An efficient hydrogenation catalytic model hosted in a stable hyper-crosslinked porous-organic-polymer: From fatty acid to bio-based alkane diesel synthesis

Article abstract of DOI:10.1039/c9gc03803e

In this study, a Pd-based catalytic model over a nitrogen enriched fibrous Porous-Organic-Polymer (POP) is established to execute hydrodeoxygenation of various vegetable oils in producing potential large-scale renewable diesel. Here we report a cost-effective synthesis strategy for a new microporous hypercrosslinked POP through the FeCl₃ assisted Friedel-Crafts alkylation reaction, followed by fabrication of Pd⁰-NPs (2-3 nm) using a solid gas phase hydrogenation route to deliver a novel catalytic system. This catalyst (called Pd@PPN) exhibits versatile catalytic performance for different types of vegetable oils including palm oil, soybean oil, sunflower oil and rapeseed oil to furnish long chain diesel range alkanes. The catalyst is comprehensively characterized using various spectroscopic tools and it shows high stability during five runs of recycling without leaching of Pd. Our results further reveal that a direct decarbonylation (DCN) pathway of fatty acids to produce alkanes with one fewer carbon is the dominant mechanism. Under optimized conditions, using stearic acid to represent the long linear carboxylic acids in the vegetable oils, up to 90% conversion with 83% selectivity of C17-alkane has been achieved on our fabricated catalyst. Density functional theory (DFT) calculations are performed to provide insights into the electronic properties of the catalyst, the mechanistic reaction pathway, the crucial role of the catalyst surface and the product selectivity trend. The strong interaction between the corrugated polymer-frame-structure and the Pd-NPs suggests the presence of high density step sites on the fabricated Pd-NP anchored within the cage of the polymer structure. DFT calculations also reveal the strong promotional effect of step sites and charge transfer in facilitating rate-limiting steps during the decarbonylation (DCN) pathway and removal of strongly bound intermediates formed during the process, therefore explaining the high activity of the fabricated Pd@PPN catayst for the hydrodeoxygenation (HDO) conversion to produce bio-based alkane diesel.

Full text of DOI:10.1039/c9gc03803e

View Article Online
View Journal
Green
Chemistry
Cutting-edge research for a greener sustainable future
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: C. Sarkar, S. C.
Shit, D. Q. Dao, J. Lee, N. H. Tran, R. Singuru, K. An, N. D. Nam, Q. V. Le, P. N. Amaniampong, A. Drif, F.
Jerome, T. H. Pham, T. T. N. Phan, D. N. Vo, N. Thanh Binh, Q. T. Trinh, M. Sherburne and J. Mondal,
Green Chem., 2020, DOI: 10.1039/C9GC03803E.
Volume 20
This is an Accepted Manuscript, which has been through the
Number
May 2018
Pages 1919-2160
9
7
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Cutting-edge research for a greener sustainable future
rsc.li/greenchem
Accepted Manuscripts are published online shortly after acceptance,
before technical editing, formatting and proof reading. Using this free
service, authors can make their results available to the community, in
citable form, before we publish the edited article. We will replace this
Accepted Manuscript with the edited and formatted Advance Article as
soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes to the
text and/or graphics, which may alter content. The journal’s standard
Terms & Conditions and the Ethical guidelines still apply. In no event
shall the Royal Society of Chemistry be held responsible for any errors
or omissions in this Accepted Manuscript or any consequences arising
from the use of any information it contains.
ISSN 1463-9262
PAPER
Paul T. Anastas et al.
The Green ChemisTREE: 20 years after taking root with the 12
principles
rsc.li/greenchem

Page 1 of 20
Green Chemistry
View Article Online
DOI: 10.1039/C9GC03803E
Mondal, John; Indian Institute of Chemical Technology, Inorganic and
Physical Chemistry

Please do not adjust margins
Green Chemistry
Page 2 of 20
View Article Online
DOI: 10.1039/C9GC03803E
ARTICLE
Efficient Hydrogenation Catalytic Model Hosted in Stable Hyper-
crosslinked Porous-Organic-Polymer: From Fatty Acids to Bio-based
Alkanes Diesel Synthesis
Chitra Sarkar,#,a Subhash Chandra Shit,#,a Duy Quang Dao,#,b Jihyeon Lee,c Ngoc Han Tran,d Ramana
Singuru,a Kwangjin An,c Nguyen Dang Nam,b Quyet Van Le,b Prince Nana Amaniampong,e Asmaa
Drif,f Francois Jerome,e,f Pham Thanh Huyen,g Thi To Nga Phan,g Dai-Viet N. Vo,h Nguyen Thanh Binh,i
Quang Thang Trinh*,b,j Matthew P. Sherburne*,k,l and John Mondal*,a
Received 00th January 20xx,
Accepted 00th January 20xx
DOI: 10.1039/x0xx00000x
Correspondence to: qttrinh@ntu.edu.sg; trinhquangthang@duytan.edu.vn (Q.T.T)
mpsherb@berkeley.edu (M.P.S); johncuchem@gmail.com (J.M)
In this study, Pd-based catalytic model hosted over nitrogen enriched fibrous Porous-Organic-Polymer (POP) is established
to execute hydrodeoxygenation of various vegetable oils in producing potential large-scale renewable diesel. Here we report
a cost effective synthesis strategy of a new microporous hypercrosslinked POP through the FeCl3 assisted Friedel-Crafts
alkylation reaction, followed by fabrication of Pd0-NPs (2-3 nm) with solid gas phase hydrogenation route to deliver a novel
catalytic system. This catalyst (called Pd@PPN) exhibits versatile catalytic performance for different types of vegetable oils
including palm oil, soybean oil, sunflower oil and rapeseed oil to furnish long chain diesel range alkanes. The catalyst is
comprehensively characterized by various spectroscopic tools and shows high stability during five runs of recycling without
the leaching of Pd occuring. Our results further reveal that direct decarbonylation (DCN) pathway of fatty acid to produce
alkanes with one carbon less is the dominated mechanism. At optimized conditions, using stearic acid to represent the long
linear carboxylic acids in the vegetable oils, up to 90 % conversion with 83 % selectivity of C17-alkane has been achieved on
our fabricated catalyst. Density functional theory (DFT) calculations are performed to provide insights into the electronic
properties of the catalyst, the mechanistic reaction pathway, the crucial role of catalyst surface and the product selectivity
trend. The strong interaction between corrugated polymer-frame-structure and the Pd-NPs suggests there is the presence
of high density step sites on the fabricated Pd-NP anchored within the cage of polymer structure. DFT calcualtions also reveal
the strong promotional effect of step sites and charge transfer in facilitating rate-limiting steps during the decarbonylation
(DCN) pathway and removal of strongly bound intermediates formed during the process, therefore explain the high activity
of the fabricated Pd@PPN catayst for the hydrodeoxygenation (HDO) conversion to produce bio-based alkanes diesel.
a. Catalysis & Fine Chemicals Division, CSIR-Indian Institute of Chemical Technology,
Uppal Road, Hyderabad-500007, India.
Introduction
b. Institute of Research and Development, Duy Tan University, 03 Quang Trung,
Danang 550000, Viet Nam.
In recent decades, the use of renewable energy resources such
as biodiesel has increasingly gained attention owing to its ability
to reduce CO2 emission and other greenhouse gases from the
consumption of fossil fuels.1-3 Various alternative energy
sources like solar, wind, and nuclear energy are developing
nowadays, however the utility of solar, wind energy have the
severe limitation by the surrounding environment and nuclear
energy application suffers from drawbacks like residue
c. School of Energy & Chemical Engineering, Ulsan National Institute of Science and
Technology (UNIST), Ulsan 689-798, Korea.
d. Department of Aquatic Sciences and Assessment, Swedish University of
Agricultural Sciences SLU, Box 7050, SE-750 07 Uppsala, Sweden.
e. CNRS Research federation INCREASE, 1 rue Marcel Doré, TSA 41105, 86073 Poitiers,
France.
f. Institut de Chimie des Milieux et Matériaux de Poitiers (IC2MP), Université de
Poitiers, CNRS, 1 rue Marcel Doré, TSA 41105, 86073 Poitiers, France.
g. School of Chemical Engineering, Hanoi University of Science and Technology
(HUST), 1 Dai Co Viet, Hanoi, Viet Nam
deposition
and radiation pollution.4 Among renewable
h. Center of Excellence for Green Energy and Environmental Nanomaterials
(CE@GrEEN), Nguyen Tat Thanh University, 300A Nguyen Tat Thanh, District 4, Ho
Chi Minh City 755414, Vietnam
resources, biomass-based fuels are considered as potential
alternative to conventional fossil derived liquid fuels.5-8 The
main component of natural oils and fats are triglycerides and
free fatty acids comprised of longer linear chain hydrocarbon in
the range of C4-C24.9 The non-edible oils derived from jatropha,
neem, karanja trees as well as waste cooking oil which are the
main feedstock of fatty acids are readily available in developing
i. Faculty of Chemistry, VNU University of Science, Vietnam National University,
Hanoi, 19 Le Thanh Tong, Hanoi, Vietnam
j. Cambridge Centre for Advanced Research and Education in Singapore (CARES), 1
Create Way, Singapore 138602.
k. Materials Science and Engineering Department, University of California, Berkeley,
Berkeley, California 94720, United States of America.
l. Berkeley Educational Alliance for Research in Singapore (BEARS),1 CREATE Way,
138602 Singapore
#
C.S., S.C.S. and D.Q.D. are equally contributed
Electronic Supplementary Information (ESI) available: [details of any supplementary
information available should be included here]. See DOI: 10.1039/x0xx00000x
J. Name., 2013, 00, 1-3 | 1

Please do not adjust margins
Page 3 of 20
Green Chemistry
ARTICLE
Journal Name
countries and economically viable compare to edible oil. The exhibited tremendous catalytic activity for directVHiewDOArticolef Olonlninge
afore-mentioned renewable feedstock are the most desirable chain fatty acids to hydrocarbons.29 A DvOeIr:y10e.1f0fi3c9i/eCn9tGCR0u3-8H0A3PE
sources for the future transportation fuel due to their high catalyst for the HDO of various types of oils (Jatropha oil, Palm
energy density as well as structural similarity with fossil fuels.4 oil, Waste cooking oil) to their respective long chain alkane for
Unfortunately, high viscosity and low volatility of natural oils biodiesel industry was developed by Guo et al.30 Chen and co-
suffer the serious disadvantage of using directly to the engine workers introduced zeolite-supported ruthenium catalysts to
such as injector coking, carbon deposits and lubricant execute HDO of biodiesel-related fatty acid methyl esters to
thickening.10, 11 Researchers are now struggling to develop the diesel-range alkanes.31 HDO of waste cooking oil was
sustainable method to upgrade biodiesel from these renewable investigated by Zheng and co-workers over unsupported
feedstock.12, 13 To overcome the drawback of direct applications CoMoS catalysts but the activity decreased associated with the
of natural oils to the engine, oxygen content reduction from the gradual loss of sulfur.32 An important finding of cascade reaction
fuel would readily ameliorate fuel stability, thus allowing to of terpene limonene dehydrogenation and stearic acid HDO
create better utilisation potential.14 Cracking of triglycerides through p-cymene accompanying H2 formation over Pd-
performed by zeolite would reduce the oxygen moieties, but Ni/HZSM-5 catalyst was recently reported by Zhao et al.33
create significant portion of carbon lose, responsible for energy Although all recent works demonstrated high catalytic
loss.15 The homogeneous base catalysis of fatty acids lead to the efficiency for green diesel production, classical issues such as
formation of fatty acid methyl esters (FAME) through sulphide leaching, catalyst deactivation and high-cost bimetallic
transesterification, but FAME are considered as “drop in” fuel catalyst remain unsolved and there is a need to develop
solution compared to conventional fuels due to their low energy catalysts with higher activity, selectivity and stability for more
density, higher cloud and smoke point along with poor cold flow realistic applications.
properties.16 Contrarily, high oxygen content of FAME make it Currently, Porous Organic Polymers (POPs) have garnered
unsuitable for diesel engine and during the process the significant interest within the scientific fraternity in the area of
situation became out of control by the deactivation of the interdisciplinary research in catalysis,34 gas separation35 and
catalyst through the saponification reaction of free fatty acids.11 CO2 capture.36 Owing to their exceptionally large surface area
Hence, deoxygenation of the fatty acid and triglycerides by and tuneable pore-size, the cross-linked 3D cage-like structure
removing oxygen content in the form of CO, CO2, and H2O is the can easily accommodate metal or metal oxide nanoparticles as
best proof of concept reaction to upgrade non-edible oil and active centers.37-39 The molecular connectivity and chemical
waste cooking oil in biofuel industry, although the atom environment of POP framework can be readily tuned and
economy of the reaction is off. First generation biodiesel is tailored by altering synthetic strategy and selection of various
produced directly from food crops (i.e. vegetable oil) and geometry and size of monomers. While monomers exhibit
contains fatty acid methyl esters (FAMEs) and its essential role in the structural design of POP, the interfacial
hydrocarbons.17 However, the first generation biodiesels shows interaction between monomer and metal nanoparticles
severe drawbacks, such as high oxygen and acidity content, enhances the stability of the structure and the catalytic
which hinders its ability to be utilised for large-scale activity.39-41 In addition, our research group has developed a
applications compared to fossil fuels.16, 18-20 To overcome these series of POP supported catalysts which exhibited great
limitations, second-generation bio-fuels which are basically free catalytic performance in biomass conversion, revealing
of oxygen and compatible with petroleum diesel are being structural benefits for hydrogen activation and easy mass
43
developed. Green biodiesel has notable features, such as higher transfer through the porous network.42, Such robust POP
heating value, oxidation stability, energy density, and cetane framework could be used as a platform to host Pd NPs with no
number comparable to conventional biodiesel.21-23 In general, aggregation and activate the nanoparticles by electron-
the production of green biodiesel from fatty acids is known to donating interaction. The system proved to be suitable for
involve various deoxygenation methods,24 such as hydrogenation of oils in water under mild conditions though
decarboxylation
(DCX),
decarbonylation
(DCN)
and inhibiting leaching and catalyst deactivation.
hydrodeoxygenation (HDO). These pathways have been In this work, we designed and developed a cost-free scalable
reported to be incredibly efficient deoxygenation methods for synthesis of nitrogen enriched Porous Organic Polymeric
the removal of oxygen moieties from oxygen containing network (PPN) by one-pot FeCl3 assisted Friedel-Crafts
compounds to form alkane and hydrogen (H2).17-21
In nature, two types of fatty acids are available and exist as α,α'
alkylation reaction of triphenylamine (TPA) monomer and
dibromo-p-xylene as cross-linkers. The active Pd@PPN
̵
saturated and unsaturated acids. The number of carbon atoms catalyst with homogenous dispensability of Pd-NPs (~2-3 nm
in the fatty acid chains often ranges substantially from C8 to size) was prepared by simple solid phase H2 reduction route at
C24, but C12, C16, and C18 fatty acids are the most abundant in high temperature. The Pd@PPN catalyst has been employed for
nature.25-27 Saha et al. reported silica supported Ir-ReOx the deoxygenation of stearic acid using green solvent (H2O)
bimetallic catalyst for the HDO of waste cooking oil and rather than organic solvents, which made the permanent
vegetable oil to diesel range alkane with 79-85% yield efficacy.28 footprint in the biofuel industry as the utilization of organic
In another study, Bhaumik and co-workers also introduced Ru solvent not only requires separation processes but also creates
nanoparticles decorated on new porous organic network which environmental hazards.30 This catalyst also showed versatile
J. Name., 2013, 00, 1-3 | 2
Please do not adjust margins

Green Chemistry
Please do not adjust margins
Page 4 of 20
Journal Name
ARTICLE
catalytic performance over different types of vegetable oils to more efficient synthetic protocols. For example, the
View Article Online
DOI: 10.1039/C9GC03803E
furnish long chain alkanes. We have achieved a nearly Deposition-Precipitation technique is more efficient for better
quantitative yield in comparison with the theoretical values for metal loading, in which the deposition from a precursor solution
possible long-chain alkanes based on the weight of the is facilitated through a change of pH, temperature, or
converted oils. In order to elucidate reaction pathway, we evaporation… so that metal compounds are formed with a low
conducted control experiments using model substrates as solubility. Then the precipitation on the support can be nicely
probe molecule involving hydrogenation-hydrodehydration and achieved when the surface free energy of small nuclei is
direct decarbonylation/decarboxylation steps. Density decreased or the precipitate is stabilized through the
Functional Theory (DFT) calculations were employed to study introduction of the support.45
the interaction between the Pd-NPs and the polymer structure,
evaluate their electronic properties and to elucidate the high
activity of our catalyst for the HDO conversion of long linear-
chain carboxyl acid.
Methodology
Synthesis of triphenylamine based Porous Organic Polymer Network
(PPN)
Porous Organic Polymer Network (PPN) has been synthesized Figure 1: Schematic illustration for the synthesis of hyper-crosslinked
nitrogen rich porous organic polymer network (PPN) and the
corresponding Pd loaded polymer (Pd@PPN).
via FeCl3 assisted template free Friedel-Crafts alkylation of
triphenylamine (TPA) in the presence of α,α'-dibromo-p-xylene
follwing the previous literature report by Bhaumik et.al.44 In a
typical synthetic procedure, 100 mL oven dried round bottom
flask was charged with TPA (245 mg, 1 mmol) and α,α'-dibromo-
p-xylene (396 mg, 1.5 mmol) and throughly mixed together in
20 mL of anhydrous 1,2-dichloroethane (DCE) solvent.
Anhydrous FeCl3 (487 mg, 3 mmol) was then added and the
mixture was continously stirred for additional 10 h at room
temperature under N2 atmosphere. The resulting mixture was
further heated to reflux for another 14 h. After that, the
reaction was cooled down to room temperatrure and the
bluish-black precipitate was dipped into acetone, collected and
washed with methanol by centrifugation. To remove iron
completely from the polymer, the product was further washed
with methanol for three days in a soxhlet apparatus. Finally, the
bluish-black precipitate was dried in an oven at 100°C over night
and designated as PPN.
Catalytic Hydrogenation of Vegetable Oils over Pd@PPN Catalyst
Catalytic hydrogenations of various vegetable oils were
conducted in a 100 mL Parr autoclave high-pressure reactor
with a built-in pressure gauge setup having the range of 0-12
MPa and maximum temperature limit of 573 K. In a typical
procedure, the pre-reduced Pd@PPN catalyst with the weight
of 0.02 g was suspended into 50 mL of water and loaded into
the reactor. The reactor was subsequently purged with H2 gas
for three times before the reaction to ensure the catalysts were
still in their pre-reduced states, and then was charged with
stearic acid (100 mg, 0.35 mmol) and 20 mL of water to initiate
the reaction. The dissolved O2 or air from the reactor was
evacuated using a vacuum pump for 15 min at room
temperature. During the reaction, H2 pressure and the
autoclave temperature were kept at 30 bar and 150 °C,
respectively. The catalytic reaction was continued for the
desired reaction time at a stirring speed of 1000 rpm. The
reaction was monitored by sampling at regular time intervals,
with analysis of the samples by an Agilent 6980 gas
chromatograph (GC) equipped with a flame ionization detector
(FID) and an SE-54 capillary column (30 mm × 0.32 mm × 1.0 μm)
with a stationary phase GC. The products were quantified using
n-Eicosane as an internal standard followed by comparison with
known standards with a standard deviation less than 2%. After
each catalytic reaction, the catalyst was isolated by simple
centrifugation technique followed by washing with methanol
several times and dried for overnight. After that, the catalyst
was reduced again under solid-phase H2 technique at 200oC for
4 h and further used for next catalytic test.
Fabrication of Pd0-NPs on the PPN
Ultrafine Pd-NPs with an average size of 2-3 nm were fabricated
on the PPN by solid-phase high temperature hydrogenation
route (Figure 1). In a typical synthesis, 200 mg of PPN was added
to 100 mL dichloromethane solution of Pd(OAc)2 (80 mg) and
continued for stirring at 60 °C for 24 h. After cooling to room
temperature, the dichloromethane solvent was removed using
the rotary evaporator and the resulting brown precipitate was
filtrated and washed thoroughly with dichloromethane and
MeOH to wipe out all the unreacted Pd-precursor. The obtained
solid was dried at 80 °C for 12 h to give an as-synthesized brown
solid; the latter was treated in a stream of H2/N2 (10% H2, 100
mL/min) at 200 °C for 4 h. The resulting black solid was isolated
and designated as Pd@PPN. It is noted that due to this
procedure, a certain amount of Pd precursor has been wiped
out during the filtration and wash. Our future work will be
dedicated to develop the metal nanoparticles encapsulated by
Porous-Organic-Polymer with different metal loading following
Computational details
To investigate the structure of the Pd NPs interacting with the PPN
frame-structure and gain theoretical insights into the activities of the
fabricated Pd-based catalyst for the HDO conversion of fatty acids,
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 3
Please do not adjust margins

Please do not adjust margins
Page 5 of 20
Green Chemistry
ARTICLE
Journal Name
reduced-cluster model has also been conveniently usVeiedwinArtaiclereOcnelinnet
DOI: 10.1039/C9GC03803E
study by Zhou et al. to investigate the light-driven methane dry
we performed Density Functional Theory (DFT) calculations based on
periodic boundary conditions and plane-wave pseudopotential
implemented in the Vienna ab-initio simulation package (VASP),
developed at the Fakultät für Physik of the Universität Wien.46, 47
Projector augmented wave (PAW) method48 employed with the
Generalized Gradient Approximation (GGA) and a cut-off energy of
450 eV were used to describe the interaction between valence
electrons and the ions. The optimized Becke88 functional (optB88)
coupled with the non-local vdW-DF correlation (optB88-vdW)
developed by Klimeš et al.49-52 were used to correctly capture the van
der Waals interaction between aromatic compounds and carboxylic
acids with Pd surfaces for all DFT calculations. As reported in earlier
studies, this level of theory was one the most appropriate to describe
the interaction between aromatic compounds on transition metal
surface and predict accurately the adsorption energies compare to
experimental data.52-54 Furthermore, since the van der Waals
interaction between carboxylic acids and the transition metal surface
is important,55, 56 optB88-vdW functional is also a suitable choice for
the investigation of carboxylic acids conversion on Pd@PPN catalyst.
We have evaluated the performance of this chosen optB88-vdW
functional in computing the adsorption energies for benzene and
formic acid on several transition metals, and found the observed
values are in very good agreement with experimental data, as are
presented in Table T7 and Figure S11 of the SI. Transition states of
the reactions occurred during the HDO conversion were located
using the Climbing-Nudged Elastic Band (Cl-NEB) method.57
Frequency calculations confirmed the nature of the transition states
with only one imaginary vibrational frequency.
reforming on Cu-Ru nano-particles.63 This reduced model has 332 Pd
atoms (called Pd332 cluster) and only the Γ point was used to sample
the Brillouin zone for DFT calculations on this model. All VASP
calculations for this reduced Pd332 cluster model were performed
with the invoking of the first-order Methfessel−Paxton smearing
technique with a width of 0.1 eV, the electronic minimization was
done using the mixture of the Davidson and RMM-DIIS algorithms
and the conjugate-gradient algorithm was used to relax the ions into
their instantaneous ground-state. To study the electronic properties
of Pd@PPN material, the Bader charges of all the atoms were
computed based on the charge density grid algorithm developed in
Henkelman et al.64, 65 Furthermore, the analysis of the Density of
states (DOS) projected on the d-band for Pd atoms in different
structures was also performed to understand the interaction
between Pd and the porous polymer frame structure.
Theoretical investigation into the structure and electronic
properties of Pd@PPN. In order to understand the structure of Pd
NPs fabricated on the PPN frame-structure, DFT calculations using
the optB88-vdW functional were conducted to study the interaction
between one full unit monomer of the polymer structure (called
umPPN molecule) and the Pd-NPs. The structure of the umPPN unit
molecule is shown in Figure 2a and it is a very good representative
molecule to study the interaction between Pd nano-particles and the
Porous Polymer frame-structure since it contains both two monomer
units for the polymer: TPA and p-xylene. To study the adsorption of
umPPN on Pd NPs, we used a cluster approach to model the Pd
particle with the size of 3 nm which was detected from experimental
XRD (see Figure 3b) and TEM (see Figure 5d-f) analysis. Firstly, we
built a structure representative for a 3 nm NPs with truncated
octahedral shape involving 1289 Pd atoms in total (Fig. 2b). This
cluster model was used for 3 nm size Ru particle in explaining the
high activity of B5 step sites during the ammonia synthesis by
Honkala et al.58 and later was widely employed to study the activity
of Pd NPs for different reactions such as methane activation,59, 60 CO
adsorption61 and conversion.62 However, due to the symmetric
structure of this cluster and to make the calculations more efficient,
we then used the simplified model including the upper half with only
four top-most layers of the 3 nm truncated octahedral cluster NPs in
the unit cell of 25 Å × 25 Å × 25 Å, now is called as the reduced cluster
(Fig. 2c). The illustration of how to create the reduced cluster model
from the full size model is presented in Figure S12, SI. The similar
Figure 2: Models used for DFT investigation on the adsorption of TPA
and the conversion of fatty acid on Pd NPs. (a) The structure of one
full unit monomer of the polymer (called umPPN molecule) used to
represent the Porous Polymer frame-structure. (b) Full cluster model
of the 3 nm Pd NPs with truncated octahedral shape, including 1289
Pd atoms. (c) Reduced cluster model of 332 Pd atoms used to study
the interaction of TPA with Pd NPs. (d) Model of p(48) Pd(111) slabs
with 4 missing rows on the top layer. Different step sites (B5 and F4)
are highlighted. (e) Model of terrace sites on p(44) Pd(111) slabs.
J. Name., 2013, 00, 1-3 | 4
Please do not adjust margins

Green Chemistry
Please do not adjust margins
Page 6 of 20
Journal Name
ARTICLE
We have checked the adsorption energies of the TPA molecule on studying the reaction mechanism. The activity of the normal terrace
View Article Online
58
DOI: 10.1039/C9GC03803E
thicker clusters (including five layers) and they are almost identical sites which was dominated on larger-size particles was evaluated
to that on the four-layer cluster used herein (Figure S13. SI).66 It is
worthy to note that in our study, all VASP calculations were
performed on CPU-mode only. Running VASP on Graphics Processing
Unit (GPU) will accelerate the VASP electronic structure calculations
significantly, as was documented in earlier reports67-69 and with the
release of modern graphics-processing unit architecture, VASP could
run up to 10X faster compared to CPU-only systems, enabling usage
of computationally more demanding and more accurate methods in
the same time.
using a slab model of p(44) Pd(111) surface (Fig. 2e). For all the slab
models in this study, two topmost layers and the adsorbates were
allowed to fully relaxed, while the atoms in the remaining bottom
layers were fixed at their bulk optimized lattice constant of 3.94 Å.7,
76
All the DFT calculations for the study of butanoic acid conversion
on Pd terrace and step sites were also performed employing the
optB88-vdW functional. The Brillouin zone was sampled with the
value of a 3×3×1 Monkhorst–Pack grid for both the p(4×8) with 4
missing rows and the p(4×4) models. Repeated slabs in those two
models were both separated by a vacuum of 20 Å above the topmost
layer to minimize their interactions.
Investigation into the Hydrodeoxygenation (HDO) conversion of
fatty acids on Pd@PPN catalyst. To evaluate the HDO reactions of
fatty acids on Pd catalyst, butanoic acid was chosen as the model
compound. As mentioned by Lu et al.70, the computed gas-phase
reaction energies for the decarbonylation (DCN) and decarboxylation
(DCX) reactions of carboxylic acids to alkanes are converged and
almost identical for reactants with carbon chain length larger than
three. Evaluation of some electronics properties of C4, C5, C6 and
C18 acids found that they gave similar results, justifying the
appropriateness of using butanoic acid to represent stearic acid in
our DFT calculations (see details in the SI). It should be mentioned
here that for locating the transition states during the reaction of
butanoic conversion by the Climbing-Nudged Elastic Band (Cl-NEB)
method, typically a series of intermediate states (six to ten
intermediate states) distributed along the initial reaction path
connecting a reactant and a product state are simultaneously
optimized while restricting atomic motions to hyperplanes
perpendicular to the reaction path, making it not computationally
efficient for large systems such as the reduce Pd332 cluster.
Therefore, to make the computational evaluation more efficient in
studying the HDO conversion of butanoic acid, we then used two
other models to evaluate the activity of the fabricated Pd@PPN
catalysts. The first model is the periodic three layers p(48) Pd(111)
with 4 missing rows on the top layer (Fig. 2d). This model exposes
two most popular step sites on it, the B5 and the F4 step sites, which
are highlighted in Fig. 2d. Note that the B5 step sites are created at
the intersection of a (111) and a (100) facet, and the F4 step sites are
formed at the intersection between two (111) facets. This model also
presents terrace sites next to those B5 and F4 steps sites, and is an
excellent representative for the reduced Pd332 cluster in the Fig. 2c.
The use of this model was very helpful and successfully applied for
earlier theoretical studies in explaining the formation of nano-Co-
islands and CO activation during the Fischer-Tropsch synthesis under
realistic conditions71, 72 and evaluating the activity of step sites in the
conversion of methane,2, 73 toluene52 and amine-coupling reaction.74,
Results and discussion
Material Characterization
Herein, we defined a truly cost effective synthesis strategy of a new
microporous hyper-crosslinked organic polymer networks PPN
through the FeCl3 assisted Friedel-Crafts alkylation reaction of
triphenylamine (TPA) monomer and α,α' ̵
dibromo-p-xylene.44 The
resulting material is rich in nitrogen content, which can be used as
anchors to introduce palladium nanoparticles. The treatment of PPN
with Pd(OAc)2 followed by solid-phase reduction with 11%
H2 /N2 mixture at high temperature gives the Pd@PPN
materials.77 The texture and chemical properties of as-synthesized
PPN and Pd@PPN materials were characterized by wide angle X-ray
scattering (WAXS), 13C CP/MAS solid-state NMR spectroscopy, X-ray
photoelectron spectroscopy (XPS), Transmission Electron
Microscopy (TEM), high angle annular dark field scanning
transmission electron microscopy (HAADF-STEM), energy-dispersive
X-ray (EDX) elemental mapping, field emission scanning electron
microscopy (FE-SEM) and nitrogen physisorption studies.
The elemental analysis confirmed that the PPN contains 83.65% C,
5.67% H and 4.21 wt.% of N, respectively. In strong contrast,
Pd@PPN material contains 78.23% C, 5.12% H and 3.91 wt.% of N,
respectively. Pd content in the Pd@PPN material is 0.459 mmol/g as
measured by Inductive Couple Plasma Atomic Emission Spectroscopy
(ICP-AES) technique. A little inconsequential difference in the
experimental
C and N content (%) from the corresponding
theoretical values in the elemental (C, H, N) analysis could be
explained by the accommodation of guest molecules in the porous
channel of the host PPN, a very widespread phenomenon of porous
materials.78 The thermal stability of the as-synthesized PPN polymer
investigated by thermogravimetric analysis (TGA) under N2
atmosphere in the temperature range 25 to 800 °C is presented in
the Figure S1, SI. At low-temperature region (< 100 °C), the negligible
weight loss could be attributed to the evaporation of surface water
and intercalated solvent molecules inside the polymeric cage. A
sharp weight loss at 430 °C is observed, corresponds to the
degradation of the organic framework. High thermal stability of this
polymer could be attributed to the high hyperbranched cross-linking
and the rigid polymeric networks in the molecular skeleton. FT-IR
analysis of both as-synthesized PPN and Pd@PPN were conducted
and the results are presented in the Figure S2, SI. The C-N stretching
75
We have also compared the adsorption energy and initial
activation barriers of butanoic acid on B5 and F4 step sites of the
reduce Pd332 cluster and on B5 and F4 step sites of the p(4x8) with 4
missing row model using the optB88-vdW functional, and the data
showed that those obtained values are very close to each other
(Table T9 and Fig S14, SI), justifying the appropriateness of using the
p(4x8) with 4 missing row model to represent the Pd332 cluster in
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 5
Please do not adjust margins

Please do not adjust margins
Page 7 of 20
Green Chemistry
ARTICLE
Journal Name
of the tertiary amines in the organic backbone unit of the as- XRD pattern of Pd@PPN clearly showed the crystalline nature of the
View Article Online
DOI: 10.1039/C9GC03803E
synthesized polymer reappeared as characteristic stretching Pd-NPs loaded in the porous polymer. The characteristic Braggs
vibrations of PPN material at 1278 cm-1. The distinctive C-H out-of- angles at 2θ = 39.9°, 46.5°, 67.8° correspond to Pd (111), Pd (200), Pd
plane bending vibrations of the p-disubstituted benzene rings is (220) lattice planes, indicative for the face centred cubic (fcc) Pd-
observed at 819 cm-1 while the bands at 1604 and 1494 cm-1 can be crystal. The result is in good agreement with previous literature
attributed to the strong stretching vibrations of the aromatic C=C (JCPDS no. 7440-05-3).79 The Pd-NPs size was estimated to be of
bonds in the aromatic rings. The formation of the extended PPN 3.0±0.5 nm, computed using the Debye-Scherrer equation for the
network was additionally evidenced by the existence of highest intensity diffraction peak of the (111) crystalline lattice plane.
characteristic Ph-CH2 unit stretching modes at 2918 cm-1. The FT-IR The result is very consistent with the particle size obtained from TEM
spectrum of the Pd@PPN material confirms that the structural and HAADF-STEM analysis later.
bonding in the PPN framework remains intact after metal loading.
To get the clear picture about the permanent porosity of the as-
synthesized PPN and Pd@PPN materials, N2 gas adsorption study
were performed at 77 K. As shown in Fig. 3c, the N2 adsorption-
desorption isotherms of PPN exhibited high N2 uptake at very low
P/P0 region, followed by gradual uptake with a very broad hysteresis
loop at the high P/P0 region. This typical type IV isotherm suggests
the coexistence of micro and interparticle void spaces within the PPN
framework. The broad hysteresis loop in the high-pressure region
could be explained due to the swelling of polymeric network upon
the gas adsorption, for which irreversible gas uptake can take place.
However, flexible expansion of restricted access pores (closed pores)
with the increase of P/P0 could be also the key factor for the
appearance of large hysteresis loop, in good agreement with the
previous report by Bhaumik et.al in their nitrogen-rich porous
organic polymer.80 The Brunauer-Emmett-Teller (BET) surface areas
and the total pore volumes were found to be (1165 m2/g, 1.033
cm3/g) and (679 m2/g, 0.825 cm3/g), for the PPN and Pd@PPN
materials, respectively. The preservation of the isotherms shape for
the N2 adsorption/desorption with no hysteresis loop of Pd@PPN in
comparison with the parent PPN material shows a considerable
diminishment in the BET surface area with both partial pore
occupancy and mass loading of Pd NPs. The Non-Local-Density-
Functional-Theory (NLDFT) fitting of the adsorption branches model
of both the PPN and Pd@PPN materials (Figure S3, SI) exhibited the
narrow pore size distributions with the pores dimension
predominately in the microporous range (1.54 nm & 1.28 nm).
Figure 3: (a) 13C CP/MAS solid state spectra, (b) Wide angle powder
X-ray diffraction patterns; (c) N2 adsorption/desorption isotherms.
The solid state 13C CP/MAS NMR spectra of the as-synthesized PPN
and Pd@PPN materials (Figure 3a) confirm the molecular
connectivity of the structural integrities and chemical surroundings
in the region of the different carbon nuclei. A weak signal at δ = 43.8
ppm in the spectrum of PPN is accorded to the carbon atom of
methylene linker. The appearance of a strong signal at δ = 134.5 ppm
confirms the non-substituted carbon atom of the aromatic rings,
whereas the concomitant appearance of distinguishable signals
centred at δ = 143.2 ppm and 153.2 ppm are accountable for the
phenyl ring and N-substituted aromatic carbon atom of the
triphenylamine unit. The Pd@PPN material exhibited chemical shifts
quite analogous to the PPN, which unambiguously proved that the
structural integrity of their rigid polymer backbone unit after
palladium implantation is successfully retained, as proposed by
Bhaumik and co-workers.44 However, the peaks appeared in the MAS
NMR spectrum of Pd@PPN with somewhat poor intensity and
overlapped peaks indicated the complex nature of the metal loaded
material. The wide-angle powder X-ray diffraction pattern (Fig. 3b) of
as-synthesized PPN exhibited a weak broad diffraction peak at 2θ =
20.2°, which could be attributed to the π-stacking of the aromatic
building blocks in small domains. This pattern shows the amorphous
nature of our PPN material, obviously caused by the random
polymerization of monomer unit upon heating conditions. Powder
Figure 4: FE-SEM images of (a, b) PPN and (c, d) Pd@PPN materials.
The morphology of the as-synthesized PPN and Pd@PPN
materials was investigated by FE-SEM analysis technique (Figure
4). The FE-SEM images of the PPN material mainly consists of
homogeneously distribution of self-assembled irregular-shaped
J. Name., 2013, 00, 1-3 | 6
Please do not adjust margins

Green Chemistry
Please do not adjust margins
Page 8 of 20
Journal Name
ARTICLE
nanofiber-like morphology (Fig. 4a) with the average length of cavities and large Pd-NPs (2.89 nm) are distributed on the
View Article Online
DOI: 10.1039/C9GC03803E
each nanofiber found in the range of 233-333ꢀnm. Similar to external surface of PPN support owing to the inhibition of
PPN material, the Pd@PPN exhibits long-range nanofiber-like enormous growth of Pd-NPs by nucleation in the porous
morphology with longer folding fibers (length in the range of channel under solid phase reduction method. This is in good
406-436ꢀnm) (Fig. 4c). Energy dispersive X-ray (EDX) analysis agreement with the previous finding report by Wang and co-
from the representative FE-SEM image (Figure S4, SI) of the workers.77 Brighter sphere-like spots of diameter 3.4 ± 0.4 nm
particular selected area consists of spectra (marked with yellow (HAADF-STEM image, Figure S5, SI) indicate the existence of the
box), signifying relative abundance of each detected element in metal phases, which are homogenously decorated on the PPN
the specific area. The elemental composition of each element support. Moreover, elemental mapping analysis (Fig. 5) in a
(C, N & Pd) in weight (%) and atomic (%) of Pd@PPN was selected area of Pd@PPN (marked by white square box),
provided in Fig. S4, SI.
demonstrates that the C (red), O (violet), N (green) and Pd (blue)
are nearly distributed at the same location of the PPN phase
boundaries.
Figure 6: Deconvoluted X-ray photoelectron spectra for the Pd-3d
binding energies of Pd@PPN.
In order to evaluate the oxidation state of Pd, XPS analysis of Pd 3d
binding energies of the Pd@PPN material was carried out. The
appearance of characteristic binding energy peaks of in Pd-3d XPS
spectrum (Figure 6) further supports the incorporation of palladium
nanoparticles onto the porous surface of PPN polymer material. The
high resolution Pd 3d spectrum could be deconvoluted into two sets
of spin–orbit doublets corresponding to Pd 3d5/2 and Pd 3d3/2 core-
level binding energies. The binding energy peaks at 334.68 and
339.58 eV are attributed to the Pd (0) states, which confirms the
existing of metallic Pd phase dispersed on the external surface of the
PPN polymer material after the treatment in a stream of H2/N2.
Additionally, the shake-up binding energy peaks at 336.37 and
341.86 eV are attributable to the existence of Pd (II) species in the
Pd@PPN polymer material, indicative of the PdO species formation
due to the surface oxidation of nano-palladium, a very common
phenomenon in noble metals. Besides, XPS analysis also reveals a
strong interaction between the well-dispersed isolated Pd species
with the frame-structure of PPN through the electron transfer
between nitrogen atoms in the PPN and Pd NPs, as was also reported
in earlier study.66 Indeed, a positive binding energy shift by +1.2 eV
in the high resolution fitted N 1s XPS spectra (Figure S6, SI) for
Pd@PPN in comparison with the as-synthesized PPN has been
observed, and it was attributed to the charge transfer of electrons
from N atoms to Pd.81 It also agrees very well with the electronic
properties analysis from DFT calculations later.
Figure 5: TEM images of PPN (a-c) & Pd@PPN (d-f) at low
magnification, (g-i) at high magnification, respectively. HAADF-STEM
image of the Pd@PPN with the elemental mapping is also provided.
To examine the distribution, shape, location with mean particle
size of Pd NPs on the POP polymer matrix, transmission electron
microscopy (TEM) analysis of the PPN and Pd@PPN has been
performed and the data are shown in Figure 5. TEM images of
the PPN material clearly shows that it consists of thin
nanosheets-like morphology having diameter of 33 - 53 nm and
length of 207 - 247 nm (Figs. 5a-c) as composed by randomly
self-assembled multiple irregular-shaped nanotubes. TEM
images of Pd@PPN (Figs. 5d-f) show homogenous spherical Pd
NPs with average diameter of 3.2±0.5 nm (Fig. 5f). It is visible
from TEM image (Fig. 5d) that PPN fibers of 23-nm-diameter
and 284-nm-length are self-assembled to produce a 3D cage like
structure in which the small Pd-NPs are strongly anchored. A
dual Pd-NPs size distribution with the mean diameter of about
1.86 ± 0.36 and 2.89 ± 0.25 nm is found (Fig. 5g). It is assumed
that small size Pd-NPs (1.86 nm) can easily occupy the interior
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 7
Please do not adjust margins

Please do not adjust margins
Page 9 of 20
Green Chemistry
ARTICLE
Journal Name
Catalytic conversion of Vegetable Oils over Pd@PPN Catalyst. pressure) later. Besides, it could be observed that the formation
View Article Online
DOI: 10.1039/C9GC03803E
of C17 alkane is more favorable than C18 alkane on our
Triglycerides and saturated/unsaturated fatty acids are
fabricated Pd@PPN catalyst, demonstrating the preference of
industrially used in bio diesel and soap preparation. We
decarboxylation/decarbonylation
pathways
over
the
selected stearic acid as the model compound in order to check
the catalytic activity of Pd@PPN catalyst to generate diesel
additives. Liquid phase hydrogenation of stearic acid gives rise
to the n-heptadecane (C17 alkane), n-octadecane (C18 alkane),
n-octadecanol (C18-OH) and a very small amount of other
cracking products. From literature, the process of alkane
formation may involve three different routes.19, 30, 70, 82, 83 The
first pathway is the initial hydrogenation of fatty acid to alcohol
along with the subsequent hydro-dehydration and produces the
corresponding alkane with the same number of carbon as the
initial fatty acid. The other two remaining routes involves the
direct decarboxylation pathway (CO2 removal) of fatty acid and
the decarbonylation pathway (CO removal), both produce the
alkane with one carbon less in the hydrocarbon chain. Previous
studies reported that the hydrogenation-hydrodehydration
pathway is more feasible at low temperature and low pressure,
while the decarbonylation/decarboxylation pathways are more
feasible at high temperature and high pressure.30
hydrogenation-hydrodehydration mechanism and therefore
only the decarboxylation/decarbonylation mechanisms will be
evaluated by DFT simulations in later sections.
Firstly, we investigate the distribution of reactant and products
of the catalytic hydrogenation of stearic acid over Pd@PPN
catalyst as a function of time and the results are presented in
Figure 7a. In this investigation, the hydrogenation reaction of
stearic acid (100 mg, 0.350 mmol) was carried out in water (70
mL) over Pd@PPN catalyst in a stainless-steel reactor under 20
bar hydrogen pressure, at 150 °C temperature for 16 h (the
choice of reaction parameters such as temperature and
pressure are also investigated and optimized in Figs. 7b,c). The
reaction was monitored by GC-FID technique based on actual
authentic sample. At the initial 1 h of reaction, only 7%
conversion of stearic acid with significant amount of stearyl
alcohol (45%), a moderate amount of C18-alkane (15%
selectivity) and very negligible quantity of C17-alkane and other
cracking products with 5% and 2% selectivity were achieved. At
4 h of reaction, the conversion of stearic acid was 30% and the
selectivity of stearyl alcohol reaches its maximum of 78%, while
the selectivity of C18-alkane and C17-alkane were 9% and 19%,
respectively, indicating the formation of stearyl alcohol as a
primary product via hydrogenation at the first step of the
reaction. With the advancement of the reaction, the selectivity
of the stearyl alcohol significantly dropped to 7% with the
increase of C17 selectivity (78%) along with negligible amount
of C18 (9%) furnished after 6 h course of the reaction. After 10
h of reaction, C17 alkane is the major product with high
selectivity (81%), while small amount of C18 alkane (10%
selectivity) and insignificant amount of stearyl alcohol (1%
selectivity) were observed. The diminishment of the stearyl
alcohol selectivity was observed after 12 h with 89% of stearic
acid conversion and 82% C17 alkane selectivity. We also noticed
that 90% of stearic acid conversion along with 83% C17
selectivity was observed with the prolongation of reaction time
up to 14h, and is not much improved after that. Therefore, the
optimized reaction time was chosen of 14 h for screening the
effect of other reaction parameters (temperature and H2
Figure 7: Evolutions of reactant and product distributions for stearic
acid hydrogenation over Pd@PPN catalyst as a function of time (a);
temperature (b) and H2 pressure (c). Reaction conditions: stearic acid
(100 mg, 0.350 mmol), Pd@PPN (20 mg, 2.6 mol % of Pd feed), water
(70 mL). The designation “other products” denotes the cracking
products, smaller alcohols and esters.
J. Name., 2013, 00, 1-3 | 8
Please do not adjust margins

Green Chemistry
Please do not adjust margins
Page 10 of 20
Journal Name
ARTICLE
To investigate the effect of reaction temperature on the selectivity alcohol), other cracking products (≤ C16 alkanes),VieawnAdrtieclsetOenrlinies
of stearyl alcohol, C17 and C18 alkanes, a series of reaction were tabulated in the Table 1. It is noteworthDyOtIh: 1a0t.1a0l3l9t/hCe9GcCat0a3l8y0s3tEs
carried out under different temperatures (ranging from 80 to 150 °C) were prepared following a similar catalyst synthesis procedure.
as shown in the Fig. 7b. It could be seen that both the conversion of Data in Table 1 demonstrated that Pd@PPN catalyst exhibited
stearic acid and the selectivity of C17-alkane are increased with the best performance among all the examined Pd catalysts.
increasing temperature. This observation suggests that high reaction Pd@SiO2, Pd@C, Pd@TiO2, and Pd@ZrO2 catalysts furnished
temperature favours the formation of C17 alkane via the the desired C17 alkane product with the selectivity of 8.7%,
decarbonylation/decarboxylation pathway. Srifa et al. reported that 41.3%, 10.5% and 33.8%, respectively. The poor selectivity of
C17 alkane was formed from the intermediate stearyl alcohol by the C17 alkane in comparison with C18 alkane for Pd@SiO2 catalysts
decarbonylation/decarboxylation reaction, which are supposed to revealed that hydrodehydration reaction pathway is promoted
be equilibrium with stearyl alcohol during the HDO process.84 In with the surface acidity of the SiO2 support. Similar finding on
constrast, increasing temperature decreases the selectivity of the the surface acidity promotion of hydrodehydration route with
stearyl alcohol gradually from 55 % (at 80 °C) to 4 % (at 150 °C). It is Ru@TpPON catalyst was reported by Bhaumik and co-
supposed that the production of C18 alkane via stearic acid workers.29 In strong contrast, we achieved higher C17 alkane
hydrogenation by the alcohol pathway is an endothermic reaction, selectivity for Pd@PPN in comparison with the Pd@C catalyst
while the decarbonylation/decarboxylation to C17 alkane is an (Entry 2, Table 1). This result could be attributed to the electron
exothermic reaction.30 The high temperature is generally preferred rich Pd-NPs, which may activate surface bound H2 on the high
to execute the first, while low temperature is convenient for the catalytic performance.87 Pd@ZrO2 possessed poor activity and
second pathway.29 Almost complete conversion of stearic acid (90%) gave alcoholic products (entry 4, Table 1), possibly due to the
with 83% selectivity on C17 alkane, about 5% of C18 alkane and poor dispersion of nanoparticles, weaker metal-support
insignificant amount of stearyl alcohol (C18-OH) and cracking interaction, low surface area, and low mass transfer.
products were generated at 150 °C.
Furthermore, in order to investigate the structural sensitivity of
Finally, the influence of H2 pressure (from 5 bar to 30 bar) on the the Pd@PPN catalyst for the HDO of carboxylic acids, we have
catalytic conversion of stearic acid was also studied and the result is synthesized another catalyst using the liquid phase reduction
presented in the Fig. 7c. It is important to note that when the method (details of this catalyst synthesis are provided in the SI).
reaction was conducted under Ar atmosphere (no H2 pressure), we By this way, we were managed to synthesize another Pd NP
observed only 0.2 % of C17 product. When the reaction was hosted in the Porous-Organic-Polymer (called Pd@PPN-1)
conducted under the H2 pressure of 5 bar, only 26% conversion of where the size of Pd-NPs was ~8.9 nm as evidenced from the
stearic acid and 77 % selectivity of stearyl alcohol (C18-OH), 18% of TEM images analysis (Fig. S7, SI). For this catalyst, we achieved
C18 alkane and 5% of C17 alkane were obtained. However, the only 63.7% conversion of stearic acid towards 51.3% selectivity
product distribution is significantly changed at higher pressure (Fig. of C-17 (Entry 6, Table 1), much lower than the smaller Pd@PPN
7c). When the reaction was carried out under 15 bar H2 pressure, catalyst with the particle size of 3 nm. The observed results
72% conversion of stearic acid was achieved with 69% selectivity of show that the HDO conversion of carboxylic acids on Pd@PPN
C17 alkane while the selectivity of stearyl alcohol (C18-OH) and C18 catalysts is structural sensitive, and is also consistent with the
alkane were significantly reduced to 15% and 12%, respectively. An high activity of step sites presented in large density on the
abrupt increment in C17 alkane selectivity observed after 15 bar smaller size Pd@PPN, which will be discussed later.
pressure suggests that decarboxylation/decarbonylation pathways
are encouraged at high H2 pressure, in good agreement with the
previous literature reports.85 At the optimized H2 pressure of 20 bar,
Table 1: Comparison of stearic acid conversions over different
Palladium-based catalysts
Selectivity (%)
C18 C18 Others
alkane alkane alcohol
90 % conversion of stearic acid along with selectivity of 83% for C17
alkane, 5% for C18 alkane and 10 % for other product was acquired.
When the H2 pressure was increased to 30 bar, the complete
conversion of stearic acid was observed, however the selectivity of
C17 alkane was reduced to 72% along with the increment of cracking
products with 18% selectivity, indicating the degrading of
decarbonylated product at higher pressure via cracking reactions,
which was also reported in earlier study.31 H2 pressure also takes part
in improving the catalyst against deactivation, as it prevent the
oligomerization of unsaturated hydrocarbons by hydrogenation
them to the corresponding alkanes, thus hinder the formation of
coke in the reaction medium.86
Conversion
(%)
Entry Catalyst
C17
1
2
Pd@SiO2
Pd@C
56.8
63.2
8.7
29.5
9.3
9.5
5.1
9.1
7.3
41.3
3
4
5
Pd@TiO2
Pd@ZrO2
Pd@PPN
49.7
64.5
89.6
10.5 37.6
1.5
48.7
0
0
0
0
5.1
10.6
83.5
6.1
6*
Pd@PPN-1
63.7
51.3 12.4
0
0
Reaction conditions: Stearic acid (100 mg, 0.350 mmol), Pd-catalyst
(20 mg, 2.6 mol% of Pd feed), water (70 mL), 150°C, H2 (20 bar), 14 h.
*The size of Pd-NPs in Pd@PPN-1 is ~8.9 nm.
To understand the role of PPN to the activity of Pd-NPs, a series
of Pd-NPs based catalytic model hosted on the various supports
including SiO2, C, TiO2, and ZrO2 were investigated under the
same above-mentioned optimized reaction conditions. The
evolution of reactant and product distributions with the
selectivity of C17 alkane, C18 alkane, C18 alcohol (stearyl
As different vegetable oils comprise unique compositions of
fatty acids and impurities, we selected four typical
commercialized oils, i.e. palm oil, soybean oil, sunflower oil, and
rapeseed oil. These oils are subjected for hydrogenation
treatment with Pd@PPN catalyst under optimized reaction
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 9
Please do not adjust margins

Please do not adjust margins
Page 11 of 20
Green Chemistry
ARTICLE
Journal Name
conditions (Figure 8). The fatty acid compositions of these oils to the diffusion of particles out of the interior caviVtiieewsAartnicdle Othnleinier
are listed in the respective Table T2, SI. Corresponding alcohols deposition over external facet. Increasing ofDavOeI:ra10g.e10p3a9r/tCic9lGeCsi0z3e8a0n3dE
and esters intermediates of the present saturated fatty acids its severe aggregation after 5th cycle (Fig. 9d) caused the drastic drop
were identified at the initial reaction course, followed by in stearic acid conversion from 90.5% to 76.8%. The surface and
progressive increase yield of respective alkanes with the further morphology of the PPN nanotube was unchanged after the 4th and
advancement of catalytic reaction. Mechanistically, 5th catalytic cycle although the width of the nanotube was a little bit
unsaturated fatty acids (oleic acid, linoleic acid) consisting of the enlarged after each cycle compare to the fresh one. The PPN
double bond can be easily transformed into their corresponding structure proved to be stable with no associated structural
saturated analogue under the present reaction conditions. Acyl degradation. To examine any leaching that took place during the
C-O bond cleavage is likely to occur resulting in alcohol reaction, we conducted a hot-filtration test under optimized reaction
intermediate followed by hydrogenolysis. Another competitive conditions. ICP-AES analysis results suggest the Pd content in the
decarbonylation/decarboxylation route of the aldehyde reused Pd@PPN catalyst after the fourth and fifth catalytic runs are
intermediates leading to the formation of alkanes with one 0.413 and 0.397 mmol/g, respectively, which validated that that
carbon atom less is also associated.14
there was an insignificant leaching of Pd species. In addition, the ICP-
AES results of both the used catalyst and the solution after reaction
(with consideration of experimental error) definitely revealed that
the Pd-NPs were barely leached in this reaction system owing to
strong attachment with highly cross-linked interconnected a three-
dimensional rigid framework at the molecular level.
Figure 8: Hydrogenation of various vegetable oils over Pd@PPN
catalyst. Reaction conditions: Vegetable Oils (0.2 g), Water (70
mL), 150 °C, 20 bar H2 pressure, Time 14 h, Catalyst (100 mg,
13.07 mol % of Pd feed).
Stability testing.
Figure 9: TEM images of used Pd@PPN catalyst after 4th (a-c) and 5th
We performed recyclability study of the newly designed Pd@PPN
catalyst to evaluate stability under optimized reaction conditions
considering stearic acid as model compound (Figure S8, SI). After
each use, the black powder Pd@PPN was isolated by centrifugation
followed by washing with methanol (three times) and acetone (two
times), drying at 100 °C in an oven for 8 h. A significant drop in stearic
acid conversion from 90.5% to 76.8% was observed after five runs.
Besides, a significant decrease of the C17 alkane selectivity from
85.8% to 69.4% occurred along with the gradual enhancement in the
stearyl alcohol (C18-OH) selectivity from 10% to 25% (Fig. S8, SI). A
drop in the efficiency of Pd@PPN catalyst could be attributed to the
accumulation of carbonaceous deposits on the catalyst surface or
slight trapping of residual reactants or products inside the pores,
thus blocking accessibility to the active sites. The location,
distribution, and size of the Pd-NPs of the reused catalytic systems
have been investigated by conducting TEM analysis at 4th and 5th
catalytic cycles, respectively in Figure 9.
(d-f) catalytic runs.
Formation of high density step sites on Pd@PPN
To gain more information on the interaction between Pd NPs and the
polymer structure, we studied the interaction of the monomer unit
of the polymer, called umPPN molecule, onto the surface of Pd NPs
using the reduced-cluster model of 3 nm particle size (Fig. 2c). Note
that this size of 3 nm was also characterized using TEM image as
shown in Fig. 5d. As mentioned earlier in the literature, the
interaction between benzene ring and Pd(111) surface was
determined as the combination between covalent bond and long-
range vdW interaction, making the very strong binding of aromatic
rings on flat Pd surface.54 Six different adsorption configurations of
umPPN with the Pd332 cluster are presented in Figure 10 and the
binding energies relative to the adsorption energy of the most stable
configuration are indicated. In the most stable configuration in
Fig.10a, umPPN binds to the Pd332 cluster via the interaction of 5
aromatics rings (2 rings from each TPA fragment and 1 ring from the
p-xylene fragment). In this configuration, two TPA fragments of
umPPN both are adsorbed at two corner sites of the Pd332 cluster. If
umPPN binds to the Pd332 cluster via the interaction of only 4
Dual size distribution of Pd NPs over carbonaceous PPN nanotube
was observed after 4th catalytic run with the uniform distribution in
the range of 3.2 - 4.0 nm and 7.4 - 11.4 nm, respectively (Fig. 9a-c).
Surprisingly, after 5th catalytic run Pd-NPs size increases enormously
to 15 - 19.5 nm and 24 - 31 nm (Fig. 9d-f), which could be attributed
J. Name., 2013, 00, 1-3 | 10
Please do not adjust margins

Green Chemistry
Please do not adjust margins
Page 12 of 20
Journal Name
ARTICLE
aromatics rings in Fig. 10b (2 rings from the first TPA fragment at the consistently the size of 3nm for Pd NPs anchoring on the polymer
View Article Online
corner, 1 ring from the p-xylene fragment and only 1 ring from the substrate. Besides, the strong interactionDoOfI:P1d0.1c0l3u9s/tCe9rGwCi0th380th3Ee
other TPA fragment), the binding energy is +56 kJ/mol less stable frame structure of the polymer at the step-edge and corner binding
than configuration (a). In case umPPN still binds to the Pd332 cluster modes could be a factor generating high density of step sites for the
via the interaction of 4 aromatics rings but the first TPA adsorbs at fabricated Pd NPs. Due to the non-planar structure of TPA fragments
steps site instead of corner site (Fig. 9c), the binding energy is +71 (three benzene rings are tilted forming corrugated structures) and
kJ/mol less stable than configuration (a). If the benzene rings from p- the very strong binding of aromatic rings to Pd terrace facets, the as-
xylene and from the second TPA fragment do not participate into the synthesized Pd NPs could be nucleated at corrugated sites formed by
interaction, the binding energies become significantly weaker by TPA fragments, and then grew along the chain of benzene rings and
+299 kJ/mol and +323 kJ/mol, where umPPN binds to the Pd332 finally formed the NPs structure with lots of surface corrugations and
cluster via the interaction of only 2 aromatics rings (Figs. 9e,f). Finally, high density of step sites. It is important to note that although the
the least stable configuration is when umPPN molecule adsorbs on monomer unit prefers adsorption at the corner and step edges, the
flat terrace sites via the coordination of only one aromatic ring of one steric effect could prevent the monomers adsorb in close vicinity to
TPA fragment (Fig. 9d), and the computed binding energy for this each other. Therefore, the coverage of the monomer at step edge
configuration is +359 kJ/mol weaker than in Fig.9a. It could be might not be high enough to block all step sites, and there is still a
observed that the more benzene rings of umPPN interacting with the large portion of free step sites on the Pd NPs (details explanation is
Pd332 cluster, the more stable is the interaction. In the most stable presented in Fig. S16, SI). The presence of step sites with high density
configuration in Fig. 9a, the computed binding energy is ~-587 consequently could be one of the important factors contributing to
kJ/mol. From the top-view of configuration in Fig. 9a, one full unit the high activity of the Pd@PPN for the HDO conversion of fatty
monomer umPPN of the polymer structure nicely covers and wraps acids. The next sections will therefore evaluate the activity of those
around one exposed facet of the Pd332 cluster, which could explain step sites in the HDO reactions.
Figure 10. Different configurations for the interaction between 3 nm Pd nano-particle and the umPPN unit molecule. The side view and top
view are presented for each configurations. Binding energies relative to the most stable configuration (configuration a) are also shown. F4,
B5 step sites and corner site are highlighted by big yellow, red and green balls, respectively.
accepted that there are three possible mechanisms for the HDO of
carboxylic acids: (i) the decarbonylation (DCN) pathway where
oxygen atoms are removed from the initial carboxylic acids reactant
in the form of CO, (ii) the decarboxylation (DCX) where oxygen atoms
Mechanistic study on the conversion of butanoic acid on Pd@PPN.
HDO conversion of butanoic acid on terrace sites of Pd(111). The
mechanism for HDO reaction of linear chain acid on Pd(111) surface
has been comprehensively studied in literature.70, 82, 83 It is widely
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 11
Please do not adjust margins

Please do not adjust margins
Page 13 of 20
Green Chemistry
ARTICLE
Journal Name
are removed in the form of CO2 and (iii) the reductive deoxygenation 91 kJ/mol (TS4-1, Fig. 11), releasing CO and forming the C3
View Article Online
DOI: 10.1039/C9GC03803E
(RDO) where oxygen atoms are removed in the form of water (also hydrocarbon product. However, the most feasible initial activation of
called as hydrogenation-hydrodehydration pathway). In the first two butanoic acid on Pd(111) surface is the OH activation with a barrier
cases, the number of carbon atoms in the products is one atom less of 63 kJ/mol (TS1-2, Fig. 11), which initiated the DCX pathway.
than in the reactant molecule while for the last case, the number of Subsequent conversion of butanoic acid processed along the DCX
carbon atoms of the product remained identical as the initial mechanism includes the Cα-H activation at the 2nd step after initial
reactant. There is a general agreement from experimental studies in OH dissociation (barrier energy of 131 kJ/mol, TS2-1) and the
literature that Pd-based catalyst favors the DCX and DCN pathway of decarboxylation reaction step 3 with the barrier of 99 kJ/mol (TS3-1,
carboxylic acids over the RDO pathway in producing the product with Fig. 11), releasing CO2 and forming the C3 hydrocarbon product. As
one less carbon atom.88-91 Besides, to facilitate the RDO mechanism, could be seen, the Cα-H activations (TS1-1 and TS2-2) play an
more hydrogen is consumed, however its supply is being in short in important role in both DCX and DCN pathways, agrees very well with
refineries and facing growing demand for other competitive the data reported earlier and is also consistent with experimental
82, 83
applications such as CO2 reduction, fuel cell… and therefore observations in the literature.10,
It is important to note that
developing the catalyst that can facilitate the RDO pathway is less regardless of the more favourable pathway, both the DCX and DCN
desirable.92-94
pathways produce the same alkane product with one less carbon
There is a known competition between the DCX and DCN pathways than from the carboxylic acid feedstock. Furthermore, if the HDO is
on Pd(111) surface. Using propanoic acid as the model compound processed via the DCN pathway, the Water Gas Shift reaction could
and performed DFT simulations on Pd(111) surface, Lu et al. reported also convert the produced CO to CO2, the same products of the DCX
that the DCX and DCN pathways were initiated by the OH activation pathway, therefore it is very challenging to distinguish between DCX
and Cα-H activation, respectively.70 By evaluating the detailed and DCN pathway by experiments.
microkinetic modelling, the authors reported that the DCN
mechanism is more feasible than the DCX one. The initial Cα-H
activation and the subsequent dehydroxylation scission of the CO-
OH bond were identified as the rate-determining steps during the
DCN pathway.70, 82 Actually, the crucial role of Cα-H activation as the
precursor to facilitate the DCN pathway was also confirmed in the
kinetics isotope study of deuterium-labelled propanoic acid HDO
over Pd catalyst by Lugo-Jose et al.83 Besides, the removal of strong-
bound intermediates or products such as unsaturated hydrocarbons
(via hydrogenation) and CO was also stated as one of the “bottle-
necks” hindering the rate of conversion.82
To setup the reference for evaluating the activity of steps sites during
the HDO on Pd@PPN catalyst, we revisited the activation and
conversion of fatty acid on Pd terrace sites of the p(4×4) Pd(111)
periodic slab model via both DCX and DCN pathways using butanoic
acid as the model compound. It is important to note that in this study
we followed the most possible mechanism for the DCX and DCN
pathways that were identified and reported for HDO of propionic
acid on Pd(111) surface in Luo et al.70, 82 The scheme of reactions,
together with the transition states and activation barriers for the
reactions along those pathways are presented in Figure 11. Note that
the reactions along this scheme are the most feasible occurring at
each steps, and other less kinetically feasible reactions are shown in
SI. Butanoic acid adsorbs on Pd(111) surface with an adsorption
energy of -83 kJ/mol (Figure S17a). Three possible initial activations
of butanoic acid in Fig. 11 could open three different pathways of the
HDO. If butanoic acid is processed via the initial Cα-H activation with
the barrier of 117 kJ/mol (TS1-1, Fig. 11), then the HDO conversion
follows the DCN pathway. Subsequent reactions along the DCN
pathway are the Cβ-H activation at the 2nd step with a barrier of 81
kJ/mol (TS2-1), followed by the subsequent dehydroxylation in step
3 with activation barrier of 112 kJ/mol (TS3-1). The DCN reaction is
then occurred in the next step 4 with computed activation barrier of
Figure 11. Hydrodeoxygenation (HDO) of butanoic acid on Pd(111)
surface via three different pathways. Reactions along the
Decarbonylation (DCN) pathway are indicated by blue arrows, while
reactions along the Decarboxylation (DCX) pathway are indicated by
red arrows. The reductive deoxygenation (RDO) pathway is indicated
by green arrow.
J. Name., 2013, 00, 1-3 | 12
Please do not adjust margins

Green Chemistry
Please do not adjust margins
Page 14 of 20
Journal Name
ARTICLE
The third pathway for the HDO of butanoic acid is the reductive sites is only 23 kJ/mol compared to a much bigger gap of 54 kJ/mol
View Article Online
DOI: 10.1039/C9GC03803E
deoxygenation (RDO) pathway, which produce the final alkane on normal terrace site of Pd(111) surface. Therefore, Pd step sites
product with the same carbon atom as in carboxylic acid resource. promote the initial Cα-H activation much stronger than the initial OH
This pathway is initially processed by the hydrogen-shuttling activation and will have great impact to facilitate the preference of
mechanism, where water acts as a hydrogen shuttle to transfer the DCN pathway over the DCX pathway.
adsorbed H on Pd(111) surface to the OH group of butanoic acid and
facilitating the dehydration, forming H2O. The activation barrier for
this reaction is 126 kJ/mol (TS1-3, Fig.11). However, it should be
mentioned that this type of H-shuttling reaction will suffer from extra
entropic energy penalty for locating the water molecule to the exact
positon in the transition state, as was reported by Gunasorria et al.95,
making it less favourable compare to the DCX and DCN pathways and
will only be feasible at high enough pressure of hydrogen. This
observation could explain the high selectivity towards the DCX and
DCN pathways obtained from the HDO of carboxylic acids obtained
in our study. There are other reactions could occur during the initial
step of RDO pathways, but with higher barriers and therefore less
competitive and are described in Figures S18 and S19 of the SI.
HDO conversion of butanoic acid on Pd step sites. The model of
p(4x8) with 4 missing rows on the top layer is used to evaluate the
activity of Pd step sites. It is worth to mention that in this model,
there are the presence of extended terrace sites next to the B5 and
F4 step sites and is an excellent representative for the 3 nm particle
model. This model has more active sites and is therefore more
appropriate than the use of periodic Pd(211) surface in literature. We
have evaluated the activities of B5 steps site and F4 steps site for few
reactions of butanoic acid and found that their activities are quite
similar (Figure S20, SI). Therefore, we only present the data observed
at B5 step sites to discuss the activity of step sites in HDO reaction.
The HDO energy profiles for butanoic acid conversion on Pd terrace
site and step site are presented in Figure 13, while the structure of
the transition states along those pathways are shown in Figure 12.
Firstly, we investigated the initial activation of butanoic acid.
Butanoic acid adsorbs at B5 step site with an adsorption energy of -
95 kJ/mol (Fig. S20a, SI), which is 22 kJ/mol stronger than it is on flat
terrace site. Initial OH activation at B5 step site is slightly promoted
with the barrier of 56 kJ/mol (TS1-2, Fig. 12b), compared to 63 kJ/mol
The subsequent conversion of butanoic acid along the DCX pathway
at flat terrace site (Fig. 11). However, the initial Cα-H activation is
much stronger promoted at step site, and the barrier of this
sites in step 2; (e) Cα-H activation on lower terrace sites in step 2; (f)
activation is brought down from 117 kJ/mol at terrace site to only 79
kJ/mol at step site (TS1-1, Fig. 12a). Interestingly, the hydrogen-
conversion of butanoic acid along the DCN pathway at Pd B5 step
shuttling reaction that initiates the RDO pathway in Fig. 11 is not
structural-sensitive, and the activation barrier for this reaction on Pd
B5 step site is 121 kJ/mol (TS1-3, Fig. 12c). Another competitive two-
step Hydrogenation-dehydroxylation mechanism for the RDO
pathway, which is not feasible on terrace site (Figure S19, SI) but
strongly promoted on step site, is discussed in Figure S23, SI.
Nevertheless, the barrier of that reaction is still higher than the
hydrogen-shuttling mechanism. From all possible activations in the
initial step, the OH activation is still the lowest barrier reaction,
however, the Cα-H activation now has the barrier energy much closer
to the OH activation. The difference between barrier energy for the
initial OH activation and Cα-H activation for reaction occurred at step
Figure 12. Activation and HDO conversion of butanoic acid via DCX
and DCN pathways at Pd B5 step sites. (a) Initial Cα-H activation; (b)
Initial O-H activation; (c) Initial hydrogenation-dehydration reaction.
after the initial O-H activation: (d) Cα-H activation on upper terrace
Subsequent decarboxylation reaction (-CO2) in step 3. The
sites after initial Cα-H activation: (g) subsequent Cβ-H activation in
step 2; (h) CO-OH scission in step 3; (i) decarbonylation (-CO) reaction
in the 4th step.
Next, we evaluate the subsequent reactions along the DCX pathway
for the butanoic acid HDO at step site, which is processed via the
initial OH activation, follow the similar mechanism occurred on
normal terrace sites of Pd(111) surface as showed in Fig. 11.
Surprisingly, almost subsequent activations in step 2 of the DCX
pathway at step sites become more difficult than they are on normal
terrace sites of Pd(111) surface. The product of initial OH activation,
butyrate, adsorbs on step site in the bidentate configuration and
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 13
Please do not adjust margins

Please do not adjust margins
Page 15 of 20
Green Chemistry
ARTICLE
Journal Name
coordinate to the step site via both two O atoms as shown in Fig. The energy profiles for the conversion of butanoic acid via DCX and
View Article Online
DOI: 10.1039/C9GC03803E
S21b. Due to the stronger binding of butyrate on step sites (39 kJ/mol DCN pathways on terrace site and step sites are presented in Figure
stronger than its binding energy on terrace site), a noticeable energy 13. Overall, it could be seen that Pd step sites strongly promote the
penalty is needed to bend the adsorbed molecule closer to the rate-limiting step during the DCN pathway (initial Cα-H activation
surface for further activations of C-H bonds or C-C scission, resulting TS1-1 and later dehydroxylation reaction TS4-1) and also in the same
in higher activation barriers for almost all reactions in step 2 of the time inhibits the competitive DCX pathway. Although both the DCX
DCX pathway. As shown in Figs. 12d,e, Cα-H activation has the higher and DCN pathway could all produce the same alkane product, we
barrier of 145 kJ/mol if processed by upper-terrace sites next to B5 could infer from the energy profiles presented in Fig. 13 that the DCN
site (TS2-2, Fig. 12d), or 158 kJ/mol if processed by lower-terrace pathway of fatty acids HDO process on the Pd@PPN material might
sites next to B5 site (Fig. 12e). Other reactions in step 2 are also have become more dominated than the DCX pathway, however it is still
higher barriers than they are on terrace site (Fig. S21c,d, SI). The very difficult to be validated by experiment. It is also worth noting
same trend is also found in the decarboxylation reaction releasing that the density of step sites is a function of particle size, and is
CO2 in step 3 (barrier of 111 kJ/mol, TS3-2, Fig. 12f), which is 12 reduced drastically when particle size is larger than 3 nm (the density
kJ/mol higher than the reaction on normal terrace site of Pd(111). of B5 step sites on 8 nm particle is only 15% of the density of B5 step
The high barriers of reactions in step 2 and the increment in sites on 3 nm particle).58 For other catalysts in Table 1 that also can
activation barrier for the subsequent decarboxylation reaction make facilitate the DCN/DCX pathways to produce alkane with one less
the reactions along the DCX mechanism more sluggish and therefore, carbon atom from the carboxylic acid resource such as Pd@C (entry
the DCX pathway is kinetically more hindered at step sites.
2) and Pd@PPN-1 (entry 6), their particle sizes are also in the range
The DCN pathway is initiated by the Cα-H activation of butanoic acid, of ~9 nm (Figure S7, SI). Therefore, the density of step sites on the
which is strongly promoted at step site. Similar to the mechanism fabricated Pd@PPN catalyst (particle size 3 nm) is much larger and
occurred on normal terrace site of Pd(111) surface, subsequent resulting in the significant higher conversion of stearic acid and very
reactions along the DCN pathway are also kinetically feasible and high selectivity of C-17 hydrocarbon product on our Pd@PPN catalyst
slightly promoted at step sites. The step 2 consists of the Cβ-H compared to other Pd-based catalysts as presented in Table 1. The
activation with activation energy of 67 kJ/mol (TS2-1, Fig. 12g), which fact that (i) there is the presence of high density of step sites on
is 14 kJ/mol lower than on normal terrace site. After the Cβ-H Pd@PPN material and (ii) both of those rate-limiting reactions are all
activation in step 2, the product has the conjugated structure CH3- promoted at Pd step sites could result in very high activity of
CH=CH-COOH and the dehydroxylation reaction from this molecule Pd@PPN for the HDO reaction.
in step 3 can be processed with barrier of 89 kJ/mol (TS3-1, Fig. 12h),
forming the ketene-type product CH3-CH=CH-C=O. The subsequent
decarbonylation reaction in step 4 releasing CO and hydrocarbon
product with one carbon atom less is also very feasible with
activation barrier of only 77 kJ/mol (TS4-1, Fig. 12i).
Electronic properties of Pd@PPN.
To gain more understanding about the charge transfer between Pd
NP and the polymer frame structure, the Bader chargers of all atoms
in the isolated umPPN molecule and for the adsorption of umPPN on
Pd332 cluster are calculated. The charges of several Pd atoms on the
Pd332 cluster coordinated with adsorbed umPPN in its most stable
configuration as shown in Fig. 10a, and the changes of the Density of
states projected on d-orbital (PDOS) for the corner Pd site interacting
with umPPN referenced to the clean Pd NP are shown in Figure 14.
In the isolated umPPN molecule, the charges of N atoms are -1.24.
However, when umPPN is adsorbed on the Pd332 cluster in its most
stable configuration (Fig.10a), the computed charges on N atoms are
-1.12, which is more positive (less negative) than the charge of N
atom in the lone umPPN. Since the XPS chemical shift is very sensitive
to the charge transfer96, 97, the more positive charge induced on N
atoms in Pd@umPPN than in the lone umPPN structure therefore is
excellently consistent with the positive shift of N 1s core-level
binding energy by +1.2 eV observed in the experimental XPS spectra
for Pd@PPN in comparison with the as-synthesized PPN (Fig. S6, SI).
More information about the interaction between Pd-NP and the
polymer frame structure could be revealed when we analysed the
charge of Pd atoms and their Projected Density of States (PDOS). The
charge of the Pd corner site on the clean Pd-NP is -0.043, but is
changed significantly to +0.144 after interacting with the umPPN
molecule, as is shown in Fig. 14a. The withdrawal of electron from
the Pd corner site is consistent with PDOS of it on clean Pd-NP and
Figure 13. Energy profiles for HDO conversion of butanoic acid on
terrace sites (Pd(111)) and on step sites (B5). Reactions along the
DCN and DCX pathways are indicated in blue and red colours,
respectively. Reactions on terrace sites and step sites are indicated
by solid lines and dash lines, respectively.
J. Name., 2013, 00, 1-3 | 14
Please do not adjust margins

Green Chemistry
Please do not adjust margins
Page 16 of 20
Journal Name
ARTICLE
on Pd@umPPN as shown in Fig. 14b, evidenced from the reduce from factor contributing to the high activity and high selectivity of our
View Article Online
92% to 88% for the d-orbital occupancy of clean Pd-NP and fabricated catalyst for the HDO of carboDxyOliIc: 10a.c1i0d3s9./CIn9dGeCe0d3,80th3Ee
Pd@umPPN, respectively and inducing positive charge for Pd corner addition of one e- improved markedly both the reaction kinetics and
site on Pd@umPPN. The PDOS of Pd corner site is broaden and selectivity of Pd-catalyst for formic acid decomposition was reported
shifted further from the Fermi-level (EF) upon interacting with the recently in Wang et al.102 More importantly, the negative charge
umPPN structure was argued by Yildirim et al. as a characteristic induced on Pd atoms upon the interaction with the polymer frame
indication for this kind of strong chemisorption enhanced by van der structure has substantial influence on adsorption and therefore, on
Waals interaction between aromatic rings and transition metals.98
reactivity. Besides the rate determining steps that were identified for
the DCN/DCX pathways (the C-H activations and dehydroxylation
reactions), the hydrogenation of hydrocarbons and the removal of
strongly bound intermediates such as CO and unsaturated
hydrocarbons were also reported as the “bottle-necks” of the HDO
82
conversion.70, Hydrogenation of unsaturated hydrocarbons was
generally considered as surface insensitive reactions,103-105 so even
with high density of step sites in our Pd@PPN catalyst, this reaction
is still not much facilitated. On another side, the adsorption energies
of all alkene and alkylidyne intermediates are very sensitive to the
surface charge of coordinated metal atom, and therefore the charge
transfer effect will have greater influence to the kinetic performance
of the hydrogenation reaction. Alkylidyne intermediates are bound
very strongly on the surface of transition metals and are generally
considered as a poison to the catalyst during the hydrogenation.100,
101, 106 When the charge of surface Pt atom is changed from -0.005 to
-0.029, the computed adsorption energies for ethylene and
ethylidyne on Pt(111) surface were reduced by 19.2 and 15.4kJ/mol,
respectively, consequently increase the hydrogenation rate by an
order of magnitude as was reported in Gunasooriya et al.100
Furthermore, CO binding energy is also weaker on more negative
surface atom, although with smaller magnitude than those above-
mentioned unsaturated hydrocabons.99, 100 Therefore due to a net
charge transfer from the umPPN molecule to the Pd-NP inducing
more negative charge of surface Pd atoms makes the removal of the
strongly bound intermediates formed during the HDO process
become easier and enhance the activity of our fabricated catalyst. It
should be noted that when the coverage of umPPN is bigger in
realistic cases, the charge transfer to Pd NP is then stronger and
results in higher promotional effect to the catalytic activity. More
insights into the electronic interaction between umPPN molecule
interacts with smaller Pd23 cluster done by Gaussian package via the
natural population analysis (NPA) and natural bond orbital (NBO)
analysis charge calculations are presented in Tables T4, T5 and T6, SI.
Finally, it is important to mention that the interaction and charge
transfer between the polymer frame structure and Pd NP is stronger
for the smaller Pd NP than for larger particle size. According to our
previous study,66 the 3 nm Pd NP fabricated from the solid-phase
hydrogenation procedure (Pd@PPN) was formed anchoring inside
the cage of the polymer frame structure and therefore the degree of
interaction between the polymer frame structure and Pd NP is
maximized. Meanwhile, the 9 nm Pd particle fabricated from the
liquid-phase reduction procedure (Pd@PPN-1) was formed
anchoring outside the cage of the polymer frame structure and the
charge transfer between the polymer frame structure and the Pd-
PPN is minimized. That could explain nicely the higher activity of the
Pd@PPN catalyst (particle size 3 nm) than the Pd@PPN-1 catalyst
(particle size ~9 nm) for the HDO conversion of stearic acid obtained
in Table 1 and demonstrate again the importance of charge transfer
in facilitating the HDO activity.
Figure 14: (a) Bader charge of Pd atoms on the Pd332 cluster
coordinated with adsorbed umPPN. (b) Density of states projected
on d-orbital (PDOS) for the corner Pd site of Pd332@umPPN (blue
color) and clean Pd332 cluster (red color).
Furthermore, we revealed that there was a net charge of 0.422e-
transferred from the umPPN molecule to the Pd-NP, inducing
negative charges on Pd atoms in adjacent positions to adsorbed
umPPN molecule (except the corner site). As seen in Fig. 14a, two Pd
atoms next to adsorbed umPPN has the charge of -0.094 and -0.072
for the nearer and further atoms, respectively, while the charge of
them on the clean Pd-NP is -0.035. Since the charge transfer effect
was reported to play very important role in modifying the activity of
the catalysts,97, 99-101 it is likely that the charge transfer from umPPN
to Pd NP inducing negative charge on Pd atoms could be another
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 15
Please do not adjust margins

Please do not adjust margins
Page 17 of 20
Green Chemistry
ARTICLE
Journal Name
reference no: HRDG/YSA-19/02/21(0045)/2019 for the financial
Conclusions
View Article Online
support. Q.T.T. and M.P.S. acknowledge tDhOeIf:i1n0a.1n0c3i9a/lCs9uGpCp0o3r8t0b3Ey
the Singapore National Research Foundation under its Campus
for Research Excellence and Technological Enterprise (CREATE)
program through the Cambridge Center for Carbon Reduction
in Chemical Technology (C4T) and eCO2EP programs. The
computational work for this research is performed with the
resources of the National Supercomputing Centre, Singapore
(NSCC) under the project ID 12000902. K.A thanks for the
In our present investigation, we have designed and developed
Pd NPs (2-3 nm size range) encapsulated nitrogen rich porous
organic polymer (Pd@PPN) and employed it as robust catalyst
towards the conversion of saturated/unsaturated fatty acids to
diesel ranged long chain biofuel additive. FeCl3 aided Friedel-
Crafts alkylation between triphenylamine (TPA) monomer and
α, α' ̵dibromo-p-xylene as cross-linkers successfully furnished
self-assembled PPN nanofibers having average length 233-333
nm. This catalyst also exhibited versatile catalytic performance
over different types of vegetable oils including palm oil,
soyabean oil, sunflower oil and rapeseed oil to furnish long
chain diesel range alkanes which definitely established
superiority and versatility of Pd@PPN catalyst. It was found
from the catalytic details study that decarbonylation route to
produce C17 alkane (with one carbon atoms less) is more
preferred over the hydrogenation-hydrodehydration route. Our
catalyst also shows excellent stability and minimum catalyst
leaching after fourth and fifth catalytic run, could be attributed
to the strongly bound Pd NPs on highly cross-linked porous
organic framework. Two reasons that could explain the good
performance of the Pd@PPN catalyst in the HDO of carboxylic
acids are revealed. The presence of high density of step sites on
the fabricated Pd-NPs anchored in the cage of PPN polymer was
predicted from Density Functional Theory (DFT) investigation
into the interaction between Pd-NPs and the polymer frame-
structure, which promotes the rate-determining steps during
the DCN pathway of carboxylic acids and was one of the crucial
factors contributing into the high activity of the catalyst.
Besides, the charge transfer from the polymer frame structure
to the Pd NP induced negative charge of Pd atoms and
facilitated the removal of strongly bound intermediates
(involving CO and other unsaturated hydrocarbons), which
usually acts as poisons to the catalyst during the HDO process,
and was identified as another important factor that enhance
the stability and activity of the catalyst. Those two factors
support
by
Basic
Science
Research
Program
(2018R1A1A1A05079555) and Technology Development
Program to Solve Climate Changes (2017M1A2A2087630) of the
National Research Foundation of Korea (NRF) funded by the
Ministry of Science and ICT, and MOTIE (KIAT_N0001754) and
UNIST (1.180082.01). D.Q.D acknowledges the financial support
by Vietnam National Foundation for Science and Technology
Development (NAFOSTED) under grant number 103.03-
2018.366. N.T.B thanks the support from the National
Foundation for Science and Technology Development in the
framework of project coded 104.05-2017.39. P.N.A, A.D and F.J.
are grateful to the CNRS, the University of Poitiers, the Région
Nouvelle Aquitaine for financial support. The International
Consortium on Eco-conception and Renewable Resources (FR
CNRS INCREASE 3707) and the chair “TECHNOGREEN” are also
acknowledged for their funding.
Notes
‡Supplementary Information: Characterization description and
results. Experimental data: Thermogravimetric analysis (TGA);
Fourier transformed infrared (FT-IR) spectra; Pore-size distributions;
FE-SEM images; HAADF-STEM images; N1s XPS spectra; Fatty acids
compositions of various vegetable oils; TEM images of different Pd-
based catalysts; Recycling efficiency testing; HOMO and LUMO
distributions; Computational data: Thermodynamics properties, NPA
charges and NBO analysis from DFT simulations; Justification on
therefore explain the unprecedented performance of the choosing level of theory; Justification on choosing the models;
Pd@PPN catalyst during the Hydrodeoxygenation (HDO)
conversion of the long linear chain carboxylic acid to
hydrocarbons product and could open up a new efficient way
for bio-diesel production.
Comparison of adsorptions and reactions on B5 vs F4 step sites;
Reactions along RDO pathway on Pd(111) terrace sites and Pd B5
step sites; Other competitive reactions along the DCX and DCN
pathways for butanoic acid HDO conversion on Pd(111) terrace sites
and Pd B5 step sites; Higher coverage of TPA adsorption on Pd322
cluster.
Conflicts of interest
There are no conflicts to declare.
Corresponding authors
qttrinh@ntu.edu.sg; trinhquangthang@duytan.edu.vn (Q.T.T);
mpsherb@berkeley.edu (M.P.S);
johncuchem@gmail.com; johnmondal@iict.res.in (J.M)
Acknowledgements
C.S. and S.C.S wish to thankfully acknowledge the Council of
Scientific and Industrial Research (CSIR)-University grant
commission (UGC), New Delhi, for their respective junior &
senior research fellowships. J.M. thanks the Department of
Science and Technology, India, for the DST-INSPIRE Faculty
Research project grant (GAP-0522) at the CSIR-IICT, Hyderabad
and kindly acknowledges CSIR-YSA Research Grant with the
References
1.
G. W. Huber, S. Iborra and A. Corma, Chemical Reviews, 2006,
106, 4044-4098.
J. Name., 2013, 00, 1-3 | 16
Please do not adjust margins

Green Chemistry
Please do not adjust margins
Page 18 of 20
Journal Name
ARTICLE
2.
3.
4.
5.
6.
Q. T. Trinh, A. Banerjee, Y. Yang and S. H. Mushrif, The Journal 29. S. Mondal, R. Singuru, S. Chandra Shit, T. Hayashi, S. Irle, Y.
View Article Online
of Physical Chemistry C, 2017, 121, 1099-1112.
J. J. Varghese, Q. T. Trinh and S. H. Mushrif, Catalysis Science &
Technology, 2016, 6, 3984-3996.
W. Li, Y. Gao, S. Yao, D. Ma and N. Yan, Green Chemistry, 2015,
17, 4198-4205.
P. N. Amaniampong, Q. T. Trinh, K. Li, S. H. Mushrif, Y. Hao and
Y. Yang, Catalysis Today, 2018, 306, 172-182.
P. N. Amaniampong, Q. T. Trinh, B. Wang, A. Borgna, Y. Yang
Hijikata, J. Mondal and A. Bhaumik, ACS Sustainable Chemistry
DOI: 10.1039/C9GC03803E
& Engineering, 2018, 6, 1610-1619.
30. G. Xu, Y. Zhang, Y. Fu and Q. Guo, ACS Catalysis, 2017, 7, 1158-
1169.
31. J. Chen and Q. Xu, Catalysis Science & Technology, 2016, 6,
7239-7251.
32. H. Zhang, H. Lin, W. Wang, Y. Zheng and P. Hu, Applied Catalysis
B: Environmental, 2014, 150-151, 238-248.
and S. H. Mushrif, Angewandte Chemie International Edition, 33. J. Zhang and C. Zhao, ACS Catalysis, 2016, 6, 4512-4525.
2015, 54, 8928-8933.
34. Y. Zhang and S. N. Riduan, Chemical Society Reviews, 2012, 41,
2083-2094.
35. S. Mondal, S. K. Kundu and A. Bhaumik, Chemical
Communications, 2017, 53, 2752-2755.
7.
8.
Q. T. Trinh, B. K. Chethana and S. H. Mushrif, The Journal of
Physical Chemistry C, 2015, 119, 17137-17145.
P. N. Amaniampong, Q. T. Trinh, K. De Oliveira Vigier, D. Q. Dao,
N. H. Tran, Y. Wang, M. P. Sherburne and F. Jérôme, Journal of 36. R. Gomes, P. Bhanja and A. Bhaumik, Chemical
the American Chemical Society, 2019, 141, 14772-14779. Communications, 2015, 51, 10050-10053.
M. Snåre, I. Kubičková, P. Mäki-Arvela, D. Chichova, K. Eränen 37. R. Dawson, A. I. Cooper and D. J. Adams, Progress in Polymer
and D. Y. Murzin, Fuel, 2008, 87, 933-945. Science, 2012, 37, 530-563.
10. R. W. Gosselink, S. A. W. Hollak, S.-W. Chang, J. vanꢀHaveren, K. 38. X. Zou, H. Ren and G. Zhu, Chemical Communications, 2013, 49,
9.
P. deꢀJong, J. H. Bitter and D. S. vanꢀEs, ChemSusChem, 2013, 6,
1576-1594.
11. S. Lestariꢁ, P. Mäki-Arvela, J. Beltramini, G. Q. M. Lu and D. Y.
Murzin, ChemSusChem, 2009, 2, 1109-1119.
3925-3936.
39. M. Nihei, H. Ida, T. Nibe, A. M. P. Moeljadi, Q. T. Trinh, H. Hirao,
M. Ishizaki, M. Kurihara, T. Shiga and H. Oshio, Journal of the
American Chemical Society, 2018, 140, 17753-17759.
12. C. Zhao, T. Brück and J. A. Lercher, Green Chemistry, 2013, 15, 40. V. M. Suresh, S. Bonakala, H. S. Atreya, S. Balasubramanian and
1720-1739.
T. K. Maji, ACS Applied Materials & Interfaces, 2014, 6, 4630-
13. J. Dupont, P. A. Z. Suarez, M. R. Meneghetti and S. M. P.
4637.
Meneghetti, Energy & Environmental Science, 2009, 2, 1258- 41. Q. Fang, S. Gu, J. Zheng, Z. Zhuang, S. Qiu and Y. Yan,
1265.
Angewandte Chemie International Edition, 2014, 53, 2878-
2882.
42. K. Dhanalaxmi, R. Singuru, S. Mondal, L. Bai, B. M. Reddy, A.
14. T. Sooknoi, T. Danuthai, L. L. Lobban, R. G. Mallinson and D. E.
Resasco, Journal of Catalysis, 2008, 258, 199-209.
15. N. Taufiqurrahmi and S. Bhatia, Energy & Environmental
Science, 2011, 4, 1087-1112.
Bhaumik and J. Mondal, ACS Sustainable Chemistry
Engineering, 2017, 5, 1033-1045.
&
16. G. W. Huber, P. O’Connor and A. Corma, Applied Catalysis A: 43. R. Singuru, K. Dhanalaxmi, S. C. Shit, B. M. Reddy and J. Mondal,
General, 2007, 329, 120-129. ChemCatChem, 2017, 9, 2550-2564.
17. A. Srivastava and R. Prasad, Renewable and Sustainable Energy 44. S. Mondal, B. C. Patra and A. Bhaumik, ChemCatChem, 2017, 9,
Reviews, 2000, 4, 111-133. 1469-1475.
18. S. N. Naik, V. V. Goud, P. K. Rout and A. K. Dalai, Renewable and 45. P. Munnik, P. E. de Jongh and K. P. de Jong, Chemical Reviews,
Sustainable Energy Reviews, 2010, 14, 578-597. 2015, 115, 6687-6718.
19. E. Santillan-Jimenez and M. Crocker, Journal of Chemical 46. G. Kresse and J. Hafner, Physical Review B, 1993, 47, 558-561.
Technology & Biotechnology, 2012, 87, 1041-1050.
20. S. H. Mushrif, V. Vasudevan, C. B. Krishnamurthy and B.
47. G. Kresse and J. Furthmüller, Computational Materials Science,
1996, 6, 15-50.
Venkatesh, Chemical Engineering Science, 2015, 121, 217-235. 48. P. E. Blöchl, Physical Review B, 1994, 50, 17953-17979.
21. M. M. K. Bhuiya, M. G. Rasul, M. M. K. Khan, N. Ashwath, A. K. 49. J. Klimeš, D. R. Bowler and A. Michaelides, Journal of Physics:
Azad and M. A. Hazrat, Renewable and Sustainable Energy
Reviews, 2016, 55, 1129-1146.
22. T. Kalnes, T. Marker and R. Shonnard David, Journal, 2007, 5.
Condensed Matter, 2009, 22, 022201.
50. J. Klimeš, D. R. Bowler and A. Michaelides, Physical Review B,
2011, 83, 195131.
23. P. N. Amaniampong, Q. T. Trinh, J. J. Varghese, R. Behling, S. 51. O. Mohan, Q. T. Trinh, A. Banerjee and S. H. Mushrif, Molecular
Valange, S. H. Mushrif and F. Jérôme, Green Chemistry, 2018,
20, 2730-2741.
24. S. De, B. Saha and R. Luque, Bioresource Technology, 2015, 178,
108-118.
Simulation, 2019, 45, 1163-1172.
52. Q. T. Trinh, A. V. Nguyen, D. C. Huynh, T. H. Pham and S. H.
Mushrif, Catalysis Science & Technology, 2016, 6, 5871-5883.
53. H. Yildirim, T. Greber and A. Kara, The Journal of Physical
Chemistry C, 2013, 117, 20572-20583.
25. A. E. Atabani, A. S. Silitonga, H. C. Ong, T. M. I. Mahlia, H. H.
Masjuki, I. A. Badruddin and H. Fayaz, Renewable and 54. J. Carrasco, W. Liu, A. Michaelides and A. Tkatchenko, The
Sustainable Energy Reviews, 2013, 18, 211-245. Journal of Chemical Physics, 2014, 140, 084704.
26. D. Kim, D. R. Vardon, D. Murali, B. K. Sharma and T. J. 55. S. E. M. Putra, F. Muttaqien, Y. Hamamoto, K. Inagaki, I.
Strathmann, ACS Sustainable Chemistry & Engineering, 2016, 4,
1775-1784.
Hamada and Y. Morikawa, The Journal of Chemical Physics,
2019, 150, 154707.
27. M. Hajjari, M. Tabatabaei, M. Aghbashlo and H. Ghanavati, 56. C. T. Campbell, Accounts of Chemical Research, 2019, 52, 984-
Renewable and Sustainable Energy Reviews, 2017, 72, 445-464. 993.
28. S. Liu, T. Simonetti, W. Zheng and B. Saha, ChemSusChem, 57. G. Henkelman, B. P. Uberuaga and H. Jónsson, The Journal of
2018, 11, 1446-1454.
Chemical Physics, 2000, 113, 9901-9904.
This journal is © The Royal Society of Chemistry 20xx
J. Name., 2013, 00, 1-3 | 17
Please do not adjust margins

Please do not adjust margins
Page 19 of 20
Green Chemistry
ARTICLE
Journal Name
58. K. Honkala, A. Hellman, I. N. Remediakis, A. Logadottir, A. 84. A. Srifa, K. Faungnawakij, V. Itthibenchapong and S.
View Article Online
Carlsson, S. Dahl, C. H. Christensen and J. K. Nørskov, Science,
2005, 307, 555-558.
Assabumrungrat, Chemical EngineeriDnOgI:J1o0u.1r0n3a9l,/C290G1C50,3820738E,
249-258.
59. Y.-H. Chin, C. Buda, M. Neurock and E. Iglesia, Journal of the 85. S. Akbayrak and S. Özkar, ACS Applied Materials & Interfaces,
American Chemical Society, 2013, 135, 15425-15442. 2012, 4, 6302-6310.
60. Y.-H. Chin, C. Buda, M. Neurock and E. Iglesia, Journal of the 86. S. Lestari, P. Mäki-Arvela, H. Bernas, O. Simakova, R. Sjöholm,
American Chemical Society, 2011, 133, 15958-15978.
61. I. V. Yudanov, A. Genest, S. Schauermann, H.-J. Freund and N.
Rösch, Nano Letters, 2012, 12, 2134-2139.
62. B. T. Loveless, C. Buda, M. Neurock and E. Iglesia, Journal of the
American Chemical Society, 2013, 135, 6107-6121.
J. Beltramini, G. Q. M. Lu, J. Myllyoja, I. Simakova and D. Y.
Murzin, Energy & Fuels, 2009, 23, 3842-3845.
87. R. Arrigo, M. E. Schuster, Z. Xie, Y. Yi, G. Wowsnick, L. L. Sun, K.
E. Hermann, M. Friedrich, P. Kast, M. Hävecker, A. Knop-
Gericke and R. Schlögl, ACS Catalysis, 2015, 5, 2740-2753.
63. L. Zhou, J. M. P. Martirez, J. Finzel, C. Zhang, D. F. Swearer, S. 88. I. L. Simakova, O. A. Simakova, A. V. Romanenko and D. Y.
Tian, H. Robatjazi, M. Lou, L. Dong, L. Henderson, P.
Christopher, E. A. Carter, P. Nordlander and N. J. Halas, Nature
Energy, 2020, 5, 61-70.
Murzin, Industrial & Engineering Chemistry Research, 2008, 47,
7219-7225.
89. S. Lestari, P. Mäki-Arvela, I. Simakova, J. Beltramini, G. Q. M. Lu
and D. Y. Murzin, Catalysis Letters, 2009, 130, 48-51.
90. J. P. Ford, J. G. Immer and H. H. Lamb, Topics in Catalysis, 2012,
55, 175-184.
64. W. Tang, E. Sanville and G. Henkelman, Journal of Physics:
Condensed Matter, 2009, 21, 084204.
65. G. Henkelman, A. Arnaldsson and H. Jónsson, Computational
Materials Science, 2006, 36, 354-360.
91. L. Boda, G. Onyestyák, H. Solt, F. Lónyi, J. Valyon and A.
Thernesz, Applied Catalysis A: General, 2010, 374, 158-169.
66. J. Mondal, Q. T. Trinh, A. Jana, W. K. H. Ng, P. Borah, H. Hirao
and Y. Zhao, ACS Applied Materials & Interfaces, 2016, 8, 92. W. Jin, L. Pastor-Pérez, D. Shen, A. Sepúlveda-Escribano, S. Gu
15307-15319. and T. RamirezꢀReina, ChemCatChem, 2019, 11, 924-960.
67. M. Hacene, A. Anciaux-Sedrakian, X. Rozanska, D. Klahr, T. 93. S. Kim, E. E. Kwon, Y. T. Kim, S. Jung, H. J. Kim, G. W. Huber and
Guignon and P. Fleurat-Lessard, Journal of Computational
Chemistry, 2012, 33, 2581-2589.
68. M. Hutchinson and M. Widom, Computer Physics
Communications, 2012, 183, 1422-1426.
J. Lee, Green Chemistry, 2019, 21, 3715-3743.
94. Q. Bu, H. Lei, A. H. Zacher, L. Wang, S. Ren, J. Liang, Y. Wei, Y.
Liu, J. Tang, Q. Zhang and R. Ruan, Bioresource Technology,
2012, 124, 470-477.
69. S. Maintz, B. Eck and R. Dronskowski, Computer Physics 95. G. T. K. K. Gunasooriya, A. P. van Bavel, H. P. C. E. Kuipers and
Communications, 2011, 182, 1421-1427. M. Saeys, ACS Catalysis, 2016, 6, 3660-3664.
70. J. Lu, S. Behtash and A. Heyden, The Journal of Physical 96. Q. T. Trinh, K. F. Tan, A. Borgna and M. Saeys, The Journal of
Chemistry C, 2012, 116, 14328-14341. Physical Chemistry C, 2013, 117, 1684-1691.
71. A. Banerjee, A. P. van Bavel, H. P. C. E. Kuipers and M. Saeys, 97. Y. Lykhach, S. M. Kozlov, T. Skála, A. Tovt, V. Stetsovych, N.
ACS Catalysis, 2015, 5, 4756-4760.
72. A. Banerjee, A. P. van Bavel, H. P. C. E. Kuipers and M. Saeys,
ACS Catalysis, 2017, 7, 5289-5293.
Tsud, F. Dvořák, V. Johánek, A. Neitzel, J. Mysliveček, S. Fabris,
V. Matolín, K. M. Neyman and J. Libuda, Nature Materials,
2016, 15, 284-288.
73. J. Xu and M. Saeys, The Journal of Physical Chemistry C, 2009, 98. H. Yildirim and A. Kara, The Journal of Physical Chemistry C,
113, 4099-4106. 2013, 117, 2893-2902.
74. C. Sarkar, S. Pendem, A. Shrotri, D. Q. Dao, P. Pham Thi Mai, T. 99. Y. P. G. Chua, G. T. K. K. Gunasooriya, M. Saeys and E. G.
Nguyen Ngoc, D. R. Chandaka, T. V. Rao, Q. T. Trinh, M. P. Seebauer, Journal of Catalysis, 2014, 311, 306-313.
Sherburne and J. Mondal, ACS Applied Materials & Interfaces, 100. G. T. K. K. Gunasooriya, E. G. Seebauer and M. Saeys, ACS
2019, 11, 11722-11735. Catalysis, 2017, 7, 1966-1970.
75. R. Singuru, Q. T. Trinh, B. Banerjee, B. Govinda Rao, L. Bai, A. 101. A. S. Crampton, M. D. Rötzer, C. J. Ridge, F. F. Schweinberger,
Bhaumik, B. M. Reddy, H. Hirao and J. Mondal, ACS Omega,
2016, 1, 1121-1138.
U. Heiz, B. Yoon and U. Landman, Nature Communications,
2016, 7, 10389.
76. Q. T. Trinh, J. Yang, J. Y. Lee and M. Saeys, Journal of Catalysis, 102. P. Wang, S. N. Steinmann, G. Fu, C. Michel and P. Sautet, ACS
2012, 291, 26-35. Catalysis, 2017, 7, 1955-1959.
77. L. Li, H. Zhao, J. Wang and R. Wang, ACS Nano, 2014, 8, 5352- 103. R. A. Van Santen, Accounts of Chemical Research, 2009, 42, 57-
5364. 66.
78. H. Zhong, C. Liu, H. Zhou, Y. Wang and R. Wang, Chemistry – A 104. C.-K. Tsung, J. N. Kuhn, W. Huang, C. Aliaga, L.-I. Hung, G. A.
European Journal, 2016, 22, 12533-12541.
79. P. Puthiaraj and K. Pitchumani, Green Chemistry, 2014, 16,
4223-4233.
80. S. Mondal, J. Mondal and A. Bhaumik, ChemCatChem, 2015, 7,
3570-3578.
81. H. Zhong, C. Liu, Y. Wang, R. Wang and M. Hong, Chemical
Science, 2016, 7, 2188-2194.
Somorjai and P. Yang, Journal of the American Chemical
Society, 2009, 131, 5816-5822.
105. R. Pestman, W. Chen and E. Hensen, ACS Catalysis, 2019, 9,
4189-4195.
106. P. S. Cremer, X. Su, Y. R. Shen and G. A. Somorjai, Journal of the
American Chemical Society, 1996, 118, 2942-2949.
82. J. Lu, S. Behtash, M. Faheem and A. Heyden, Journal of
Catalysis, 2013, 305, 56-66.
83. Y. K. Lugo-José, S. Behtash, M. Nicholson, J. R. Monnier, A.
Heyden and C. T. Williams, Journal of Molecular Catalysis A:
Chemical, 2015, 406, 85-93.
J. Name., 2013, 00, 1-3 | 18
Please do not adjust margins

Please do not adjust margins
Green Chemistry
Page 20 of 20
View Article Online
DOI: 10.1039/C9GC03803E
ARTICLE
Table of Contents: Graphical Abstract
Abstract: Novel design Pd-based catalyst hosted over nitrogen enriched fibrous Porous-Organic-Polymer exposes a
high density of step sites and exhibits versatile catalytic performance over different types of vegetable oils to furnish
long chain diesel-range alkanes.
J. Name., 2013, 00, 1-3 | 19

Page 21 of 20
Green Chemistry
Table of Contents: Graphical Abstract
View Article Online
DOI: 10.1039/C9GC03803E
Abstract: Novel design Pd-based catalyst hosted over nitrogen enriched fibrous Porous-Organic-Polymer exposes a
high density of step sites and exhibits versatile catalytic performance over different types of vegetable oils to furnish
long chain diesel-range alkanes.