Remarkable catalytic activity of polymeric membranes containing gel-trapped palladium nanoparticles for hydrogenation reactions

Article abstract of DOI:10.1016/j.cattod.2020.04.027

Polymeric flat-sheet membranes and hollow fibers were prepared via UV photo-initiated polymerization of acrylic acid at the surface of commercial polyether sulfones (PES) membranes. These polymeric materials permitted to immobilize efficiently palladium nanoparticles (PdNP), which exhibited a mean diameter in the range of 4?6 nm. These materials were synthesized by chemical reduction of Pd(II) precursors in the presence of the corresponding support. We successfully applied the as-prepared catalytic materials in hydrogenation reactions under continuous flow conditions. Flat sheet membranes were more active than hollow fibers due to the flow configuration and defavorable operating conditions. Actually, various functional groups (i.e. C[dbnd]C, C[tbnd]C and NO₂) were reduced in flow-through configuration, under mild conditions (between 1.4 and 2.2 bar H₂ at 60 °C, using 3.2 mol% of Pd loading), archiving high conversions in short reaction times (12?24 s).

Full text of DOI:10.1016/j.cattod.2020.04.027

CatalysisTodayxxx(xxxx)xxx–xxx
Contents lists available at ScienceDirect
Catalysis Today
journal homepage: www.elsevier.com/locate/cattod
Remarkable catalytic activity of polymeric membranes containing gel-
trapped palladium nanoparticles for hydrogenation reactions
Melissa López-Viverosa,*, Isabelle Favierb, Montserrat Gómezb, Jean-François Lahittea,
Jean-Christophe Remigya,*
a Laboratoire de Génie Chimique, Université de Toulouse, CNRS, INPT, UPS, Toulouse, France
b Laboratoire Hétérochimie Fondamentale et Appliquée, UMR CNRS 5069, Université de Toulouse 3 – Paul Sabatier, 118 route de Narbonne, 31062, Toulouse Cedex 9,
France
A R T I C L E I N F O
A B S T R A C T
Keywords:
Polymeric flat-sheet membranes and hollow fibers were prepared via UV photo-initiated polymerization of ac-
rylic acid at the surface of commercial polyether sulfones (PES) membranes. These polymeric materials per-
mitted to immobilize efficiently palladium nanoparticles (PdNP), which exhibited a mean diameter in the range
of 4−6 nm. These materials were synthesized by chemical reduction of Pd(II) precursors in the presence of the
corresponding support. We successfully applied the as-prepared catalytic materials in hydrogenation reactions
under continuous flow conditions. Flat sheet membranes were more active than hollow fibers due to the flow
configuration and defavorable operating conditions. Actually, various functional groups (i.e. C]C, C^C and
NO2) were reduced in flow-through configuration, under mild conditions (between 1.4 and 2.2 bar H2 at 60 °C,
using 3.2 mol% of Pd loading), archiving high conversions in short reaction times (12−24 s).
Catalytic membrane
Palladium nanoparticles
Hydrogenations
Flat sheet membranes
Hollow fiber contactor
1. Introduction
unsaturated organic compounds [18–27] including the reduction of 4-
nitrophenol [28–32], among others). In polymeric catalytic mem-
Palladium is the metal of preference for many catalyzed organic
reactions, due to its ability in promoting a wide range of transforma-
tions (C-halide and CeH bond activations, carbonylations (CO activa-
tion), hydrogenations (H2 activation), etc.) [1]. This versatile reactivity
is mainly attributed to the palladium ability for making different types
of well-defined metallic species: molecular complexes, small clusters,
nanoparticles, and extended surfaces [2,3]. Thus, appropriate tuning on
the nature of ligands or stabilizers, and reaction conditions, leads to
structures at molecular and nanometric scale [4]. Furthermore, sup-
ported palladium catalysts have been largely studied and efficiently
applied for synthesis of both fine [5] and bulk chemicals [6]. Especially,
working under continuous flow conditions represents a sustainable tool
in terms of safety, diminution of reaction time, process intensification
and scale-up [7,8]. In this frame, polymeric catalytic membranes have
proven to be outstanding tools for catalytic reactions in the fine che-
mical industry, generally working under mild conditions (< 150 °C)
[9–12]. Efficient catalytic membranes containing metal nanoparticles
(MNP) have been reported for different reactions, mainly involving
noble metals such as Pd, Au and Ag (CeC coupling reactions [13,14],
hydrogenation of nitrates in water [15–17], hydrogenation of
branes, the catalyst can be incorporated in all the membrane (during or
after its synthesis) or in a grafted polymeric layer at the surface of the
membrane which stabilized the MNP by steric or electro-steric effect.
This membrane modification is carried out by photo-polymerization in
order to obtain polymer chains covalently linked to the membrane
surface. The use of a cross linker during this step permits to control the
porosity and swelling properties of the layer, leading to a polymeric gel.
MNP can be prepared via an ex-situ or in-situ method. The in-situ method
is based in two steps, an impregnation of the grafted layer by a metallic
salt solution following by a reduction step. This last strategy is used in
this work permitting the use of commercially available membranes,
leading to a high experiment repeatability and high local MNP con-
centration at the membrane surface without aggregation. This high
local concentration is the key aspect to obtain high conversions and the
most relevant advantage comparing to other catalytic supports [33]. As
they can be produced as either flat sheet or hollow fiber membranes,
high production capacity in the order of hundred or thousand ton/year/
m3 are achievable, in particular with hollow fiber membrane module
[28].
Regarding the possible configurations for catalytic membrane
⁎ Corresponding authors.
E-mail addresses: melilopezv@gmail.com (M. López-Viveros), remigy@chimie.ups-tlse.fr (J.-C. Remigy).
https://doi.org/10.1016/j.cattod.2020.04.027
Received 1 October 2019; Received in revised form 17 March 2020; Accepted 6 April 2020
0920-5861/©2020ElsevierB.V.Allrightsreserved.
Pleasecitethisarticleas:MelissaLópez-Viveros,etal.,CatalysisToday,https://doi.org/10.1016/j.cattod.2020.04.027

M. López-Viveros, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
reactors, an important increase of the catalytic efficiency of the mem-
branes has been already achieved before by our group when using
membranes in forced flow-through configuration for the Suzuki-
Miyaura CeC cross-coupling reaction compared to batch configuration
under the same conditions [33,34]. Thus, we aim presenting here the
results obtained in hydrogenation reactions using polymeric catalytic
membranes, which contain immobilized PdNP, under a forced flow-
through configuration. Moreover, we report innovative results on the
hydrogenation of nitroarenes using a polymeric hollow-fiber contactor.
2. Materials and methods
2.1. Materials
Scheme 1. Experimental set-up for continuous flow hydrogenation reactions
Microfiltration PES flat sheet membranes MicroPES® were pur-
chased from Membrana-3 M (Wuppertal, Germany) with a nominal
pore size of 0.2 μm. Microfiltration MicroPES® hollow fibers were
purchased from Membrana-3 M (Wuppertal, Germany), nominal pore
using flat sheet membranes in flow-through configuration.
For PdNP characterization, grafted membrane samples were first
embedded in resin and then thin slices (80 nm) were cut with a mi-
crotome. Micrographs of PdNP were obtained by Transmission Electron
Microscopy (TEM, JEOL Jem 1400). For each membrane, at least 10
slices were analyzed. Size distribution of PdNP were determined using
Image J software [38].
size of 0.2 μm, inner diameter of 300
40 μm and wall thickness of
100
25 μm. All chemical reagents were obtained from Sigma-Aldrich
and used without further purification.
2.2. Methods
2.2.3. Catalytic experiments
2.2.1. Membrane surface modification
Surface of flat sheet PES membranes (12.4 cm2) was functionalized by
UV photo-grafting radical polymerization using a Heraeus TQ 150 lamp
with a quartz filter. Membranes were dipped into monomer solutions
containing 5 wt% of acrylic acid as monomer and 1.5 wt% of diethylene-
di-acrylate as crosslinker (0.1 mol crosslinker /mol of monomer) and then
irradiated for 3 min (total energy dose received 6.3 J cm−2).
2.2.3.1. Continuous flow hydrogenations using catalytic flat-sheet
membranes. Reactions were carried out in flow-through configuration
with the as-prepared catalytic flat sheet membranes. Solutions
containing the substrate were pre-saturated with H2 using a hollow-
fiber contactor. The Gas/Liquid hollow-fiber contactor was designed to
assure H2 saturation of solutions [39].
The experimental set-up developed to perform the hydrogenation
reactions in continuous mode is presented in Scheme 1.
Surface of hollow PES membranes was functionalized by UV photo-
grafting radical polymerization as previously described [35,36] using
two high power Mercury lamps, UVA-PRINT LE, Hoenle UV France,
with a line rate of 10 m min−1 and an intensity of the lamps of 110 %
(total energy dose received 21.3 J cm-2). Monomer solutions containing
20 wt% of acrylic acid as monomer and 0.85 wt% of N,N’-methylenebis
(acrylamide) as crosslinker (0.027 mol crosslinker /mol of monomer).
The change on the cross-linker was simply due to follow a previously
established protocol for hollow fibers. Although both cross-linkers have
different nature, the polymer gel grafted on the membrane presented
the same swelling behavior in water and ethanol (probably due to si-
milar short distances between C = C, for diethylene-diacrylate: the
C = C to C = C distance is 1.012 nm; for bisacrylamide, the C = C to
C = C distance is 0.859 nm, according to molview modelling).
A home-made stainless-steel filtration cell, equipped with a mag-
netic stir bar, was placed on a hot-plate at 65 °C. The flow rate of
ethanol solution containing the substrate was controlled by a gear
pump (Scheme 1, (1)). The solution was sent into to a home-made
hollow fiber contactor module in counter-current to saturate the solu-
tion with H2 using a hydrogen generator (Schmidlin-FDBS). The home-
made hollow fiber contactor module contained 15 fibers, an effective
length of 0.3 m and 4.0 10−3 m of diameter. The H2 saturated solution
(Scheme 1, (2)) went into the filtration cell where it flowed through the
membrane. 25 ml of permeate were taken for further analyses after
attaining steady-state (established when passing 4 times the total dead
volume: reactor volume + connection pipes volume, ca. 4⨯85 mL).
The reaction time is expressed as the residence time
actant inside the membrane using the following equation:
of the re-
To evaluate the functionalization of the grafted membranes, ATR-
FTIR was performed using a Thermo-Nicolet Nexus 670 spectro-
photometer in the region of 400 to 4000 cm−1 and SEM images were
obtained with Tabletop Microscope Phenom XL after cryofracturation.
L
F
=
S
(1)
with L as the membrane thickness (110·10−4 cm), F the permeate
2.2.2. In-situ synthesis of palladium nanoparticles and their
characterization
flow rate (mL s-1) and S the membrane surface area (12.4 cm2). The
membrane porosity was taken as 0.8 [40].
The synthesis of PdNP was carried out based on the intermatrix
method [37]. Grafted membranes were immerged at room temperature
in a 0.02 M aqueous solution of [Pd(NH3)4]Cl2, H2O for 18 h. PdNP
were then formed by reduction of Pd(II) ions through immersion of the
membrane in a 0.1 M aqueous solution of NaBH4 for 1 h. Membranes
were thoroughly washed (30 ml of ethanol for 1 h each time) and
stocked in ethanol before use.
If only the grafted layer of thickness l, where the palladium nano-
particles are found, is taken into account for the calculation, the re-
sidence time is defined as:
l
F
*
=
S
(2)
Palladium content was determined by Inductively Coupled Plasma
Optical Emission Spectrometry analyses (ICP-OES, Ultima 2, Horoba
Jobin Yvon). The ICP-OES detection limit for palladium is 3 ppb.
Palladium loading is expressed as μg of Pd per cm²; for flat sheet
membranes, the mean value was obtained for at least 3 membrane
samples of 1 cm² of surface; for hollow fibers, the mean value was
obtained for at least 3 membrane samples of 2 cm length.
Permeates (Scheme 1, (3)) were analyzed by gas chromatography (GC)
using an Agilent GC6890 instrument with a flame ionization detector
(FID), coupled to a Perkin Elmer Clarus MS56 mass spectrometer (MS),
with a SGE BPX5 column composed by 5% of phenylmethylsiloxane and
by 1H NMR using a Bruker Avance-III 300 (300 MHz) spectrometer. De-
cane was used as internal standard. The palladium content in the filtered
solutions and purified product was determined by ICP-OES.
2

M. López-Viveros, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Scheme 2. Experimental set-up for the continuous flow hydrogenation reac-
Fig. 1. ATR-FTIR spectra of grafted (-) and pristine (—) MicroPES® flat sheet
tions using a catalytic hollow fiber contactor.
membranes.
2.2.3.2. Continuous flow hydrogenations using G/L catalytic hollow fiber
membrane contactor. The flow of the solution containing the substrate
was controlled by a gear pump (Scheme 2, (1)) and sent it to the home-
made G/L catalytic hollow fiber contactor module where the reaction
takes place. Reaction solutions passed in the outer side of the hollow
fibers while the H2 produced by the hydrogen generator (Schmidlin-
FDBS) passed in the inner side of the hollow fibers. Reactions were
performed at room temperature. 15 ml of permeate (Scheme 2, (2))
were taken for further analyses after attaining steady-state (established
corresponds to the asymmetric OeH stretching for the modified mem-
brane, attributed to both PAA and water trapped into the highly hy-
drophilic PAA layer. Spectrum of the grafted hollow fiber membrane
exhibited the same absorption bands that the flat-sheet grafted mem-
brane (see Fig. S1 in the Supplementary Information), pointing to the
same type of functionalization.
Membranes were also analyzed by SEM to estimate the thickness of
the grafted layers, thus corroborating the surface functionalization of
the pristine membrane. Actually, SEM images of the cross-section of the
modified membrane clearly showed a grafted PAA layer at the mem-
brane surface for both flat sheets and hollow fibers (Fig. 2).
It could be observed that both membrane supports (flat sheet and
hollow fiber) show highly porous structures. For the flat sheet mem-
brane (Fig. 2a), the mean thickness of the PAA grafted layer was esti-
mated to be ca. 3.1 μm and for the hollow fiber membrane (Fig. 2b), ca.
5.6 μm.
when passing
4
times the total dead volume: reactor
volume + connection pipes volume, ca. 4⨯5 ml).
For the G/L catalytic hollow fiber contactor, the reaction time is
expressed as the residence time
of the solution inside the mem-
G/L
brane using the following equation:
2
Lcont [rCont
nfib re2]
V
Ql
Cont
=
=
G/L
Ql
(3)
Where VCont is the volume of the contactor shell side (m3) , Ql the vo-
lumetric flow rate of the liquid phase (m3⋅s−1), Lcont the lengh of the
module (0.1 m), rCont is the radius of the module (0.002 m), nfib the
number of fibers (8), re the external diameter of fibers (3.0⋅10-4 m).
Products were analyzed as described in Section 2.2.3.1.
3.2. In-situ synthesis of palladium nanoparticles and their characterization
The PAA grafted gels on both supports, flat sheets and hollow fibers,
swelled by aqueous solutions, offered a large space between crosslinked
network which served for nucleation and further growth of PdNP
2.2.3.3. Catalytic hydrogenations under batch configuration. Reactions
were carried on in a Top Industrie autoclave. The catalytic membrane
was maintained on a glass support. The substrate (0.25 mmol) was
dissolved in 25 ml of ethanol. Then, the reactor was purged with argon
and pressurized with H2 (20 bar), heated at 60 °C in an oil bath and
magnetically stirred for 16 h. Products were analyzed as described in
Section 2.2.3.1.
3. Results and discussion
3.1. Membrane surface modification
ATR-FTIR analyses permitted to elucidate the functional groups of
the PES-based membranes after photo-grafting polymerization of ac-
rylic acid. In Fig. 1, the spectra of the flat-sheet membranes before and
after photo-polymerization are presented.
The most representative bands of the PES membrane were
1578 cm−1 and 1486 cm−1 which correspond to C]C bond stretching
of the aromatic groups of PES membrane [41]. Other characteristic
absorption bands attributed to SO2 (i.e. asymmetric and symmetric
stretching, 1300 and 1150 cm−1 respectively) were also observed. The
analysis performed on the modified membrane showed an absorption
band at 1740 cm−1 attributed to the C]O vibration of the carboxylic
acid group of polyacrylic acid (PAA). A broad peak below 2100 cm−1
Fig. 2. SEM cross-section micrographs of: a) functionalized flat-sheet mem-
brane and b) functionalized hollow fibers.
3

M. López-Viveros, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Fig. 3. TEM micrographs of PdNP in the grafted PAA layer on a) Flat sheet membrane and b) hollow fiber.
[[42]]. The in-situ synthesis of PdNP within the PAA was carried out in
aqueous medium by the treatment of the metal precursor [Pd
(NH3)4]Cl2 followed by the addition of the reducing agent, NaBH4
[43,44]. The as-prepared materials were characterized by TEM (Fig. 3).
For both supports, flat sheet membranes and hollow fibers, the forma-
tion of PdNP was evidenced showing a good dispersion and quasi
spherical shape. These PdNP are efficiently immobilized on the mem-
branes within the PAA grafted gel.
conversion after recycling 3 times (x12 s) or 36 s of residence time;
entry 2) with a low chemoselectivity, mainly obtaining isomerization
products (i.e. internal alkenes), what indicates that this catalytic system
selectively hydrogenate terminal but not internal C]C bonds; actually,
for limonene, the hydrogenation was selective towards the reduction of
the exocyclic C]C bond but with a low conversion (8% conversion for
24 s of residence time; entry 3). In agreement with this behavior, the
internal C]C bond of ethyl oleate could not be reduced (entry 4). The
hydrogenation of the C^C bond of diphenyl acetylene mainly led to the
semi-hydrogenation product, i.e. (Z)-stilbene (53 % conversion in 24 s
with a (Z)-stilbene/1,2-diphenylethane ratio = 80/20; entry 5). This
catalytic system was also efficient for the reduction of NO2 of ni-
trobenzene (41 % conversion in 12 s, entry 6); but, for nitrobenzene
containing an electron-withdrawing group, such as 4-nitrobenzonitrile,
the activity dramatically decreased (21 % conversion, entry 7). Un-
fortunately, the reduction of aromatic aldehydes and ketones (such as
benzaldehyde and acetophenone), and nitrile derivatives (benzonitrile)
did not work under flow conditions even after recirculating the
permeate, in contrast to batch configuration.
Regarding PdNP incorporated in the grafted flat sheets (Fig. 3a),
two zones could be identified: (i) Zone 1 corresponds to the grafted gel
where PdNP are mostly well-dispersed with only few aggregates and (ii)
Zone 2 corresponds to the membrane pores where PdNP form large
aggregates. Without considering the aggregates, the mean diameter of
PdNP was estimated to be of 5.5
1.8 nm. Palladium loading was of
68 μg cm−2 (i.e. 0.008 mmol Pd in a flat sheet membrane of 12.4 cm2).
For the hollow fibers (Fig. 3b), the mean diameter of particles was
found 4.8
1.4 nm. PdNP in the hollow fiber appeared more homo-
geneously dispersed than in the flat sheet membrane. Palladium content
was 520 μg⋅ cm−2. The module prepared contains 8 fibers of 10 cm, and
then the total amount of palladium in the hollow fiber module is
3.93 mg (i.e. 0.037 mmol Pd).
No Pd leaching was observed in any of the filtered solutions (i.e. Pd
content below 3 ppb from ICP-OES analyses), which indicates that
PdNP are suitably stabilized by the PAA grafted gel, permitting an ef-
ficient recycling.
3.3. Catalytic experiments
With the aim of studying the effect of hydrogen pressure, hydro-
genations in batch mode configuration were carried out. For this pur-
pose, benzaldehyde, acetophenone and ethyl oleate were selected
(Table 2). Reactions were performed under similar conditions than
those described in flow conditions, but under 20 bar H2 pressure for
16 h (cH2, sol = 0.073 M; i.e. 10–15 times more than under flow-through
configuration). In this case, conversions were calculated based on the
substrate (limiting reagent).
3.3.1. Continuous flow hydrogenations using flat sheet membranes
For each reaction, a freshly prepared membrane was used. The total
palladium (0.008 mmol)/substrate (0.25 mmol) ratio was 0.032 (i.e.
3.2 mol% Pd), considering the whole catalytic material.
Hydrogenations in flow-through configuration were performed for
substrates containing different functional groups (Table 1): C]C bond
for both conjugated and non-conjugated compounds (entries 1–4), C^C
bond (entry 5) and NO2 group (entries 6 and 7).
Full hydrogenation of benzaldehyde to benzylic alcohol was
achieved (entry 1, Table 2); for acetophenone, moderate conversion
was obtained (46 % conversion, entry 2). However, a low conversion
was attained for ethyl oleate (10 % conversion, entry 3, Table 2). These
results proved the catalytic efficiency of the flat sheet membranes at
high pressure. Traces of Pd were detected by ICP-OES (less than 60 ppb
in the extracted organic products).
For the hydrogenation of alkene substrates, full conversion in 24 s
was obtained for the indene hydrogenation, selectively reducing the
conjugated C]C bond to the aromatic ring (entry 1), leading to a
turnover frequency ca. three times higher than other conjugated sub-
strates described in the literature [8]. However, for non-conjugated
C]C bonds, such as 1-dodecene, lower conversion was achieved (26 %
4

M. López-Viveros, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Table 1
Pd-catalyzed hydrogenation reactions in flow-through configuration in ethanol at 60 °C.
Entry Substratea
Product
PH2 [bar] (cH
[M])
Reaction timec (Liquid flow
Conv. [%] (selectivity [%])d
[TOF (s−1)]g
b
2, sol
rate [mL/min])
1
1.4 (0.0051)
8 s⨯3 (2.0)
100 (100) [1.3]
2
3
2.2 (0.0080)
1.4 (0.0051)
12 s⨯3 (0.8)
8 s⨯3 (2.0)
26 (7)e [0.2]
8 (100) [0.1]
4
5
–
1.4 (0.0051)
2.2 (0.0080)
8 s⨯3 (2.0)
8 s⨯3 (2.0)
0 (0)
53 (80)f [0.7]
6
7
2.0 (0.0073)
2.2 (0.0080)
12 s (0.8)
12 s (0.8)
41 (100) [1.07]
21 (100) [0.55]
a
b
c
0.25 mmol of substrate in 25 mL of ethanol (substrate concentration of 0.01 mol L−1). The total palladium to substrate molar ratio was 0.032.
H2 absolute pressure.
Residence time of substrates in the catalytic membrane. 8 s x 3 indicating 3 successive filtrations of 8 s each.
Conversions were calculated based on H2 concentration (see Supplementary Information) and determined by GC–MS and 1H NMR using decane as internal
d
standard.
e
f
g
Main products correspond to internal alkenes formed by isomerization of the terminal C]C bond.
In addition to (Z)-stilbene, 1,2-diphenylethane was also obtained (ca. 20 %).
TOF = turnover frequency; TOF = conversion/(Pd load*time).
Table 2
Pd-catalyzed hydrogenation reactions in batch reactor, in ethanol at 60 °C.
Entry
Substratea
Product
Conversionb [%] (yield [%])
100 (100)
1
2
3
46 (46)
10 (100)
a
b
0.25 mmol of substrate in 25 ml of ethanol (substrate concentration of 0.01 molL−1). The total palladium to substrate molar ratio was 0.032.
Conversions and yields determined by GC–MS and 1H NMR using decane as internal standard· Conversions were calculated based on the substrate.
5

M. López-Viveros, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
Table 3
Pd-catalyzed hydrogenation reactions using a G/L catalytic hollow fiber contactor in ethanol at room temperature.
Entry Substratea
Product
PH2 [bar] (cH
[M]) Reaction time (liquid flow rate [mL/min]) Conv.b [%] (selectivity[%])c TOFd (s−1
)
2, sol
1
1.4 (0.0051)
1.3 min (0.8)
23 (100)
0.72
2
a
b
c
d
1.4 (0.0051)
2.2 min (0.5)
22 (100)
0.41
0.15 mmol of substrate in 15 ml of ethanol (substrate concentration =0.01 mol L−1). Pd/substrate = 0.246.
Determined by GC–MS and 1H NMR using decane as internal standard. Conversions were calculated based on the substrate.
Selectivity towards the aniline derivative.
TOF = turnover frequency; TOF = conversion/(Pd load*time).
As indicated in Table 1, short reaction times were required to obtain
hollow fiber contactor, promising results were obtained for the hydro-
genation of nitro-arenes.
high conversions for some of the evaluated substrates (i.e. between 12 s
and 24 s) under sustainable conditions in terms of temperature (60 °C),
hydrogen pressure (1.4–2.2 bar) and solvent (ethanol). Even more, if we
In order to enlarge the applications scope for both flow modes, we
are currently working in tuning experimental conditions (e.g. increasing
temperature and/or residence time by decreasing liquid flow rate or
increasing H2 pressure; tuning the design of the contactor module with
higher packing density).
assume that the catalytic activity is only attributed to PdNP trapped on
*
the PAA grafted layer, the reaction time
is in fact smaller (i.e. τ*
between 0.34 and 0.68 s). This remarkable activity is mainly due to the
high local concentration of Pd in relation to substrate (i.e. 120 mol of Pd
per mol of substrate), while having a relatively low total Pd loading in
the membrane (i.e. 0.008 mmol of Pd). The total palladium to substrate
molar ratio was 0.032 considering the whole catalytic material.
CRediT authorship contribution statement
Melissa López-Viveros: Methodology, Investigation, Formal ana-
lysis, Conceptualization, Writing - original draft, Visualization, Writing
-
review & editing. Isabelle Favier: Methodology, Investigation,
3.3.2. Continuous flow hydrogenations using G/L catalytic hollow fibers
contactor
Conceptualization, Writing - review & editing. Montserrat Gómez:
Conceptualization, Validation, Formal analysis, Writing - review &
editing. Jean-François Lahitte: Conceptualization, Funding acquisi-
tion, Validation, Writing - review & editing. Jean-Christophe Remigy:
Conceptualization, Formal analysis, Validation, Supervision, Funding
acquisition, Writing - review & editing.
A G/L catalytic hollow fiber contactor module was settled as a proof
of concept for hydrogenation reactions. The objective was to develop
simpler, more compact and efficient process by performing in the same
device the hydrogenation reaction and the continuous supply of hy-
drogen to the catalyst surface to circumvent the low solubility of hy-
drogen in ethanol. The contact of the reacting phases should improve
using hollow fibers in contactor mode, thanks to their high specific
surface [45]. In this case, the limiting reagent is the substrate to be
hydrogenated, therefore conversions were calculated based on the
substrate concentration. The total palladium to substrate molar ratio
was 0.246.
Declaration of Competing Interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influ-
ence the work reported in this paper.
Encouraging results were obtained for the hydrogenation of NO2
group using the G/L catalytic hollow fiber contactor (Table 3). Actually,
the activity observed is as efficient as reported data using PdNPs im-
mobilized on imidazolium-functionalized polymers. [46]
Acknowledgements
The authors acknowledge the financial support of the French
Ministry of Higher Education and Research. I. F. and M. G. thanks the
Centre National de la Recherche Scientifique (CNRS) and the University
Toulouse 3 - Paul Sabatier for financial support. The contribution of
MSc. Mohammed-Lamine Ouinten for the development of the Python
code used for the design of the G/L hollow fiber modules is also ac-
knowledged.
However, it was not possible to fairly compare catalytic perfor-
mance of the G/L hollow fiber contactor (entry 1, Table 3) and the flat
sheet membrane (entry 6, Table 1), since reaction conditions greatly
differ and most of them are not favorable for the G/L catalytic hollow
fiber (lower temperature, H2 pressure, higher palladium content).
4. Conclusions
Appendix A. Supplementary data
To sum up, new polymeric materials were prepared for the im-
mobilization of metal nanoparticles, in this case PdNP. These as-pre-
pared catalytic materials were applied in hydrogenations of different
functional groups (NO2, C]C, C^C) under flow-through (flat sheet
membranes) and G/L contactor (hollow fibers) configurations. For the
first configuration, reactions worked under smooth conditions: low
hydrogen pressure (1.4–2.2 bar) and 60 °C for short residence times
(< 1 min), using relatively low Pd loading (3.2 mol%). For ketones and
aldehydes, hydrogenations were not efficient in flow-through config-
uration, but flat sheet membranes were active under batch conditions
by applying harsher conditions (20 bar H2, 16 h). For the G/L catalytic
Supplementary material related to this article can be found, in the
online version, at doi:https://doi.org/10.1016/j.cattod.2020.04.027.
References
[1] J. Tsuji, John Wiley, S. Sons (Eds.), Palladium Reagents and Catalysts, John Wiley &
Sons, Sussex, 2004.
[2] B.M. Trost, Metal Catalyzed Allylic Alkylation: Its Development in the Trost
Laboratories, Tetrahedron 71 (35) (2015) 5708–5733.
[3] I. Favier, D. Madec, E. Teuma, M. Gomez, Palladium nanoparticles applied in or-
ganic synthesis as catalytic precursors, Curr. Org. Chem. 15 (18) (2011) 3127–3174.
6

M. López-Viveros, et al.
CatalysisTodayxxx(xxxx)xxx–xxx
[4] R. Ciriminna, V. Pandarus, F. Béland, Y.-J. Xu, M. Pagliaro, Heterogeneously cat-
alyzed alcohol oxidation for the fine chemical industry, Org. Process Res. Dev. 19
(11) (2015) 1554–1558.
[25] D. Fritsch, K. Kuhr, K. Mackenzie, F.D. Kopinke, Hydrodechlorination of
Chloroorganic Compounds in Ground Water by Palladium Catalysts: Part 1.
Development of Polymer-Based Catalysts and Membrane Reactor Tests, Catal.
Today 82 (1–4) (2003) 105–118.
[5] H.-U. Blaser, A. Indolese, A. Schnyder, H. Steiner, M. Studer, Supported palladium
catalysts for fine chemicals synthesis, J. Mol. Catal. A Chem. 173 (1–2) (2001) 3–18.
[6] N. Pernicone, M. Cerboni, G. Prelazzi, F. Pinna, G. Fagherazzi, An investigation on
Pd/C industrial catalysts for the purification of terephthalic acid, Catal. Today 44
(1–4) (1998) 129–135.
[26] L. Brandão, L.M. Madeira, A.M. Mendes, Propyne hydrogenation in a continuous
polymeric catalytic membrane reactor, Chem. Eng. Sci. 62 (23) (2007) 6768–6776.
[27] D. Fritsch, K.V. Peinemann, Catalysis with homogeneous membranes loaded with
nanoscale metallic clusters and their preparation, Catal. Today 25 (3–4) (1995)
277–283.
[7] L. Vaccaro, D. Lanari, A. Marrocchi, G. Strappaveccia, Flow approaches towards
sustainability, Green Chem. 16 (8) (2014) 3680–3704.
[28] R. Hilke, N. Pradeep, P. Madhavan, U. Vainio, A.R. Behzad, R. Sougrat, S.P. Nunes,
K. Peinemann, Block copolymer hollow Fiber membranes with catalytic activity and
PH-Response, Appl. Mater. Interfaces 5 (2013) 7001–7006.
[8] A. Reina, I. Favier, E. Teuma, M. Gómez, A. Conte, L. Pichon, Hydrogenation re-
actions catalyzed by colloidal palladium nanoparticles under flow regime, AIChE J.
(2019).
[29] H. Wang, Z. Dong, C. Na, Hierarchical carbon nanotube membrane-supported gold
nanoparticles for rapid catalytic reduction of p -Nitrophenol, ACS Sustain. Chem.
Eng. 1 (7) (2013) 746–752.
[9] I.F.J. Vankelecom, Polymeric membranes in catalytic reactors, Chem. Rev. 102
(2002) 3779–3810.
[10] S.S. Ozdemir, M.G. Buonomenna, E. Drioli, Catalytic Polymeric Membranes:
Preparation and Application, Appl. Catal. A Gen. 307 (2) (2006) 167–183.
[11] A. Brunetti, P.F. Zito, L. Giorno, E. Drioli, G. Barbieri, Membrane reactors for low
temperature applications: an overview, Chem. Eng. Process. Process Intensif. 124
(May 2017) (2018) 282–307.
[30] A. Mashentseva, D. Borgekov, S. Kislitsin, M. Zdorovets, A. Migunova, Comparative
catalytic activity of PET track-etched membranes with embedded silver and gold
nanotubes, Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact. Mater.
Atoms 365 (2015) 70–74.
[31] Y. Liu, M. Li, G. Chen, A New Type of Raspberry-like Polymer Composite Sub-
Microspheres with Tunable Gold Nanoparticles Coverage and Their Enhanced
Catalytic Properties, J. Mater. Chem. A 1 (3) (2013) 930–937.
[12] R. Poupart, D. Grande, B. Carbonnier, B. Le Droumaguet, Porous Polymers and
Metallic Nanoparticles: A Hybrid Wedding as a Robust Method toward Efficient
Supported Catalytic Systems, Prog. Polym. Sci. 96 (2019) 21–42.
[13] L.F. Villalobos, P. Neelakanda, M. Karunakaran, D. Cha, K.V. Peinemann, Poly-
Thiosemicarbazide/Gold Nanoparticles Catalytic Membrane: In-Situ Growth of
Well-Dispersed, Uniform and Stable Gold Nanoparticles in a Polymeric Membrane,
Catal. Today 236 (2014) 92–97.
[32] B. Domènech, M. Muñoz, D.N. Muraviev, J. Macanás, Catalytic membranes with
palladium nanoparticles: from tailored polymer to catalytic applications, Catal.
Today 193 (1) (2012) 158–164.
[33] Y. Gu, P. Bacchin, J. Lahitte, J. Remigy, I. Favier, M. Gómez, D.L. Gin, R.D. Noble,
Catalytic membrane reactor for suzuki‐miyaura C−C cross‐coupling: explanation
for its high efficiency via modeling, AIChE J. 63 (2) (2017) 698–704.
[34] Y. Gu, I. Favier, C. Pradel, D.L. Gin, J.F. Lahitte, R.D. Noble, M. Gómez, J.C. Remigy,
High catalytic efficiency of palladium nanoparticles immobilized in a polymer
membrane containing poly(Ionic liquid) in suzuki-miyaura cross-coupling reaction,
J. Memb. Sci. 492 (2015) 331–339.
[14] V.W. Faria, D.G.M. Oliveira, M.H.S. Kurz, F.F. Gonçalves, C.W. Scheeren, G.R. Rosa,
Palladium nanoparticles supported in a polymeric membrane: an efficient phos-
phine-free “Green” catalyst for suzuki-miyaura reactions in water, RSC Adv. 4 (26)
(2014) 13446–13452.
[15] O.M. Ilinich, E.N. Gribov, P.A. Simonov, Water denitrification over catalytic
membranes: hydrogen spillover and catalytic activity of macroporous membranes
loaded with Pd and Cu, Catal. Today 82 (1–4) (2003) 49–56.
[35] A. Akbari, S. Desclaux, J.C. Rouch, J.C. Remigy, Application of nanofiltration
hollow fibre membranes, developed by Photografting, to treatment of anionic dye
solutions, J. Memb. Sci. 297 (1–2) (2007) 243–252.
[16] O.M. Ilinitch, F.P. Cuperus, L.V. Nosova, E.N. Gribov, Catalytic Membrane in
Reduction of Aqueous Nitrates: Operational Principles and Catalytic Performance,
Catal. Today 56 (1–3) (2000) 137–145.
[36] P.T. Nguyen, E. Lasseuguette, Y. Medina-gonzalez, J.C. Remigy, D. Roizard,
E. Favre, A dense membrane contactor for intensified CO 2 gas / liquid absorption
in post-combustion capture, J. Memb. Sci. 377 (1–2) (2011) 261–272.
[37] D.N. Muraviev, J. Macanás, J. Parrondo, M. Muñoz, A. Alonso, S. Alegret,
M. Ortueta, F. Mijangos, Cation-exchange membrane as nanoreactor: intermatrix
synthesis of platinum-copper core-shell nanoparticles, React. Funct. Polym. 67 (12
SPEC. ISS) (2007) 1612–1621.
[17] X. Zhu, K. Choo, J. Park, Nitrate removal from contaminated water using poly-
electrolyte-enhanced ultrafiltration, Desalination 193 (August 2005) (2006)
350–360.
[18] S.R. Hogg, S. Muthu, M.O. Callaghan, J. Lahitte, M.L. Bruening, Wet air oxidation of
formic acid using nanoparticle-modified polysulfone hollow fibers as gas − liquid
contactors, ACS Appl. Mater. Interfaces 4 (2012) 1440–1448.
[38] C.A. Schneider, W.S. Rasband, K.W. Eliceiri, NIH image to ImageJ: 25 years of
image analysis, Nat. Methods 9 (2012) 7.
[19] F. Peirano, T. Vincent, F. Quignard, M. Robitzer, E. Guibal, Palladium supported on
chitosan hollow Fiber for nitrotoluene hydrogenation, J. Memb. Sci. 329 (1–2)
(2009) 30–45.
[39] A. Gabelman, S.T. Hwang, Hollow Fiber membrane contactors, J. Memb. Sci. 159
(1–2) (1999) 61–106.
[20] R. Van der Vaart, V.I. Lebedeva, I.V. Petrova, L.M. Plyasova, N.A. Rudina,
D.I. Kochubey, G.F. Tereshchenko, V.V. Volkov, J. Van Erkel, Preparation and
characterisation of palladium-loaded polypropylene porous hollow fibre mem-
branes for hydrogenation of dissolved oxygen in water, J. Memb. Sci. 299 (1–2)
(2007) 38–44.
[40] R.W. Baker, S.P. Nunes, K. Peinemann (Eds.), Membrane Technology, Wiley, 2006.
[41] S. Belfer, R. Fainchtain, Y. Purinson, O. Kedem, Surface Characterization by FTIR-
ATR Spectroscopy of Polyethersulfone Membranes-Unmodifed, Modified and
Protein Fouled. 172 (2000) 113–124.
[42] K. Karami, M. Ghasemi, N. Haghighat Naeini, Palladium nanoparticles supported on
polymer: an efficient and reusable heterogeneous catalyst for the Suzuki cross-
coupling reactions and aerobic oxidation of alcohols, Catal. Commun. 38 (2013)
10–15.
[21] C. Liu, Y. Xu, S. Liao, D. Yu, Y. Zhao, Y. Fan, Selective hydrogenation of Propadiene
and propyne in Propene with catalytic polymeric hollow-fiber reactor, J. Memb. Sci.
137 (1–2) (1997) 139–144.
[22] D. Singh, M.E. Rezac, P.H. Pfromm, Partial Hydrogenation of Soybean Oil Using
Metal-Decorated Integral-Asymmetric Polymer Membranes: Effects of Morphology
and Membrane Properties, J. Memb. Sci. 348 (1–2) (2010) 99–108.
[23] L. Gröschel, R. Haidar, A. Beyer, H. Cölfen, B. Frank, R. Schomäcker, Hydrogenation
of propyne in palladium-containing polyacrylic acid membranes and its char-
acterization, Ind. Eng. Chem. Res. 44 (24) (2005) 9064–9070.
[43] D.B. Pacardo, M.R. Knecht, Pd Nanoparticles in C-C Coupling Reactions, (2014), pp.
112–156.
[44] M. Kralik, A. Biffis, Catalysis by metal nanoparticles supported on functional or-
ganic polymers, J. Mol. Catal. A Chem. 177 (1) (2001) 113–138.
[45] R. Dittmeyer, K. Svajda, M. Reif, A review of catalytic membrane layers for gas /
liquid reactions, Top. Catal. 29 (May) (2004) 3–27.
[24] A. Bottino, G. Capannelli, A. Comite, R. Di Felice, Polymeric and ceramic mem-
branes in three-phase catalytic membrane reactors for the hydrogenation of me-
thylenecyclohexane, Desalination 144 (1–3) (2002) 411–416.
[46] S. Doherty, J.G. Knight, T. Backhouse, A. Bradford, F. Saunders, R.A. Bourne,
T.W. Chamberlain, R. Stones, A. Clayton, K. Lovelock, Catal. Sci. Technol. 8 (5)
(2018) 1454–1467.
7