
1
1
0
1
65
51
78
61
54
47
49
36
28
80
66
59
a
1
2
1
3
Reaction conditions: ethyl levulinate (3 mmol), nitro compound (1 mmol), hydrogen
o
a
(
10 bar), 50 mg catalyst (0.05 mol%), S/C= 2000, at 120 C, 48 h; EL (1 mmol),
b
nitrobenzene (1 mmol);
5%Pt-Al O (50g) 1g (EL(30wt%):NB(26wt%):Dioxane
2
3
o
18a c
(44wt%) at 150 C, under 51.7 bar H and 7 hours.
GC Yield
2
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
2
1
1
00
8
6
4
2
0
0
0
0
0
0
6
12
18
24
Time (h)
Figure 1. Catalytic activity of 0.2wt% Pt/TiO for the formation of imine by reaction
2
o
between EL and aniline in absence of H at 120 C. [] EL conversion; [] imine yield.
2
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
2
2
1
00
8
6
4
2
0
0
0
0
0
0
6
12
18
24
30
36
42
48
Time (h)
Figure 2. Catalytic activity of 0.2wt% Pt/TiO -NT for the formation of imine by reaction
2
o
between EL and aniline in absence of H at 120 C. [] EL conversion; [] imine yield.
2
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
2
3
100
80
60
40
20
0
0
4
8
12
16
20
24
Time (h)
Figure 3. Results of the one pot process in the synthesis of pyrrolidone starting from
o
nitrobenzene and ethyl levulinate in the presence of 0.2wt% Pt/TiO , at 120 C and 10
2
bar H . [] nitrobenzene conversion; [] EL conversion; [▼] pyrrolidone yield (1); [♦]
2
imine yield (2); [] by products (3+4) yield.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
2
4
-
1
(
c)
b)
1525 cm
(
-
1
1350 cm
(a)
0,9
0,8
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0,0
1
000
1200
1400
1600
1800
-
1
cm
8
00
1200
1600
2000
2400
2800
3200
3600
-
1
cm
Figure 4. Infrared spectra of 0.2wt% Pt/TiO catalysts: (a) fresh, (b) after one-pot
2
reaction, (c) fresh in which nitrobenzene (5wt%) was added.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
2
5
1
00
8
6
4
2
0
0
0
0
0
0
6
12
18
24
Time (h)
Figure 5. Results for the formation of pyrrolidone by reductive amination of EL with
aniline using as catalyst 0.2wt% Pt/TiO (previously used in hydrogenation of
2
o
nitrobenzene), at 120 C and 10 bar H . [] EL conversion; [] imine yield; [▼]
2
pyrrolidone yield
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
2
6
100
80
60
40
20
0
0
4
8
12
16
20
24
Time (h)
Figure 6. Yield of pyrrolidone obtained in the reductive amination of aniline with
o
EL using 0.2wt% Pt/TiO as catalyst, at 120 C and 10 bar H in where small amounts of
2
2
nitrobenzene were added. [] 0wt% NB; [] 0.1 wt% NB; [] 0.2 wt% NB; [▼] 0.4
wt% NB; [◄] 0.8 wt% NB; [►] 1 wt% NB.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
2
7
100
%
%
Conv. Levulinate
Yield Imine
80
60
40
20
0
0
4
8
12
16
20
24
Time (h)
Figure 7. Catalytic activity of used 0.2wt% Pt/TiO catalyst in the formation of imine
2
o
between EL and aniline in absence of H at 120 C.(▲)EL conversion; (■) imine yield
2
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
2
8
100
80
60
40
20
0
0
1
2
3
4
5
6
Time (h)
Figure 8. Yield of pyrrolidone by imine hydrogenation using 0.2 wt% Pt/TiO as catalyst
2
in absence of solvent at 120 oC and 10 bar H2. Fresh catalyst: [] Imine conversion;
[
] Pyrrolidine Yield; Used catalyst: [◊] Imine conversion; [∆] Pyrrolidine Yield.
This article is protected by copyright. All rights reserved.

ChemSusChem
10.1002/cssc.201601333
2
9
100
80
60
40
20
0
0
6
12
18
24
30
36
42
48
Time (h)
Figure 9. Results of synthesis of 5-methyl-N-phenyl-2-pyrrolidone in one pot, starting
o
from nitrobenzene and EL in the presence of Pt/TiO -NT catalyst, at 120 C, 10 bar H
2
2
and in absence of solvent. [] nitrobenzene conversion; [] EL conversion; [] imine
yield; [▼] pyrrolidone yield.
This article is protected by copyright. All rights reserved.