Selective defunctionalization of citric acid to tricarballylic acid as a precursor for the production of high-value plasticizers

Article abstract of DOI:10.1039/d0gc02298e

Strong concerns about the toxicity and endocrine disrupting properties of widespread phthalate plasticizers stimulate the demand for safe and preferably biobased alternatives. Citric acid forms in this respect an excellent and abundant platform chemical for the production of valuable plasticizers. Here, we report a new and direct synthesis route for propane-1,2,3-tricarboxylic acid (PTA) from citric acid via a sequential one pot dehydration-hydrogenation process. This saturated triacid can serve as a basis for the production of tricarballylate esters via esterification, which have been shown to possess excellent plasticizing properties in vinyl resins. In the presence of a solid acid H-Beta zeolite and Pd/C hydrogenation catalyst, yields up to 85% of PTA were obtained under mild reaction conditions and in water as a green solvent. Partial dealumination of the H-Beta zeolite by citric acid could be counteracted by reincorporating aluminium into the framework of the recycled H-Beta zeolite through realumination, regenerating a significant fraction of the initial activity of the catalytic system. The success of the realumination procedure was verified via MAS NMR spectroscopy.

Full text of DOI:10.1039/d0gc02298e

View Article Online
View Journal
Green
Chemistry
Cutting-edge research for a greener sustainable future
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: W. Stuyck, J.
Verduyckt, A. Krajnc, G. Mali and D. E. De Vos, Green Chem., 2020, DOI: 10.1039/D0GC02298E.
Volume 20
This is an Accepted Manuscript, which has been through the
Number
May 2018
Pages 1919-2160
9
7
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Cutting-edge research for a greener sustainable future
rsc.li/greenchem
Accepted Manuscripts are published online shortly after acceptance,
before technical editing, formatting and proof reading. Using this free
service, authors can make their results available to the community, in
citable form, before we publish the edited article. We will replace this
Accepted Manuscript with the edited and formatted Advance Article as
soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes to the
text and/or graphics, which may alter content. The journal’s standard
Terms & Conditions and the Ethical guidelines still apply. In no event
shall the Royal Society of Chemistry be held responsible for any errors
or omissions in this Accepted Manuscript or any consequences arising
from the use of any information it contains.
ISSN 1463-9262
PAPER
Paul T. Anastas et al.
The Green ChemisTREE: 20 years after taking root with the 12
principles
rsc.li/greenchem

Page 1 of 13
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02298E
Selective defunctionalization of citric acid to tricarballylic acid as a
precursor for the production of high-value plasticizers
Wouter Stuyck, a Jasper Verduyckt, a Andraž Krajnc, b Gregor Mali b and Dirk E. De Vos *a
a.
Centre for Membrane Separations, Adsorption, Catalysis and Spectroscopy for Sustainable Solutions (cMACS), KU Leuven,
Celestijnenlaan 200F, 3001 Leuven (Belgium).
b. National Institute of Chemistry, Hajdrihova 19, 1001 Ljubljana (Slovenia).
Abstract
Strong concerns about the toxicity and endocrine disrupting properties of widespread phthalate plasticizers
stimulate the demand for safe and preferably biobased alternatives. Citric acid forms in this respect an excellent and
abundant platform chemical for the production of valuable plasticizers. Here, we report a new and direct synthesis
route for propane-1,2,3-tricarboxylic acid (PTA) from citric acid via a sequential one pot dehydration-hydrogenation
process. This saturated triacid can serve as a basis for the production of tricarballylate esters via esterification, which
have been shown to possess excellent plasticizing properties in vinyl resins. In the presence of a solid acid H-Beta
zeolite and Pd/C hydrogenation catalyst, yields up to 85% of PTA were obtained under mild reaction conditions and
in water as a green solvent. Partial dealumination of the H-Beta zeolite by citric acid could be counteracted by
reincorporating aluminium into the framework of the recycled H-Beta zeolite through realumination, regenerating
a significant fraction of the initial activity of the catalytic system. The success of the realumination procedure was
verified via MAS NMR spectroscopy.
Introduction
With a production of 380 million metric tons (Mt) in 2015,1 and a projected growth to 1 124 Mt by the year 2050,2 it
is clear that plastics are indispensable for our current and future modern society. On account of the increasing
production and consumption of plastics, the production of plastic additives is forecasted to grow in a similar fashion.3
The largest market share of plastic additives is covered by plasticizers,4 which impart flexibility and processability to
the resulting plastic product by lowering its glass transition temperature.5 Over 90% of all plasticizers are used in
PVC-formulations, with the majority of these plasticizers (65%-70%) consisting of ortho-phthalates.6 Growing
concerns on these ortho-phthalates, due to their tendency to bioaccumulate and endocrine disrupting properties,7
have recently caused severe restrictions on these compounds.5 The generation of alternative plasticizers has mainly
focussed on developing safe compounds that have a similar structure as these ortho-phthalates, including
cyclohexanoates, trimellitates and terephthalates.5 However with increasing interest in renewable resources and
low toxicity of plasticizers, bio-based plasticizers have recently gained much attention. In this respect, epoxidised
(and acetylated) vegetable oils8 and isosorbide diesters9 have been shown to possess good plasticizing properties
for PVC. In addition, effective plasticizers based on cinnamic acid,4 cardanol10 and 2,5-furandicarboxylic acid
(FDCA)11,12 have been developed as well. However, the limited availability of cinnamic acid and cardanol, the
chemical modifications required for cardanol and the complexities in the synthesis of FDCA from hexoses, somewhat
hamper the applicability of these renewable resources for plasticizer production.
Another class of biobased plasticizers consists of compounds that are based on citric acid. With its branched
structure, three carboxylic functional groups and hydroxyl group, it may serve as a general basis to synthesize
valuable plasticizers. Moreover, citric acid is a widely abundant renewable resource, since it can be produced via an
efficient fermentation of biological waste, resulting in an annual production of 2.3 million tons in 2016.13 Citric acid
itself is too polar to be compatible with PVC formulations. Therefore it is made more apolar via an esterification to
for instance triethyl citrate (TEC) and tributyl citrate (TBC), often followed by an acetylation or butyrylation of the
tertiary hydroxyl group to acetyl triethyl citrate (ATEC), acetyl tributyl citrate (ATBC) and butyryl trihexyl citrate
(BTHC), respectively. When solely esterified, the resulting plasticizers still possess a prominent hydroxyl group, which
makes them prone to migration out of the polymer matrix.14 After acetylation (ATBC) or butyrylation (BTHC), these
plasticizers possess a quality that is comparable to the widely used di-2-ethylhexyl phthalate (DEHP). However, ATBC
often remains more prone to leaching compared to DEHP.15,16 In addition, the derivatization of the tertiary hydroxyl
group requires the use of anhydrides and strong Brønsted acid catalysts,17 thus involving harsh acidic conditions and
stoichiometric acetic or butyric acid waste. Moreover, the use of petrochemically based anhydrides implicates a loss
1

Green Chemistry
Page 2 of 13
View Article Online
DOI: 10.1039/D0GC02298E
of the biobased aspect of the final plasticizer. An alternative approach is to remove this tertiary hydroxyl group of
citric acid, prior to esterifying the resulting tricarballylic acid (propane-1,2,3-tricarboxylic acid, PTA) with the desired
alcohol. The resulting tricarballylate plasticizers have shown excellent plasticizing efficiency in polyvinyl chloride
(PVC) resins, with plasticizing properties such as tensile strength, modulus at 100 %, elongation at break and brittle
point comparable to and in some cases even better than those of DEHP.18 Furthermore, these plasticizers also
possess excellent light stability as well as a more satisfactory spew rating compared to abovementioned citrate
plasticizers19 and can be expected to leach less from the polymer matrix due to their lower polarity. In addition to
plasticizers for PVC, tricarballylates are also promising plasticizer candidates for polylactic acid, a biobased plastic
that has gained much attention over the years. Due to the high density of aliphatic ester bonds of both, they are
expected to be highly compatible with each other.20 However, a desirable direct route from the abundant citric acid
to PTA has not yet been described.21 Recently, Verduyckt et al. optimised the route from citric acid towards
methylsuccinic acid; PTA was observed in some cases as a side product with a maximal yield of 7%.22
Due to the large number of oxygen functionalities in biomass, deoxygenation routes have been a major topic of
interest when considering these renewable resources for the production of valuable chemicals.23 To perform
biomass dehydration under aqueous conditions, traditionally strong liquid Brønsted acids are applied.24-27
Concerning the low recyclability of these catalysts and the generation of large quantities of salt waste during product
separation, heterogeneous acid catalysts have been proposed to replace strong liquid Brønsted acids. In this respect,
zeolites have been proven to be active and stable under aqueous conditions28,29 and have shown good applicability
for the cyclo-dehydration of alditols30-33 and the conversion of levulinic acid to methyltetrahydrofuran.34 In this work,
we present the successful production of PTA via a sequential dehydration-hydrogenation of citric acid under mild
conditions, in the presence of a solid acid H-Beta zeolite and Pd/C catalyst, water as a green solvent and H2 as a clean
reductant. Furthermore we report on the stability and recyclability of the H-Beta zeolite catalyst.
Experimental section
Materials
All chemicals were used as received: citric acid monohydrate (Sigma-Aldrich, ≥99.5%), tricarballylic acid (Sigma-
Aldrich, ≥99%), methylsuccinic acid (Sigma-Aldrich, 99%), trans-aconitic acid (Sigma-Aldrich, 98%), mesaconic acid
(Sigma-Aldrich, 99%), itaconic acid (TCI, 99.0%), methacrylic acid (Acros Organics, 99.5%), HCl (VWR, 37% in H2O),
H2SO4 (Acros, 95% in H2O), Al(OH)3 (Merck, >99.0%), zirconium n-propoxide (TCI Europe, 70 wt% in n-propanol),
isopropanol (Acros Organics, 99.5%), zirconium(IV) acetylacetonate (abcr GmbH, 98%) H2 (Air Liquide, N40), N2 (Air
Liquide, α1), Pd/C (5 wt%, Johnson Matthey), Amberlyst 15 hydrogen form wet (Sigma Aldrich), zeolite H-ZSM-5
(Zeolyst, CBV 3024E), zeolite H-MOR 6.8 (Norton), zeolite H-MCM-22 14 (China Catalyst Holding Co., Ltd.), zeolite
H-Y 2.7 (China Catalyst Holding Co., Ltd.), zeolite H-Y 17.5 (China Catalyst Holding Co., Ltd.), zeolite H-Y 30 (Zeolyst,
CBV 760), zeolite H-Beta 10.8 (PQ corporation, CP 811 BL-25), zeolite H-Beta 12.5 (Zeolyst, CP814e), zeolite H-Beta
32.5 (Zeocat, PB/65 H), zeolite H-Beta 75 (Clariant, HCZB 150), 2-methyltetrahydrofuran (Fischer Scientific, >99%),
maleic acid (TCI, >99%), D2O (Sigma Aldrich, 99.9% D).
Catalytic reactions
In a typical reaction, a glass vial (5.5 mL) was loaded with a solution of citric acid in deionized water (2 mL, 0.1 M), a
hydrogenation catalyst (0.5 mol% Pd/C), an acid catalyst (H-Beta 10.8, 0.1 g) and a magnetic stirring bar. A teflon
stopper was placed on the glass vial, partially closing it, after which it was placed in a stainless steel reactor (11 mL).
The reactor was subsequently sealed, purged five times with N2, purged three times with H2 and pressurized with
10 bar H2. Next, the reactor was placed in a heating block at 150 °C under continuous magnetic stirring at 750 rpm.
After 20 hours the reactor was cooled in an ice bath to quench the reaction. The remaining gas was evacuated at
room temperature and the solid catalysts were separated from the liquid reaction mixture by centrifugation. The
reaction mixture was analysed with 1H-NMR.
Product analysis and identification
To determine the product distribution and the amount of unreacted citric acid and citrate, the reaction mixtures
1
were analysed by H-NMR. For sample preparation, 300 µL crude reaction mixture was mixed with 300 µL of a
2

Page 3 of 13
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02298E
solution of maleic acid in D2O (0.1 M) and transferred to an NMR tube. Maleic acid (δ = 5.998 ppm, s, 2H) will serve
1
as an external standard for the quantification of the different compounds of the reaction mixture (ESI). H-NMR
spectra were recorder on a Bruker Avance 400 MHz spectrometer equipped with a sample case and a 5 mm PABBO
BB/19F-1H/D probe with z-gradients and ATM accessory for Automatic Tuning and Matching. The broad singlet of
the solvent water (around δ = 4.65 ppm) was suppressed by applying a zgpr pulse program: p1 9.75 µs; plw1 15 W;
plw9 5.7 10^(-0.5) W; ds 4; ns 32; d1 12 s; aq 2.55 s; sw 16 ppm; o1P on the resonance signal of water, automatically
determined and selected. For correct quantitative analysis, the relaxation time (d1) of the pulse program was
1
adjusted to 12 s. Complete product identification by H- and 13C-NMR can be retrieved from previous work of our
group.22 To distinguish between the amount of citric acid and citrate, the crude reaction mixture (200 µL) was
1
acidified with a solution of HCl in water (100 µL, 2 M) prior to H-NMR analysis. Description of this method and
interpretation of the 1H-NMR spectra are provided in the ESI.
The calculated conversions (X) in this paper comprise the sum of the amount of citric acid and citrate that has reacted
under the applied reaction conditions on a molar base. The distribution of the remaining substrate, citric acid and
citrate, is also shown. The different selectivities (S) mentioned in this work are calculated as follows:
푌푃푇퐴
푆푃푇퐴
=
=
푋
푌푀푆퐴
푋
푆푀푆퐴
푌퐴푐푒푡푖푐 퐴푐푖푑 + 푌퐴푐푒푡표푛푒
푆퐹푟푎푔푚푒푛푡푎푡푖표푛
=
푋
Where PTA and MSA stand for propane-1,2,3-tricarboxylic acid and methylsuccinic acid respectively. Y represents
the yield of the respective compounds, always on a molar base.
Recycle experiments and zeolite realumination
To recycle the catalyst, both solid catalysts were separated from the liquid fraction in the glass vial by means of
centrifugation (4000 rpm, 10 min). The liquid fraction is then removed, after which the solid catalysts were washed
with water. This procedure is repeated three times. At last, the catalyst is dried overnight (16 hours) in a vacuum
oven (30 °C, 30 mbar). Both catalysts were reused directly in a new catalytic experiment, without further activation
or regeneration steps. To calcine the H-Beta zeolite prior to reuse, this catalyst first has to be separated from the
hydrogenation catalyst (Pd/C) in the crude reaction mixture. This was achieved via an extraction, where the addition
of 2-methyltetrahydrofuran to the crude reaction mixture resulted in a clear separation of both catalysts over both
liquid phases: H-Beta remains in the aqueous phase, while Pd/C migrates to the organic phase. Next, the aqueous
phase was transferred into a glass vial and H-Beta was obtained after multiple washing and centrifugation steps,
followed by overnight drying in a vacuum oven (cfr. supra) and calcination in a shallow bed in a porcelain crucible
under static air (550 °C for 4 h, 2 °C/min).
Realumination of zeolite H-Beta was achieved following a procedure recently published by Zhao et al.35 To obtain a
sufficient amount of dealuminated H-Beta zeolite to conduct the realumination, the two recycle experiments prior
to the realumination procedure were performed on a larger scale in a 600 mL Parr reactor. For both recycle
experiments, the H-Beta catalyst to citric acid ratio was the same as under standard reaction conditions; For the first
recycle experiment, the reactor was loaded with a solution of citric acid in deionized water (200 mL, 0.1 M), 0.5 mol%
Pd/C and with 10 g of H-Beta catalyst. For the second recycle experiment, the reactor was loaded with 160 mL of a
0.1 M citric acid in deionized water solution, 0.5 mol% Pd/C and 8 g of the recycled H-Beta catalyst. Both recycle runs
were conducted at 150 °C and with 10 bar H2 under continuous mechanical stirring (750 rpm). After each recycle
run, the H-Beta catalyst was recovered through liquid extraction, dried and calcined (cfr. supra). To realuminate the
H-Beta catalyst, 1 g of the twice dealuminated dry zeolite H-Beta was mixed with an aqueous Al(NO3)3 solution
(20 mL, 0.1 M), after which the resulting mixture was sealed in a stainless steel autoclave. The autoclave was heated
at 165 °C for 72 h. After quenching the autoclave, the material is obtained, washed and dried (110 °C, 6 h) prior to
3

Green Chemistry
Page 4 of 13
View Article Online
DOI: 10.1039/D0GC02298E
calcination in a shallow bed in a porcelain crucible under static air (550 °C for 4 h, 2°C/min). The resulting zeolite is
named realuminated H-Beta.
Solid-state NMR spectroscopy
27Al, 29Si magic-angle spinning (MAS) and 1H-29Si cross-polarization (CP)MAS NMR experiments were carried out on
a 600 MHz Varian VNMRS spectrometer, equipped with a 3.2 mm HX CPMAS probe. Larmor frequencies for 27Al and
29Si nuclei were 156.19 MHz and 119.08 MHz, respectively. In all experiments, the sample rotation frequency was
20 kHz. 27Al MAS NMR measurements employed a short 1.0 µs excitation pulse; 5000 scans were collected and the
delay between the scans was 0.15 s. In 29Si MAS NMR measurements 90-degree excitation pulse of 2.8 µs and a
repetition delay of 60 s were used; the number of scans was 500. 1H-29Si cross-polarization MAS (CPMAS) experiment
used ramped-amplitude cross-polarization scheme with a duration of 5 ms and high-power proton decoupling;
repetition delay was 3 s and the number of scans was 10 000. The applied formulas for the calculations of the total
and framework Si/Al ratios of all three zeolite samples can be retrieved from the ESI.
Results and discussion
Sequential dehydration-hydrogenation of citric acid to PTA
To obtain PTA from citric acid, a sequential dehydration-hydrogenation has to be performed (Scheme 1). First citric
acid (1) is dehydrated to aconitic acid (2), after which this unsaturated compound has to be hydrogenated to obtain
PTA (3). This first dehydration step is an equilibrium reaction and is enhanced by applying temperatures above
200 °C.36 However, a high temperature also promotes the spontaneous decarboxylation reaction of aconitic acid (3)
to itaconic acid (4), which can isomerise to mesaconic (5) and citraconic acid (6).36,37 Under reductive conditions
these unsaturated diacids can be reduced to methylsuccinic acid (MSA) (7),22 or they can further decarboxylate to
methacrylic (8) and crotonic acid (9), which in turn can be reduced to isobutyric (10) and butyric acid (11).22,37 Next
to high temperatures, both decarboxylation steps can also be catalyzed by metals like Ru, Pt or Pd.22,38
Scheme 1. Reaction scheme of the dehydration-hydrogenation of citric acid with undesirable decarboxylation side reactions.
For maximizing the yield of PTA (3) from citric acid (1), the decarboxylation of aconitic (2) to itaconic acid (4) is
undesired, since it implies an irreversible loss of CO2. To suppress this spontaneous decarboxylation, citric acid was
reacted at a low temperature of 140 °C, under reductive conditions in presence of 0.5 mol% Pd (Table 1, entry 1).
Even if good selectivities for PTA (3) were obtained (86%), the conversion of citric acid remained low (7%), since the
dehydration to aconitic acid is slow at this temperature (2). Besides PTA and a small amount of MSA (7), a small
fraction of fragmented products was observed. These consist of acetone (12) and acetic acid (13) and originate from
radical fragmentation, which might be facilitated by non-reduced Pd species at the surface of the Pd/C catalyst.22
Other side products like itaconic acid (4) (or its isomers mesaconic (5) and citraconic acid (6)) and products resulting
from the second decarboxylation (8-11) were not observed in any of the experiments. Itaconic acid (4) is readily
4

Page 5 of 13
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02298E
reduced to MSA (7) and is therefore less prone to a second decarboxylation under the applied low temperature
reductive conditions.22,38
To promote the dehydration of citric acid (1) to aconitic acid (2), an acid catalyst was added in following experiments
(Table 1, entries 2-11) under the same conditions as in entry 1. To avoid harsh acidic conditions, limit salt waste and
facilitate catalyst and product separation, an emphasis was made on heterogeneous solid acids. In this respect,
Amberlyst-15, a strong acid sulphonated polystyrene (entry 2) and different zeolites in their protonated form (entries
3-12) were evaluated. In the case of Amberlyst-15, no improvement compared to the reaction without acid catalyst
was noticed. In the presence of zeolite acid catalysts, a significant increase in activity was observed together with an
increase in selectivity towards the desired product PTA (3) (except for entry 12). The product distribution shown in
Table 1 comprises also citrate, besides citric acid. It was proven via 27Al NMR and ICP-OES that part of the Al of the
zeolites forms complexes with citric acid (cfr. infra). The 10-membered ring (MR) zeolites H-MCM-22 and H-ZSM-5
(entries 3 & 4) were clearly outperformed by the 12 MR zeolites H-Mor (entry 5), H-Y (entries 6-8) and H-Beta (entries
9-12), indicating that the acid sites are more readily accessible in 12 MR zeolites. The better activity of H-MCM-22
(2D pore system, entry 4) compared to H-ZSM-5 (3D pore system, entry 3), at a similar Si/Al ratio, can be assigned
to the more accessible acid sites in the 12 MR hemisupercages on the external surface of H-MCM-22.39 Comparing
12 MR zeolites, it is clear that H-Beta and H-Y (3D pore system) are superior to H-Mor (2D pore system). For H-Y
(entries 6-8) and H-Beta (entries 9-12), different Si/Al ratios were tested, where a higher Si/Al represents less but
stronger Brønsted acid sites and more hydrophobic zeolite pores.40,41 In both cases a lower Si/Al ratio resulted in a
higher conversion, evidencing that the number of acid sites is more important than the strength of these Brønsted
acid sites. This is illustrated by the proposed reaction mechanism in Scheme 2, where a higher amount of Al and thus
Brønsted acid sites in the zeolite, will lead to more catalytic sites where the dehydration of citric acid (1) can take
place. In contrast to previous research on the cyclodehydration of alditols in presence of zeolites, the hydrophobic
effect of higher Si/Al zeolites does not seem to play a key role in the dehydration of citric acid (1) to aconitic acid
(2).31-33
Table 1. Sequential dehydration - hydrogenation of citric acid. a
Distribution [%]f
Selectivity [%]
Amount Pd/Cd
Conc.e
[M]
PH2
Acid Cat.b
-
Si/Al cat.c [g] [mol%]
[bar]
T [°C] Citric acid Citrateg
X [%]h
7
PTAi
86
75
97
97
88
95
93
92
96
96
MSAj
Fragm.k
1
2
3
4
5
6
7
8
9
-
-
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
0.5
3
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
6
140
140
140
140
140
140
140
140
140
140
140
140
140
140
140
140
150
160
150
150
93
93
89
77
72
15
54
78
71
63
86
92
7
0
9
< 1
4
Amberlyst-15
H-ZSM-5
H-MCM-22
H-Mor
H-Y
-
0.02
0.02
0.02
0.02
0.02
0
7
21
3
15
14
6.8
2.7
1
10
18
24
53
34
20
28
29
14
8
< 1
< 1
3
5
3
5
9
33
12
2
4
2
H-Y
17.5 0.02
30 0.02
5
2
H-Y
5
3
H-Beta
10.8 0.02
12.5 0.02
32.5 0.02
1
1
2
10 H-Beta
11 H-Beta
12 H-Beta
13 H-Y
8
2
2
< 1
< 1
53
20
30
22
18
5
8
4
75
0.02
0.1
85
95
97
99
85
96
90
90
94
11
< 1
1
4
2.7
40
47
44
42
80
95
75
71
5
14 H-Beta
15 H-Beta
16 H-Beta
17 H-Beta
18 H-Beta
19 H-Beta
20 H-Beta
10.8 0.1
12.5 0.1
10.8 0.1
10.8 0.1
10.8 0.1
10.8 0.1
10.8 0.1
33
26
36
3
1
1
< 1
1
14
3
0.5
0.5
0.1
0.5
< 1
< 1
1
< 1
4
9
22
21
9
8
6
< 1
5

Green Chemistry
Page 6 of 13
View Article Online
DOI: 10.1039/D0GC02298E
21 H2SO4
-
-
-
-
-
2 eq.l
8 eq.l
0.1
0.5
0.1
0.1
0.1
0.1
0.1
0.4
0.8
10
10
10
10
10
10
10
150
150
150
150
150
150
150
88
79
71
35
94
12
30
-
15
21
29
32
6
94
95
68
94
95
93
92
2
4
22 H2SO4
0.5
-
2
3
23 SO42-/ZrO2
24 Al(OH)3
25 Zr-Beta
26 H-Beta
27 H-Beta
0.5
-
24
5
8
0.5 eq.l
0.5
33
-
< 1
3
0.1
0.5
1
10.8 0.1
0.125
0.063
10
6
79
65
6
< 1
< 1
10.8 0.1
7
a Reaction conditions: water (2 mL), 20 h. b Acid catalyst. c Amount of acid catalyst. d mol% of Pd (5 wt.% Pd/C) with respect to
e
f
citric acid. Concentration of substrate citric acid. Distribution of the reactants, consisting of citric acid and Al-citrate, after
reaction. The sum of both represents the total amount of unreacted substrate. g The amount of citric acid that is present as Al-
citrate determined by 1H-NMR after acidifying the mixture (See ESI). h Conversion i.e. the amount of citric acid and citrate that
has reacted. i Propane-1,2,3-tricarboxylic acid. j Methylsuccinic acid. k Fragmentation products, which are represented by acetone
and acetic acid. l Equivalents with respect to citric acid.
Scheme 2. Proposed reaction mechanism for the subsequent dehydration-hydrogenation of citric acid on H-Beta and Pd/C
respectively.
Next, the amount of acid catalyst was varied for the three best performing catalysts: H-Y 2.7, H-Beta 10.8 and
H-Beta 12.5 (entries 13-15). An increase in the amount of both H-Beta zeolites resulted in a higher conversion and
selectivity for PTA (3), without further significant improvement above 5 wt% (ESI, Table S2). Increasing the amount
of H-Y 2.7 resulted in a significant decrease in conversion, although high selectivities for PTA (3) were maintained.
This reduction can be assigned to the lower stability of aluminum-rich H-Y in aqueous acidic conditions due to
extensive dealumination, as suggested by the large amount of Al-citrate identified in the 1H-NMR spectra
(entry 13).28,42-44 Increasing the amount of Pd/C did not result in an increase in conversion and induced a reduction
in selectivity towards PTA (3), presumably due to Pd-catalyzed decarboxylation of aconitic acid (2) to itaconic acid
(4) (entry 16).22,38 The lack of rise in conversion with an increased amount of Pd also indicates that not the
hydrogenation, but rather the dehydration reaction is the limiting step in the transformation of citric acid (1) to PTA
(3). The highest yield of PTA (3) obtained thus far is 46% in presence of 0.1 g H-Beta 10.8 and 0.5 mol% Pd (entry 14);
this catalyst combination was used to further optimize the reaction conditions.
Increasing the reaction temperature had a marked effect on the conversion of citric acid (1), reaching 80% at 150 °C
(entry 17) and 95% at 160 °C (entry 18). Next to accelerating the dehydration of citric acid (1) to aconitic acid (2), an
increase in reaction temperature also enhances the undesired decarboxylation of aconitic acid (2) to itaconic acid
(4), as shown by the increased selectivity for MSA (7) at higher temperatures. However, selectivities of at least 90%
for PTA (3) were obtained at both temperatures. To further investigate the conversion of citric acid and product
formation at these elevated temperatures, time profiles were recorded at 150 °C and 160 °C (Figure 1). Both profiles
have a similar shape; after 10 h at 160 °C a similar PTA yield (79%) was observed as after 20 h at 150 °C (77%). From
the profile at 150 °C it is also clear that Al citrate is formed almost instantaneously and that its concentration remains
approximately constant over time. Thus, even if partial dealumination of H-Beta 10.8 has already taken place at the
onset of the reaction, the catalytic system remains active, indicated by the linear increase in PTA (3) production over
time. From the profile at 150 °C it also seems that Al citrate might be a dead end product, since its concentration
remains more or less unaffected over time. However, when the reaction temperature is raised to 160 °C, Al citrate
6

Page 7 of 13
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02298E
is converted as well but only after most of the citric acid (1) is converted. This indicates that Al citrate is less reactive
than citric acid (1) but not a dead end product.
Figure 1. Time profiles of the sequential dehydration-hydrogenation of citric acid at 150 °C (left) and 160 °C (right). Each series of
data points at a specific time corresponds to a different batch reaction. The error bars at 150 °C and 20 h represent the standard
deviation with a sample size of n = 4. Reaction conditions: Citric acid (0.2 mmol), water (2 mL), 0.1 g H-Beta 10.8, 0.5 mol% Pd
(5 wt% Pd/C), 10 bar H2. Legend: Citric acid (orange), Citrate (yellow), PTA (blue), Side products* (green). * Side products consist
of MSA and less than 1% of fragmented products in all cases.
Both lowering the amount of Pd/C (entry 19) and H2-pressure (entry 20) resulted in a slightly lower conversion of
citric acid (1) and lower selectivity towards PTA (3). This indicates that a sufficient, but still industrially relevant
amount of Pd/C and H2-pressure are required to direct the unsaturated aconitic acid (2) to the desired hydrogenated
product PTA (3) and avoid decarboxylation to itaconic acid (4). Further variations of both parameters are included in
Table S3 and Table S4 in the ESI.
Homogeneous Brønsted acid catalysts were ineffective to catalyze the dehydration of citric acid (1) (entries 21 &
22); a conversion of only 21% was reached in presence of 8 eq. H2SO4. The superiority of H-Beta 10.8 compared to
H2SO4 might be ascribed to the stronger acidity of the Brønsted acid sites in the zeolite. However, when a solid
superacid (SO42-/ZrO2) was applied under the same conditions (entry 23), again a low conversion and in this case also
a lower selectivity for PTA (3) were observed. From these results, it can be concluded that catalysts with
predominantly strong Brønsted acidity are not effective, which could already be perceived from the performance of
Amberlyst 15 (entry 2). To further investigate the role of the acid catalyst, 0.5 eq. of Al(OH)3 was applied as a
homogeneous Lewis acid catalyst (entry 24). Although large amounts of Al citrate were observed, a conversion of
35% was reached and high selectivities were obtained, indicating that Lewis acids can also facilitate the dehydration
of citric acid (1) in acidic aqueous media. Applying a Lewis acid Zr-Beta zeolite on the other hand, did not result in an
effective conversion of citric acid (1) (entry 25).
These findings suggest that next to its hydrothermal stability and adequate amount and accessibility of acid sites,
zeolite H-Beta 10.8 has another particular feature that makes it more suitable than other homogeneous and
heterogeneous acids for the dehydration of citric acid (1). Next to strong Brønsted acidity, H-Beta also possesses
Lewis acidity which can originate both from octahedrally coordinated extra-framework Al-species as well as from
“perturbed” tetrahedrally coordinated framework Al-species.45-47 The latter are generated on hydrolysed
tricoordinated Al species (Al(SiO)3OH) which under the applied aqueous conditions of this work will be
predominantly present in (pseudo)-octahedral symmetry.46-50 Considering the significant amount of Al citrate
formation during the catalytic reactions in presence of H-Beta and the decrease of activity upon its recycling (cfr.
7

Green Chemistry
Page 8 of 13
View Article Online
DOI: 10.1039/D0GC02298E
infra), it is clear that a notable amount of Al will be leached from the zeolite. During this dealumination process,
octahedrally coordinated (framework or extra-framework) Al species are formed (Lewis acids) at the expense of
tetrahedrally coordinated Al species, which are associated with Brønsted acid sites. It therefore seems that when a
citric acid (1) molecule is coordinated to such a Lewis acid, it is more prone to dehydration when it encounters a
strong Brønsted acid site in its close vicinity. Since about half of the Al sites of low Si/Al Beta zeolites (10 < Si/Al < 18)
are present as Al pairs, of which at least 50% face the same zeolite channel, combined presence of Lewis and
Brønsted acid sites is highly likely to occur in zeolite H-Beta channels.49,51 Here we propose that this combined
interaction, together with previously cited characteristics of H-Beta 10.8, accounts for the superiority of this solid
acid for the dehydration of citric acid (1) to aconitic acid (2).
A final set of experiments included the increase of the citric acid concentration (entries 26 & 27), to mimic a more
industrially relevant aqueous citric acid stream. For both experiments the amount of both catalysts was kept
constant (0.1 g H-Beta 10.8 and 2.1 mg Pd/C), resulting in a decrease of the catalyst/substrate ratio at higher
concentrations of citric acid. At these elevated concentrations of citric acid, the acidity of the aqueous medium
increases, which will facilitate the dehydration reaction as well. Starting from 0.8 M citric acid (1), an acceptable
conversion of 65% and high selectivities for PTA (3) of 92% were obtained, ensuring the industrial relevance of this
catalytic system.
Recyclability, dealumination and realumination of zeolite H-Beta 10.8
The observation of significant quantities of Al citrate in the reaction mixtures with zeolite catalysts, suggests an
inevitable loss of Al through dealumination. To investigate whether this dealumination affects the stability and
activity of the catalyst, recycling tests were performed (Figure 2 & Table S5 ESI). Directly reusing the catalytic system
(both H-Beta and Pd/C) in consecutive recycle runs resulted in a significant decrease of the activity of H-Beta, as
evidenced by the decrease in citric acid (1) conversion (80% to 45%) and PTA (3) yield (77% to 42%). This gradual
decay in activity can be attributed to loss of acid sites due to dealumination. This is illustrated by the formation of
additional citrate in the first and second reuse experiments (Figure 2 & Table 5 ESI). When zeolite H-Beta was
separated from Pd/C through a liquid extraction with 2-methyltetrahydrofuran (Me-THF) and calcined at 550 °C prior
to the following catalytic run, the decreases in citric acid (1) conversion (80% to 55%) and PTA (3) yield (77% to 53%)
were less pronounced and high selectivities for PTA (3) remained (> 95%) for both recycle runs. A possible
explanation for this increase in citric acid (1) conversion and PTA (3) yield compared to directly reusing both catalysts,
is that some of the (partially) hydrolysed octahedral framework and extraframework aluminum is reverted to
tetrahedral framework aluminum upon calcination of H-Beta, thereby delaying the dealumination in the subsequent
recycling experiment.
In an endeavour to restore the activity of the zeolite catalyst, a twice reused and calcined H-Beta (further
denominated as dealuminated H-Beta) was realuminated with Al(NO3)3 following a recently published procedure of
Zhao et al.35 Applying this realuminated H-Beta catalyst under the same reaction conditions resulted in a clear
increase in citric acid conversion (72%) and PTA yield (68%). Although the high selectivity of PTA is again maintained
(> 95%), the activity of the fresh H-Beta 10.8 catalyst could not be fully restored. To monitor the crystallinity of the
different zeolite samples, X-ray diffractograms of the fresh, dealuminated and realuminated H-Beta catalyst were
acquired (Figure S5 ESI). The XRD diffractogram of the fresh H-Beta 10.8 shows diffraction peaks characteristic for
zeolite Beta framework structure at 2θ of 7.6°, 13.5°, 14.4°, 21.3° and 22.5°. For the dealuminated H-Beta sample, a
decrease in peak intensities is noticed, evidencing some loss of crystallinity. Upon realumination, the loss of
crystallinity is rather limited.
8

Page 9 of 13
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02298E
Figure 2. Recycling of the catalytic system. (a) Direct reuse, without calcination of H-Beta. (b) H-Beta is calcined at 550 °C prior to
reuse. Reaction conditions: 0.1 g H-Beta, 0.5 mol% Pd/C, citric acid 0.2 mmol, water 2 mL, 10 bar H2, 150 °C, 20 h. Legend: Citrate
(diamond), citric acid (horizontal lines), PTA (white), side products* (dark). The percentages denote the yield of PTA. The error
bars for the fresh H-Beta catalyst represent the standard deviation with a sample size of n = 4.* Side products consist of MSA and
less than 1% of fragmented products in all cases.
Since XRD gives no detailed local insight, solid state NMR was invoked to study the loss and reincorporation of Al
upon dealumination and realumination respectively in more detail. 29Si MAS NMR spectra (Figure 3) and 27Al MAS
NMR spectra (Figure 4) were recorded for the fresh H-Beta 10.8, for the dealuminated H-Beta, and for the
1
realuminated H-Beta catalysts. To reliably assign the resonance lines of the 29Si MAS NMR spectra, H-29Si CPMAS
NMR spectra were recorded as well (ESI). The NMR spectra were normalized indirectly in such a way that the total
integrals of the 29Si MAS NMR spectra of the three zeolites were equal. This approach is reasonable, since it can be
assumed that the silicon content in the samples changes very little during de- and realumination. After this
normalization, 27Al MAS NMR spectra offer a solid quantitative comparison of the total amount of Al within the
samples and of the ratio of framework and extra-framework Al species (Figure 4). The framework- and total Si/Al
ratio of the three zeolite samples can be estimated from deconvolution and integration of the 29Si MAS NMR and
27Al MAS NMR spectra whilst taking into account the amount of extra-framework Al from 27Al MAS NMR spectra for
the calculations of the total Si/Al (Table 2; calculations can be found in ESI). After two consecutive catalytic runs, the
framework Si/Al of the H-Beta zeolite increased from 11.4 to 39.5, indicating significant loss of Al from the zeolite
framework through dealumination. Through realumination, the framework Si/Al decreases again to 17.7, illustrating
that Al is reincorporated into the zeolite framework. However, the initial framework Si/Al could not be fully restored.
Table 2. Normalized Al-distribution from 27Al MAS NMR spectra and estimated framework- and total Si/Al ratio
Zeolite catalyst
AlNorm-Fr
94.0
27.1
AlNorm-EF
6.0
0.4
Si/AlFr
11.4
39.5
17.7
Si/AlTot
10.8
38.9
Fresh H-Beta 10.8
Dealuminated H-Beta
Realuminated H-Beta
60.6
4.9
16.4
The chemical shifts that are observed in the 27Al MAS NMR spectra are in agreement with previously reported
chemical shifts for H-Beta zeolites.52-56 The peaks in the region between -30 and 10 ppm are assigned to extra-
framework and framework octahedral aluminum. For the fresh H-Beta 10.8, these peaks correspond to a
superposition of a sharp signal at around 0 ppm and a broad peak of aluminum that experiences a large quadrupolar
3+
interaction. The former corresponds to aluminum species in highly symmetrical environment similar to Al(H2O)6
,
while the latter is assigned to a more hydrolyzed, distorted framework octahedral Al species.53-56 This distorted
octahedral framework aluminum is attached more loosely to the zeolite structure56 and is thus more accessible for
citric acid, resulting in a significant decrease of this octahedral Al species in the dealuminated zeolite (Figure 4). A
3+
small amount of Al(H2O)6 octahedral species remains present in the dealuminated zeolite due to continuous
hydrolysis of tetrahedral framework Al under the applied aqueous citric acid conditions. The resonance between 50
and 65 ppm is typical of tetrahedral framework Al. For the fresh zeolite H-Beta 10.8, this resonance consists of a
superposition of two contributions: one peak at 54 ppm, which corresponds to aluminum atoms on T1 and T2 sites
9

Green Chemistry
Page 10 of 13
View Article Online
DOI: 10.1039/D0GC02298E
in the H-Beta framework, and one peak at 57 ppm which can be assigned to aluminum atoms on T3-T9 sites.53,54
Under the applied reaction conditions, Al seems to be preferentially removed from these T3-T9 sites since the
contribution of the peak at 57 ppm is severely lowered for the dealuminated sample, while the peak at 54 ppm
remains more preserved (Figure 4). The increased resistance towards dealumination of framework Al on T1 and T2
positions is in agreement with previously reported dealumination of zeolite H-Beta under steaming conditions53,54
and acid leaching.57,58
During realumination, most of the aluminum is reincorporated into the zeolite H-Beta framework, as indicated by a
larger increase of the contribution of Al to the resonance of framework Al compared to the resonance of extra-
framework Al. Nonetheless, the distribution of Al over the different framework T sites after realumination does not
resemble the distribution that was observed in the original sample. This is also reflected in the 29Si MAS NMR spectra,
where the contributions of the Si(OSi)3(OAl) and Si(OSi)2(OAl)2 peaks are significantly different in the fresh and
realuminated zeolite (Figure 3). Realumination also changes the shapes (and widths) of the signals that belong to
tetrahedrally coordinated Al; the resonance of the realuminated sample is not as sharp as the resonance of the
original sample. These findings suggest that during realumination Al does not incorporate integrally into the sites
from which it was removed during the dealumination process. A possible explanation for this dissimilarity can be
found in the recycling procedure of the zeolite catalysts, where a calcination at 550 °C is performed between two
recycle runs. This calcination might have induced a loss of the silanol nests that were originally generated by removal
of Al during the sequential dehydration-hydrogenation of citric acid, resulting in an irreversible change of the zeolite
structure. However, upon realumination a considerable amount of the catalytic activity was restored (Figure 2),
illustrating that this realumination technique is a suitable approach to recycle the applied zeolite H-Beta catalyst and
in general the whole catalytic system.
Figure 3. 29Si MAS NMR spectra and their decompositions into contributions of Si from different environments of the fresh H-Beta
10.8 catalyst (bottom), the twice reused and calcined, dealuminated H-Beta (middle) and the realuminated H-Beta catalyst (top).
The black curve represents the measured NMR spectrum and the red dashed curve represents the modeled spectrum.
10

Page 11 of 13
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02298E
Figure 4. Comparison of the 27Al MAS NMR spectra of the fresh H-Beta 10.8 (bottom), the twice reused and calcined,
dealuminated H-Beta (middle) and the realuminated H-Beta catalyst (top). The amount of Al present as tetrahedral framework
Al and octahedral framework (broad signal extending between 0 and -25 ppm) and extra-framework (narrow signal at 0 ppm) Al
for the different H-Beta zeolites is noted above the respective ppm-regions relative to the amount of Al present in the fresh H-
Beta zeolite.
Conclusion
In conclusion, a direct synthesis route for propane-1,2,3-tricarboxylic acid (PTA) from the abundant citric acid was
successfully developed via a sequential dehydration-hydrogenation process. By applying a low Si/Al H-Beta solid acid
zeolite in combination with 0.5 mol% Pd/C, the dehydration of citric acid to aconitic acid was achieved at low reaction
temperatures, thereby suppressing the undesired decarboxylation of aconitic acid and boosting the selectivity for
PTA. Under relatively mild reaction conditions (160 °C, 10 bar H2) and in water as a green solvent, a yield of 85% PTA
was obtained after 20 h. Further investigation on the role of the acid catalyst elucidated that Lewis acidity also
facilitates the dehydration of citric acid in aqueous acidic media. From these findings we concluded that next to
hydrothermal stability and an adequate amount and accessibility of acid sites, the combined interaction of Brønsted
and Lewis acid sites in the zeolite channels accounts for the superiority of low Si/Al H-Beta zeolites for the
dehydration of citric acid. Finally, recycling experiments were performed, where the gradual decay in citric acid
conversion upon following reuses was attributed to loss of acid sites due to dealumination. We have shown that
upon realumination a considerable amount of Al was reincorporated into the zeolite framework. This resulted in an
increase of the conversion of citric acid while a high selectivity to PTA was preserved, illustrating that this
realumination technique is a suitable approach to recycle the catalytic system.
Author information
Corresponding Author
E-mail: dirk.devos@kuleuven.be.
ORCID
Wouter Stuyck: 0000-0002-2887-0580.
11

Green Chemistry
Page 12 of 13
View Article Online
DOI: 10.1039/D0GC02298E
Funding
W.S. and J.V. are grateful to the FWO for their PhD fellowships. G. M. and D.E.D.V. acknowledge FWO for
research project funding (G0781118N, G0D0518N); D.E.D.V. thanks KULeuven as well for long-term
structural funding through Metusalem (METU14/04). A. K. and G. M. acknowledge the financial support
from the Slovenian Research Agency (research core funding No. P1-0021 and project No. N1-0079).
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
The authors also thank Jannick Vercammen for catalyst synthesis and Karel Duerinckx and prof. Dimitrios
Sakellariou for assistance with 1H- and 27Al NMR measurements respectively.
References
1
2
R. Geyer, R. J. Jambeck and K. Lavender Law, Science advances, 2017, 3, e170082.
Ellen MacArthur Foundation, The New Plastics Economy: Rethinking the future of plastics, 2016, Retrieved
from: https://www.ellenmacarthurfoundation.org/. Accessed on April 2, 2020.
Markets and Markets, Plastics Additives Market by Type, Plastics and by Application – Global Trends &
Forecasts to 2021, 2016, Retrieved from: https://www.marketsandmarkets.com/. Accessed on April 2, 2020.
M. Bocqué, C. Voirin, V. Lapinte, S. Caillol and J.-J. Robin, Journal of Polymer Science, Part A: Polymer Chemistry,
2016, 54, 11-33.
P. Walters, D. F. Cadogan and C. J. Howick, Plasticizers in Ullmann’s Encyclopedia of Industrial Chemistry, Wiley-
VCH Verlag, Germany, 2020, pp. 1-27.
European plasticisers, Plasticisers, March 2020, Retrieved from: https://www.plasticisers.org/. Accessed on
April 2, 2020.
3
4
5
6
7
K. J. Groh, T. Backhaus, B. Carney-Almroth, B. Geueke, P. A. Inostroza, A. Lennquist, H. A. Leslie, M. Maffine, D.
Slunge, L. Trasande, A. M. Warhurst and J. Muncke, Science of the Total Environment, 2019, 651, 3253-3268.
C. Bueno-Ferrer, M.C. Garrigós and A. Jiménez, Polym. Degrad. Stab., 2010, 95, 2207-2212.
B. Yin and M. Hakkarainen, J. Appl. Polym. Sci., 2011, 119, 2400-2407.
8
9
10 A. Greco, D. Brunetti, G. Renna, G. Mele and A. Maffezzoli, Polymer Degradation and Stability, 2010, 95, 2169-
2174.
11 Z. Yu, J. Zhou, J. Zhang, K. Huang, F. Cao and P. Wei, J. Appl. Polym. Sci., 2014, 40938.
12 M. Matos, R. A. Cordeiro, H. Faneca, J. F. J. Coelho, A. J. D. Silvestre and A. E. Sousa, Materials, 2019, 12, 2336.
13 I. E. Sweis and B. C. Cressey, Toxicology Reports, 2018, 5, 808-812.
14 E. Snejdrova and M. Dittrich, Pharmaceutically Used Plasticizers in Recent Advances in Plasticizers, IntechOpen,
2012, pp. 45-68.
15 F. Welle, G. Wolz and R. Franz, Pharma International, 2005, 3, 17-21.
16 SCENIHR, Preliminary report on the safety of medical devices containing DEHP-plasticized PVC or other
plasticizers on neonates and other groups possibly at risk, European Commission 2007.
17 Acetylation: a) S. Ting, CN101353305B, 2007. b) B.L. Song, G. Zhang, Y.Z. Wang and W.J. Yanli, CN101402571A,
2008. c) G. Kia, Z. Yang, Z. Pengwei, H. Wei, L. Xiaolin, CN102351696B, 2011. d) L. Jianzhong, Z. Yuepeng, L.
Hiulai, CN102633640B, 2012. e) C. Xingjian, CN106928065A, 2017. Butyrylation: a) J.F. Day, US20060094894A1,
2004.
18 F. C. Magne, R. R. Mod, Ind. Eng. Chem., 1953, 45, 1546-1547.
19 R. J. Reid, C. Fulton, W. M. Smith Jr. C. Falls and B. H. Werner, US2802802, 1957.
20 G. E. Hilbert, Premium-quality plasticizers made from aconitic acid, a byproduct of sugarcane in Annual Reports
of Department of Agriculture, United States, 1952, 62-64.
21 Wypych G., Handbook of Plasticizers 3rd edition, ChemTec Publishing, Toronto, 2017.
22 J. Verduyckt and D. E. De Vos, Chem. Sci., 2017, 8, 2616-2620.
23 B. M. Stadler, C. Wulf, T. Werner, S. Tin and J. G. de Vries, ACS Catal., 2019, 9, 8012-8067.
24 G. Flèche and M. Huchette, Starch, 1986, 38, 26-30.
12

Page 13 of 13
Green Chemistry
View Article Online
DOI: 10.1039/D0GC02298E
25 V. E. Tarabanko, M. Yu. Chernyak, S. V. Aralova and B. N. Kuznetsov, React. Kinet. Catal. Lett., 2002, 75, 117-
126.
26 B. Danon, G. Marcotullio and W. de Jong, Green Chem., 2014, 16, 39-54.
27 K. Bock, C. Pedersen and H. Thøgersen, Acta Chemica Scandinavica B, 1981, 35, 441-449.
28 R. M. Ravenelle, F. Schüβler, A. D’Amico, N. Danilina, J. A. van Bokhoven, J. A. Lercher, C. W. Jones and C. Sievers,
J. Phys. Chem. C, 2010, 114, 19582-19595.
29 Okuhara T., Chem. Rev., 2002, 102, 3641-3666.
30 A. J. Sanborn, US7420067 B2, 2008.
31 H. Kobayashi, H. Yokoyama, B. Feng and A. Fukuoka, Green Chem., 2015, 17, 2732-2735.
32 R. Otomo, T. Yokoi and T. Tatsumi, 2015, Applied Catalysis A: General, 2015, 505, 28-35.
33 H. Yokoyama, H. Kobayashi, J. Hasegawa and A. Fukuoka, ACS Catal., 2017, 7, 4828-4834.
34 T. Mizugaki, K. Togo, Z. Maeno, T. Mitsudome, K. Jitsukawa and K. Kaneda, ACS Sustainable Chem. Eng., 2016,
4, 682-685.
35 R. Zhao, L. Xu, S. Huang and W. Zhang, Catal. Sci. Technol., 2019, 9, 5676-5685.
36 M. Carlsson, C. Habenicht, L. C. Kam, M. J. Antal Jr., N. Bian, R. J. Cunningham and M. Jones Jr., Ind. Eng. Chem.
Res., 1994, 33, 1989-1996.
37 R. T. Arnold, O. C. Elmer and R. M. Dodson, J. Am. Chem. Soc., 1950, 72, 4359-4361.
38 J. Le Nôtre, S. C. M. Witte-van Dijk, J. van Haveren E. L. Scott and J. P. M. Sanders, ChemSusChem, 2014, 7, 2712-
2720.
39 J. Rigoreau, S. Laforge, N.S. Gnep, M. Guisnet, Journal of Catalysis, 2005, 236, 45-54.
40 N. Y. Chen, J. Phys. Chem., 1976, 80, 60-64.
41 A. Corma, Current Opinion in Solid State & Material Science, 1997, 2, 63-75.
42 T. Ennaert, J. Geboers, E. Gobechiya, C. M. Courtin, M. Kurttepeli, K. Houthoofd, C. E. A. Kirschhock, P. C. M. M.
Magusin, S. Bals, P. A. Jacobs and B. F. Sels, ACS Catal., 2015, 5, 754-768.
43 L. Xiao, J. Mao, J. Zhou, X. Guo and S. Zhang, Applied Catalysis A: General, 2011, 393, 88-95.
44 D. Barthomeuf, Stud. Surf. Sci. Catal., 1997, 105, 1677-1706.
45 I. Kiricsi, C. Flego, G. Pazzuconi, W. O. Parker Jr., R. Millini, C. Perego and G. Bellussi, J. Phys. Chem., 1994, 98,
4627-4634.
46 G. H. Kuehl and H. K. C. Timken, Microporous and Mesoporous Materials, 2000, 35-36, 521-532.
47 O. Bortnovsky, Z. Sobalík, B. Wichterlová and Z. Bastl, Journal of Catalysis, 2002, 210, 171-182.
48 E. Bourgeat-Lami, P. Massiani, F. Di Renzo, P. Espiau and F. Fajula, Applied Catalysis, 1991, 72, 139-152.
49 J. Dedecek, E. Tabor and S. Sklenak, ChemSusChem, 2019, 12, 556-576.
50 J. Brus, L. Kobera, W. Schouferger, M. Urbanová, P. Klein, P. Sazama, E. Tabor, S. Sklenak, A. V. Fishchuk and J.
Dedecek, Angewandte Chemie, 2015, 54, 541-545.
51 O. Bortnovsky, Z. Sobalík and B. Wichterlová, Microporous and Mesoporous Materials, 2001, 46, 265-275.
52 J. Pérez-Pariente, J. Sanz, V. Fornés and A. Corma, Journal of Catalysis, 1990, 124, 217-223.
53 J. A. van Bokhoven, D. C. Koningsberger, P. Kunkeler, H. van Bekkum and A. P. M. Kentgens, J. Am. Chem. Soc.,
2000, 122, 12842-12847.
54 S. M. Maier, A. Jentys and J. A. Lercher, J. Phys. Chem. C, 2011, 115, 8005-8013.
55 A. Vjunov, J. L. Fulton, T. Huthwelker, S. Pin, D. Mei, G. K. Schenter, N. Govind, D. M. Camaioni, J. Z. Hu and J.
A. Lercher, J. Am. Chem. Soc., 2014, 136, 8296-8306.
56 Z. Zhao, S. Xu, M. Y. Hu, X. Bao, C. H. F. Peden and J. Hu, J. Phys. Chem. C, 2015, 119, 1410-1417.
57 A. Omegna, M. Vasic, J. A. van Bokhoven, G. Pirngruber and R. Prins, J. Phys. Chem. Chem. Phys., 2004, 6, 447-
452.
58 D. M. Roberge, H. Hausmann and W. F. Hölderich, J. Phys. Chem. Chem. Phys., 2002, 4, 3128-3135.
13