Enabling Process Intensification by 3 D Printing of Catalytic Structures

Article abstract of DOI:10.1002/cctc.201700829

Small-scale, intensified chemical reactors (i.e., process intensification) mediated by structured catalysts substantially diminishes the advantages of large-scale gas-to-liquid (transport fuels) process plants and can be realized at low capital costs, minimum energy consumption, and zero/small CO₂ footprints. Current structured-catalysts approaches are complex and expensive; therefore, simple methods are crucial that are capable of depositing a desired geometry of catalysts into engineered channels. Herein, we developed printable composition by incorporating nickel and molybdenum ions into water-soluble PVA and starch; the subsequent pyrolysis of organic compounds resulted into three-dimensional carbon scaffold with micro/macro interconnected pores (d_pore, 6.5 ?; d_pore, 100 μm) containing up to 25 wt % catalyst loading. 2 D (TEM, SEM) and 3D (X-ray computed tomography) microstructural analyses and catalytic tests (conversion of syngas to alcohols) were performed for 3 D printed catalysts and compared with conventional pelleted catalysts. At a high feed flow rate (6000 h^?1), CO conversion is rapidly reduced to 16 mol % for pelleted catalysts, whereas 3 D printed catalysts converted 35 mol % of CO, with the same catalyst loading.

Full text of DOI:10.1002/cctc.201700829

Accepted Article
Title: Enabling process intensification via 3D printing of catalytic
structures
Authors: Muxina Konarova, Waqas Aslam, Lei Ge, Qing Ma, Fengqiu
Tang, Victor Rudolph, and Jorge Norberto Beltramini
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201700829
Link to VoR: http://dx.doi.org/10.1002/cctc.201700829
A Journal of
www.chemcatchem.org

10.1002/cctc.201700829
ChemCatChem
FULL PAPER
1 Enabling process intensification via 3D printing of catalytic
2 structures
3
4
Muxina Konarova[a], Waqas Aslam[a], Lei Ge[b], Qing Ma[a], Fengqiu Tang[b], Victor Rudolph[b] and Jorge
Norberto Beltramini[a] *
5
6
Abstract: Small-scale, intensified chemical reactors (i.e. Process
intensification) mediated by structured catalysts substantially
diminishes the advantages of large-scale gas to liquid (transport
fuels) process plants and can be realized at low capital costs,
minimum energy consumption, and zero/small CO2 footprints.
Current structured catalysts approaches are complex and expensive,
simple methods are crucial that are capable of depositing a desired
geometry of catalysts into engineered channels. Herein, we
developed printable composition by incorporating nickel and
molybdenum ions into water soluble PVA and starch; subsequent
pyrolysis of organic compounds resulted into three dimensional
carbon scaffold with micro/macro interconnected pores (dpore-6.3Å;
precisely deposit catalysts into engineered channels, which
match diffusion and transport timescales. Raise of additive
manufacturing (3D printing) enables formation of a highly
reproducible two-dimensional layer, which is then developed into
3D structures by overprinting sequential layers [5]. It has been
applied to fabricate of sensors, ceramics, metallic scaffolds,
microvascular networks, self-healing materials and tissue
engineering scaffolds due to its ability to orderly assemble nano-
structures and precisely pattern micro-sized objects in complex
3D architectures[6]. 3D printing can be adopted for the precise
engineering of catalytic structures and allows the catalyst to be
placed into the desired geometry. By tuning the size and length
of the flow channels, the bulk residence time and pressure drop
over the reactor can be precisely set. The catalyst material
properties and diffusion lengths can be controlled by
manipulating layer height (nozzle diameter) and porosity of the
catalyst bulk layers[7]. 3D printed catalysts have been
demonstrated in literatures for catalysts used in Ullmann
reactions[8], Al2O3 synthesis[9], catalytic reaction-ware for
chemical synthesis[10] and platinum metal patterning[11], while
reports on gas phase reactions over 3D printed catalysts are
scarce. 3D printing for the design of complex catalytic monolith
structures is not widespread[12], mainly because it is difficult to
obtain printable catalyst composition that has less non-catalytic
components such as plasticizers, binders, and mechanical
strength additives and hard to achieve high loadings of
nanoparticles without forming large agglomerated particles.
Therefore an optimum printable composition or technique must
allow incorporation of catalyst nanoparticles (or in-situ
nanoparticle formation) and result to a large fraction (>80 wt. %)
of catalytic material in the 3D catalytic structure. Influence of 3D
printed structures on the catalyst activity; product selectivity,
active particle dispersion, and internal micro/nano structure and
meso/macro porosity are not yet fully investigated and remain a
research challenge. Herein, we report the use of extrusion
based 3D printing techniques for the fabrication of structured
NiMo-oxide catalysts that were tested in the catalytic synthesis
of syngas to higher alcohols. NiMoO2 was chosen due to its
suitability to form corresponding sulfides or carbides by either
subjecting to hydrogen sulphide or methane, respectively. The
printable catalyst composition was designed by incorporating
aqueous solutions of metal precursors into PVA and starch.
Subsequent pyrolysis of organic components resulted into
porous carbon scaffold, which served as a catalytic support. By
modifying synthesis parameters, we fabricated cellular carbon
structures with open porosity, in which catalyst particles are
uniformly dispersed across the carbon scaffold, as observed 2D
structures by TEM, SEM. X-ray CT 3D images revealed micro-
cracks in samples with a large channel size (500 µm), which can
be eliminated by re-designing monolith configuration/geometrical
dpore-100µm) containing up 25 wt.% catalyst loading. 2D (TEM, SEM)
and 3D (X-ray CT) microstructural analyses and catalytic tests
(conversion of syngas to alcohols) were carried out for 3D printed
catalysts and compared with conventional pelleted catalysts. At high
feed flow rate (6000 h-1), CO conversion is rapidly reduced to 16
mol.% for pelleted catalysts, whereas 3D printed catalysts converted
35 mol. % of CO, with the same catalyst loading.
Introduction
Process intensification offers innovative solutions for production
of renewable fuels and chemicals, and its successful
implementation will lead to sustainable chemical industry with
maximum productivity realized at the minimal energy
[1]
consumption, capital investments and CO2 footprints
.
In
modern chemical industry, the use of Velocys’s mircochannel
reactors is an example of process intensification in which bio-
syngas is converted to renewable fuels via Fischer –Tropsch
synthesis[2]. However, widespread application of this technology
is hampered by complex design of the reactor, high fabrication
cost and low volume of catalyst per unit volume of reactor [3]
.
A possible solution for rapid growth of process intensification is
to design structured, mini-channel catalytic reactors [4], and
[a]
[b]
Dr Beltramini J., Dr Konarova M., Mr Aslam W., Dr Ma Q.
Australian Institute of Bioengineering and Nanotechnology
The University of Queensland
St Lucia, QLD, 4072, Australia
Fax: (+61 7 3346 3973)
E-mail: j.beltramini@uq.edu.au
Prof Rudolph V., Dr Ge L., Dr Tang F.
School of Chemical Engineering
The University of Queensland
St Lucia, QLD, 4072, Australia
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/cctc.200xxxxxx.
1
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700829
ChemCatChem
FULL PAPER
shape. To our knowledge, this is the first report providing
detailed structural analysis of 3D printed catalysts and
highlighting advantages and shortfalls of extrusion based 3D
printing methods. Furthermore, during syngas reaction the
catalytic activity of 3D printed catalysts was not affected with an
increase of gas hourly velocity (34 mol.% at 6000 h-1) whereas
conventional pelleted catalysts showed significant drop in CO
conversion (16 mol. % at 6000 h-1). This study expands catalyst-
engineering approaches, as traditional preparation methods and
forms of conventional catalysts (pellets) are grossly sub-optimal
a
b
1µm
10µm
for modern highly active nano-structured catalysts[13]
.
Furthermore, 3D printing enables an effective control over
heat/mass transfer properties of structured catalytic reactors[14]
.
Figure 2. SEM images of (a) NiMo/PVA internal microstructure at low
magnification; (a-inset) fractured area to observe subsurface morphology; (b)
catalyst particles taken at high magnification (x11k) of image (a).
Results and Discussion
The microstructure composed of round and honeycomb-
patterned cage network arised from strong interactions between
the PVA hydroxyl groups and starch molecules. Previously, Xu
et al[15] observed similar porous microstructure during the cross-
linking of PVA with PEG/cellulose nanofiber. Subsequent
pyrolysis of PVA/starch resulted into interconnected porous
carbon scaffold resembling solid cellular structure. In this
structure, catalyst particles are randomly dispersed in a form of
large agglomerates. High magnification images revealed catalyst
particles with a porous architecture formed by aggregation of
primary catalyst particles with size of ~100 nm (Fig 1b).
Catalytic monoliths with cylindrical shape (d=15; h=80 mm) were
printed due to its analytical convenience for testing catalytic
performances in tubular reactors (Fig S1). Optical images were
taken before and after the thermal decomposition of binders.
Fig 1 shows optical images of monolith with square and
hexagonal channels.
b
a
To enhance catalyst dispersion in the carbon scaffold, a cross-
linking agent (maleic acid) was added to PVA-water mixture
prior to addition of catalyst precursors and starch.
Corresponding images of the NiMo/PVA-MA are given in Fig 3.
b*
a
*
b
a
10µm
1µm
Figure 3. SEM images of (a) NiMo/PVA-MA internal microstructure at low
magnification; (a-inset) fractured area to observe subsurface morphology;
(b)microstructure taken at a high magnification (x10k) of image (a).
Figure 1. Optical images of 3D printed NiMo catalysts; (a) square patterned
channels, (b)-honeycomb patterned channels; (a*) and (b*) are the
corresponding images of NiMo catalysts after pyrolysis of organic constitutes.
Internal microstructure shows round-patterned cellular structure
with cell sizes ranging from 10 to 60 µm. Compared to previous
NiMo/PVA (Fig 2) , the cells are densely packed with carbon
particles (~3 µm). These carbon particles are entrapped within
the cells and their surface is covered by (~100 nm) catalyst
particles (Fig 3b, catalysts-white dots). Catalyst particles are
homogenously dispersed on the carbon scaffold indicating that
the cross-linking of PVA with maleic acid is critical for dispersing
catalyst particles and preventing particles agglomeration. In both
cases, catalyst particles are dispersed on the surface and inner
walls of the channels. Dispersion of catalyst particles is
homogenous throughout the carbon network particularly in
By measuring the size of the monolith before and after the heat-
treatment, ~53% volume reduction and ~70% weight loss was
recorded in both cases. The monolith maintained its structural
integrity despite of the substantial weight loss and volume
shrinkage. In square-patterned monolith, channel size were
reduced from 500 to 300 µm after the transformation of binders;
however, hexagonal-patterned monolith exhibited a non-uniform
shrinkage of the channels; therefore, square-patterned monolith
was used for further studies. NiMo monolith obtained by PVA-
Starch possesses hierarchal porous structure. Internal
microstructure taken at fractured area is given in Fig 2.
2
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700829
ChemCatChem
FULL PAPER
PVA/MA cross-linked NiMo-3D (Fig S2, EDX-spectra). NiMo-
precursors were loaded into commercial activated carbon-pellets
by incipient-wetness impregnation to highlight the structural
differences between conventional and 3D printed monolith
catalysts. Fig 4 shows SEM of NiMo/AC pellets taken at cross-
sections of a cylindrical pellet (d=0.8 mm).
Although the 3D printed monoliths exhibited uniform distribution
of catalyst particles, available BET surface area is relatively
small. In attempt to enhance surface area without breaking the
3D monolith structure, the catalysts were impregnated with KOH
aqueous solution and excess water was removed by freeze-
drying. The same previously reported KOH-activation procedure
was followed[16]. Activation of carbon materials with KOH is well-
known procedure; difficulty is associated with separating the
slurry mixture (K2CO3+Carbon) during the washing/removal of
potassium ions. However, the 3D printed monolith retained its
geometrical shape after the activation/heat-treatment making the
washing step is much more convenient. NiMo/PVA-MA showed
poor uptake of KOH aqueous solution associated with its low
porosity. After the freeze-drying white powder (dry KOH) was
separated from the catalytic 3D monolith and deposited at the
bottom of flask. Therefore, only NiMo/PVA was further
characterized and Fig 5 shows SEM images of NiMo/PVA after
the KOH activation.
a
b
10µm
10µm
Figure 4. SEM images of NiMo/AC pellets (a) surface morphology of pellets
(a-inset)-low magnification of the pellets; (b) SEM back-scattering images
taken at a cross-section of the cylindrical pellets (bright spots correspond to
catalyst particles ).
Back scattering images (Fig 4b) were taken to show catalyst
dispersion on the surface and inner walls of the carbon pellets.
Spherical catalyst particles with size of 300 nm were found only
on the surface and internal pores were mostly vacant. These
images highlight that during incipient-wetness impregnation,
diffusion of catalyst precursors into internal pores are hindered
by the pore geometry, which results to an agglomeration of large
particles on the surface. In this microstructure, most of available
pores remain unfilled due to the blockage of surface pores by
catalyst particles; this consequently reduces catalyst loading and
density of catalytic active sites. Table 1 shows surface are
measurements and chemical compositions of the 3D printed
NiMo monoliths and NiMo/AC pellets. Nitrogen physisorption
measurements revealed BET surface area of 120 and 6 m2/g for
3D printed NiMo/PVA and NiMo/PVA-MA 3D monolith,
respectively. Low surface area of NiMo/PVA-MA is attributed to
its dense microstructure as observed by SEM (Fig 3), indicating
the presence of inaccessible pores within the microstructure.
Among them, NiMo/AC pellets exhibited the large BET surface
area (228 m2/g) due to the remaining un-filled pores on the inner
walls of the pellets, which was depicted by SEM (Fig 4.) All
samples contain around ~75wt % of carbon, 21.3 wt. % Mo and
4.5 wt.% Ni (3/1 molybdenum/nickel mol ratio), slightly larger
carbon content was found for samples maleic acid cross-linked
samples (NiMo/PVA-MA 3D monolith).
a
b
10µm
1µm
Figure 5. SEM images of NiMo/PVA-3D monolith after KOH activation
Internal microstructure of NiMo/PVA-KOH 3D monolith consists
of cellular frameworks with open pores. High dispersion of the
spherical catalyst particles with diameters ranging between 100-
200 nm were observed. The distribution of catalyst particles is
homogenous across the porous carbon surface compared to
non-activated samples (Fig 2b). Apparent changes in internal
structure is caused by reactive and oxidizing nature of KOH
which triggered decomposition of amorphous carbon via C-C
bond cleavage producing CO/CO2 volatiles and open pores.
Simultaneously, large catalyst agglomerates, observed before
activation (Fig 2b), were broken down into spherical particles
(dp~100 nm) and uniformly dispersed across the carbon network.
Breakage of large agglomerates (~600 nm) into small particles
(~100 nm) might be caused by migration/diffusion of catalyst
particles from/into surface of 3D monolith during the
impregnation with strong alkaline KOH solution.
Table 1. Summary of Nitrogen physisorption measurements and ICP/CHNS
elemental analysis.
Pore size distribution (PSD) of 3D printed monolith before and
after KOH activation was measured by nitrogen physisorption.
KOH activation step increased both BET surface area and total
pore volume from 120 m2/g and 0.2 ml/g (NiMo/PVA) to
640 m2/g and 0.36 ml/g (NiMo/PVA/KOH), respectively. For the
PSD calculations, two-dimensional non-local density functional
theory model for carbon with heterogeneous surfaces (2D-
NLDFT-HS) was used (SAIEUS, Micromeritics). NLDFT models
have been used in previous reports to define porous structures
of various carbon materials[17]. Corresponding experimental and
Sample Name
BET
Surface
area, m2/g
Average
pore
size, nm
Carbon,
wt%
Ni,
wt%
Mo, wt%
NiMo/PVA
120
6
3.4
72.6
4.71
3.84
4.13
21.01
17.50
20.82
NiMo/PVA/MA
5.3
77.67
68.62
NiMo/AC
pellets
228
2.67
3
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700829
ChemCatChem
FULL PAPER
simulated isotherms along with the resulting PSD before and
after KOH activation are given in Fig 6. Bimodal pore size
distribution with pore widths centering at 6.1 and 15 Å was
observed before the activation. The pore volume around
Dpore= 6.3 Å significantly increased due to the formation of
additional micropores induced by KOH activation (Fig 6).
sample is
a
hierarchal porous catalyst with microporous-
macroporous architecture as evidenced from broad micropores
region from (dp=6.3 Å) and bi-modal macropore distribution (~10
and 100 µm). Such catalytic architecture enables fast chemical
reactions as the micropores provide superior mass transfer
properties and the macropores enhance active site accessibility
to feed molecules[1,
.
Table S1 (Supplementary file)
18]
summarizes the changes in surface area, average (micro,
macro) pore sizes and total pore volume occurred after the KOH
activation and ICP/Carbon elemental analysis.
0.14
500
Exp: NiMo/PVA
NLDFT (Heterogeneous slit)
0.12
Exp: NiMO/PVA/KOH
400
Further observations were carried out by TEM to illustrate the
0.10
300
structure at nano-scale. Fig 8 shows TEM images of
NiMo/PVA/KOH 3D monolith prepared by suspending crushed
powder in ethanol. Particles with a diameter of ~50 nm are
dispersed in carbon network and catalyst particles are coated by
thin graphitic carbon layers as evidenced by an interlayer
spacing of ~0.35nm. This suggests that precursor metal ions are
encapsulated by PVA polymeric chains during the initial mixing,
further carbonization led to the formation of the graphitic carbon
0.08
100
0.06
10-4
10-3
10-2
10-1
100
0.04
0.02
0.00
Relative Pressure, p/p0
shell[19]
.
Having such microstructure is advantageous as the
carbon layer prevents particles agglomeration by separating the
small crystallites and retains the surface area of catalysts during
a catalytic reaction.
0
10
20
30
40
50
Pore Width, Å
Figure 6. Nitrogen adsorption experimental and simulated isotherms (inset;
blue line corresponds to NLDFT fitting curves) and pore size distribution
curves of NiMo/PVA (black triangle) and NiMo/PVA/KOH (red cross).
3D-NiMo/KOH
3D-NiMo
0.25
0.20
0.15
0.10
0.05
0.00
Figure 8. TEM images of NiMo/PVA/KOH catalysts; (inset) taken at high
magnifications.
0.01
0.1
1
10
100
Pore Radius, µm
The key important factor in 3D printed catalysts whether the
printed structure can hold its original shape, after removing the
organic binders. Since none of the samples contains inorganic
binders (clay, calcium salts, glass fibers) to enhance mechanical
strength of the monolith and all sample were treated by KOH to
produce open porosity, which lead to carbon loss from the
structure; it was expected that some structural strength might be
lost. To visualize overall structure and identify minor cracks, 3D
X-ray CT images were constructed with 1600 images acquired
as the sample was rotated 180° at each vertical location (Fig.9).
Images obtained from vertical cross-section of the 3D printed
samples are given in Fig. 9. Sample with a channel diameter of
500 µm (honeycomb) shows arrays of radial cracks, the length
of the cracks is relatively short enabling the catalysts to maintain
their original monolithic shape. Similar radial cracks were not
observed in samples with a channel diameter of 200 µm
(square-patterned). However, this sample (Fig 9a) show
irregular channel size along the vertical direction, lower part of
the cylinder has larger channel size than the upper parts; this
could be associated with non-uniform dispersion of KOH
Figure 7. Pore size distribution curves obtained from Hg-porosimetry
measurements NiMo/PVA (black circle) and NiMo/PVA/KOH (red triangle).
Hg-porosimetry was measured to evaluate macro-porosity of
samples before and after KOH activation. Large mercury uptake
was observed from the KOH activated samples. Both samples
exhibit large pore size ~90.8 µm; however, additional pores with
size of 6.98 and 10.6 µm were formed as a result of KOH
activation (Fig 7). The development of new pores is due to the
reaction of potassium hydroxide with amorphous carbon [23].
Comparison of pore sizes of two samples indicate that large
pores (≥100 µm) were formed as a result of decomposition of
organic molecules (Starch, PVA) and more large pores were
produced along with small pores (~6-10 µm) as a result of KOH
activation. Porosity measurements revealed that KOH activated
4
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700829
ChemCatChem
FULL PAPER
resulting to large pore opening in some parts. Comparison of X-
ray CT images of 3D printed catalysts revealed that 3D catalytic
structure with a channel diameter of 500 µm or higher will have
less rigid structure, which can be eliminated by varying the
geometries/channel sizes.
microstructure show irregular pore size and low catalyst
dispersion that led to high internal diffusion resistances and the
poor intra-particle mass transfer.
100
100
2400 h-1
3600 h-1
6000 h-1
4800 h-1
90
80
70
60
50
40
30
20
10
0
1200 h-1
NiMo/AC pellets
NiMo/3D monolith
80
60
40
20
0
a)
b)
20
40
60
80
100
120
140
160
180
200
TOS, hr
1000
2000
3000
4000
5000
6000
Gas hourly space velocity, h-1
Figure 10. CO conversion as a function of gas hourly space velocity for
NiMo/PVA/KOH-3D monolith (red triangle) and NiMo/AC pellets (black circle);
(inset) CO conversion as function of time on stream at various feed flow rates.
Figure 9. X-ray CT images of (a) NiMo/PVA/KOH with honeycomb patterns
and (b) NiMo/PVA/KOH with square patterns
Table
2 shows product selectivity for NiMo/PVA/KOH 3D
monolith and NiMo/AC pellets. Methane is one of the
undesirable products during the CO hydrogenation, which
produced in both AC pellets and 3D monolith product streams.
Catalytic performances of the NiMo/PVA/KOH 3D monolith were
tested for CO hydrogenation reaction at 60 bar, 280°C at various
gas flow rates. As a control test, NiMo/AC-pellets were also
tested under the same conditions but the samples were
impregnated with K2CO3 prior to the catalytic tests. The addition
of potassium salts to the pelleted catalysts was done; because,
ICP analysis results of KOH-activated NiMo/PVA/KOH 3D
monolith revealed the presence of 1.23 wt.% of potassium. Fig
10 compares the catalytic activity of NiMo/PVA/KOH 3D
monolith with that of NiMo/AC pellets (both contain ~1.2 wt.%
potassium), highlighting that the structured catalyst are not
severely affected by the contact times. For a given gas hourly
velocity, catalytic reaction were conducted at least for 40 hours
(Fig.10 inset). Average CO conversion vs gas hourly space
velocity is given in Fig 10. Both catalysts show similar CO
conversion rates at low flow rates; however, NiMo/AC pellets
rapidly lose its activity with the increase of gas hourly space
velocity and only 16 mol.% of CO is converted at 6000 h-1. In the
case of NiMo/PVA/KOH 3D monolith, significant decrease in CO
conversion was not observed, indicating its ability to convert
Table 2. CO conversion and Selectivity of NiMo/PVA/KOH-3D
monolith and NiMo/AC pellets catalysts a.
Catalysts
NiMo-3D
35
NiMo/AC pellets
X CO b[mol %]
16
24.7
18.6
7.2
HCs
9.1
C1-OH
C2-OH
C3-OH
31.4
34.6
6.8
Sc [mol %]
1.39
0
Other
Oxy.
0.4
feed molecules under high reactant flow rates.
A major
difference between two samples is their geometrical shape. In
the case of 3D printed monolith, we purposely printed the
geometric channels to eliminate pressure drop and produced
channel frames/walls with porous cellular microstructure to
promote fast diffusion of feed/product molecules from catalytic
active centers (Fig 5). Thus, this catalyst architecture with
interconnected micropores shortens diffusion path lengths and
enhances the overall heterogeneous reaction rate [18]. In contrast,
the pelleted catalysts are randomly packed; its internal
Alcohol
73.2
47.2
27.1
26.5
[mol %]
C2-OH in total alcohol
[a] Test conditions: P=60 bar, GHSV=6000 h-1, H2/CO=2. [b] CO
conversion. [c] C-based, selectivity of products CO2 included
5
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700829
ChemCatChem
FULL PAPER
heating rate of 0.5°C/min, then the furnace temperature was increased at
rate of 2°C/min to 700°C and then held for 3 hours. The heat-treatment
experiments were carried out under nitrogen atmosphere. As a control
experiment, activated carbon pellets (Norit, φ=0.8 mm) were impregnated
with aqueous solutions of ammonium molybdate tetrahydrate and nickel
(II) nitrate hexahydrate followed by drying and heat-treatment at 500°C
for 4 hours, under nitrogen atmosphere. Samples were activated by KOH
as follows: 3D printed monolith was immersed into KOH aqueous
solution (C/KOH=1/3 mass ratio) and then freeze-dried at 198K. The
dried 3D monolith was heat-treated at 700°C for 1 hour under nitrogen
flow. The activated 3D monolith was washed with water and then freeze-
dried.
Selectivity to methane with 3D monolith catalysts are remarkably
smaller (~9 mol%) compared to NiMo/AC pellets (25 mol%). The
reduced selectivity of methane formation observed in the
NiMo/PVA/KOH 3D monolith can be due to the fast transport of
products from the catalytic active centers, as entrapped
oxygenates may undergo further hydrogenation reactions
caused by the high diffusion resistance of products in the
NiMo/AC pellets. The 3D monolith lowers the extent of
secondary reactions as extreme operational conditions (short
residence time/high flow rate) can be realized with negligible
change in CO conversion. Catalytic results obtained at high gas
hourly velocity shows the possibility of developing highly
compact chemical reactors with mini-channel architecture.
Printing catalysts directly onto the mini-channels would provide
optimum conditions for process intensification in which tight
control over fluid and heat transport can be achieved.
Catalyst characterisations
The internal morphology of the samples was observed using a high-
resolution Transmission Electron Microscope (TEM) (200 kV, JEOL). N2
absorption/desorption isotherms of samples at -196°C were obtained
using
a TriStar II Micrometrics Surface Analyzer. The pore size
distributions of monolith catalysts were measured by mercury intrusion
porosimetry (MIP, Micromeritics AutoPore IV9520) up to 60,000psi. A
CHNS elemental analyser (Thermo Electron Corp. Flash EA-1112
Series) was used to determine the carbon content. X-ray computed
tomography (CT) studies were employed to detect and map the pore
Conclusions
structure of 3D printed catalyst.
A ZEISS Xradia Versa X-ray
microscopes (XRM500) was used (voltage = 80 kV, Power: 7W) to image
the samples. Data collection was carried out with 1600 images acquired
as the sample was rotated 180° at each vertical location. Scanned
images of the samples were reconstructed, visualised and normalised
using the Avizo (FEI Visualization Sciences Group). The bulk
composition of the catalyst was determined by inductively coupled
plasma (ICP) analysis with a Varian Vista Pro ICP-OES instrument after
digesting the catalyst using a Milestone Ethos 1 microwave digester.
Carbon content of samples was determined by an elemental CHNS/O
analyser (FlashEA 1112 series, Thermo Electron Corporation). The
catalytic tests were performed at 280°C, 60 bar in a high-pressure fixed
bed reactor setup. The stainless-steel fixed bed reactor (i.d. 10 mm;
length 100 mm) with a thermocouple positioned inside was inserted in a
furnace. The catalyst was reduced under hydrogen atmosphere at 400°C
for 3 hours before the reaction. In all catalytic tests, premixed syngas
with H2/CO ratio of 2/1 and 4% of N2 (internal standard) was used as
feed. The mass flow controller (MFC, Bronkhorst High-Tech B.V)
regulated the inlet gas flow rate. The weight of catalyst samples loaded in
the reactor is maintained as 2.5 g both for pelleted and 3D monolith. The
feed flow rate was varied as 50, 100, 150, 200 and 250 ml/min,
respectively. Gaseous products leaving the reactor passed through the
condenser to obtain the liquid fraction of products. The outlet gas stream
composition was analysed by gas chromatography equipped with a
thermal conductivity detector (Shimadzu GC-2014). Liquid samples were
discharged from the condenser every 12 hours and analysed by GC
equipped with a flame ionization detector (Shimadzu GC-8A). Conversion
In this study, we demonstrated 3D printing of catalysts using the
ubiquitous 3D printing method (solid free forming). From the
anaylsis of microstructure and catalytic properties, it is shown
that uniformly dispersed catalyst particles (dp-50nm) on the
porous carbon scaffold can be obtained. Catalytic tests revealed
high CO conversion rates for NiMo-3D monolith due to favorable
diffusion paths created by pore channels with various sizes
(dpore=6.3 Å; dpore=10-100µm). Pelleted catalysts showed
decrease in CO conversion (16 mol.%) at high feed flow rate
(6000 h-1),whereas 3D printed catalysts converted 35 mol. % of
CO with the same catalyst loading. This catalyst printing
approach enables to develop a highly compact, structured
catalytic reactors designed for intensive chemical reactions.
Various geometric structures can be printed to match channel
passages providing optimal mass and heat transfer properties.
Furthermore, 3D printing methods are best suited to highly
porous carbon supported catalysts; depending on the polymer
used as carbon source various microstructural features can be
obtained
Experimental Section
of syngas and selectivity of products were calculated as previously[20]
.
Preparation of catalyst composition for 3D printing
To aqueous solution of (10 wt. %) PVA (87-90% hydrolyzed, average
mol. wt. 30,000-70,000), 15 mL (40 wt. %) aqueous solution of
ammonium molybdate tetrahydrate (Sigma-Aldrich, 83.0% MoO3 basis)
and 10 mL (38 wt. %) of nickel (II) nitrate hexahydrate (Sigma-Aldrich)
were added. The mixture was heated at 100°C until homogenous
concentrate were formed. Following this, a thickener-starch (Starch-Rice,
Sigma-Aldrich) was added to produce extrudable paste (PVA/starch =
1/4). To enhance plasticity of the extrudate, 5mL of diethyl ether were
added for every gram of the paste. The same procedures were followed
in the experiments with maleic acid (MA) addition used as a cross-linking
agent for PVA molecules at the ratio of PVA/MA : 2/1. 3D monolith was
printed using a Hyrel 3D (Engine SR) printer equipped with EMO-25
extruder (cold flow) with a nozzle diameter of 500 µm. The printed
samples were dried at room temperature conditions for 24 hours, prior to
subjecting to heat-treatment. Samples were first heated up to 290°C at a
Acknowledgements
M.K acknowledges Advance Queensland Research Fellowship
scheme for funding this research project. Authors thank Mrs
Ying Yu from Centre for Microscopy and Microanalaysis (UQ) for
her assistance with high-resolution electron microscopy work.
6
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700829
ChemCatChem
FULL PAPER
Authors also thank ANFF-Q members who facilited in carrying [10] P. J. Kitson, M. D. Symes, V. Dragone, L.
out some the 3D printing modelling/slicing works.
Cronin, Chemical Science 2013, 4, 3099-
3103.
Keywords: structured catalysts • 3D printing • alcohol synthesis•
porous carbon • 3D structure visualization
[11] P. Shah, Y. Kevrekidis, J. Benziger,
Langmuir 1999, 15, 1584-1587.
[12] S. Govender, H. Friedrich, Catalysts 2017,
7, 62.
[1]
[2]
J. Gascon, J. R. van Ommen, J. A. Moulijn,
F. Kapteijn, Catalysis Science &
Technology 2015, 5, 807-817.
[13] a) S. Mitchell, N.-L. Michels, K. Kunze, J.
Pérez-Ramírez, Nat Chem 2012, 4, 825-
831; b) S. Mitchell, N.-L. Michels, J.
Perez-Ramirez, Chemical Society Reviews
2013, 42, 6094-6112.
H. J. Robota, L. A. Richard, S. Deshmukh,
S. LeViness, D. Leonarduzzi, D. Roberts,
Catalysis Surveys from Asia 2014, 18,
177-182.
[14] a) A. Tonkovich, D. Kuhlmann, A. Rogers,
J. McDaniel, S. Fitzgerald, R. Arora, T.
Yuschak, Chemical Engineering Research
and Design 2005, 83, 634-639; b) in
Process Intensification for Green
Chemistry, John Wiley & Sons, Ltd, 2013,
pp. i-xvi.
[3]
[4]
a) S. S. Ail, S. Dasappa, Renewable and
Sustainable Energy Reviews 2016, 58,
267-286; b) B. Todic, V. V. Ordomsky, N.
M. Nikacevic, A. Y. Khodakov, D. B.
Bukur, Catalysis Science & Technology
2015, 5, 1400-1411.
[15] Z. Xu, J. Li, H. Zhou, X. Jiang, C. Yang, F.
Wang, Y. Pan, N. Li, X. Li, L. Shi, X. Shi,
RSC Advances 2016, 6, 43626-43633.
a) R. M. de Deugd, F. Kapteijn, J. A.
Moulijn, Topics in Catalysis 2003, 26, 29-
39; b) A. C. a. J. A., Moulijn, Structured
Catalysts and Reactors, CRC Press, Boca
Raton,, 2005; c) A. Holmen, H. J. Venvik,
R. Myrstad, J. Zhu, D. Chen, Catalysis
Today 2013, 216, 150-157.
[16] D. Lozano-Castelló, M. A. Lillo-Ródenas,
D. Cazorla-Amorós, A. Linares-Solano,
Carbon 2001, 39, 741-749.
[17] a) G.-P. Hao, G. Mondin, Z. Zheng, T.
Biemelt, S. Klosz, R. Schubel, A.
[5]
[6]
a) J. R. Raney, J. A. Lewis, Mrs Bulletin
2015, 40, 943-950; b) B. Derby, Annual
Review of Materials Research 2010, 40,
395-414.
Eychmüller, S. Kaskel, Angewandte
Chemie International Edition 2015, 54,
1941-1945; b) G.-P. Hao, W.-C. Li, D.
Qian, G.-H. Wang, W.-P. Zhang, T. Zhang,
A.-Q. Wang, F. Schüth, H.-J. Bongard, A.-
H. Lu, Journal of the American Chemical
Society 2011, 133, 11378-11388; c) J.
Jagiello, J. P. Olivier, Adsorption 2013, 19,
777-783; d) E. A. Ustinov, D. D. Do, V. B.
Fenelonov, Applied Surface Science 2007,
253, 5610-5615.
a) J. A. Lewis, J. E. Smay, J. Stuecker, J.
Cesarano, Journal of the American
Ceramic Society 2006, 89, 3599-3609; b)
S. Lee, T. Boeltken, A. K. Mogalicherla, U.
Gerhards, P. Pfeifer, R. Dittmeyer, Applied
Catalysis A: General 2013, 467, 69-75.
[7]
[8]
X. Zhou, C.-j. Liu, Advanced Functional
Materials, 1701134-n/a.
[18] C. M. A. Parlett, K. Wilson, A. F. Lee,
Chemical Society Reviews 2013, 42, 3876-
3893.
C. R. Tubío, J. Azuaje, L. Escalante, A.
Coelho, F. Guitián, E. Sotelo, A. Gil,
Journal of Catalysis 2016, 334, 110-115.
[19] W. KICIŃSKI, M. BYSTRZEJEWSKI, M.
H. RÜMMELI, T. GEMMING, Bulletin of
Materials Science 2014, 37, 141-150.
[9]
K. A. M. Seerden, N. Reis, J. R. G. Evans,
P. S. Grant, J. W. Halloran, B. Derby,
Journal of the American Ceramic Society
2001, 84, 2514-2520.
[20] M. Konarova, F. Tang, J. Chen, G. Wang,
V. Rudolph, J. Beltramini, ChemCatChem
2014, 6, 2394-2402.
7
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700829
ChemCatChem
FULL PAPER
8
This article is protected by copyright. All rights reserved.

10.1002/cctc.201700829
ChemCatChem
FULL PAPER
9
This article is protected by copyright. All rights reserved.