Sustainable Continuous Flow Valorization of Γ-Valerolactone with Trioxane to Α-Methylene-Γ-Valerolactone over Basic Beta Zeolites

Article abstract of DOI:10.1002/cssc.201900418

The need for more sustainable products and processes has led to the use of new methodologies with low carbon footprints. In this work, an efficient tandem process is demonstrated for the liquid-phase catalytic upgrading of lignocellulosic biomass-derived γ-valerolactone (GVL) with trioxane (Tx) to α-methylene-γ-valerolactone (MeGVL) in flow system using Cs-loaded hierarchical beta zeolites. The introduction of mesopores along with the presence of basic sites of mild strength leads to MeGVL productivity 20 times higher than with the bulk beta zeolite, reaching 0.325 mmol min^?1 g_cat^?1 for the best-performing catalyst, the highest value reported so far. This catalyst proves stable upon reuse in consecutive cycles, which is ascribed to the partial depletion of the basic sites. The obtained MeGVL is subjected to visible-light-induced polymerization, resulting in a final material with similar properties to the widely used poly(methyl) methacrylate.

Full text of DOI:10.1002/cssc.201900418

Accepted Article
Title: Sustainable Continuous Flow Valorization of γ-Valerolactone with
Trioxane to α-Methylene-γ-Valerolactone over Basic Beta Zeolite
Authors: Majd Al-Naji, Begoña Puértola, Baris Kumru, Daniel Cruz,
Marius Bäumel, Bernhard Schmidt, Nadezda Tarakina, and
Javier Pérez‐Ramírez
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.201900418
Link to VoR: http://dx.doi.org/10.1002/cssc.201900418
A Journal of
www.chemsuschem.org

10.1002/cssc.201900418
ChemSusChem
FULL PAPER
Sustainable Continuous Flow Valorization of -Valerolactone with
Trioxane to -Methylene- -Valerolactone over Basic Beta Zeolites
Majd Al-Naji,*[a],† Begoña Puértolas,[b],† Baris Kumru,[a] Daniel Cruz,[a] Marius Bäumel,[a]
Bernhard V.K.J. Schmidt,[a] Nadezda V. Tarakina,[a] and Javier Pérez-Ramírez[b]
Abstract: The necessity of more sustainable products and processes
has led to the use of new methodologies which result in a low carbon
footprint. Here, we demonstrated an efficient tandem process for the
liquid-phase catalytic upgrading of lignocellulosic biomass-derived -
valerolactone (GVL) with trioxane (Tx) to -methylene--valerolactone
(MeGVL) in flow using Cs-loaded hierarchical beta zeolites. The
introduction of mesopores along with the presence of basic sites of
mild strength led to 20-times higher MeGVL productivity compared to
substitute the fossil-based methyl methacrylate (MMA), a
monomer produced on a large scale for the production of
poly(methyl methacrylate), the latter having numerous
applications in the automotive, electronics, and medical sectors.[1,
2, 16]
The non-catalytic synthesis of MeGVL using GVL as
precursor was patented in 1953 using GVL and ethyl formate in
ethanol in the presence of granulated sodium sand.[17] In the
absence of catalyst, a two-step process using ethyl formate and
paraformaldehyde and tetrahydrofuran or dimethylsulfoxide as
solvents was also reported.[18] Even though the yield of MeGVL
obtained was ca. 90%, the use of non-environmentally-friendly
organic solvents along with the moderate kinetics make their
practical implementation unfeasible. More recently, the catalytic
synthesis of MeGVL was approached via the base-catalyzed -
methylation reaction in batch using either aqueous formaldehyde
(FA) solution, i.e., formalin, or paraformaldehyde as FA source
(Scheme 1). The use of a base enables the nucleophilic attack of
FA on the -position of GVL, a step that is essential for the
subsequent conversion into MeGVL. The MeGVL yield was
limited to 36% using an homogeneous base, i.e., K2CO3 and
CsCO3, in the presence of paraformaldehyde and
2-methyltetrahydrofuran (MeTHF) or supercritical CO2 as
solvents.[19, 20] A step forward towards the continuous operation
using different rare earth metals supported on silica was also
reported.[21] Although GVL was converted to a large extent, i.e.,
60%, no information regarding the selectivity towards the desired
MeGVL is available.
−1
the bulk beta zeolite reaching 0.325 mmol min−1 gcat for the best-
performing catalyst, the highest value reported so far. This catalyst
proved stable upon reuse in consecutive cycles, which was ascribed
to the partial depletion of the basic sites. The obtained MeGVL was
subjected to visible-light-induced polymerization resulting into a final
material with similar properties to the widely used poly(methyl)
methacrylate.
Introduction
The necessity for the green manufacture of fine chemicals and
polymers motivate the search of novel processes based on
renewable and sustainable resources.[1-7] In particular,
lignocellulosic biomass consisting of cellulose, hemicellulose, and
lignin fractions, represents one of the most abundant and
economical starting materials.[1] The prospective of lignocellulosic
biomass as a substrate for the efficient production of building
blocks such as levulinic acid, GVL, 5-hydroxymethylfuran (HMF)
In this work, we investigated the development of a sustainable
tandem process in flow for MeGVL synthesis from lignocellulosic
biomass-derived GVL in the presence of Tx using tailored
Cs-loaded hierarchical beta zeolite catalysts (Scheme 1). The
use of Tx instead of FA, the latter generally available in aqueous
solutions, is believed to increase the yield of MeGVL as the
presence of water inhibits the -methylation reaction. The
decomposition of Tx in the temperature range 483-503 K leads to
the formation of 3 molecules of FA that are essential to initiate the
desired transformation.[22-24] The comparative impact of the
porous, acidic, and basic properties of the zeolites upon Cs
loading on the yield of MeGVL is quantified, providing insights into
the enhanced selectivity observed over the hierarchical zeolites.
The synthesized MeGVL over the best-performing material was
subjected to visible-light induced polymerization to produce
and dimethyl furan (DMF) has been already demonstrated at the
8-13]
lab-scale.[4,
Among them, GVL produced via the
hydrodeoxygenation of lignocellulosic biomass, exhibits excellent
solvent properties and is one of the main precursors in the
synthesis of high-value fine chemicals and polymers, such as
pentenoic acid, valeric acid and MeGVL.[1, 4, 14, 15] In this regard,
MeGVL is a type of methacrylic monomer with the potential to
[a]
Dr. M. Al-Naji, B. Kumru, D. Cruz, M. Bäumel, Dr. B.V.K.J. Schmidt,
Dr. N.V. Tarakina
Department of Colloid Chemistry
Max Planck Institute of Colloids and Interfaces
Am Mühlenberg 1, 14476 Potsdam (Germany)
E-mail: majd.al-naji@mpikg.mpg.de
[b]
Dr. B. Puértolas, Prof. J. Pérez-Ramírez
poly(MeGVL) using graphitic-carbon nitride (g-CN) as catalyst.[25-
Department of Chemistry and Applied Biosciences, ETH Zurich
Institute for Chemical and Bioengineering
28]
Vladimir-Prelog-Weg 1, 8093 Zurich (Switzerland)
[†]
These authors have contributed equally
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900418
ChemSusChem
FULL PAPER
Results and Discussion
(Si/Al = 140) and mordenite (Si/Al = 10) zeolites was lower than
that over the beta (Si/Al = 150) zeolite. Increasing the Si/Al ratio
to 220 led to a further increase of the GVL conversion, which
could be associated to the presence of stronger acid sites in this
sample in comparison to beta-150. It is worth noticing that
trioxane was fully converted in all cases. However, the yield of
MeGVL was very low for all the materials, and the major formation
of ring-opening compounds such as 4-pentonoic acid and butene
as side products was observed (Figure S2). The latter was only
noticeable in the case of the beta zeolites as the lower GVL
conversion obtained when ZSM-5 and mordenite zeolites were
used impeded the quantification of the products formed. The
selectivity towards the desired MeGVL product increases upon
the introduction of Cs (Figure 1b, Figure S2, and Table S1).
Comparison between the performance of the Cs-containing
sample and the parent material revealed the key role of Cs in
catalyzing the -methylation of GVL to MeGVL. The formation
rate of MeGVL slightly decreases upon increasing the Si/Al ratio
of the framework for the Cs-loaded zeolites, i.e., 0.133 and
GVL valorization over zeolites with different framework
topologies
−1
0.126 mmolMeGVL min−1 gcat for the 5Cs/beta-150 and 5Cs/beta-
220 materials, respectively. The rest of the products formed
account for 4-pentenoic acid and methyl 4-pentenoate as the
main compounds, the latter likely formed upon the reaction
between 4-pentenoic acid and the methanol formed from the
decomposition of FA at high temperatures (Figure S1). In
contrast to the commercial materials, the formation of butane, 2-
butanol, 2-pentene, and pentanal are suppressed, which
suggests the occurrence of a different reaction mechanism in the
presence of Cs. In view of the slightly higher productivity of
MeGVL, beta zeolite with Si/Al 150 and cesium loading of 5 wt.%
was selected for further investigation.
Scheme 1. Traditional pathway for the synthesis of MeGVL from GVL in the
presence of FA from formalin or paraformaldehyde (left). The new tandem route
demonstrated in this work that uses trioxane as formaldehyde source (right).
Post-synthetic modification of beta zeolites by alkaline
treatment and cesium loading
To date, the availability of studies tackling the valorization of GVL
into MeGVL over heterogeneous catalysts is scarce. Accordingly,
we firstly assessed the performance of commercial zeolites in
acidic form with different framework topology and Si/Al ratio in the
-methylation of GVL at 568 K in flow (Figure 1a and Figure S1).
After 1 h on stream, the conversion of GVL obtained over ZSM-5
The parent beta (Si/Al = 150) zeolite was subjected to an alkaline
treatment, attaining, after ion exchange to its acidic form, a
hierarchical analog coded H-beta-150. The composition and
textural properties are summarized in Table 1. As expected, the
Figure 1. a) Rate of conversion of GVL and b) rate of formation of MeGVL and side products over zeolites with different framework topologies and Si/Al ratios prior
to and after cesium loading. Reaction conditions: mcat = 3 g, mGVL = 5.1 g, mTx = 10.1 g, VMeTHF = 500 cm3, Qreactant = 0.3 cm3 min−1
Vreactor = 4.15 cm3.
,  = 13.8 min, and
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900418
ChemSusChem
FULL PAPER
Table 1. Characterization data of the catalysts.
[b]
[c]
[c]
[d]
Cs loading[a]
(wt.%)
Vpore
Vmicro
Smeso
SBET
Cryst.
()
Catalyst
Si/Al
(cm3 g–1
0.39
0.74
0.30
0.59
0.45
0.68
0.47
0.37
0.75
)
(cm3 g–1
0.23
0.17
0.13
0.12
0.12
0.16
0.11
0.09
-
)
(m2 g–1
150
351
132
253
151
266
192
152
-
)
(m2 g–1
)
beta-150
150
115
-
582
677
388
491
383
565
395
318
550
100
74
45
44
19
38
45
2
H-beta-150
-
5Cs/beta-150
146
110
111
116
109
114
-
5
5Cs/H-beta-150
4.9
4.9
4.8
9.6
22
4.7
5Cs/H-beta-150 used 1st cycle
5Cs/H-beta-150 reg. after 3rd cycle
10Cs/H-beta-150
23Cs/H-beta-150
5Cs/SiO2
-
[a] ICP-OES. [b] Volume adsorbed at p/p0 = 0.99. [c] t-plot method. [d] BET method.
desilication led to the development of additional mesoporous area,
which was accompanied by a drop in crystallinity and microporous
volume. In addition, the Si/Al ratio decreased with the consequent
increase of the amount of charge-compensating H+ cations. The
incorporation of different amounts of Cs, i.e., 5, 10, and 23 wt.%
by dry impregnation led to a further decrease of the textural
properties and crystallinity as the amount of Cs increases
(Table 1 and Table S1). These samples are called xCs/H-beta-
150, where x refers to the Cs loading.
The basicity of the hierarchical beta zeolites with different Cs
loadings was assessed by CO2-TPD (Figure 2a). The parent and
the hierarchical zeolites produced no significant desorption,
confirming their negligible basicity. In contrast, two major peaks
centered at 410 and 550 K became evident in the curves of the
5Cs/H-beta-150 zeolite, indicating the presence of basic sites with
a moderate character. A fraction of Cs is expected to be present
at ion exchange positions to balance the framework charge
caused by the isomorphous substitution of silicon with aluminum
atoms and possess very mild basicity.[29] Another part is believed
to substitute the protons in the silanol groups present in defect-
rich regions (i.e., the external surface of the crystal or the
mesopore walls), leading to stronger basic sites similar to those
observed for high-silica faujasites.[30] Its occurrence was
corroborated by XPS that revealed the presence of CsOH species
on the catalyst surface (Figure 2b). Additionally, the occurrence
of Cs2O species was identified in the XPS spectrum, which could
also contribute to the basicity. Regardless their speciation, Cs
species are believed to be highly dispersed over the zeolite
surface in line with STEM and XRD results that evidenced the
absence of Cs-rich areas and Cs-containing nanoparticles
(Figure 3 and Figure S3), respectively. Acidity assessment by
the infrared spectroscopy of pyridine adsorbed revealed the
absence of acid sites in this sample, which is in line with the ion
exchange of the protonic sites by cesium cations during the dry
impregnation step. Increasing of the Cs loading led to the
decrease of the density of mild basic sites and the formation of
stronger basic sites (Figure 2a) in line with the appearance of a
desorption peak centered at ca. 850 K. These sites could be the
result of the agglomeration of different Cs-containing species as
revealed by the micrographs of the sample prepared with the
highest Cs loading (Figure S4). These species could contain
Figure 2. a) CO2 desorption profiles of selected catalysts. b) Cs 3d core level
spectra of 5Cs/H-beta-150 in fresh form.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900418
ChemSusChem
FULL PAPER
Figure 3. ADF-STEM images of 5Cs/H-beta-150 zeolite and EDX spectra at the location indicated in the micrograph.
CsO2, Cs2O2, Cs2O3, Cs2O4, Cs2O, Cs7O, Cs metal, and any
combination thereof that could be formed under a particular
oxygen environment during calcination,[31] and whose
identification is beyond the scope of the present study. In addition,
a partial contribution from CsCl to the strong basicity cannot be
discarded. XRD patters of the 23Cs/H-beta-150 catalyst
(Figure S3) evidenced the characteristic reflections of this
crystalline phase suggesting that it was not fully transformed into
CsOx species upon calcination.
observations.[29] Therefore, the external surface area of the
hierarchical sample decreased to a much lower extent than that
of the bulk material (Table 1 and Table S2). It is worth mentioning
that while the external surface area decreased upon reaction, the
micropore volume remained unchanged, which might initially
suggest that the investigated reaction does not require a
microporous material. To discard this hypothesis, 5 wt.%Cs-
loaded SiO2 was synthesized using a SiO2 support with the same
average pore size as the H-beta-150 material. The results
revealed that the 5Cs/SiO2 material was less active than the
5Cs/H-beta-150 material (Figure 4a), thus highlighting the
beneficial role of the microporous structure in the reaction.
The increase of the Cs loading over the H-beta-150 material lead
to the increase of the amount of GVL converted at the expense of
the decrease of the formation rate of MeGVL (Figure 4b). The
latter is maximized over the sample with 5 wt.% Cs loading, which
could be attributed to the presence of a higher density of mild
basic sites along with the increased dispersion of the Cs species
over the zeolite surface (Figure 2a and Figure 3). The presence
of strong basic sites in the case of the samples with 10 and
23 wt.% led to lower amount of MeGVL formed with similar rate of
formation of side products (Figure 4b). Comparison of the best-
performing catalyst with the benchmark CsX zeolite led to 3-times
higher formation rate of MeGVL. Additionally, CsX favored the
conversion of GVL to a larger extent leading to the formation of
an increase amount of side and polymerization products, mostly
related to the occurrence of a higher density of basic sites with
strong basicity over this sample.
Catalytic performance of Cs-loaded hierarchical zeolites
The mesoporous beta zeolites prepared through the
post-synthetic treatments were tested as catalysts prior to and
after Cs incorporation in the -methylation of GVL under the same
conditions as those applied to the screening of commercial
zeolites presented above. Introduction of mesoporosity led to an
increase of the conversion of GVL, which is accompanied by an
increase in the formation of MeGVL (Figure 4a). Addition of
5 wt.% Cs to H-beta-150 led to a substantial increase in the
amount of GVL converted along with the enhanced formation of
−1
MeGVL, i.e., from 0.028 to 0.325 mmolMeGVL min−1 gcat
.
Additionally, the rate of formation of undesired compounds
decreased. In the case of the hierarchical H-beta-150 zeolite, the
presence of 4-pentenoic acid, methyl-4-pentenoate, butene, 2-
pentanol, 2-pentene is detected whereas only 4-pentenoic-acid
and methyl-4-pentenoate are detected upon Cs incorporation
(Table S1, Figure S5). In the case of the H-beta-150 material, the
formation of undesired products could be attributed to the
presence of acid sites of different strength that could catalyze the
unselective conversion and/or the decomposition of GVL.
However, the absence of acid sites in the 5Cs/H-beta-150
material led to a decrease formation of side products. Additionally,
owing to the higher surface area, the hierarchical material could
accommodate the carbon deposits associated to the occurrence
of polymerization reactions more efficiently, in line with previous
Optimization of the reaction conditions
The reaction conditions were adjusted, in terms of temperature,
Tx/GVL molar ratio, and residence time (), for the best catalyst,
5Cs/H-beta-150, to maximize the formation of MeGVL (Figure 5).
With respect to temperature, two additional tests were conducted
at 553 K and 583 K (Figure 5a). Under the milder conditions, a
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900418
ChemSusChem
FULL PAPER
drop in activity was observed. At the higher temperature, the
amount of GVL converted was higher but the rate of formation of
MeGVL substantially decreased. The dependence of the rate of
GVL converted with the temperature is expected since the
reaction kinetics increase with the temperature (Figure S6).
Therefore, the use of 568 K is selected and the role of the Tx/GVL
ratio was investigated under this condition (Figure 5b). In this
case, two lower points were assessed, i.e., 0.5 and 1, as a higher
Tx/GVL was unfeasible due to the limited solubility of Tx in
2-methyltetrahydrofuran. In both cases, an activity depletion was
determined, which was higher at the lower Tx/GVL. The better
performance of the catalyst at Tx/GVL molar ratio of 2 suggests
that the excess of FA generated from the decomposition of Tx is
required for the efficient conversion of GVL. Applying 568 K and
a Tx/GVL molar ratio of 2, the residence time was decreased to
5.9 and 8.3 min and increased to 41.5 min. As shown in Figure 5c,
both the rate of conversion of GVL and the rate of formation of
MeGVL reached a maximum for a residence time of 13.8 min. For
shorter residence times, there is not enough contact between the
substrates and the catalyst to ensure an efficient performance. In
contrast, higher reaction times might lead to an increased
Figure 4. a) Rate of formation of MeGVL and side products and of conversion
of GVL over bulk and hierarchical beta zeolites prior to and after cesium loading.
The performance of CsX and Cs-loaded SiO2 are included for comparative
purposes. b) Rate of conversion of GVL and of formation of MeGVL and side
products over hierarchical beta zeolites with different Cs loadings. Reaction
conditions: mcat = 3 g (bulk zeolite) and 2 g (hierarchical zeolite), mGVL = 5.1 g,
Figure 5. Impact of a) temperature, b) Tx/GVL molar ratio, and c) residence
time on the rate of formation of MeGVL and side products and of conversion of
GVL over 5Cs/H-beta-150 zeolite. Reaction conditions: mcat = 2 g,
mGVL = 5.1 g,
mTx = 10.1 g,
VMeTHF = 500 cm3,
Qreactant = 0.3 cm3 min−1
,
mTx = 10.1 g, VMeTHF = 500 cm3, Qreactant = 0.3 cm3 min−1
Vreactor = 4.15 cm3.
,  = 13.8 min, and
 = 13.8 min and Vreactor = 4.15 cm3.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900418
ChemSusChem
FULL PAPER
formation of carbon deposits on the catalyst surface owing to the
higher contact time with the reactants.
Catalytic deactivation, regeneration, and reuse
The stability of the 5Cs/H-beta-150 zeolite was evaluated in a 6-
h test (Figure 6). The catalyst performance was retained during
the first 2 h, after which the zeolite started to deactivate and
retained ca. 30% of the initial yield of MeGVL at the end of the run,
in line with previous results by Lemonidou et al.[32] Coke formation
and the strong adsorption of polymeric derivatives are believed to
encompass the main routes for catalyst deactivation in the
-methylation of GVL. Characterization of the best-performing
catalyst (5Cs/H-beta-150) by TGA confirmed the deposition of a
conspicuous amount of coke, i.e., 30 wt.% (Figure S7). This
weight loss is divided in three different regions: region 1 (from
295 K to 500K) in which the weight loss of 4.9 wt.% corresponds
to the desorption of physisorbed water, and regions 2 (from 575 K
to 905 K) and 3 (from 910 K to 1150 K) in which the weight loss
of 26 wt.% is assigned to the depletion of coke and large
carbonaceous species on the catalyst. The porosity analysis
uncovered the full preservation of the micropore volume and the
halving of the mesopores (Table 1).
Figure 7. Withdrawn sample after reaction (left), sample after polymerization
reaction (middle) using g-CN and the resulting polymer PMeGVL_S after
filtration of g-CN and drying at 333 K (right).
catalysts (Figure 6). Herein, the activity of 5Cs/H-beta-150 zeolite
was monitored over 2 consecutive runs between which the
catalyst was thermally-treated in air to simulate a typical
regeneration procedure. During the second test, the rate of
formation of MeGVL slightly decreased compared to the first cycle,
but both the MeGVL productivity and the formation of side
prodcuts during the third cycle was fully retained (Figure 6b).
Characterization of the regenerated sample after 3 cycles
revealed that the textural properties and crystalline structure are
mostly preserved (Figure S3 and Table 1). CO2-TPD analysis
evidenced a decreased of the basic sites, especially of the milder
type (Figure 2a). Since no Cs leaching was detected during the
reaction (Table 1), Cs restructuring into different species upon
regeneration is believed to be the origin of the changes in basicity
determining the steady catalyst performance upon cycling.
The implementation of reaction-regeneration cycles has been
proposed as a strategy to extend the long-term use of the
Visible-light-induced MeGVL polymerization
As a proof of concept, MeGVL synthesized using 5Cs/H-beta-150
was subjected to visible-light-induced polymerization using
graphitic carbon nitride (g-CN) as catalyst. This approach
promises a sustainable polymerization route that ensures ease
catalyst separation, reusability, and avoids the deoxygenation of
the monomer mixture. g-CN was utilized together with radical
mediator (triethylamine) as otherwise polymerization would
proceed on the g-CN surface. Additionally, and for comparison
purposes, commercial MeGVL was subjected to similar reaction
conditions (which will be denoted as PMeGVL_R). After
polymerization, precipitated and dried samples were investigated
via 1H-NMR and size exclusion chromatography (SEC). SEC
results indicate polymer formation from the synthesized MeGVL
(which will be denoted as PMeGVL_S) with a number of average
molecular weight (Mn) of 4200 g mol-1 (~35 joined monomer units)
and 6800 g mol-1 (~60 joined monomer units) for PMeGVL_R
(Figure S8). In addition, molecular dispersities (Ɖ) of each
polymer were found to be 1.92, which is a typical value for free
radical polymerization mechanism. Furthermore, 1H-NMR spectra
of PMeGVL_S proved the suppression of vinylic hydrogen peaks
(5.72 ppm and 6.03 ppm) which indicate the successful
polymerization reaction and solvent removal (Figure S9). The
significant difference upon comparison of PMeGVL_R and
PMeGVL_S is the appearance of the polymer, where PMeGVL_R
is glass-like polymer with respect PMeGVL_S that found to be
brown and non-glass-like polymer (Figure 7). This observation is
attributed to the presence of the side products in the withdrawn
MeGVL synthesized sample. These results present the high
Figure 6. Rate of formation of a) MeGVL and of b) side products (s.p.) over
5Cs/H-beta-150 upon three subsequent catalytic cycles. Reaction conditions:
mcat = 2 g,
mGVL = 5.1 g,
mTx = 10.1 g,
VMeTHF = 500 cm3,
Qreactant = 0.3 cm3 min-1,  = 13.8 min, and Vreactor = 4.15 cm3.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900418
ChemSusChem
FULL PAPER
Catalyst preparation
potential of utilizing biomass-derived compounds such as MeGVL
as bio-based building blocks through a green polymerization
process via visible light.
Commercial Beta zeolites (beta-150, beta-220), where 150 and 220 refer
to the nominal Si/Al ratio according to the manufacturer’s specifications,
were treated in an aqueous alkaline solution (0.2 M NaOH and
0.01 M CTABr) at 338 K for 30 min (30 cm3 gzeolite−1) using an EasymaxTM
102 reactor system (Mettler Toledo). The resulting slurry was quenched in
ice-water, filtered, and the isolated solids washed extensively with
deionized water and dried at 338 K. The resulting materials were
converted into the H-form by three consecutive ion exchanges in aqueous
ammonium nitrate solution (0.1 M, 338 K, 8 h, 100 cm3 gzeolite−1) followed
Conclusions
In this study, we have developed a sustainable process for the
-methylene--valerolactone synthesis from -valerolactone
using trioxane as a formaldehyde source in a continuous-flow
system over a tailored Cs-loaded hierarchical beta catalyst. After
an initial screening of commercial aluminosilicates with different
topologies and Si/Al ratios, we identified beta zeolite with Si/Al
ratio of 150 as a potential catalyst for this reaction. Based on
previous studies that identified the occurrence of basic sites as
beneficial for the selectivity towards the desired product,
incorporation of Cs led to a substantial increase of the activity
compared to the bulk material. The performance was further
improved by the subsequent alkaline treatment to produce a
hierarchical zeolite followed by the optimization of the Cs loading.
The best-performing catalyst exhibited enhanced selectivity to -
by calcination in static air at 823 K for 5 h using a heating rate of 5 K min−1
.
Along the manuscript, the code H denotes the hierarchical materials. For
the preparation of graphitic carbon nitride (g-CN) 1 g of cyanuric acid and
1 g melamine were mixed with 40 cm3 of deionized water and stirred
overnight. After centrifugation at 6000 rpm for 10 min, a precipitate was
dried at 333 K under vacuum overnight and afterwards transferred into a
capped crucible and calcined under N2 atmosphere at 823 K for 4 h using
a heating rate of 2.3 K min−1
.
Cs catalysts supported on Beta, H-Beta, ZSM-5, and mordenite zeolites,
and SiO2 were prepared via incipient wetness impregnation. The
appropriate amount of CsCl was dissolved in deionized water targeting 5,
10, and 23 wt.% Cs loading followed by dropwise addition of the resulting
solution into the zeolite support until slurry was formed. Subsequent drying
was performed at 333 K in static air for 12 h. CsX was obtained by
subjecting zeolite X to 3 subsequent ion-exchange treatments in aqueous
solutions of CsCl (0.1 M, 8 h, 298 K, 100 cm3 gzeolite−1). The catalysts in
powder form were using hydraulics press (CARVER 3851CE) with 10
metric ton, crushed and sieved in the range of 250-500 m to ensure a
homogeneous flow of the reactant through the catalyst bed during the
reaction. Prior the reaction, the catalyst was dried in static air at 393 K for
1 h (5 K min−1) and calcined at 773 K for 5 h (5 K min−1).
methylene--valerolactone
with
a
productivity
of
0.325 mmol min−1 gcat−1. This is attributed to the enhanced
accessibility achieved by the introduction of mesoporosity along
with the presence of basic sites of mild strength. Evaluation of this
material in consecutive reaction-regeneration tests revealed that
the catalyst performance slightly decreased in the second cycle,
but the activity is retained during the third run, which was
associated to the decrease of the basicity. The catalyst structure
was preserved and the metal phase proved stable against
leaching. As a proof of concept, the synthesized -methylene--
Catalyst Characterization
valerolactone
was
subjected
to
visible-light-induced
The metal content was determined by ICP-OES using an Optima 8000
ICP-OES from Perkin Elmer. Prior the analysis, the samples (0.01 g) were
dissolved in a solution consisting of HCl (333 L) and HNO3 (167 L).
polymerization, reaching a final material with similar properties to
poly(methyl) methacrylate. These findings pave the path for a
large scale production of -methylene--valerolactone from
cellulosic biomass, which can be readily polymerized towards a
new and green generation of polymers.
Powder XRD measurements were performed on
a Bruker D8
diffractometer. The X-rays source was Cu-K ( = 0.154 nm) equipped
with a scintillation counter-Scinti-Detector. The diffraction data was
recorded in the 4-70° 2 range with an angular step of 0.05° and a
continuing time of 2 s per step. Ar sorption at 77 K was undertaken in
Micromeritics TriStar II instrument. Prior to the measurement, the samples
were evacuated at 573 K for 3 h. Temperature-programmed desorption of
carbon dioxide (CO2-TPD) was performed in a Micromeritics Autochem II
2920 analyzer. First, the solid (0.1 g) was loaded into a U-shaped quartz
micro-reactor, pretreated in He (20 cm3 min−1) at 423 K for 2 h, and cooled
to 323 K. Thereafter, CO2 was chemisorbed at 323 K in 50 consecutive
pulses with a 5 vol.% CO2/He mixture (1 cm3 STP per pulse) followed by
purging with He (10 cm3 STP min−1) at the same temperature for 90 min.
Desorption of CO2 was monitored in the range of 323-973 K using a
heating rate of 10 K min−1 and a He flow of 20 cm3 STP min−1.The
scanning transmission electron microscopy (STEM) study of selected
samples was performed using a double Cs corrected JEOL JEM-
ARM200F (S)TEM operated at 80 kV and equipped with a cold-field
emission gun, with a high-angle silicon drift Energy Dispersive X-ray (EDX)
detector (solid angle up to 0.98 steradians with a detection area of
100 mm2), and a Gatan Imaging Filter (GIF Quantum). For STEM
observation, samples were put onto a lacey carbon film supported on a Cu
grid. Annular dark field scanning transmission electron microscopy (ADF -
STEM) images were collected at a probe convergence semi-angle of
25 mrad. XPS measurements were carried out under ultra-high vacuum
(UHV) 1.5×10−8 Pa (CISSY equipment) equipped with SPECS XR 50 X-
Experimental Section
Materials
-Valerolactone (≥99%), 4-pentenoic acid (≥98%), cyanuric acid (98%),
isopropanol (anhydrous, 99.9%), melamine (99%), methyl 4-pentenoate
(≥95%), 2-butanol (>99.5%), 2-pentene (99%), pentanal (>97%),
tetrahydrofuran (THF, GC grade), dioctyl ether (99%), sodium hydroxide
(97%), cesium chloride (≥99%), triethylamine (98%), ammonium nitrate
(≥98%), silica gel (SiO2, grade 62), and and CsCl (99.9%) were purchased
from Sigma-Aldrich. Cetyltrimethylammonium bromide (CTABr, 99%) was
provided by Acros Organics. Lithium bromide (LiBr, anhydrous 99%) was
purchased from Alfa Aesar. -methylene--valerolactone (stabilized with
HQ, >97%) was obtained from TCI Chemicals. 2-methyltetrahydrofuran
(MeTHF) (99%) was supplied from Carl Roth GmbH & Co. BETA zeolites
(Si/Al = 150, 220) were provided by Clariant (HCZB 150) and TOSOH
(980HOA), respectively. ZSM-5 (Si/Al = 140, CBV28014) and mordenite
(Si/Al = 10, CBV21A) zeolites were provided by Zeolyst International.
Zeolite X (Si/Al = 1.4) was purchased from Sigma Aldrich. All chemicals
were used as received without further purification.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900418
ChemSusChem
FULL PAPER
ray gun with Al K radiation (1486.6 eV) and combined with lens analyzer
module (CLAM). The binding energy scale was calibrated using Au 4f7/2
(84.0 eV) band as reference. The binding energy and quantitative analysis
were performed utilizing “peakfit” and “Igor” software and using Lorenzian-
Gaussian functions and Shirley background deletion in photoemission
spectra. Thermogravimetric analysis (TGA) was performed via TG 209
Libra from Netzsch. Typically, 50 mg of used catalyst sample was dried at
423 K for 2 h. Then, the sample was heated to 1273 K with a heating rate
of 10 K min−1. Fourier transform infrared spectroscopy (FTIR) of adsorbed
where cGVL or i refer to the concentration of GVL and products in mmol cm−3
,
0
respectively in the withdrawn sample, mcat is catalyst mass in g and NGVL
is the initial molar flow of GVL in cm3 min−1
.
Visible-light-induced MeGVL polymerization
Commercial MeGVL and the synthesized MeGVL from GVL using
5Cs/H-beta-150 at 568 K were subjected to polymerization reactions using
g-CN. In the first case, 1.0 cm3 of commercial MeGVL was mixed with
10 mg of g-CN and 0.05 cm3 of trimethylamine. The resulting mixture was
subsequently illuminated with visible light for 3 h at room temperature after
which an increase in viscosity was observed. g-CN was removed by
centrifugation and the remaining solution poured into methanol to
precipitate the poly(MeGVL) which will be considered as a reference
polymer (PMeGVL_R). The white precipitate was filtered and dried
overnight at 333 K. Polymerization of the synthesized MeGVL was
conducted after evaporation of the MeTHF. A 1 cm3 of MeGVL was mixed
with 10 mg of g-CN, 1 cm3 of isopropyl alcohol and 0.05 cm3 of
trimethylamine followed by illumination with visible light for 16 h at room
temperature. Similarly, to the case where the commercial MeGVL was
used, an increase in viscosity was detected. After centrifugation, the
mixture was added into deionized water to precipitate the poly
MeGVL_Sample which is denoted (PMeGVL_S). The resulting precipitate
was filtered and dried overnight at 333 K. It is important to note that
increasing drying temperature influences darkening of the polymer product
and that commercial MeGVL polymerization with similar conditions show
considerably low yield. Size exclusion chromatography (SEC) of the
PMeGVL_R and PMeGVL_S was conducted in THF with 0.05 M of LiBr
and BSME as internal standard using a PSS GRAM 100/1000 column
(8×300 mm, 7 m particle size) with a PSS GRAM pre-column (8×50 mm),
a Shodex RI-71 detector and a calibration with PMMA standards from PSS.
PMeGVL_S were analyzed by 1H NMR using a Bruker Avance 400
spectrometer. Prior to the analysis, 1 cm3 of dimethyl sulfoxide-d6 (DMSO)
was added to 200 l of PMeGVL_S.
pyridine was conducted in a Bruker IFS 66 spectrometer (650-4000 cm−1
,
4.0 cm−1 optical resolution, co-addition of 32 scans). Self-supporting
wafers of catalyst (5 ton m−2, 0.03 mg, 1 cm2) were degassed under
vacuum (103 mbar) for 4 h at 693 K, prior to adsorbing pyridine at room
temperature. Gaseous and weakly adsorbed molecules were
subsequently removed by evacuation at 673 K for 30 min. The total
concentrations of Brønsted and Lewis acid sites were calculated from the
band area of adsorbed pyridine at 1545 and 1454 cm−1, using a previously
determined extinction coefficient of Brønsted
Lewis = 2.94 cm mmol−1
=
1.67 cm mmol−1 and

.
Catalytic evaluation
The -valerolactone (GVL) upgrading to -methylene--valerolactone
(MeGVL) in the presence of trioxane as a formaldehyde (FA) source was
studied in a homemade continuous-flow fixed-bed reactor (Figure S1A)
comprising (i) an HPLC pump equipped with a pressure sensor (Jasco PU-
2080 plus), (ii) a two-sides opened oven equipped with a heat controller
(Model # 4848 from Parr Instrument Company), and (iii) a sampling unit
equipped with a back pressure regulator and a cooling trap. To ensure an
efficient heat transfer from the oven to the fixed-bed reactor, an aluminum
cylinder with 3 different holes was placed inside the oven (Figure S1B),
i.e., a 1/16” hole as pre-heating unit to heat the reactant to the desired
reaction temperature before its get in contact with catalyst bed, a 1/4” hole
to place the fixed-bed reactor (HPLC blank column from Supelco was used
as tubular reactor - inner diameter = 4.6 mm, outer diameter = 1/4” and
length = 25 cm) and a third hole for the thermocouple (Model # A472E5
from Parr Instrument Company).
In a typical experiment, the reaction solution (5.1 g of GVL and 10.1 g of
trioxane which equal to molar ratio of 2) in 500 cm3 of MeTHF) was fed via
the HPLC pump at a rate of 0.3 cm3 min−1 through the pre-heating unit prior
to the catalytic reactor (mcatalyst = 3 g for bulk materials and 2 g for the
hierarchical counterparts). The reactor temperature and pressure were
kept at ambient for 15 min, after which the temperature was increased to
the target value, i.e., 553 K, 568 K or 583 K, and the system pressurized
to 6 MPa, allowing an average residence time () of 13.8 min. Samples
(2 cm3) were collected once the steady state was reached (after
ca. 45 min). To 1 cm3 of the samples, 50 L of dioctyl ether as internal
standard was added. After the reaction, the catalyst was washed with
acetone (45 cm3) and dried at 333 K for 12 h in static air. Regeneration of
the catalyst was conducted via thermal treatment in static air at 823 K for
5 h (5 K min−1). Product analysis was conducted using an off-line gas
Acknowledgements
The authors are grateful for the financial support from Max-Planck
Society. Scientific discussions, suggestions, feedback and
support from Prof. Dr. Markus Antonietti is gratefully
acknowledged. Also, thanks to the Electrical and Mechanical
Workshops at Max Planck Institute of Colloids and Interfaces for
their technical contributions to our research project. Thanks are
also due to Irina Shekova, Jessica Brandt, Katharina Otte, Marlies
Gräwert and Antje Völkel from Max Planck Institute of Colloids
and Interfaces for XRD, ICP-OES, SEC and TGA measurements,
respectively. Dr. Iver Lauermann from Helmholtz Zentrum Berlin
PVComB is acknowledged for XPS measurements. Dr. Silke
Sauerbeck from Clariant is greatly acknowledged for supplying
the zeolite sample. The authors are thankful to Alaa Al-Naji for the
idea and design of the graphical abstract.
chromatograph
diameter = 0.25 mm, length = 30 m and film = 0.25 m) coupled to a mass
spectrometer (Agilent GC 6890, Agilent MSD 5975) and gas
equipped
with
a
HP-5MS
column
(inner
a
chromatograph (Agilent GC 6890) equipped with a FID detector and a HP-
5 column (inner diameter = 0.25 mm, length = 30 m and film = 0.25 m).
The temperature program starts at 323 K (hold time = 2 min) and raises up
to 573 K (30 K min−1, hold time = 2 min). The rates of GVL conversion
(rGVL) and formation of the reaction products i (ri) were calculated according
to the following equations:
Keywords: basic catalysis • flow chemistry • lignocellulosic
biomass • -methylene--valerolactone • -valerolactone
[1]
[2]
[3]
D. Esposito, M. Antonietti, Chem. Soc. Rev. 2015, 44, 5821-5835.
N. Brun, P. Hesemann, D. Esposito, Chem. Sci. 2017, 8, 4724-4738.
P. C. Bruijnincx, B. M. Weckhuysen, Angew. Chem. 2013, 125, 12198-
12206; Angew. Chem. Int. Ed. 2013, 52, 11980-11987.
cGVL
−1
cat
rGVL mmolGVL min−1
g
=
Eq. (1)
Eq. (2)
(
)
m
NG0VL
(
)
cat
ci
NG0VL
r mmol min−1
g
=
−1
(
)
i
i
cat
[4]
A. Corma, S. Iborra, A. Velty, Chem. Rev. 2007, 107, 2411-2502.
m
(
)
cat
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900418
ChemSusChem
FULL PAPER
[5]
A. J. Ragauskas, C. K. Williams, B. H. Davison, G. Britovsek, J. Cairney,
[18] A. Tanaka, K. Yamashita, Agric. Biol. Chem. 2014, 42, 1585-1588.
[19] Z. Vobecka, C. Wei, K. Tauer, D. Esposito, Polymer 2015, 74, 262-271.
[20] L. E. Manzer, K. W. Hutchenson, US Patent No. 6649776 B2, 2003.
[21] L. E. Manzer, Appl. Catal., A 2004, 272, 249-256.
[22] K. Doi, E. Togano, S. S. Xantheas, R. Nakanishi, T. Nagata, T. Ebata, Y.
Inokuchi, Angew. Chem. 2013, 125, 4476-4479; Angew. Chem. Int. Ed.
2013, 52, 4380-4383.
[23] G. Wang, C. Sararuk, Z. X. Li, C. S. Li, AIChE J. 2018, 4, 1359-1372.
[24] A. W Schnizer, G. J Fisher, US Patent No. US2714616A, 1955.
[25] M. Shalom, S. Inal, C. Fettkenhauer, D. Neher, M. Antonietti, J. Am.
Chem. Soc. 2013, 135, 7118-7121.
C. A. Eckert, W. J. Frederick, J. P. Hallett, D. J. Leak, C. L. Liotta, J. R.
Mielenz, R. Murphy, R. Templer, T. Tschaplinski, Science 2006, 311,
484-489.
E. L. Kunkes, D. A. Simonetti, R. M. West, J. C. Serrano-Ruiz, C. A.
Gärtner, J. A. Dumesic, Science 2008, 322, 417-421.
B. O. De Beeck, M. Dusselier, J. Geboers, J. Holsbeek, E. Morré, S.
Oswald, L. Giebeler, B. F. Sels, Energy Environ. Sci. 2015, 8, 230-240.
http://www.nrel.gov/docs/fy04osti/35523.pdf_accessed on 18.12.2018.
J. A. Dumesic, D. A. Simonetti, E. L. Kunkes, US Patent No. 0116449 A1,
2013.
[6]
[7]
[8]
[9]
[10] B. Girisuta, L. P. B. M. Janssen, H. J. Heeres, Ind. Eng. Chem. Res. 2007,
46, 1696-1708.
[26] B. Kiskan, J. Zhang, X. Wang, M. Antonietti, Y. Yagc, ACS Macro Lett.
2012, 1, 546-549.
[11] J. Q. Bond, D. M. Alonso, D. Wang, R. M. West, J. A. Dumesic, Science
2010, 327, 1110-1114.
[27] B. Kumru, V. Molinari, M. Shalom, M. Antonietti, B. V. K. J. Schmidt, Soft
Matter. 2018, 14, 2655-2664.
[12] D. Albani, Q. Li, G. Vilé, S. Mitchell, N. Almora-Barrios, P. T. Witte, N.
López, J. Pérez-Ramírez, Green Chem. 2017, 19, 2361-2370.
[13] T. Ennaert, J. Van Aelst, J. Dijkmans, R. De Clercq, W. Schutyser, M.
Dusselier, D. Verboekend, B. F. Sels, Chem. Soc. Rev. 2016, 45, 584-
611.
[14] I. T. Horváth, H. Mehdi, V. Fábos, L. Boda, L. T. Mika, Green Chem. 2008,
10, 238-242.
[15] M. Al-Naji, A. Yepez, A. M. Balu, A. A. Romero, Z. Chen, N. Wilde, H. Li,
K. Shih, R. Gläser, R. Luque, J. Mol. Catal. A: Chem. 2016, 417, 145-
152.
[28] B. Kumru, V. Molinari, R. Dünnebacke, K. G. Blank, B. V. K. J. Schmidt,
Macromol. Rapid Commun. 2018, DOI: 10.1002/marc.201800712.
[29] G. M. Lari, B. Puértolas, M. S. Frei, C. Mondelli, J. Pérez-Ramírez,
ChemCatChem 2016, 8, 1507-1514.
[30] T. C. Keller, M. Položij, B. Puértolas, H. V. Thang, P. Nachtigall, J. Pérez-
Ramírez, J. Phys. Chem. C 2016, 120, 4954-4960.
[31] Y. Ono, H. Hattori in Solid Base Catalysis, Springer, 2012, pp. 180.
[32] A. A. Lemonidou, L. López, L. E. Manzer, M. A. Barteau, Appl. Catal., A
2004, 272, 241-248.
[16] C. Wei, D. Esposito, K. Tauer, Polym. Degrad. Stab. 2016, 131, 157-168.
[17] W. J. McGraw, US Patent No. 2624723, 1953.
This article is protected by copyright. All rights reserved.

10.1002/cssc.201900418
ChemSusChem
FULL PAPER
Entry for the Table of Contents
FULL PAPER
Sustainable tandem flow process:
Majd Al-Naji*, Begoña Puértolas, Baris
Kumru, Daniel Cruz, Marius Bäumel,
Bernhard V. K. J. Schmidt, Nadezda V.
Tarakina, Javier Pérez-Ramírez
MeGVL
was
synthesized
from
lignocellulosic biomass-derived GVL in
the presence of trioxane using tailored
Cs-loaded hierarchical beta zeolite
catalysts. The obtained MeGVL was
Page No. – Page No.
subjected
to
visible-light-induced
Sustainable
Valorization of -Valerolactone with
Trioxane
Valerolactone
Zeolites
Continuous
Flow
polymerization reaction, reaching a
final material with similar properties to
poly(methyl) methacrylate.
to
over
-Methylene--
Basic Beta
This article is protected by copyright. All rights reserved.