Stabilising Ni catalysts for the dehydration-decarboxylation-hydrogenation of citric acid to methylsuccinic acid

Article abstract of DOI:10.1039/c7gc01773a

A new reaction sequence of dehydration-decarboxylation-hydrogenation to transform citric acid into methylsuccinic acid has recently been developed using Pd as a noble metal catalyst in water. In this work Ni catalysts were investigated as low cost, non-noble metal alternatives. Several home-made and commercial catalysts were screened for this reaction. Citric acid was very reactive and full conversions were readily obtained in all cases. However, the selectivity to methylsuccinic acid was initially low, since typical Ni catalysts were not stable and therefore not able to hydrogenate the formed CC double bonds. Due to the lower hydrogenation activity of Ni compared to Pd, new side products appeared. Particularly, hydration of the CC double bonds made the reaction network more complex in this case. Fortunately, the formation of all hydration products-even the rather stable lactone, β-carboxy-γ-butyrolactone-was eventually shown to be completely reversible. Three routes were then studied to stabilise Ni catalysts and to enable the Ni catalyzed conversion of citric acid to methylsuccinic acid; partial neutralisation of the acid reactant, adding Fe to Ni/ZrO₂ or to the reaction mixture and coating Ni particles with carbon, all proved to stabilise Ni and all resulted in high to very high methylsuccinic acid yields. Furthermore, the role of Fe was unravelled by performing reference reactions with different Fe compounds and by in depth characterisation of the NiFe/ZrO₂ catalyst. Finally, the reaction conditions were optimised using the carbon-coated Ni nanoparticles and kinetic profiles were recorded to confirm the extended reaction network.

Full text of DOI:10.1039/c7gc01773a

View Article Online
View Journal
Green
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: J. Verduyckt, A.
Geers, B. Claes, S. Eyley, C. Van Goethem, I. Stassen, S. Smolders, R. Ameloot, I. F.J. Vankelecom, W.
Thielemans and D. E. De Vos, Green Chem., 2017, DOI: 10.1039/C7GC01773A.
Volume 18 Number
7
7
April 2016 Pages 1821–2242
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Cutting-edge research for a greener sustainable future
www.rsc.org/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
author guidelines.
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, outlined
in our author and reviewer resource centre, 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
CRITICAL REVIEW
G. Chatel et al.
Heterogeneous catalytic oxidation for lignin valorization into valuable
chemicals: what results? What limitations? What trends?
rsc.li/green-chem

Page 1 of 9
Pleas eG dr eo e nn o Ct ha de jmu si st tmr yargins
View Article Online
DOI: 10.1039/C7GC01773A
Journal Name
ARTICLE
Stabilising Ni catalysts for the dehydration-decarboxylation-
hydrogenation of citric acid to methylsuccinic acid
a
a
a
b
a
a
Jasper Verduyckt, Anton Geers, Birgit Claes, Samuel Eyley, Cédric Van Goethem, Ivo Stassen,
a
a
a
b
a
Simon Smolders, Rob Ameloot, Ivo Vankelecom, Wim Thielemans, and Dirk E. De Vos *
A new reaction sequence of dehydration-decarboxylation-hydrogenation to transform citric acid into methylsuccinic acid,
was recently developed using Pd as noble metal catalyst in water. In this work Ni catalysts were investigated as cheap,
non-noble metal alternative. Several home-made and commercial catalysts were screened for this reaction. Citric acid was
very reactive and full conversions were readily obtained in all cases. However, the selectivity to methylsuccinic acid was
initially low, since typical Ni catalysts were not stable and therefore not able to hydrogenate the formed C=C double
bonds. Due to the lower hydrogenation activity of Ni compared to Pd, new side products appeared. Especially hydration of
the C=C double bonds made the reaction network more complex in this case. Fortunately, the formation of all hydration
products – even the rather stable lactone, β-carboxy-γ-butyrolactone – was eventually shown to be completely reversible.
Three routes were then studied to stabilise Ni catalysts and to enable the Ni catalyzed conversion of citric acid to
Received 00th January 20xx,
Accepted 00th January 20xx
DOI: 10.1039/x0xx00000x
www.rsc.org/
methylsuccinic acid; partial neutralisation of the acid reactant, adding Fe to Ni/ZrO or to the reaction mixture and coating
2
Ni particles with carbon, all proved to stabilise Ni and all resulted in high to very high methylsuccinic acid yields.
Furthermore, the role of Fe was unravelled by performing reference reactions with different Fe compounds and by in
depth characterisation of the NiFe/ZrO
2
catalyst. Finally, the reaction conditions were optimised using the carbon-coated
Ni nanoparticles and kinetic profiles were recorded to confirm the extended reaction network.
sugarcane, is increasingly being applied for the sustainable
1
4
Introduction
production of this biobased compound. Citric acid has
already been used as platform chemical for the synthesis of
methacrylic acid, 2- and 3-methylbutyrolactone, 3-methyl-
tetrahydrofuran and 3-(2’-hydroxyethyl)tetrahydrofuran.1
In addition, we recently developed the one-step conversion of
Methylsuccinic acid is a versatile dicarboxylic acid, which can
be used as a building block for the manufacture of solvents for
5–18
1
2
3–6
cosmetics, binders, and biodegradable polyesters.
Incorporating methylsuccinic acid in e.g. polybutylene
succinate enhances the biodegradability and the toughness of
the polyester and simultaneously reduces the viscosity of this
19
citric acid to methylsuccinic acid. Citric acid was first
dehydrated towards aconitic acid, subsequently
decarboxylated to itaconic acid and finally hydrogenated to
methylsuccinic acid. By using mild hydrogenation conditions
and water as the solvent, carboxylic acid hydrogenation could
be avoided. Moreover, the rates of decarboxylation and
hydrogenation were carefully balanced to prevent further
decarboxylation to e.g. methacrylic acid, and premature
hydrogenation to propane-1,2,3-tricarboxylic acid. Besides,
radical fragmentation of citric acid was proven to be minimal
under more acidic conditions. Altogether these insights lead to
high methylsuccinic acid yields; a yield of 84% was e.g.
achieved by reacting citric acid in water for 40 min, using
4
polyester’s solutions. Nowadays, methylsuccinic acid can only
be produced via hydrogenation of itaconic acid.7 Recently,
Palkovits and co-workers reported an electrocatalytic system
that enables the direct reduction of itaconic acid in a
fermentation broth at room temperature, yielding up to 98%
–10
11
methylsuccinic acid. However, in 2011 only 41.4 thousand
tons of itaconic acid were produced via fermentation.12
The annual production volume of citric acid, on the other
13
hand, was
2
million tons in 2016.
The solid-state
fermentation of agricultural waste streams from e.g. corn or
0
.5 mol% Pd/C as hydrogenation catalyst with 8 bar H
2
at
a.Centre for Surface Chemistry and Catalysis, Department of Microbial and
Molecular Systems, KU Leuven – University of Leuven, Leuven Chem&Tech,
Celestijnenlaan 200F, Post Box 2461, 3001 Heverlee, Belgium. E-mail:
dirk.devos@kuleuven.be
Renewable Materials and Nanotechnology Research Group, Department of
Chemical Engineering, KU Leuven – University of Leuven, Campus Kulak Kortrijk,
225°C.
The first step in the new reaction sequence of dehydration-
decarboxylation-hydrogenation is acid-catalyzed, while the
decarboxylation occurs spontaneously at temperatures higher
than 150°C – although this step might also be aided by Pd. So
intrinsically, the only metal-catalyzed step in this sequence is
the hydrogenation of itaconic acid (or its isomers), an α,β-
b.
Etienne Sabbelaan 53, 8500 Kortrijk, Belgium.
Electronic Supplementary Information (ESI) available: experimental details, N
physisorption data, XRD data of Ni/ZrO and Ni Fe /ZrO , XPS data, recycling
2
2
5
5
2
experiments and product identification. See DOI: 10.1039/x0xx00000x
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 1
Please do not adjust margins

Pleas eG dr eo e nn o Ct ha de jmu si st tmr yargins
Page 2 of 9
ARTICLE
Journal Name
4
unsaturated carboxylic acid. Two metals are generally known applied several times for the dry and steam reforVmiewinAgrticolef OCnHline.
DOI: 10.1039/C7GC01773A
to be suitable to catalyze the C=C hydrogenation of these α,β- Besides, Ni particles encapsulated by carbon species have been
unsaturated carboxylic acids, namely Pd and Ni.20 This reported for electrocatalytic applications in sensors, batteries
prompted us to investigate Ni catalysts for the conversion of and oxygen evolution reactions (e.g. in fuel cells).33–35 To the
citric acid to methylsuccinic acid.
Ni is a cheap, readily available non-noble metal, which far not been reported as suitable chemocatalysts.
makes it an interesting compound for the production of In this research we aim at identifying Ni-based catalysts
best of our knowledge, carbon-coated Ni nanoparticles have so
industrial catalysts. However, the acidic hydrothermal and reaction conditions that enable the conversion of citric
conditions and the presence of citric acid as a chelating acid to methylsuccinic acid, while minimising the leaching of
compound render the use of Ni as hydrogenation catalyst for Ni.
this reaction untrivial. Nevertheless, research by Lercher and
21
co-workers encouraged us to start this investigation. They
showed that Ni nanoparticles are stable during the
0
Results and discussion
hydrothermal hydrodeoxygenation of phenol; Ni/HZSM-5 is
Screening of Ni-based catalysts and extending the reaction
network
stable in dilute H
3 4
PO and even in 15 wt% acetic acid if the
21
catalyst is pre-reduced. In the latter conditions only around
4
% Ni was leached into solution. Moreover, a passivated Initially, nine Ni-based catalysts were screened for their
Ni/SiO
catalyst was almost completely reduced within 10 min hydrogenation activity to transform citric acid into
under the aqueous reaction conditions (200°C, 35 bar H
indicating that reduction is faster than dissolution of Ni
2
), methylsuccinic acid (Table 1, entries 1-9). Among these
II catalysts were four home-made materials – synthesised via
2
21
3 2
incipient wetness impregnation with Ni(NO ) , see ESI for
species. Ni/HZSM-5 was also used for the hydrogenolysis of
tetrahydrofurfuryl alcohol to 1,5-pentanediol and tetrahydro- experimental details – and five commercial catalysts (entries
1-4 and 5-9, respectively). All these catalysts were pre-reduced
this case, the Ni catalyst was quite stable as well: only 5% Ni and passivated to ensure safe handling in air. Keeping in mind
was leached into solution.
the balancing of the hydrogenation and decarboxylation rate,
2
2
2
pyran under hydrothermal conditions (250°C, 34 bar H ). In
Furthermore, extensive research has been conducted to the reaction conditions were changed to compensate for the
stabilise Ni catalysts for demanding reactions, such as lower hydrogenation activity of Ni compared to Pd in
methanation of CO and CO
carbons and ethanol.
2
or dry/steam reforming of hydro- reference [19]. Citric acid was still reacted in water (2 mL,
0.1 M) for 6 h; the temperature was however decreased to
catalysts can improve the Ni dispersion, avoid sintering and 175°C to decrease the decarboxylation rate and the H
pressure was increased to 20 bar to increase the
, can also prevent hydrogenation rate. Also the catalyst-to-substrate ratio was
23–32
For instance, adding Fe to Ni
2
2
3–28
facilitate the reduction of Ni.
metal oxide layer, such as SiO
Coating Ni particles with a
or ZrO
2
, Al
2
9–32
O
3
2
sintering and coke formation.2
increased to 10 mol% for the same purpose.
This method has been
Table 1. Dehydration-decarboxylation-hydrogenation of citric acid using Ni-based catalysts.a
Conversion
Selectivity [%]
Leaching
Ni (Fe) [%]
Catalyst
Ni/MgAl
[%]
MSA b
2nd deca c
Fragmentation d
Itaconic isomers e
Isomer hydration f
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
2
O
4
>99
>99
>99
>99
>99
>99
>99
>99
97
>99
>99
>99
>99
>99
85
3
3
2
2
9
13
13
6
6
7
4
3
7
5
8
11
7
6
15
15
14
5
4
5
22
36
50
27
30
39
28
6
6
48
9
1
1
19
40
38
23
34
37
35
21
21
30
26
9
23
Ni/TiO
Ni/ZrO
Ni/Al
2
>99
>99
>99
>99
>99
2
2
O
3
13
16
11
14
12
12
10
10
6
6
4
2
3
Ni/SiO
Ni/SiO
2
-Al
2
2
O
3
4
g
h
Katalco JM 11-4R
Katalco JM 23-8
HTC Ni 500 RP j
12
48
44
5
36
69
66
81
<1
<1
368 mg/L
116 mg/L
238 mg/L
58 (43)
28 (80)
5 (93)
16 (97)
20
i
h
h
0
1
2
3
4
5
6
5 1
Ni Fe /ZrO
2
5 3
Ni Fe /ZrO
2
Ni
5
Fe
Fe
5
/ZrO
2
2
Ni
5
5
/ZrO
(Ac)
12
10
32
33
Ni-C NPs
-
<1
62
59
-
-
NiSO
4
90
a
, 6 h. b Methylsuccinic acid. c ‘2nd deca’
Reaction conditions: citric acid (0.2 mmol), catalyst (10 mol% Ni), water (2 mL), 175°C, 2 bar N
2
and 20 bar H
2
d
represents methacrylic acid, 2-hydroxyisobutyric acid, isobutyric acid and butyric acid. ‘Fragmentation’ represents acetone, acetic acid, pyruvic acid and
lactic acid. ‘Itaconic isomers’ represents itaconic acid, mesaconic acid and citraconic acid. ‘Isomer hydration’ represents 2-(hydroxymethyl)succinic
acid, β-carboxy-γ-butyrolactone and 2-hydroxy-2-methylsuccinic acid. 3 mg. The concentration of Ni in the reaction solution is given, since the precise
catalyst composition is unknown. 23 mg. 8 mg.
e
f
g
h
i
j
2
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 3 of 9
Pleas eG dr eo e nn o Ct ha de jmu si st tmr yargins
Journal Name
ARTICLE
In all cases the conversion of citric acid was very high; on appeared to be limited to 23%; this was probably due to the
View Article Online
II
DOI: 10.1039/C7GC01773A
the other hand, the selectivity towards methylsuccinic acid incorporation of oxidised Ni species in the layers of the
II
was initially rather low (entries 1-9). The different products layered double hydroxide. To make sure that Ni is not active
detected by H-NMR can be explained by the reaction network for the hydrogenation, a reference reaction was performed
1
in Scheme 1. There are some clear differences compared to using NiSO
4
as catalyst (entry 16). The result coincided with the
19
0
the reaction using Pd as catalyst. Due to the lower blank reaction (entry 15); this clearly indicates that Ni is
hydrogenation activity of Ni, even under the adapted reaction needed for the C=C hydrogenation. In these reactions citric
conditions, early hydrogenation of aconitic acid to propane- acid was converted for 85-90% and the majority of the
1
,2,3-tricarboxylic acid was never observed. On the other products were itaconic acid and its isomers, which were partly
hand, this also lead to the appearance of a new type of side hydrated to mainly β-carboxy-γ-butyrolactone. A minor
reaction, namely hydration of the C=C double bonds. This amount of the itaconic isomers was further decarboxylated to
hydration resulted in the formation of 2-hydroxy-2- methacrylic acid and around 5% was fragmented to form
methylsuccinic acid, 2-hydroxyisobutyric acid and 2- mainly acetone and acetic acid. The product distributions of
(hydroxymethyl)succinic acid; the latter can cyclise to the Ni catalyzed reactions generally correspond to the one in
β-carboxy-γ-butyrolactone. Besides, some further the blank reaction (entries 1-6 and 15). Nevertheless, besides
decarboxylation of itaconic acid and its isomers (mesaconic the minor hydrogenation to methylsuccinic acid, the
and citraconic acid) to methacrylic and crotonic acid, still conversions of citric acid are now complete, which is most
occurred. These unsaturated monocarboxylic acids might be probably due to the faster dehydration facilitated by the
hydrogenated to form isobutyric and butyric acid. Finally, also supports. Moreover, more ‘further decarboxylation’ products
acetone, acetic and pyruvic acid were produced via a radical were observed. This observation was most pronounced for
1
9
pathway.
2 4 2 3
However, Ni has a stronger tendency to Ni/MgAl O and Ni/Al O , where also more than 30% of the
20
hydrogenate pyruvic acid to lactic acid, which was observed products was lost from solution, next to the formation of 13%
more often than in the case of Pd – although still in minor of secondary decarboxylation products. Mass loss from
quantities (≤6%). The ESI gives a detailed overview of the solution primarily occurred through extensive decarboxylation
identification of the observed compounds.
In the initial screening the yield of methylsuccinic acid was the decarboxylation might be mediated by Mg and Al
always lower than 10% for the regular Ni/MeO
19
and formation of propene (Scheme 1). The acceleration of
2+
3+
36–38
x
catalysts leached into solution.
Finally, three catalysts generously
(
Table 1, entries 1-6). It appears that Ni was rapidly and supplied by Johnson Matthey, especially Katalco JM 23-8 and
completely leached into solution under the demanding HTC Ni 500 RP, showed to maintain some hydrogenation
reaction conditions. Only for Ni/MgAl the leaching activity, resulting in the formation of 12-48% methylsuccinic
2 4
O
Scheme 1. Extended reaction network for the dehydration-decarboxylation-hydrogenation of citric acid, built on the reaction network in reference [19]. The ESI gives a detailed
overview of the identification of the observed (framed) compounds. The colour codes of the different product groups will be further used in the following figures.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Pleas eG dr eo e nn o Ct ha de jmu si st tmr yargins
Page 4 of 9
ARTICLE
Journal Name
acid (entries 7-9). However, a significant amount of Ni was still acceptors. A similar result could be obtained with a NiFe
View Article Online
DOI: 10.1039/C7GC01773A
catalyst based on acetate precursors (Table 1, entry 13).
leached into solution.
Subsequently, to gather more information about the
To elucidate the peculiar role of Fe, several reference
fragmentation and hydration side reactions under these reactions were performed, including blank reactions with the
conditions, several blank reactions were performed starting support and different Fe compounds, as well as addition of
from intermediates in the reaction network (Table 2). these compounds to the regular Ni/ZrO
2
catalyst (Table 3). The
Fragmentation products were observed starting from aconitic reaction with ZrO alone confirms that this support accelerates
2
acid, itaconic isomers and methacrylic acid. This means that in (de)hydration reactions, since the citric acid conversion was
the absence of adequate hydrogenation, radical fragmentation complete and the product selectivity shifted towards the
not only proceeds directly from citric acid,19 but also from ‘isomer hydration’ products (entry 2). The different Fe
unsaturated compounds further along the reaction network. compounds seem to have some hydrogenation ability,
Furthermore, itaconic and mesaconic acid were shown to be however, also further decarboxylation and especially
converted mainly via isomerisation and hydration; 2-hydroxy- fragmentation were accelerated (entries 3-6). The latter is
0
2
-methylsuccinic acid, 2-(hydroxymethyl)succinic acid and most pronounced for fine Fe powder, which quickly dissolved
-
β-carboxy-γ-butyrolactone indeed originated from these in the citric acid solution, and for Fe(NO
itaconic isomers (entries 2 and 3). Moreover, itaconic acid is act as an additional electron acceptor. When Fe(NO
more reactive than mesaconic acid, which is probably due to Fe(CH COO) were added to a reaction mixture with Ni/ZrO
3
)
3
, where NO
3
might
3
)
3
and
3
2
2
,
the stronger delocalisation of the C=C double bond in we noticed a stabilising effect: only 3% and 11% Ni were
mesaconic acid. Finally, it is clear that 2-hydroxyisobutyric acid leached into solution, respectively (entries 7 and 8). This
can also be produced via hydration of methacrylic acid (entry means that ionic Fe species might catalyze the reduction of Ni
4
), next to via hydration of the itaconic isomers followed by with H
2
under the reaction conditions, stabilising the Ni
particles. For Fe(CH COO) this resulted in decent
methylsuccinic acid yield of 45%, whereas in the case of
Fe(NO only 13% yield was obtained. This can be explained
19
decarboxylation (Scheme 1).
3
2
a
2
Stabilising Ni/ZrO with Fe
3 3
)
Encouraged by the somewhat better stability of Katalco JM 23- by the increased fragmentation rate in this case. On the other
0
8
and HTC Ni 500 RP and inspired by previous research on hand, Fe might act as a sacrificial reductant, in this way
23–28
stabilising Ni catalysts,
Fe was added to the Ni/ZrO
2
stabilising the Ni against oxidation (only 7% leaching) and
catalyst. ZrO
co-impregnated with Ni(NO
2
was selected as stable support and Fe(NO
3
)
3
was enabling a methylsuccinic acid yield of 55% (entry 9). Sun et al.
0
3
)
2
in three different ratios: 1, 3 showed that the dissolution of Fe indeed induces a more
39
0
and 5 wt% Fe vs. 5 wt% Ni. The reduction and passivation reducing environment. In the case of Fe(CH
conditions were the same as for the regular Ni/ZrO catalyst the effect on the stability of the Ni catalyst through a (subtle)
see ESI for further details). The three new catalysts were increase in pH cannot be excluded (cfr. infra), although this
3 2
COO) and Fe ,
2
3
9
(
0
screened using the conditions in Table 1 (entries 10-12). There amount of Fe did not result in a measurable pH increase.
was a remarkable positive correlation between the amount of When Fe was dissolved in the citric acid solution, the pH
0
Fe added and the yield of methylsuccinic acid. For Ni
5
Fe
5
/ZrO
2
remained at 2.2.
a yield of 69% was already achieved. This positive trend
2
To clarify the role of Fe in the NiFe/ZrO catalysts, the
2 5 5 2
corresponded to a clear increase in the stability of Ni; the Ni Ni/ZrO and Ni Fe /ZrO catalysts were characterised by
leaching steadily decreased from >99%, when no Fe was scanning transmission electron microscopy (STEM) in
added, to only 5%, when Fe was stoichiometrically added combination with energy dispersive X-ray spectroscopy (EDX)
(
Table 1, entries 3 and 10-12). However, now a large part of and by X-ray photoelectron spectroscopy (XPS). Additionally,
the Fe leached into solution. In addition, this resulted in a clear these catalysts were also analysed by N physisorption and
increase of the fragmentation side products to around 15% powder X-ray diffractometry (XRD). The N adsorption-
due to the presence of oxidised Fe species as electron desorption isotherms showed a behaviour that is typical for a
2
2
mesoporous structure (Table S1
,
Figure S1 and S2), and the
Table 2. Reactivity of intermediates in the dehydration-decarboxylation-hydrogenation of citric acid under blank reaction conditions.a
Conversion
[%]
Selectivity [%]
Substrate
MSA b
2nd deca c
Fragmentation d
Itaconic isomers e
Isomer hydration f
1
2
3
4
Aconitic acid
Itaconic acid
Mesaconic acid
Methacrylic acid
>99
76
36
<1
<1
<1
-
4
4
7
1
1
3
3
52
44
44
38
-
g
19
52
h
i
45
59
-
a
, 6 h. b Methylsuccinic acid. c ‘2nd deca’
Reaction conditions: substrate (0.2 mmol), water (2 mL), 175°C, 2 bar N
2
and 20 bar H
2
d
represents methacrylic acid and 2-hydroxyisobutyric acid. ‘Fragmentation’ represents acetone, acetic acid and pyruvic acid.
e
f
‘
Itaconic isomers’ represents itaconic acid, mesaconic acid and citraconic acid. ‘Isomer hydration’ represents
g
2
h
-(hydroxymethyl)succinic acid, β-carboxy-γ-butyrolactone and 2-hydroxy-2-methylsuccinic acid. Mesaconic acid and citraconic acid.
i
Itaconic acid and citraconic acid. 2-Hydroxyisobutyric acid.
4
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 5 of 9
Pleas eG dr eo e nn o Ct ha de jmu si st tmr yargins
Journal Name
ARTICLE
Table 3. Reference reactions to elucidate the role of Fe.a
View Article Online
DOI: 10.1039/C7GC01773A
Leaching
Conversion
[%]
Selectivity [%]
Catalyst
Additive
MSA b 2nd deca c Fragmentation d
Itaconic isomers e
Isomer hydration f
Ni (Fe) [%]
1
2
3
4
5
6
7
8
9
-
-
-
-
85
<1
<1
2
9
4
2
4
7
12
8
9
5
7
7
4
3
62
53
39
26
25
34
6
32
41
33
15
19
31
5
-
-
g
ZrO
Fe/ZrO
-
-
-
Ni/ZrO
Ni/ZrO
Ni/ZrO
2
>99
>99
>99
>99
>99
>99
>99
>99
2
12
38
32
21
42
17
27
- (93)
- (>99)
-
-
3
0
Fe
Fe(NO
Fe(CH
Fe(NO
Fe(CH
Fe
3
)
3
3
COO)
2
5
2
)
3 3
13
45
55
2
2
3
COO)
2
3
<1
22
12
11
7 (>99)
0
a
and 20 bar H
2
, 6 h. b Methylsuccinic
Reaction conditions: citric acid (0.2 mmol), catalyst (10 mol% Me), additive (10 mol% Fe), water (2 mL), 175°C, 2 bar N
2
acid. c ‘2 deca’ represents methacrylic acid, 2-hydroxyisobutyric acid, isobutyric acid and butyric acid. ‘Fragmentation’ represents acetone, acetic acid,
nd
d
e
f
pyruvic acid and lactic acid. ‘Itaconic isomers’ represents itaconic acid, mesaconic acid and citraconic acid. ‘Isomer hydration’ represents
-(hydroxymethyl)succinic acid, β-carboxy-γ-butyrolactone and 2-hydroxy-2-methylsuccinic acid. g 24 mg ZrO
2
2
.
metal loadings (5 wt%) were too low to distinguish
characteristic metal peaks in the X-ray diffractogram (Figure
S3). In literature, the beneficial effect of Fe addition to Ni has
been ascribed to an improved dispersion and/or a facilitated
23–28
reduction.
In this work, the effect of Fe on the dispersion
of Ni could not be confirmed by STEM, since few metal
particles were observed due to the low contrast between Ni
and zirconia (Figure 1-(a) and -(e)). Simultaneous EDX analysis
did confirm that Ni and Fe are intimately mixed in the
5 5 2
Ni Fe /ZrO catalyst (Figure 1-(h) and -(i)). However, an effect
of Fe on the dispersion seems unlikely to be decisive for the
stabilisation of Ni, since Fe almost completely leaches into
solution. Finally, XPS analysis showed that the surface of the Ni
particles is indeed oxidised by the passivation procedure;
before reaction Ni is mainly present as NiO and Ni(OH)
, entries 1 and 2; Table S2 and S3 Figure S4 and S5). The low
amount of metallic Ni observed, might be due to the limited
2
(Table
4
;
40,41
information depth of the XPS analysis.
These results
indicate that the outer oxidised Ni layer should indeed be
reduced during the initial stages of the reaction,21 which can
be faster in the presence of Fe. The latter occurs in an oxidised
state, mainly as Fe
2 3
O , but the presence of other oxides cannot
2
Figure 1. (a) & (e) STEM images and (b-d) & (f-i) EDX elemental maps of Ni/ZrO (a-d)
be excluded due to interference of Ni Auger peaks with the Fe and Ni Fe /ZrO (e-i).
5
5
2
4
2
2
p signal (Table S3
, Figure S6). This means that Fe might
stabilise Ni in the NiFe/ZrO
2
catalysts by promoting the Table 4. Summary of XPS results.
reduction of Ni as catalyst in an ionic form using H
reductant.
2
as
Fraction of Ni species [%]
Catalyst
Ni/ZrO
Ni Fe /ZrO
Ni-C NPs
Ni0
NiO
Ni(OH)
2
2
Stabilising Ni/ZrO by partial neutralisation of the reactant
1
2
3
2
0.4
3.7
>99
46.4
48.8
<1
53.2
47.6
<1
In our previous research using Pd as catalyst for the direct
production of methylsuccinic acid from citric acid, we
demonstrated that the addition of base was detrimental for
the methylsuccinic acid yield due to a strong increase of
5
5
2
1
9
fragmentation reactions. Nevertheless, the influence of small respectively. These low amounts of base did not have a
amounts of NaOH and KOH on the hydrogenation activity of significant effect on the fragmentation side reactions.
Ni/ZrO
bases were added to the reaction mixture using the screening concomitant increase of the methylsuccinic acid yield were
conditions in Table 1 Figure 2). Separate pH measurements remarkable. For NaOH the yield increased from 2% to 48% and
showed for both bases that the pH of the citric acid solution further to 58%, while the Ni leaching decreased from >99% to
2
was investigated. First, 10 and 20 mol% of the mineral However, the effect on the stability of Ni/ZrO2 and the
(
slightly increased from 2.2 to 2.5 and 2.7 for 10 and 20 mol%,
14% and further to only 4% at 20 mol% NaOH. The same trend
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Pleas eG dr eo e nn o Ct ha de jmu si st tmr yargins
Page 6 of 9
ARTICLE
Journal Name
was observed for KOH, where the yield increased to 57% and adsorption-desorption isotherms reveal that the particles had
View Article Online
2
DOI: 10.1039/C7GC01773A
further to 72%, while the Ni leaching decreased to 7% and a BET surface area of only 26 m /g, which matches the nominal
further to only 2% at 20 mol% KOH. Furthermore, when the H
2
value that was anticipated for nanoparticles of the specified
2
pressure was increased from 20 bar to 30 bar under these size range (i.e. 33.7 m /g geometrical surface area for perfect
conditions, the hydrogenation rate increased (Figure 3). This spheres of 20 nm size). Hence, this indicates that the carbon
resulted in an even more enhanced yield of methylsuccinic coating does not contain significant nanoporosity (Figure S9).
acid of up to 81%. The increased pressure did not further Nevertheless, this material was tested as well using the
stabilise Ni/ZrO
remained the same as at 20 bar (7%).
2
, since the mean Ni leaching at 30 bar H
2
screening conditions in Table 1 (entry 14). Surprisingly, a high
methylsuccinic acid yield of 81% was obtained and the Ni
leaching was limited to 20%. Since the carbon layer was
impermeable, this means that the layer should contain
imperfections, such as cracks. The catalyst was therefore
subjected to a thorough analysis using transmission electron
Figure 2. Effect of small amounts of mineral bases on the hydrogenation activity of
Ni/ZrO . Reaction conditions: citric acid (0.2 mmol), Ni/ZrO (10 mol% Ni), water (2 mL),
75°C, 2 bar N and 20 bar H , 6 h.
Figure 4. X-ray diffractogram of the carbon-coated Ni nanoparticles.
2
2
1
2
2
Figure 3. Effect of small amounts of mineral bases on the hydrogenation activity of
Ni/ZrO at higher H pressure. Reaction conditions: citric acid (0.2 mmol), Ni/ZrO
10 mol% Ni), water (2 mL), 175°C, 2 bar N and 30 bar H , 6 h.
2
2
2
(
2
2
Stabilising Ni nanoparticles with a carbon coating
The previous stabilisation methods both resulted in high
methylsuccinic acid yields, however, additives – mineral bases
or Fe – were always necessary. Therefore we explored the use
of a coating to stabilise the Ni nanoparticles and avoid the use
of additives. Unsupported carbon-coated Ni nanoparticles
(
Ni-C NPs) were purchased; they were prepared via laser
43
evaporation and have an aerodynamic particle size of 20 nm.
The protective carbon layer prevented the Ni from oxidation
when in contact with air. XPS confirmed that Ni is only present
in its metallic state (Table 4, entry 3; Table S4
Moreover, from XRD it is also clear that Ni has its metallic,
face-centred cubic (fcc) structure (Figure 4).44 The N
; Figure S7).
Figure 5. TEM images of the carbon-coated Ni nanoparticles. Zones with a very thin or
even absent carbon coverage are indicated by white arrows.
2
6
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 7 of 9
Pleas eG dr eo e nn o Ct ha de jmu si st tmr yargins
Journal Name
ARTICLE
microscopy (TEM) to try to visualise some of these intermediate compounds, resulting in a methylsuccinic acid
View Article Online
DOI: 10.1039/C7GC01773A
imperfections (Figure 5). First of all, the TEM images show that yield of 91%.
the catalyst consisted of particles with a very broad size
distribution, surrounded by a carbon matrix (Figure 5-(a-c)).
Close inspection of the 2D images shows some peculiarities.
The carbon layer varied in thickness (Figure 5-(d)), and some
parts of the Ni were hardly covered by carbon (Figure 5-(e)).
The Ni-C NPs were able to catalyze the hydrogenation in
the conversion of citric acid to methylsuccinic acid without the
use of any additives; they eventually gave the highest
methylsuccinic acid yield (81%) under the screening conditions
in Table 1. Therefore, the reaction conditions were optimised
for this catalyst. First, the effect of the H
2
pressure was
investigated (Figure 6). A minimum H pressure of 20 bar
2
appears to be necessary for an adequate hydrogenation of the
intermediate itaconic isomers. At 30 bar the methylsuccinic
acid yield slightly increased to 83%. Under these conditions a
time profile was subsequently gathered (Figure 7). This kinetic
profile is consistent with the reaction network in Scheme 1. It
demonstrates that the fragmentation products are formed
from the start of the reaction in a parallel path and that
Figure 6. Effect of H
using the carbon-coated Ni nanoparticles. Reaction conditions: citric acid (0.2 mmol),
Ni-C NPs (10 mol% Ni), water (2 mL), 175°C, 2 bar N , 6 h.
2
pressure on the conversion of citric acid to methylsuccinic acid
2
‘further decarboxylation’ products are indeed only formed
after a few hours. It is also clear that the itaconic isomers are
intermediates in this reaction sequence. Moreover, the
‘
isomer hydration’ products are intermediates as well, since
their formation is shown to be completely reversible. So
eventually, also β-carboxy-γ-butyrolactone can be reconverted
to the itaconic isomers and hydrogenated to the desired
methylsuccinic acid. This is why the yield increased to 89%
after a longer reaction time of 24 h. In addition, for this
reaction the Ni leaching was only 5%. This shows that also if
the Ni particles are appropriately protected by carbon, the
catalyst’s activity remains high enough to result in a very high
yield of methylsuccinic acid. Furthermore, reusability studies
Figure 7. Time course of the dehydration-decarboxylation-hydrogenation of citric acid
using the carbon-coated Ni nanoparticles. Reaction conditions: citric acid (0.2 mmol),
showed that the methylsuccinic acid yield after 6 h gradually Ni-C NPs (10 mol% Ni), water (2 mL), 175°C, 2 bar N and 30 bar H .
2
2
decreased from 80% to 76% and further to 55% at the third
run (Figure S10), which is consistent with limited Ni leaching
and minor catalyst loss during the recycling process.
Next, the influence of the temperature was studied (Figure
8
). At 150°C the dehydration-decarboxylation of citric acid was
much slower, resulting in a conversion of only 24% after 6 h.
Conversely, when the reaction temperature was increased to
2
00°C, the reaction sequence was faster: now all the
intermediate ‘isomer hydration’ products were already
converted to methylsuccinic acid after 6 h. On the other hand,
further decarboxylation and fragmentation side reactions
clearly increased with raising the temperature from 150°C to
2
6
00°C. Therefore, the methylsuccinic acid yield at 200°C after
h was slightly lower than the yield at 175°C after 24 h (86%
vs. 89%).
Finally, lowering the amount of Ni to 5 mol% was
Figure 8. Effect of temperature on the conversion of citric acid to methylsuccinic acid
using the carbon-coated Ni nanoparticles. Reaction conditions: citric acid (0.2 mmol),
considered (Figure 9). This decrease in catalyst-to-substrate Ni-C NPs (10 mol% Ni), water (2 mL), 2 bar N
2 2
and 30 bar H .
ratio slowed down the hydrogenation and thus the reaction
sequence to methylsuccinic acid. After 6 h with only 5 mol% Ni
there were still some itaconic isomers left (8%), next to 16% of
‘
isomer hydration’ products. However, a reaction time of 24 h
sufficed to reach complete conversion of citric acid and all
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Pleas eG dr eo e nn o Ct ha de jmu si st tmr yargins
Page 8 of 9
ARTICLE
Journal Name
Author contributions
View Article Online
DOI: 10.1039/C7GC01773A
All authors have contributed to the work presented in this
manuscript. J.V. and D.E.D.V. designed the experiments and
prepared the manuscript. J.V. and A.G. prepared the catalysts,
carried out the reactions and did the product analysis. B.C.
coordinated the ICP measurements and interpreted the
results. S.E. and W.T. performed XPS analysis, interpreted the
results and edited the manuscript. C.V.G. and I.V. performed
TEM and STEM-EDX analysis and processed the results. I.S. and
2
R.A. performed N physisorption measurements, interpreted
the results and reviewed the manuscript. S.S. performed the
XRD measurements. All authors have given approval to the
final version of the manuscript.
Figure 9. Effect of a lower catalyst-to-substrate ratio on the conversion of citric acid to
methylsuccinic acid using the carbon-coated Ni nanoparticles. Reaction conditions:
citric acid (0.2 mmol), Ni-C NPs, water (2 mL), 175°C, 2 bar N and 30 bar H .
2 2
Conflicts of interest
There are no conflicts of interest to declare.
Conclusions
Acknowledgements
Citric acid can be transformed to methylsuccinic acid via the
new reaction sequence of dehydration-decarboxylation-
hydrogenation. Regular Ni catalysts are not stable under the
demanding conditions of an aqueous citric acid solution at
natural pH, and are thus not able to hydrogenate the
intermediate itaconic isomers. On the one hand, partial
neutralisation of the reactant results in a much slower
corrosion of the Ni particles and maintains the hydrogenation
J.V., C.V.G., I.S. and S.S. thank FWO and IWT for (post)doctoral
fellowships. D.E.D.V. is grateful to FWO and IWT for research
project funding, and acknowledges Belspo (IAP-PAI 7/05) and
the Flemish government (Methusalem) for financial support.
W.T. and S.E. thank FWO (G.0C60.13N), KU Leuven
(OT/14/072) and the European Union’s European Fund for
3
Regional Development (Accelerate
project, Interreg
Vlaanderen-Nederland programme) for financial support. W.T.
also thanks the Provincie West-Vlaanderen (Belgium) for his
Provincial Chair in Advanced Materials. The authors are also
grateful for Hercules project AKUL/13/19, which financed the
TEM equipment, and to Johnson Matthey for the kind
provision of industrial Ni catalysts. Finally, the assistance of
Karel Duerinckx, Werner Wouters, Arnaud Struyf and Carlos
Marquez with NMR, pressure reactors, and performing and
interpreting XRD measurements, respectively, is very much
appreciated.
activity of Ni/ZrO
2
, leading to methylsuccinic acid yields up to
8
1%. On the other hand, adding Fe to either the Ni/ZrO
2
catalyst or directly to the reaction mixture, results in a more
reducing environment for the Ni particles, since the added Fe
may act either as a sacrificial reductant (in the case of Fe0
powder), or as reduction catalyst to keep Ni in its zerovalent
2
state with H as reducing agent. This ensures a better stability
and hydrogenation activity of the Ni catalysts; however, Fe
also strongly increases the fragmentation side reactions,
limiting the maximum yield of methylsuccinic acid. Finally, also
protecting the Ni particles with a carbon layer avoids fast
corrosion of the metallic Ni. Optimising the reaction conditions
for these carbon-coated Ni nanoparticles results in an optimal
yield of 91% methylsuccinic acid without using additives, at
Notes and references
1
2
H. Richard and B. Muller, WO Pat., WO/2012/119861, 2012.
D. Mijolovic, Z. J. Szarka, J. Heimann and S. Garnier, Ger. Pat.,
DE 102011080722, 2011.
1
2
75°C with 30 bar H and 5 mol% Ni after 24 h. In all three
cases (Ni + base, Ni + Fe, Ni + C) the Ni leaching could be
minimised to around 5%. These concepts in improving the
stability of Ni catalysts may be expanded to the conversion of
other biobased compounds and further research on the
stabilisation of Ni catalysts is still necessary in order to make
them completely recyclable. This would benefit the greenness
of the Ni-based catalytic systems, also because the high
temperature catalyst synthesis procedures have a considerable
energy demand.
3
4
5
6
A. Takasu, Y. Iio, Y. Oishi, Y. Narukawa and T. Hirabayashi,
Macromolecules, 2005, 38, 1048–1050.
H. G. Chae, S. H. Park, B. C. Kim and D. K. Kim, J. Polym. Sci.,
Part B: Polym. Phys., 2004, 42, 1759–1766.
T. Xie, C. Gao, C. Wang, S. Shen and Y. Wu, Polym.—Plast.
Technol. Eng., 2014, 53, 465–471.
R. Loos, D. Mijolovic, J. Heimann and Z. J. Szarka, U.S. Pat.
Appl., US 20120245256, 2012.
7
8
9
A. Bavley and C. J. Knuth, U.S. Pat., 2,773,897, 1956.
H. Shi, W. Shen and H. Xu, SIPO Pat., CN 1609089, 2005.
Y. Wu, C. Gao, Y. Wang, C. Wang and J. Xu, SIPO Pat., CN
1
02617326, 2012.
1
1
0 Q. Huang, W. Yu, R. Lu, F. Lu, J. Gao, H. Miao and J. Xu, RSC
Adv., 2015, 5, 97256–97263.
1 F. J. Holzhäuser, J. Artz, S. Palkovits, D. Kreyenschulte, J.
Büchs and R. Palkovits, Green Chem., 2017, 19, 2390–2397.
8
| J. Name., 2012, 00, 1-3
This journal is © The Royal Society of Chemistry 20xx
Please do not adjust margins

Page 9 of 9
Pleas eG dr eo e nn o Ct ha de jmu si st tmr yargins
Journal Name
ARTICLE
1
2 Weastra, http://www.bioconsept.eu/wp-
43 US Research Nanomaterials, http://www.us-
View Article Online
nano.com/inc/sdetail/171, accessed May 29, 2017.
content/uploads/BioConSepT_Market-potential-for-
selected-platform-chemicals_report1.pdf, accessed
November 21, 2016.
DOI: 10.1039/C7GC01773A
44 L. Seinberg, S. Yamamoto, R. Gallage, M. Tsujimoto, Y.
Kobayashi, S. Isoda, M. Takano and H. Kageyama, Chem.
Commun., 2012, 48, 8237–8239.
1
3 IHS Chemical, http://blog.ihs.com/video-bon-appetit-a-taste-
of-the-citric-acid-and-monosodium-glutamate-msg-markets,
accessed November 21, 2016.
1
1
1
1
4 C. R. Soccol, L. P. S. Vandenberghe, C. Rodrigues and A.
Pandey, Food Technol. Biotechnol., 2006, 44, 141–149.
5 J. Verduyckt and D. E. De Vos, Chem. Commun., 2017, 53,
5
682–5693.
6 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.
7 H. Weyer, R. Fischer and C. Sigwart, U.S. Pat., 5,391,771,
1
995.
1
1
8 V. N. M. Rao, Eur. Pat. Appl., EP0277562, 1988.
9 J. Verduyckt and D. E. De Vos, Chem. Sci., 2017, 8, 2616–
2
620.
2
0 S. Nishimura, Handbook of heterogeneous catalytic
hydrogenation for organic synthesis, John Wiley & Sons, New
York, 2001.
2
1 Z. A. Chase, S. Kasakov, H. Shi, A. Vjunov, J. L. Fulton, D. M.
Camaioni, M. Balasubramanian, C. Zhao, Y. Wang and J. A.
Lercher, Chem. – Eur. J., 2015, 21, 16541–16546.
2 E. Soghrati, C. Choong, C. K. Poh, S. Kawi and A. Borgna,
ChemCatChem, 2017, 9, 1402–1408.
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
3 J. Ren, X. Qin, J.-Z. Yang, Z.-F. Qin, H.-L. Guo, J.-Y. Lin and Z.
Li, Fuel Process. Technol., 2015, 137, 204–211.
4 A. S. Bambal, K. S. Vecchio and R. J. Cattolica, Ind. Eng. Chem.
Res., 2014, 53, 13656–13666.
5 S. Abelló, E. Bolshak and D. Montané, Appl. Catal., A, 2013,
4
50, 261–274.
6 S. Hwang, J. Lee, U. G. Hong, J. C. Jung, D. J. Koh, H. Lim, C.
Byun and I. K. Song, J. Ind. Eng. Chem., 2012, 18, 243–248.
7 L. Huang, J. Xie, W. Chu, R. Chen, D. Chu and A. T. Hsu, Catal.
Commun., 2009, 10, 502–508.
8 L. Huang, Q. Liu, R. Chen and A. T. Hsu, Appl. Catal., A, 2011,
3
93, 302–308.
9 T. D. Gould, A. Izar, A. W. Weimer, J. L. Falconer and J. W.
Medlin, ACS Catal., 2014, 4, 2714–2717.
0 J. W. Han, C. Kim, J. S. Park and H. Lee, ChemSusChem, 2014,
7, 451–456.
1 H. Liu, C. Guan, X. Li, L. Cheng, J. Zhao, N. Xue and W. Ding,
ChemCatChem, 2013, 5, 3904–3909.
2 J. C. Park, J. U. Bang, J. Lee, C. H. Ko and H. Song, J. Mater.
Chem., 2010, 20, 1239–1246.
3 S. F. Wang, F. Xie and R. F. Hu, Sens. Actuators, B, 2007, 123,
4
95–500.
4 H. Zhang, Y. Feng, Y. Zhang, L. Fang, W. Li, Q. Liu, K. Wu and
Y. Wang, ChemSusChem, 2014, 7, 2000–2006.
5 J. Yu, Y. Zhong, W. Zhou and Z. Shao, J. Power Sources, 2017,
3
38, 26–33.
6 A. Zhang, Q. Ma, K. Wang, X. Liu, P. Shuler and Y. Tang, Appl.
Catal., A, 2006, 303, 103–109.
7 X. Wang, K. Li, X. Di Yang, L. L. Wang and R. F. Shen, J. Inorg.
Biochem., 2009, 103, 657–665.
8 J. Liu, X. Xia, Y. Li, H. Wang and Z. Li, Struct. Chem., 2013, 24,
2
51–261.
9 Y.-P. Sun, X. Li, J. Cao, W. Zhang and H. P. Wang, Adv. Colloid
Interface Sci., 2006, 120, 47–56.
0 D. Briggs and J. T. Grant, Surface analysis by Auger and X-ray
photoelectron spectroscopy, IM Publications, Chichester and
SurfaceSpectra, Manchester, 2003.
4
4
1 M. P. Seah and I. S. Gilmore, Surf. Interface Anal., 2001, 31,
8
35–846.
2 Thermo Fisher Scientific,
http://xpssimplified.com/elements/iron.php, accessed June
3
, 2017.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins