Scalable continuous flow hydrogenations using Pd/Al2O3-coated rectangular cross-section 3D-printed static mixers

Article abstract of DOI:10.1016/j.cattod.2020.07.046

A novel type of catalytic static mixers with rectangular cross-section coated with Pd/Al₂O₃ has been developed. The catalytic mixers, produced by additive manufacturing and coated with the metal catalyst, have shown excellent catalytic performance for continuous flow hydrogenation of nitroaromatics. The heterogeneous catalyst is highly stable. No loss of catalytic performance was observed after long periods of activity and ICP-MS measurements determined that essentially no Pd or Al leaches from the reactor. The rectangular cross-section reactor is readily scalable. The process developed has been showcased for the synthesis of the key intermediate of the medicine tizanidine.

Full text of DOI:10.1016/j.cattod.2020.07.046

Journal Pre-proof
Scalable continuous flow hydrogenations using Pd/Al2O3-coated
rectangular cross-section 3D-printed static mixers
´
Rene Lebl, Yutong Zhu, Derrick Ng, Christian H. Hornung, David
Cantillo, C. Oliver Kappe
PII:
S0920-5861(20)30517-4
DOI:
https://doi.org/10.1016/j.cattod.2020.07.046
CATTOD 13049
Reference:
To appear in:
Catalysis Today
Received Date:
Revised Date:
Accepted Date:
28 February 2020
16 May 2020
20 July 2020
Please cite this article as: Lebl R, Zhu Y, Ng D, Hornung CH, Cantillo D, Kappe CO, Scalable
continuous flow hydrogenations using Pd/Al2O3-coated rectangular cross-section 3D-printed
static mixers, Catalysis Today (2020), doi: https://doi.org/10.1016/j.cattod.2020.07.046
This is a PDF file of an article that has undergone enhancements after acceptance, such as
the addition of a cover page and metadata, and formatting for readability, but it is not yet the
definitive version of record. This version will undergo additional copyediting, typesetting and
review before it is published in its final form, but we are providing this version to give early
visibility of the article. Please note that, during the production process, errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal
pertain.
© 2020 Published by Elsevier.

Scalable continuous flow hydrogenations using Pd/Al2O3-coated
rectangular cross-section 3D-printed static mixers
René Lebla,b, Yutong Zhuc, Derrick Ngc, Christian H. Hornungc, David Cantilloa,b,*, C. Oliver Kappea,b,*
a
Center for Continuous Flow Synthesis and Processing (CCFLOW), Research Center Pharmaceutical
Engineering GmbH (RCPE), Inffeldgasse 13, 8010 Graz, Austria.
b Institute of Chemistry, University of Graz, Heinrichstrasse 28, 8010 Graz, Austria.
c
Centre for Industrial Flow Chemistry, CSIRO Manufacturing, Ian Wark Laboratory, Bag 10, Clayton
South, Victoria, 3169, Australia.
*E-mail addresses: david.cantillo@uni-graz.at (D.C.), oliver.kappe@uni-graz.at (C.O.K.)
Graphical abstract
Highlights





A continuous flow process for heterogeneously catalyzed hydrogenations is presented
Specially designed static mixers (SM) are produced by additive manufacturing
Catalytic Static Mixers (CSM) were obtained by coating SM with a layer of Pd/Al2O3
CSM rectangular cross-section ensured excellent heat transfer and easy scalability
The performance of CSM was evaluated using an aromatic nitro reduction as model

Abstract
A novel type of catalytic static mixers with rectangular cross-section coated with Pd/Al2O3 has been
developed. The catalytic mixers, produced by additive manufacturing and coated with the metal oxide
catalyst, have shown excellent catalytic performance for continuous flow hydrogenation of
nitroaromatics. The heterogeneous catalyst is highly stable. No loss of catalytic performance was
observed after long periods of activity and ICP-MS measurements determined that essentially no Pd
or Al leaches from the reactor. The rectangular cross-section reactor is readily scalable. The process
developed has been showcased for the synthesis of the key intermediate of the medicine tizanidine.
Keywords: catalytic hydrogenation • continuous flow • additive
manufacturing • anilines • scale up
1. Introduction
Heterogeneously catalyzed reactions account for the largest share of chemical transformations ranging
from petroleum, bulk- and fine chemical, to pharmaceutical industries [1–5]. The prevalence of these
transformations typically increases with the production volume of the respective sector, as catalytic
reactions are a major strategy to enhance process efficiency and minimize waste generation. This is
mainly the result of eliminating the need for stoichiometric reagents or reducing resources required
for product isolation, as indicated by the E factor [6–9]. Pursuing the implementation of more
sustainable processes for fine chemical and especially pharmaceutical industries, investigations
towards catalytic processes are therefore of paramount importance for the synthesis of highly
functionalized molecules. In this context, hydrogenation reactions promoted by a heterogeneous
catalyst are widely used methods to reduce a variety of functional groups including alkenes, alkynes,
imines, nitriles, nitro groups and (hetero-) aromatic moieties [10–12]. Advantages of using
heterogeneous catalyst systems for these transformations include increased stability of the catalyst,
ease of catalyst separation and recyclability. Among the many metals that are typically employed
including nickel, cobalt and other platinum group metals (ruthenium, rhodium, osmium, iridium and
platinum), palladium has been recognized as one of the most prominent catalysts for hydrogenation
reactions. To increase the active surface area of the metal catalyst, different types of catalyst supports
are utilized, including activated carbon, zeolite, BaSO4, CaCO3 and Al2O3 [5,12].
Utilization of hydrogen gas as reducing agent maximizes atom economy. For the reduction of C-C, C-O
or C-N multiple bonds, the reagent is fully incorporated into the product, without any byproduct
formation. When functional groups such as nitro groups are reduced by H2, the only resulting
byproduct is water, which can be considered benign in many cases. To optimize the respective

multiphase process between gaseous reagent, liquid substrate (dissolved in a suitable solvent) and
solid catalyst (metal immobilized on a high-surface support), different reactors such as agitated tanks,
fluidized-, packed-, and trickle beds, slurry systems and microstructured reactors are well established
[1,13–17]. The field of microstructured systems has particularly gained attention during the past
decades due to their excellent mass and heat transfer properties, high potential for process
intensification, and increased safety due to the absence of headspace [18–22].
A common approach to perform a heterogeneously catalyzed multiphase reaction on microstructured
scale is to pack a microreactor with commercially available catalyst powders with particle sizes in the
order of 10-100 µm [23–26]. Such packed bed systems are commercially available and allow for
intensified process conditions up to pressures of 100 bar as well as temperatures up to 100 °C and
above [27–32]. Current strategies to intensify hydrogenation reactions using microstructured catalysts
include coated walls [33–35], catalytic falling film reactors [36–38], monoliths [39–42], catalytic foams
[43–45] and catalytically coated 3D-printed scaffolds [46–50]. These designed porous structure
reactors and catalytic static mixers (CSMs) were developed to overcome common limitations
associated with packed bed reactors, such as high pressure drops and temperature gradients which
can be hard to control. Recently, different coatings including Cu, Ni, Pd, Pt and Au have been
immobilized on different 3D-printed base scaffolds by electroplating or cold spray techniques [51].
CSMs for heterogeneous hydrogenation reported so far have utilized circular cross-section channels
and, as mentioned above, deposition of thin layers of metal applied by electroplating or cold spray.
Importantly, the heat transfer and scalability of tube bundle reactors can be improved using
rectangular channels. This is due to their improved heat transfer, which can be more easily maintained
when the reactor dimensions are increased [52,53]. For this reason, we hypothesized that the
scalability of the CSMs could be further improved by designing mixer templates for rectangular cross-
section channels. Moreover, utilization of porous catalyst supports for the coating of the mixers could
enhance the contact surface with the reaction mixture. Herein, we present the development,
characterization and evaluation of rectangular cross-section CSMs coated with a layer of Pd/Al2O3 as
active catalyst. The CSMs were designed to fit in a commercially available and scalable reactor system
[52–57], and were produced by additive manufacturing. Additive manufacturing features several
advantages over conventional machining techniques for the generation of metal parts, such as rapid
prototyping and the utilization of lower amounts of materials and energy [58]. The flat rectangular
channel structure provided excellent heat transfer (in the order of 10 kW m-2K-1 [53]). The system was
tested for the reduction of aromatic nitro compounds to the corresponding anilines.

2. Material and methods
2.1. General remarks
Solvents and reagents were obtained from commercial vendors and used without further purification
unless noted otherwise. HPLC analysis was performed using a Shimadzu LC20 system with a C-18
column (150 mm × 4.6 mm with a particle size of 5 µm) at 37 °C. Mobile phases A (water/acetonitrile
9+1 v/v + 0.1% trifluoroacetic acid) and B (acetonitrile + 0.1% trifluoroacetic acid) were pumped with
a total flow rate of 1.5 mL/min. The following gradient program was applied: start at 3% solvent B,
increase to 5% solvent B until 3 min, increase to 30% solvent B until 7 min and finally increase to 100%
solvent B until 10 min. The conversion of 1-nitronaphthalene (1) to 1-naphthylamine (2) was
determined based on the HPLC area at λ 254 nm according to internal calibrations (see Figure A4,
1
Appendix A). NMR spectra were recorded on a Bruker 300 MHz instrument. H and 13C spectra were
recorded at 300 MHz and 75 MHz, respectively, with a chemical shift (δ) relative to TMS expressed in
parts per million. The abbreviations d, dd, and bs are used to indicate doublet, doublet of doublets and
broad signals respectively.
2.2. CSM preparation
Additive manufacturing
The rectangular static mixer was designed using SolidWorks CAD design software. The structure allows
for high surface area and facilitates multi-phase mixing. The static mixers were manufactured by
Conflux Technologies to the design provided by CSIRO [59]. The 3D printing process was performed
using selective laser melting and 316L stainless steel powder.
Pre-treatment of the mixers
The 3D-printed static mixers were firstly washed with 3 wt% hydrochloric acid for 1 h to remove
oxidized species and then sonicated in deionized water for 20 min, after which they were washed again
with acetone for 10 min and finally dried in an oven under 120 °C vacuum. All the static mixers were
pretreated with this procedure.
Pd-Alumina coating
Typically, a dip-coating slurry was prepared by adding Pd/Al2O3 powder and solvent (water) into a vial
and stirring overnight until a uniform slurry was formed.
Uncoated mixers were dipped in the slurry for a few seconds, followed by pressurized air blowing to
eliminate excess liquid and prevent channel blockage. The dipped mixers were then horizontally placed
in an open container at room temperature for a day so that most of the solvent would evaporate in
air. Finally, the mixers were placed in an oven under 120 °C vacuum to eliminate any residual solvent.

This coating and drying process was repeated 2-3 times to gain the desired catalyst layer thickness on
the CSMs.
Before and after each coating step, the mass of the mixer was recorded to calculate the loading
percentage (typically a total loading 3-5% of Pd/Al2O3 is attained). To evaluate the adherence of the
catalyst layer, the coated CSMs were subjected to sonication in water for 10 min.
2.3. CSM characterization
Morphological images of CSMs were created using scanning electron microscopy (SEM) via secondary
electron (SE)-, and back-scattered electron (BSE) modes at an accelerating voltage of 5 kV, while the
elemental mapping was conducted via energy-dispersive X-ray spectroscopy (EDX) mode, by using a
Hitachi TM3030Plus tabletop microscope. EDX data was analyzed using an AZtec analysis software.
2.4. Continuous flow setup
The liquid feed was pumped using a Knauer AZURA® P 4.1S HPLC pump. An H-Genie™ hydrogen
generator from ThalesNano Energy was utilized as hydrogen source and introduced into the reactor
using a built-in mass flow controller (MFC). The gas and liquid streams were combined in a Y connector
made from PEEK immediately before being fed into the Ehrfeld Modular MicroReaction System
(MMRS). The hydrogenation reactions were performed in an Ehrfeld Miprowa® Lab reactor (attached
to the MMRS) featuring up to 8 channels with a cross section of 1.5 mm × 12 mm and a length of
300 mm each. The channels are connected in series and can be filled individually either with
customized CSMs (length 150 mm) or initial herringbone shaped flow baffles (45° angle, strut width
1.0 mm, spacing 2.0 mm, length 300 mm) made of Hastelloy (up to 16 CSMs can be inserted in the
reactor). During all experiments, the amount of channels was reduced from 8 to 4 and they were either
filled with CSMs or flow baffles to enhance mixing between gas and liquid phase. A Huber CC 304
thermostat, featuring an integrated temperature sensor at the reactor output, was used to adjust the
temperature of the Miprowa® Lab reactor. After a stainless steel cooling coil, a Swagelok® KCB series
backpressure regulator (BPR) was used to set the pressure inside the flow reactor. The gas/liquid (g/l)
mixture was separated by a customized low volume g/l separator at ambient conditions.
2.5. General procedure for continuous flow hydrogenation
In a 250 mL volumetric flask, a 0.2 mol/L solution of 1-nitronaphthalene (1) (8.66 g) in absolute EtOH
was prepared. Before each experiment, the reactor was purged with a solvent/gas mixture under the
reaction conditions (e.g. T = 125 °C, pBPR = 25 bar, QL = 1.00 mL/min, QG = 50 mLN/min, with QL and QG
representing the flow rates for the liquid and the gaseous H2 stream respectively). The reaction was
initialized by switching the input of the liquid feed pump from solvent to the substrate solution.
Samples were collected at regular intervals at the output of the gas/liquid separator and diluted for
HPLC analysis.

2.6. Execution of residence time distribution (RTD) measurements
To determine the RTD under different conditions, a UV/Vis flow cell was connected to the liquid output
of the g/l separator. Absorption spectra were recorded using an optical fiber based Avantes AvaSpec-
ULS2048 spectrometer and the corresponding AvaLight-DHc light source including both a deuterium
and a halogen lamp. A solution of anthracene in EtOH was used as tracer and the absorbance at λ 339
nm, λ 356 nm and λ 375 nm was monitored over time.
2.7. Determination of catalyst leaching by inductively coupled plasma mass spectrometry (ICP-MS)
Prior to measurements, the samples were diluted with ultrapure water (Milli-Q) to contain 10 % (v/v)
ethanol. The concentrations of Al, Cr, Cu, Fe, Mo, Ni and Pd were determined by ICP-MS (7700x, Agilent
Technologies, Waldbronn, Germany). The instrument was equipped with a Micro Mist nebulizer (Glass
Expansion, Melbourne, Australia), a Scott type double pass spray chamber, a 2.5 mm ID quartz torch,
a sample cone made from copper with nickel tip and a skimmer cone made from nickel. Quantification
was done via external calibration. The calibration range was as follows: 0.01–10 µg/L for Al, Cr, Cu, Fe,
Mo, Ni and Pd. The calibration standards include 10% (v/v) of ethanol. Certified reference material
(NIST 1640a, Trace Elements in Natural Water) was used to evaluate the trueness of the
measurements. The following isotopes were measured: 27Al, 52Cr/53Cr, 56Fe, 60Ni, 65Cu, 98Mo, 105Pd.
3. Results and discussion
3.1. Manufacturing and characterization of CSMs
Figure 1. 3D model of rectangular static mixer.
A tailor-made static mixer design was created to retrofit to the reactor channel geometry of the Ehrfeld
Miprowa® system. Figure 1 above shows a design drawing of these static mixers. The Miprowa® Lab
reactor channels have a width of 12 mm, a height of 1.5 mm and a length of 300 mm; we decided to
fill one channel with two CSMs of 150 mm length in series. Especially the small channel height of 1.5

mm presented a challenge as most commercial metal printers have a limited resolution, determined
by the particle size used for the print, hence design features of < 0.4 mm are not desirable. This
minimum design feature of 0.4 mm was adhered to herein, with the thickness of the metal sheet being
0.45 mm and the minor axis of the elongated holes being 0.8 mm. The main design concept here was
to achieve a high degree of interfacial contact between liquid and a gas phase, continued break up of
bubbles along the pathway of the fluidic channel and to maximize contact of the two mobile phases
with the supported catalyst. Secondly, the static mixer was also optimized for a maximum ‘coat-able’
surface, in order to deposit as much catalyst onto the CSM as possible without obstructing the flow
path. The mixer contains triangular ducts above and below the sheet, which are angled towards the
side wall of the channel and away from the general flow direction, which is along the length of the
CSM. The elongated wholes within the sheet connect the triangular ducts and allow for a cross-flow
from the ducts above the sheet to the ducts below and vice versa. This results in a zig-zag type flow
pattern over the length of the CSM, changing between flow along the angled ducts and flow crossing
the ducts, which in turn is believed to maximize the interfacial contact between liquid phase, hydrogen
and solid catalyst mentioned above. The mixing elements were analyzed by scanning electron
microscopy before and after the coating process (Figure 2). BSE images showed almost full coverage
of the mixer with the Pd/Al2O3 and a good catalyst dispersion. Only small spots where the cotaing was
detached from the mixing element could be observed by the brighter color of the BSE image (Figure
2c). Additional images and EDX spectra can be found in Figures A1 to A3, Appendix A.

Figure 2. (a,b) Back scattered electron (BSE) images of SM before coating; (c,d) BSE images after coating with Pd/Al2O3.
Some of the standard 300 mm herringbone shaped flow baffles of the employed Ehrfeld hydrogenation
reactor were replaced by the thus obtained CSMs which fit into the flat rectangular channels (see
Figure 3). In total 16 CSMs with a length of 150 mm could fit into one reactor.

Figure 3. Opened hydrogenation reactor with up to 8 rectangular channels. The photograph shows two CSMs (top right)
compared to the initial herringbone flow baffles (bottom right).
3.2. Reaction parameter optimization using one CSM element
The performance of the Pd/Al2O3 CSMs was evaluated using the reduction of nitroaromatics as model.
In particular, the conversion of 1-nitronaphthalene (1) to 1-naphthylamine (2) was selected for the
study. A series of solvents was initially evaluated, namely methanol, ethanol, ethyl acetate, toluene,
tetrahydrofuran and acetonitrile. An important aspect for solvent selection for continuous flow
processes is the solubility of both the starting materials and products under the reaction conditions,
which needs to be appropriate to avoid clogging of the reactor. Safety, health and environmental score
of the solvents were also considered [60]. All tested solvents showed good solubility for amine 2
(> 2 mol/L). The solubility screening for the starting material 1 revealed that the nitroaromatic was
readily soluble in ethyl acetate, toluene, tetrahydrofuran and acetonitrile (> 4 mol/L), but it was less
soluble in methanol and ethanol (ca. 0.3 mol/L).
Preliminary hydrogenation test runs were carried out at 75 °C in the six solvents (see Table A1,
Appendix A). The conversions of substrate 1 to amine 2 ranged from 33% in tetrahydrofuran to 53% in
ethyl acetate under identical conditions. Considering the results of solubility tests for compounds 1
and 2, preliminary hydrogenation test runs and CHEM 21 recommendations [59], ethanol and ethyl
acetate were identified as most promising solvents. To avoid solubility problems, and consequent
precipitate formation, the substrate concentration was set to 0.2 mol/L for all further experiments.
The continuous flow setup was arranged according to Scheme 1, including a temperature sensor at the
output of the flow reactor (T) and an integrated pressure sensor of the HPLC pump (P). All initial
optimization studies were carried out using one CSM, which was inserted at the beginning of the 4th

rectangular reactor channel. To allow better comparison between the different optimization steps, the
following conditions were chosen as “standard conditions” for the hydrogenations using one CSM: QL
= 1.00 mL/min, QG = 50 mLN/min, 125 °C, 25 bar.
Scheme 1. Continuous flow setup for hydrogenation reaction using CSMs.
When the reaction was performed in EtOAc over a period of 4 h at 125 °C and 25 bar, a decrease of
the reaction conversion from 91% to 80% was observed (see Figure 4). Notably, the reaction selectivity
remained constant (> 99%); HPLC monitoring did not reveal the formation of any side products, but a
gradual increase in the amount of unreacted starting material 1. Using EtOH instead of EtOAc caused
less drop in conversion during the first 2 h of operation (from 91% to 87%) and then resulted in a steady
conversion within the monitored time. This effect was ascribed to the higher strength of the protic
EtOH, capable of eluting with more efficiency amine 2, which might adhere to the surface of the Al2O3
causing a decrease of the active catalyst surface. This phenomenon was temperature dependent. At
lower temperatures (e.g. 75 °C) decrease of conversion over time was observed even with EtOH, while
above 125 °C the conversion was essentially constant over long periods (vide infra). A 1:1 v/v solvent
mixture of EtOH and EtOAc did not provide satisfactory results (Figure 4). Thus, EtOH was selected as
solvent for further optimization.

Figure 4. Performance of hydrogenation reaction over 4 h in EtOH, EtOAc and EtOH/EtOAc 1:1. QL = 1.00 mL/min, QG =
50 mLN/min, T = 125 °C, pBPR = 25 bar, c (1) = 0.2 mol/L.
Next, the influence of pressure and temperature were investigated (Figure 5). As expected, increase
of both parameters had a positive effect on the reaction outcome. A significant increase of conversion
was observed from 75 °C to 125 °C (Figure 5a), which then became less pronounced at temperatures
up to 163 °C, potentially due to a decreased hydrogen adsorption on the catalyst surface. The reactor
pressure was increased up to the limit of the reactor (25 bar) pressure (Figure 5b).

(a)
(b)
Figure 5. Influence of pressure and temperature QL = 1.00 mL/min, QG = 50 mLN/min, c (1) = 0.2 mol/L. (a) Temperature
influence at 25 bar. (b) Pressure influence at 125 °C.
The g/l ratio was also investigated. In addition to the hydrogen stoichiometry, this parameter has a
profound importance as it may determine the flow regime and the mixing efficiency in the system.
Liquid flow rates ranging between 0.50 mL/min and 2.00 mL/min and gas flow rates between 25
mLN/min and 100 mLN/min were evaluated (Figure 6). Although higher conversions can be achieved at
e.g. 163 °C (Figure 5), the experiments were performed at 125 °C to obtain moderate conversions in
order to see more clearly the effect of the flow rates. For QL = 2.00 mL/min, the gas flow rate was not
set below 27 mLN/min to ensure delivery of 1.00 stoichiometric equivalents of hydrogen. Interestingly,
the gas flow rate QG had a negligible effect on the reaction conversion, whereas an almost linear
relation between QL and the conversion was observed. This effect pointed to a parallel flow regime for
the biphasic gas/liquid mixture, in which most likely the liquid stream adheres to the catalyst surface.
At low QG values the flow regime would eventually change at the late stage of the reaction, when most
of the H2 has been consumed. This hypothesis was confirmed by determining the residence time
distribution in the reactor at a constant value for QL = 1 mL and variable gas flow rates (25-
100 mLN/min) under reaction conditions (EtOH, 125 °C, 25 bar). Analogous residence times were
obtained in all cases (see Figure 7), which would not be the case for a typical segmented plug flow

regime. The conversions determined by HPLC ranged from ca. 71% at QL = 2.00 mL/min to 94% at QL =
0.50 mL/min.
Figure 6. Relationship between liquid flow rate (QL) and gas flow rate (QG) for the heterogeneous hydrogenation at 125 °C
and 25 bar using one CSM. * At a liquid flow rate of 2.00 mL/min the gas flow rate was increased from 25 mLN/min to
27 mLN/min to ensure 1.00 stoichiometric equivalents of hydrogen are delivered.
Figure 7. RTD measurement using a step experiment, followed by UV/Vis absorbance. At t = 0 min (step up) a tracer was
injected, until t = 30 min (step down) when the tracer was purged from the reactor.

3.3. Process scale up
The process could be readily scaled up by increasing the number of CSMs inserted into the reactor
from one to four (arranged in a serial fashion), resulting in a longer catalytically active channel. Thus,
the throughput of the hydrogenation could be augmented while maintaining similar contact times with
the catalyst surface. Utilizing the flow reactor equipped with 4 CSM units, a new range of flow rates
was investigated: QL = 4.00 mL/min to 8.00 mL/min and QG = 200 mLN/min to 400 mLN/min, at 125 °C
and 25 bar (Table 1). Compared to the results obtained using one CSM (Figure 6), an impressive
increase in conversion to the desired product 2 was observed at higher throughputs. For example,
using one CSM at QL = 1.00 mL/min and QG = 50 mLN/min, 85% conversion was obtained. With 4 CSMs
units and using 4-fold the flow rates (thus matching the contact time with the catalyst) 99% conversion
was achieved. This performance increase was ascribed to an improved mass transfer at the higher flow
rates utilized, pointing to a diffusion-limited reaction at low flow rates. Even at higher liquid
throughputs of 8.00 mL/min, excellent conversions up to 98% to the desired amine 2 could be obtained
(Entry i, Table 1). The influence of QL on the conversion was much higher than QG (similar to the results
observed with one CSM, Figure 6). Reactions conducted at QL = 4.00 mL/min (Entries a-c, Table 1)
showed the highest conversion and minimum deviation between the results at different gas flow rates.
At QL = 6.00 mL/min (Entries d-f, Table 1) the conversion slightly dropped, but was still above 99%. At
QL = 8.00 mL/min (Entries g-i, Table 1), the conversion was 96%-98%. The excellent results obtained
and the inherent scalability of the reactor utilized using rectangular cross-section channels [57], clearly
demonstrate the potential of the concept herein described to develop effective heterogeneous
hydrogenation protocols using CSM technology.
Table 1. Relationship between liquid flow rate (QL) and gas flow rate (QG) using a set of four CSMs in the flow setup
described in Scheme 1 at 125 °C, 25 bar.
QL
QG
HPLC conv.
[%] a)
entry
[mL/min]
[mLN/min]
a
b
c
d
e
f
4.00
4.00
4.00
6.00
6.00
6.00
8.00
8.00
8.00
200
300
400
200
300
400
200
300
400
99.9
99.6
99.7
99.0
99.4
99.3
96.7
96.2
98.0
g
h
i
a) determined at 254 nm

3.4. Synthesis of tizanidine intermediate
A potential practical application of the developed heterogeneous hydrogenation process was
showcased with the hydrogenation of 4-chloro-2-nitroaniline (3), an intermediate step in the synthesis
of the α2 agonist tizanidine (5) (Scheme 2) [61,62].
Scheme 2. Synthesis of tizanidine (5), based on the hydrogenation reaction of 3.
The hydrogenation of 3 to 4-chlorobenzene-1,2-diamine (4) was conducted using a 0.1 mol/L solution
of the nitroaromatic in EtOH abs. as feed solution. The hydrogenation was carried out using 4 CSMs in
series. Thus, to achieve full conversion 4 mL/min of liquid feed were injected into the reactor
(conditions from entry a in Table 1). The gas flowrate was set to 100 mLN/min due to the lower
concentration of starting material compared to the optimization studies. . The reactor (Scheme 1) was
set to 125 °C and 25 bar. Gratifyingly, under these conditions full conversion of the nitroarene 3 and
excellent selectivity (99%) toward the target amine 4 was observed (GC-FID). The dehalogenation side-
product o-phenylenediamine was the only impurity detected (typically <1.0%). The hydrogenation
process was performed uninterruptedly for 8 h (0.2 mol scale). The product was collected in 4 separate
fractions (2 h intervals) under N2 atmosphere and protected from light. Simple evaporation of the
solvent under reduced pressure, without any additional work-up, yielded the target amine 4 with 99%
purity (27.3 g, 96% isolated yield). Evaporation of the solvent was carried out for each of the aliquots
of the crude reaction mixture (480 mL) as it was collected from the reactor output. The ethanol
evaporated during the isolation of 4 was recycled to prepare new starting solution during the course
of the long run, allowing to the reduce the solvent consumption for the preparation of the aniline.
3.5. CSM stability, Pd leaching and coating characteristics
One of the most important characteristics of immobilized catalysts is their stability over long periods.
In this context, it should be noted that all optimizations in this study were performed using the same
CSM over two months and with a net reaction time well beyond 50 h. During that time very high
reproducibility was observed for the experiments carried out under standard conditions (T = 125 °C,
pBPR = 25 bar, QL = 1.00 mL/min, QG = 50 mLN/min). Moreover, the weight of the CSM element was

completely stable during that time (7.458 g before the first hydrogenation, compared to 7.462 g after
two months of operation).
To investigate the stability of the CSMs further, the amount of Pd leaching was determined by ICP-MS,
together with other metal constituents of the CSMs and the reactor, such as Al, Cr, Fe, Ni, Cu and Mo.
All analyzed samples (see Table A2) show a Pd concentration below 0.5 ppb and an Al concentration
below 50 ppb, indicating very high stability of the Pd/Al2O3 coating under the employed reaction
conditions. No difference was observed between a used mixer and new mixers, which were used for
the first time during the hydrogenation reactions. The concentration of other metals contained in the
stainless steel scaffold, such as Cr, Fe, Ni, Cu and Mo was below the limit of quantification (1 to 10 ppb
depending on the metal analyzed), except one sample from a hydrogenation reaction at 125 °C, which
showed a Ni content of 19 ± 2 ppb.
When a set of four new CSMs was used for the hydrogenation reaction, the concentration of product
eluted from the reactor was monitored by regular HPLC sampling. The flow rates were adapted
accordingly to QL = 4.00 mL/min and QG = 200 mLN/min. During the first run, the concentration of amine
2 increased steadily over the monitored duration of 1 h, despite quantitative conversion (see Figure
A6). During the second run, the concentration of eluted product 2 reached a constant plateau within
5 – 10 min. We assume this observed effect is based on the retention of compounds by the highly
porous Al2O3 support.
4. Conclusions
We have developed a continuous hydrogenation process using a heterogeneous palladium catalyst
immobilized on static mixing elements in a highly porous Al2O3 support. The presented catalytic static
mixers are very stable under intensified reaction conditions (> 100 °C, 25 bar), indicated by high
catalytic activity over time, constant weight and essentially no metal leaching (confirmed by ICP-MS
measurements). During the optimization process, the influence of pressure, temperature and flow
rates of both the gas and the liquid flow rates were investigated using the hydrogenation of 1-
nitronaphthalene (1) to 1-naphthylamine (2) as model. Only a minor influence of the hydrogen flow
rate was observed within the investigated range, pointing to a parallel flow regime in the reactor.
When the amount of CSMs in the reactor was increased from 1 to 4 units, the corresponding increase
in the liquid flow rate resulted in excellent conversions > 99%. This effect was ascribed to the increased
mass transfer at elevated liquid throughputs. To demonstrate the applicability of the developed
process using CSMs for sustainable hydrogenations in pharmaceutical industry, 4-chloro-2-nitroaniline
(3) was reduced to 4-chlorobenzene-1,2-diamine (4), a key intermediate in the synthesis of the drug

tizanidine. The process was operated for 8 h, yielding 96% of the desired product after simple removal
of the solvent.
The lab-scale continuous flow reactor utilized, featuring rectangular cross-section reaction channels,
has the advantage of a very fast transfer to pilot and to production scales with minimal re-optimization
of the reaction conditions. This is achieved by the ‘smart dimensioning’ concept of the reactor design,
and the excellent heat transfer than can be achieved in rectangular cross-section channels, typically
superior to e.g. circular cross-section channels [52,53]. This scale up concept has already been
demonstrated for a very exotermic g/l alkoxylation reaction involving ethylene oxide, in which
improved product quality, yield and safety were achieved by transferring the process from batch to
flow on ultimately 10.000 t/a scale [57].
Author credt statement
This is optional according to the guidelines and not a requirement!!
Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.
Acknowledgments
The CCFLOW Project (Austrian Research Promotion Agency FFG No. 862766) is funded through the
Austrian COMET Program by the Austrian Federal Ministry of Transport, Innovation and Technology
(BMVIT), the Austrian Federal Ministry for Digital and Economic Affairs (BMDW), and by the State of
Styria (Styrian Funding Agency SFG). This work was funded by the Austrian Research Promotion Agency
FFG No. 871458, within the program “Produktion der Zukunft”. The Authors thank Prof. Goessler from
the University of Graz for conducting ICP-MS measurements and Mr. Mark Greaves from CSIRO
Manufacturing for his assistance in SEM and EDX analysis
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in the online version, at
https:/doi.org/10.1016/j.cattod.xxxxxx

References
[1]
G. Ertl, H. Knözinger, F. Schüth, J. Weitkamp, eds., Handbook of Heterogeneous Catalysis,
Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2008.
https://doi.org/10.1002/9783527610044.
[2]
[3]
G. Rothenberg, Catalysis, Wiley, 2008. https://doi.org/10.1002/9783527621866.
O. Deutschmann, H. Knözinger, K. Kochloefl, T. Turek, Ullmann’s Encycl. Ind. Chem. -
Heterogeneous Catalysis and Solid Catalysts, 2. Development and Types of Solid Catalysts,
Ullmann’s Encycl. Ind. Chem. (2012) 223–269. https://doi.org/10.1002/14356007.o05.
[4]
J. Hagen, Heterogeneously Catalyzed Processes in Industry, in: Ind. Catal., Wiley-VCH Verlag
GmbH & Co. KGaA, Weinheim, Germany, 2015: pp. 261–298.
https://doi.org/10.1002/9783527684625.ch8.
[5]
[6]
P. Unnikrishnan, D. Srinivas, Heterogeneous Catalysis, in: Ind. Catal. Process. Fine Spec.
Chem., Elsevier, 2016: pp. 41–111. https://doi.org/10.1016/B978-0-12-801457-8.00003-3.
R.A. Sheldon, Catalysis: The Key to Waste Minimization, J. Chem. Technol. Biotechnol. 68
(1997) 381–388. https://doi.org/10.1002/(SICI)1097-4660(199704)68:4<381::AID-
JCTB620>3.0.CO;2-3.
[7]
[8]
[9]
R.A. Sheldon, Atom efficiency and catalysis in organic synthesis, Pure Appl. Chem. 72 (2000)
1233–1246. https://doi.org/10.1351/pac200072071233.
R.A. Sheldon, The e Factor: Fifteen years on, Green Chem. 9 (2007) 1273–1283.
https://doi.org/10.1039/b713736m.
R.A. Sheldon, Fundamentals of green chemistry: Efficiency in reaction design, Chem. Soc. Rev.
41 (2012) 1437–1451. https://doi.org/10.1039/c1cs15219j.
[10] R.L. Augustine, Selective heterogeneously catalyzed hydrogenations, Catal. Today. 37 (1997)
419–440. https://doi.org/10.1016/S0920-5861(97)00025-4.
[11] G. V. Smith, F. Notheisz, Heterogeneous Catalysis in Organic Chemistry, 1999.
[12] S. Nishimura, Handbook of heterogeneous catalytic hydrogenation for organic synthesis,
Wiley-Interscience, New York, 2001. https://www.wiley.com/en-
us/Handbook+of+Heterogeneous+Catalytic+Hydrogenation+for+Organic+Synthesis-p-
9780471396987.

[13] H.P. Hofmann, Multiphase Catalytic Packed-Bed Reactors, Catal. Rev. 17 (1978) 71–117.
https://doi.org/10.1080/03602457808080879.
[14] V. Hessel, P. Angeli, A. Gavriilidis, H. Löwe, Gas−Liquid and Gas−Liquid−Solid Microstructured
Reactors: Contacting Principles and Applications, Ind. Eng. Chem. Res. 44 (2005) 9750–9769.
https://doi.org/10.1021/ie0503139.
[15] M.P. Dudukovic, Frontiers in reactor engineering, Science 325 (2009) 698–701.
https://doi.org/10.1126/science.1174274.
[16] M.N. Kashid, A. Renken, Three-Phase Systems, in: Microstruct. Devices Chem. Process., Wiley-
VCH Verlag GmbH & Co. KGaA, Weinheim, Germany, 2014: pp. 331–350.
https://doi.org/10.1002/9783527685226.ch8.
[17] M. Duduković, P. Mills, Scale-up and multiphase reaction engineering, Curr. Opin. Chem. Eng.
9 (2015) 49–58. https://doi.org/10.1016/j.coche.2015.08.002.
[18] K.F. Jensen, Microreaction engineering — is small better?, Chem. Eng. Sci. 56 (2001) 293–303.
https://doi.org/10.1016/S0009-2509(00)00230-X.
[19] J.Y. Volker Hessel, Albert Renken, Jaap C. Schouten, Micro Process Engineering, WILEY‐VCH,
2009. https://www.wiley.com/en-
us/Micro+Process+Engineering%3A+A+Comprehensive+Handbook%2C+3+Volume+Set-p-
9783527315505.
[20] B. Gutmann, D. Cantillo, C.O. Kappe, Continuous-Flow Technology-A Tool for the Safe
Manufacturing of Active Pharmaceutical Ingredients, Angew. Chem. Int. Ed. 54 (2015) 6688–
6728. https://doi.org/10.1002/anie.201409318.
[21] C.J. Mallia, I.R. Baxendale, The Use of Gases in Flow Synthesis, Org. Process Res. Dev. 20
(2016) 327–360. https://doi.org/10.1021/acs.oprd.5b00222.
[22] M.B. Plutschack, B. Pieber, K. Gilmore, P.H. Seeberger, The Hitchhiker’s Guide to Flow
Chemistry, Chem. Rev. 117 (2017) 11796–11893.
https://doi.org/10.1021/acs.chemrev.7b00183.
[23] L. Kiwi-Minsker, A. Renken, Microstructured reactors for catalytic reactions, Catal. Today. 110
(2005) 2–14. https://doi.org/10.1016/j.cattod.2005.09.011.
[24] M. Irfan, T.N. Glasnov, C.O. Kappe, Heterogeneous catalytic hydrogenation reactions in
continuous-flow reactors, ChemSusChem. 4 (2011) 300–316.
https://doi.org/10.1002/cssc.201000354.

[25] R. Munirathinam, J. Huskens, W. Verboom, Supported catalysis in continuous-flow
microreactors, Adv. Synth. Catal. 357 (2015) 1093–1123.
https://doi.org/10.1002/adsc.201401081.
[26] J. Yue, Multiphase flow processing in microreactors combined with heterogeneous catalysis
for efficient and sustainable chemical synthesis, Catal. Today. 308 (2018) 3–19.
https://doi.org/10.1016/j.cattod.2017.09.041.
[27] R. V. Jones, L. Godorhazy, N. Varga, D. Szalay, L. Urge, F. Darvas, Continuous-flow high
pressure hydrogenation reactor for optimization and high-throughput synthesis, J. Comb.
Chem. 8 (2006) 110–116. https://doi.org/10.1021/cc050107o.
[28] T. Ouchi, C. Battilocchio, J.M. Hawkins, S. V. Ley, Process Intensification for the Continuous
Flow Hydrogenation of Ethyl Nicotinate, Org. Process Res. Dev. 18 (2014) 1560–1566.
https://doi.org/10.1021/op500208j.
[29] D.C. Fabry, S. Heddrich, E. Sugiono, M.A. Liauw, M. Rueping, Adaptive and automated system-
optimization for heterogeneous flow-hydrogenation reactions, React. Chem. Eng. 4 (2019)
1486–1491. https://doi.org/10.1039/C9RE00032A.
[30] A. Carangio, L.J. Edwards, E. Fernandez-Puertas, J.F. Hayes, M. Kucharski, G.W. Rutherford,
K.M.P. Wheelhouse, G.D. Williams, Evaluation of Sponge Metal Catalysts in a Trickle Bed
Reactor for the Continuous Hydrogenation of an Aliphatic Nitro Intermediate, Org. Process
Res. Dev. (2020) acs.oprd.9b00447. https://doi.org/10.1021/acs.oprd.9b00447.
[31] ThalesNano H-Cube Series, (2020). https://thalesnano.com/products-and-services/h-cube-
pro/ (accessed January 23, 2020).
[32] HEL GROUP FlowCAT, (2020). https://www.helgroup.com/reactor-systems/hydrogenation-
catalysis/flowcat/ (accessed January 23, 2020).
[33] J. Kobayashi, A Microfluidic Device for Conducting Gas-Liquid-Solid Hydrogenation Reactions,
Science 304 (2004) 1305–1308. https://doi.org/10.1126/science.1096956.
[34] C.H. Hornung, B. Hallmark, M.R. Mackley, I.R. Baxendale, S. V. Ley, A palladium wall coated
microcapillary reactor for use in continuous flow transfer hydrogenation, Adv. Synth. Catal.
352 (2010) 1736–1745. https://doi.org/10.1002/adsc.201000139.
[35] H. Feng, X. Zhu, R. Chen, Q. Liao, J. Liu, L. Li, High-performance gas-liquid-solid microreactor
with polydopamine functionalized surface coated by Pd nanocatalyst for nitrobenzene
hydrogenation, Chem. Eng. J. 306 (2016) 1017–1025.

https://doi.org/10.1016/j.cej.2016.08.011.
[36] K.K. Yeong, A. Gavriilidis, R. Zapf, V. Hessel, Catalyst preparation and deactivation issues for
nitrobenzene hydrogenation in a microstructured falling film reactor, Catal. Today. 81 (2003)
641–651. https://doi.org/10.1016/S0920-5861(03)00162-7.
[37] K. Kin Yeong, A. Gavriilidis, R. Zapf, V. Hessel, Experimental studies of nitrobenzene
hydrogenation in a microstructured falling film reactor, Chem. Eng. Sci. 59 (2004) 3491–3494.
https://doi.org/10.1016/J.CES.2004.04.022.
[38] T.H. Rehm, C. Berguerand, S. Ek, R. Zapf, P. Löb, L. Nikoshvili, L. Kiwi-Minsker, Continuously
operated falling film microreactor for selective hydrogenation of carbon-carbon triple bonds,
Chem. Eng. J. 293 (2016) 345–354. https://doi.org/10.1016/j.cej.2016.02.081.
[39] N. Nikbin, M. Ladlow, S. V. Ley, Continuous Flow Ligand-Free Heck Reactions Using Monolithic
Pd [0] Nanoparticles, Org. Process Res. Dev. 11 (2007) 458–462.
https://doi.org/10.1021/op7000436.
[40] P. He, S.J. Haswell, P.D.I. Fletcher, S.M. Kelly, A. Mansfield, Scaling up of continuous-flow,
microwave-assisted, organic reactions by varying the size of Pd-functionalized catalytic
monoliths, Beilstein J. Org. Chem. 7 (2011) 1150–1157. https://doi.org/10.3762/bjoc.7.133.
[41] M. Braun, U.D. Thach, B. Prelot, P. Hesemann, D. Esposito, Pd@ionosilica as heterogeneous
hydrogenation catalyst for continuous flow reductive upgrade of cinnamaldehyde, J. Chem.
Technol. Biotechnol. 92 (2017) 2229–2235. https://doi.org/10.1002/jctb.5278.
[42] A. Nagaki, K. Hirose, Y. Moriwaki, M. Takumi, Y. Takahashi, K. Mitamura, K. Matsukawa, N.
Ishizuka, J. Yoshida, Suzuki–Miyaura Coupling Using Monolithic Pd Reactors and Scaling-Up by
Series Connection of the Reactors, Catalysts. 9 (2019) 300.
https://doi.org/10.3390/catal9030300.
[43] O. Verho, K.P.J. Gustafson, A. Nagendiran, C.W. Tai, J.E. Bäckvall, Mild and selective
hydrogenation of nitro compounds using palladium nanoparticles supported on amino-
functionalized mesocellular foam, ChemCatChem. 6 (2014) 3153–3159.
https://doi.org/10.1002/cctc.201402488.
[44] J.-N. Tourvieille, R. Philippe, C. de Bellefon, Milli-channel with metal foams under an applied
gas–liquid periodic flow: Flow patterns, residence time distribution and pulsing properties,
Chem. Eng. Sci. 126 (2015) 406–426. https://doi.org/10.1016/j.ces.2014.11.059.
[45] J. Lévêque, R. Philippe, M.-L. Zanota, V. Meille, F. Sarrazin, L. Baussaron, C. de Bellefon,

Hydrodynamics and mass transfer in a tubular reactor containing foam packings for
intensification of G-L-S catalytic reactions in co-current up-flow configuration, Chem. Eng. Res.
Des. 109 (2016) 686–697. https://doi.org/10.1016/j.cherd.2016.03.017.
[46] Y. Elias, P. Rudolf von Rohr, W. Bonrath, J. Medlock, A. Buss, A porous structured reactor for
hydrogenation reactions, Chem. Eng. Process. Process Intensif. 95 (2015) 175–185.
https://doi.org/10.1016/j.cep.2015.05.012.
[47] A. Avril, C.H. Hornung, A. Urban, D. Fraser, M. Horne, J.-P. Veder, J. Tsanaktsidis, T.
Rodopoulos, C. Henry, D.R. Gunasegaram, Continuous flow hydrogenations using novel
catalytic static mixers inside a tubular reactor, React. Chem. Eng. 2 (2017) 180–188.
https://doi.org/10.1039/C6RE00188B.
[48] C.H. Hornung, X. Nguyen, A. Carafa, J. Gardiner, A. Urban, D. Fraser, M.D. Horne, D.R.
Gunasegaram, J. Tsanaktsidis, Use of Catalytic Static Mixers for Continuous Flow Gas–Liquid
and Transfer Hydrogenations in Organic Synthesis, Org. Process Res. Dev. 21 (2017) 1311–
1319. https://doi.org/10.1021/acs.oprd.7b00180.
[49] X. Nguyen, A. Carafa, C.H. Hornung, Hydrogenation of vinyl acetate using a continuous flow
tubular reactor with catalytic static mixers, Chem. Eng. Process. - Process Intensif. 124 (2018)
215–221. https://doi.org/10.1016/j.cep.2017.12.007.
[50] J. Gardiner, X. Nguyen, C. Genet, M.D. Horne, C.H. Hornung, J. Tsanaktsidis, Catalytic Static
Mixers for the Continuous Flow Hydrogenation of a Key Intermediate of Linezolid (Zyvox), Org.
Process Res. Dev. 22 (2018) 1448–1452. https://doi.org/10.1021/acs.oprd.8b00153.
[51] C.H. Hornung, S. Singh, S. Saubern, Additive Layer Manufacturing of Catalytic Static Mixers for
Continuous Flow Reactors, Johnson Matthey Technol. Rev. 62 (2018) 350–360.
https://doi.org/10.1595/205651318X696846.
[52] A.-L. Sengen, F. Herbstritt, M. Grünewald, J. Heck, Experimentelle Bestimmung des
konvektiven Wärmeübergangs in einem mikrostrukturierten Kanal, Chemie Ing. Tech. 89
(2017) 379–389. https://doi.org/10.1002/cite.201600163.
[53] S. Lier, J. Riese, G. Cvetanoska, A.K. Lesniak, S. Müller, S. Paul, L. Sengen, M. Grünewald,
Innovative scaling strategies for a fast development of apparatuses by modular process
engineering, Chem. Eng. Process. - Process Intensif. 123 (2018) 111–125.
https://doi.org/10.1016/J.CEP.2017.10.026.
[54] N. Kockmann, Modular Equipment for Chemical Process Development and Small-Scale
Production in Multipurpose Plants, ChemBioEng Rev. 3 (2016) 5–15.

https://doi.org/10.1002/cben.201500025.
[55] M.J. Begall, A. Krieger, S. Recker, F. Herbstritt, A. Mhamdi, A. Mitsos, Reducing the Fouling
Potential in a Continuous Polymerization Millireactor via Geometry Modification, Ind. Eng.
Chem. Res. 57 (2018) 6080–6087. https://doi.org/10.1021/acs.iecr.8b00206.
[56] A. Sengen, F. Herbstritt, J. Heck, M. Grünewald, Einfluss mikrostrukturierter statischer Mischer
auf den Gas/Flüssig‐Stofftransport in einem engen Rechteckkanal, Chemie Ing. Tech. 91
(2019) 614–621. https://doi.org/10.1002/cite.201800187.
[57] Millireactor in Production, Ehrfeld Mikrotechnik BTS GmbH, (2016).
https://www.chemanager-online.com/en/topics/thermal-processes/millireactor-production
(accessed February 2, 2020).
[58] W. E. Frazier, Metal Additive Manufacturing: A Review, J. Mater. Eng. Perform. 23 (2014)
1917-1928. https://doi.org/10.1007/s11665-014-0958-z.
[59] Conflux Technology, (2020). https://www.confluxtechnology.com/ (accessed February 24,
2020).
[60] D. Prat, A. Wells, J. Hayler, H. Sneddon, C.R. McElroy, S. Abou-Shehada, P.J. Dunn, CHEM21
selection guide of classical- and less classical-solvents, Green Chem. 18 (2015) 288–296.
https://doi.org/10.1039/c5gc01008j.
[61] A.J. Wagstaff, H.M. Bryson, Tizanidine - A review of its pharmacology, clinical efficacy and
tolerability in the management of spasticity associated with cerebral and spinal disorders,
Drugs. 53 (1997) 435–452. https://doi.org/10.2165/00003495-199753030-00007.
[62] Z. Chen, Green new process for preparing tizanidine hydrochloride, CN102140095A, 2010.
https://worldwide.espacenet.com/patent/search/family/044407912/publication/CN1021400
95A?q=CN102140095A.