Palladium Nanoparticles Encaged in a Nitrogen-Rich Porous Organic Polymer: Constructing a Promising Robust Nanoarchitecture for Catalytic Biofuel Upgrading

Article abstract of DOI:10.1002/cctc.201700186

Robust nanoarchitectures based on surfactant-free ultrafine Pd nanoparticles (NPs) (2.7–8.2±0.5 nm) have been developed by using the incipient wetness impregnation method with subsequent reduction of Pd^II species encaged in the 1,3,5-triazine-functionalized nitrogen-rich porous organic polymer (POP) by employing NaBH₄, HCHO, and H₂ reduction routes. The Pd-POP materials prepared by the three different synthetic methods consist of virtually identical chemical compositions but have different physical and texture properties. Strong metal–support interactions, the nanoconfinement effect of POP, and the homogeneous distribution of Pd NPs have been investigated by performing ¹³C cross-polarization (CP) solid-state magic angle spinning (MAS) NMR, FTIR, and X-ray photoelectron spectroscopy (XPS), along with wide-angle powder XRD, N₂ physisorption, high-resolution (HR)-TEM, high angle annular dark field scanning transmission electron microscopy (HAADF-STEM), and energy-dispersive X-ray (EDX) mapping spectroscopic studies. The resulting Pd-POP based materials exhibit highly efficient catalytic performance with superior stability in promoting biomass refining (hydrodeoxygenation of vanillin, a typical compound of lignin-derived bio-oil). Outstanding catalytic performance (≈98 % conversion of vanillin with exclusive selectivity for hydrogenolysis product 2-methoxy-4-methylphenol) has been achieved over the newly designed Pd-POP catalyst under the optimized reaction conditions (140 °C, 10 bar H₂ pressure), affording a turnover frequency (TOF) value of 8.51 h^?1 and no significant drop in catalytic activity with desired product selectivity has been noticed for ten successive catalytic cycles, demonstrating the excellent stability and reproducibility of this catalyst system. A size- and location-dependent catalytic performance for the Pd NPs with small size (1.31±0.36 and 2.71±0.25 nm) has been investigated in vanillin hydrodeoxygenation reaction with our newly designed Pd-POP catalysts. The presence of well-dispersed electron-rich metallic Pd sites and highly rigid cross-linked amine-functionalized POP framework with high surface area is thought to be responsible for the high catalytic activity and improvement in catalyst stability.

Full text of DOI:10.1002/cctc.201700186

Accepted Article
Title: Palladium Nanoparticles Encaged in Nitrogen Rich
Porous Organic Polymer: Constructing Promising Robust
Nanoarchitecture in Catalytic Biofuel Upgrade
Authors: John Mondal, Ramana Singuru, Karnekanti Dhanalaxmi,
Subhash Chandra Shit, and Benjaram Mahipal Reddy
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201700186
Link to VoR: http://dx.doi.org/10.1002/cctc.201700186
A Journal of
www.chemcatchem.org

ChemCatChem
10.1002/cctc.201700186
Palladium Nanoparticles Encaged in Nitrogen Rich Porous Organic
Polymer: Constructing Promising Robust Nanoarchitecture in Catalytic
Biofuel Upgrade
[
a,b]
[a, b]
[a, b]
Ramana Singuru,
Mahipal Reddy,
Karnekanti Dhanalaxmi,
and John Mondal*
Subhash Chandra Shit,
Benjaram
[
a, b]
[a, b]
[
a] R. Singuru, K. Dhanalaxmi, S. C. Shit, Prof. Dr. B. M. Reddy, Prof. Dr. J. Mondal
Inorganic and Physical Chemistry Division, CSIR-Indian Institute of Chemical Technology,
Uppal Road, Hyderabad 500007, India.
[
b] AcSIR-Indian Institute of Chemical Technology, Hyderabad-500 007, India.
E-mail: johncuchem@gmail.com; johnmondal@iict.res.in
ABSTRACT: Surfactant-free ultrafine Pd nanoparticles (NPs) (2.7-8.2±0.5 nm) based robust
nanoarchitecture have been developed by incipient wet impregnation method with subsequent
reduction of Pd(II) species encaged into the 1,3,5-triazine functionalized Nitrogen-rich
4 2
Porous Organic Polymer (POP) employing NaBH , HCHO and H reduction routes,
respectively. The as prepared Pd-POP materials following three different synthetic methods
consist of virtually identical chemical compositions but with different physical and texture
properties. Strong Metal-support interaction, nanoconfinement effect of POP and
1
3
homogeneous distribution of Pd-NPs have been investigated by performing C CP solid state
MAS NMR, FT-IR, XPS, Wide angle Powder XRD, N physisorption, HR-TEM, HAADF-
2
STEM, Energy-dispersive X-ray (EDX) mapping spectroscopic studies. The resulting Pd-
POP based materials exhibit highly efficient catalytic performance with superior stability in
promoting biomass refining (hydrodeoxygenation of vanillin, a typical compound of lignin-
derived bio-oil). Outstanding catalytic performance (~98% conversion of vanillin with
exclusive selectivity for hydrogenolysis product 2-methoxy-4-methylphenol) has been
achieved over the newly designed Pd-POP catalyst under the optimized reaction conditions
-1
(
140°C, 10 bar H
2
pressure), affording a TOF value of 8.51 h and no significant drop in
th
catalytic activity with desired product selectivity has been noticed for the 10 successive
catalytic cycles demonstrating excellent stability and reproducibility of this catalyst system.
A size- and location-dependent catalytic performance for the Pd-NPs having small size
(1.31±0.36 and 2.71±0.25 nm) has been investigated in vanillin hydrodeoxygenation reaction
with our newly designed Pd-POP catalysts. Presence of well-dispersed electron-rich metallic
1
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
Pd sites and highly rigid cross linked amine functionalized POP framework with high surface
area is thought to be responsible for the high catalytic activity and improvement in catalyst
stability.
KEYWORDS: Heterogeneous Pd Catalyst • Porous Organic Polymer • Biomass • Biofuel
Upgrade • Hydrodeoxygenation of Vanillin
INTRODUCTION: In the past few decades, production of Biofuel and chemical feedstocks
from renewable and sustainable biomass has emerged a growing scientific interest as an
[1]
alternative owing to the continuous depletion on fossil fuel sources. However the presence
of high amounts of oxygen components limited the effective utilization of bio-oils thereby
making its’ a low calorific value, immiscibility with conventional fuels and instability over
time. Therefore, upgradation of bio-oils employing deoxygenation become necessary for their
[2]
effective utilization. Lignin which is composed with the ~30 wt% of woody biomass has
been considered one of the most abundant biosources on earth. Selective modification or
deoxygenation of lignin-derived pyrolysis oil has marked serious concerns and challenging
task compared with the cellulose-derived pyrolysis oil owing to its exceptionally complicated
structure associated with the oxygen-rich subunits derived from phenol, p-coumaryl,
coniferyl, and sinapyl alcohols which are naturally attached with ether linkages with no
[3]
uniform repeating units. Several kind of approaches have been adopted including
dehydration, hydrogenolysis, decarboxylation, decarbonylation and hydrodeoxygenation for
bio-oil upgrading to a liquid transportation fuel, among which hydrodeoxygenation (HDO) is
[4]
considered to be the most effective and feasible strategy for bio-Fuel upgrading. Recently
several research groups have brought into focus the need to find effective strategies for the
development of metal nano-structures useful in potential applications in versatile research
[5]
areas. Literature survey reports demonstrate that a series of supported noble metal catalysts
like Ru, Pt and Pd) could be employed in successful deoxygenation for a wide range of
(
[6]
biomass derived compounds such as phenol, anisole, vanillin and sorbitol. In this regard,
extensive efforts have been focused for the construction of various kind of heterogeneous
catalysts to explore hydrodeoxygenation of vanillin (4-hydroxy-3-methoxybenzaldehyde), a
typically model component of pyrolysis oil derived from the lignin fraction, with three
different types of oxygenated functional groups (aldehyde, ether, and hydroxyl) and its partial
solubility in both the organic and aqueous phases just as bio-oil.
2
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
For example, Wang and co-workers have designed novel N-doped carbon-supported Pd
catalyst along with the use of some traditional supports such as TiO , MgO, CeO and γ-
2
2
2 3
Al O and investigated their catalytic performance in the aqueous phase hydrodeoxygenation
[7]
of vanillin. They have found that the excellent catalytic activity and selectivity of the
desired product could be ascribed to the unique and stable N-doped carbon nanostructure that
facilitates homogeneous distribution of Pd nanoparticles with the additional electronic
activation of the metal nanoparticles in the reaction medium. Superior catalytic performance
for the hydrodeoxygenation of vanillin has been achieved by Xiao and co-workers employing
highly active hydrophilic mesoporous sulfonated melamine-formaldehyde resin (MSMF)-
[8]
supported Pd catalyst (Pd/MSMF). Very recently, Shall et al. have developed a Pd@NH
2
-
UiO-66 catalyst where palladium nanoparticles are encapsulated within amine-functionalized
[9]
UiO-66 MOF and employed it as a synergic catalyst in hydrodeoxygenation of vanillin.
Shall et al. have also reported synthesis of highly efficient bifunctional catalyst, Pd/SO H-
3
MIL-101(Cr) with the facile immobilization of Pd-NP on a mesoporous sulfonic acid-
functionalized metal-organic framework and investigated its catalytic activity in vanillin
[10]
hydrodeoxygenation reaction.
carbon derived from metal-organic-framework has been recognized as a potential catalyst
Pd/NPC-ZIF-8) in vanillin hydrodeoxygenation reaction with remarkable catalytic
Pd nanoparticles were stabilized on the N-doped porous
(
[11]
efficiency. Pd nanoparticle size-and location-dependent catalytic activity and selectivity in
[12]
HDO of vanillin with Pd-MIL-101 catalyst was reported by Xu and co-workers. Wang and
co-workers have extensively investigated catalytic activity of carbon nanotube supported gold
[6c]
catalyst in bio-oil upgrading using vanillin hydrodeoxygenation reaction.
However,
practical applications of these recent reported supported metal catalysts have been restricted
by some obdurate associated problems owing to poor dispersion of nanoparticle on support,
weak catalyst-support interactions, uncontrollable growth, reaction monofunctionality and
leaching of metal nanoparticles causing an obvious decrease in the catalyst recyclability.
Although MOFs with huge surface area and well-defined cavities offer a great opportunity to
serve as a promising platform/host for encapsulation of catalytically active particles but their
versatile application in heterogeneous catalysis yet remains unexplored due to their moisture
sensitivity, physiochemical instability in acids or bases and the labile nature of the metal-
ligand coordination bonds. In contrast with MOF (Metal-Organic-Framework), N-doped
carbon, and carbon nanotubes, Porous Organic Polymer (POP) have intrigued incredible
degree of awareness as a fascinating candidate for spatial confinement of catalytically active
3
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
crystalline metal or metal oxide nanoparticles owing to its interconnected three-dimensional
highly rigid symmetric skeletons mechanically stable to heat, moisture, acid and base,
designable pore structure, low skeleton density and easily tuneable structural integrity by
[13]
changing straightforward effective synthesis strategy and selection of monomers.
Very
recently, POP-supported metal or metal oxide nanoparticles become very attractive candidate
[14a]
[14b]
as heterogeneous nanocatalyst in performing hydrogenation,
oxidation
and
[14c, 14d]
photocatalytic reactions.
A certain degree of chemical interaction including
coordination and electron transfer between the polymers and metal NPs can surely enhance
the catalytic activity and stability of metal NPs thereby evading agglomeration and leaching
from the POP networks.
Scheme 1. Schematic illustration showing the synthetic strategy of Pd nanoparticles
fabrication into/onto POP and its selective catalysis in biofuel upgrade reaction.
In our previous work, we have accomplished the fruitful synthesis of nitrogen rich
Porous Organic Polymer (TRIA) poly-divinylbenzene-co-triallyloxytriazine by nonaqueous
radical co-polymerization of 2,4,6-triallyoxy-1,3,5-triazine with the cross-linker
[14a]
divinylbenzene under metal-free and template-free solvothermal conditions.
Successful
4
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
synthesis of porous organic polymers employing metal-free and template-free radical
polymerization under solvothermal conditions has been widely recognised as a promising
route for synthesizing functional hierarchically porous organic polymers on the concept of
green chemistry and also to avoid metal catalyzed harsh reaction conditions. In the present
work, we take advantage of the distinctive features of TRIA-POP including huge surface area
with high thermal and mechanical stability that can easily accommodate Pd nanoparticles
within the mesoporous cages to develop efficient Pd-catalysts Pd-A, Pd-B and Pd-C,
respectively by following three different reduction routes and investigated their catalytic
performances in hydrodeoxygenation of vanillin as a model system to explore the
hydrogenation and deoxygenation routes of lignin as presented in Scheme 1. Concerning
regarding the outstanding thermal and chemical stability of the POP materials, it has been
well documented that many of the POPs could be safely exposed in the aggressive media and
wet chemical environment without disturbing framework stability and porous structure. For
example, the remarkable stability of the resulting POP materials is illustrated by the
resistance to high moisture, high acidity, and/or high basicity under catalytic reaction
[14e]
conditions because of their covalent bond-formation chemistry.
reported the synthesis of highly cross-linked porous polymers by Co
trimerization reaction and described their exceptional stability during acidic work up
Lin and co-workers have
2
(CO) -mediated
8
[14f]
procedure which was most essential to remove the unreacted cobalt species.
Structural
integrity of the organic back-bone unit, porosity, nanostructure, homogeneous distribution of
13
Pd-NPs with the corresponding oxidation state have been investigated by performing C CP
solid state MAS NMR, FT-IR, Wide angle Powder XRD, N physisorption, HR-TEM,
2
HAADF-STEM, Energy-dispersive X-ray (EDX) elemental mapping, XPS spectroscopic
studies. Impressive catalytic performance (~98% conversion of vanillin with exclusive
selectivity for hydrogenolysis product 2-methoxy-4-methylphenol) has been achieved over
the newly designed Pd-POP catalyst with tremendous recyclability after ten recycles in
succession, no sign of catalyst deactivation and negligible leaching. An enhancement in the
catalytic performance for the Pd-POP catalysts is observed compared with the conventional
and reported catalysts under optimized reaction conditions making it a potential catalyst in
Biomass refining industry. The enhanced catalytic performance of Pd-POP could be
attributed to the intrinsic synergistic effects of Pd-NPs interacting with the nitrogen rich POP
with an exceptional porous nanoarchitecture making them electron-rich metallic Pd and
hierarchical pores in the interconnected POP network, facilitates hydrogenolysis of the
5
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
intermediate to the targeted product with easy diffusion. To the best of our knowledge, this is
the first example for the development of porous organic polymer encapsulated Pd NPs as
efficient catalyst with superior stability in promoting biomass refining (hydrodeoxygenation
of vanillin, a typical compound of lignin).
Experimental Section:
Materials and methods:
Divinylbenzene, 2, 4, 6-triallyoxy-1, 3, 5-triazine, Azobisisobutyronitrile (AIBN), Pd(OAc)
2
,
Vanillin, NaBH were procured from Sigma-Aldrich, India and were used respectively
4
i
without any further modification. Acetone, MeOH, PrOH and HCHO were supplied by E-
Merck, India.
Characterization Techniques:
Powder X-ray diffraction (PXRD) patterns of different samples were recorded with a Bruker
D8 Advance X-ray diffractometer operated at a voltage of 40 kV and a current of 40 mA
using Ni-filtered Cu Kα (λ=0.15406 nm) radiation. High Resolution Transmission electron
microscopy (HR-TEM) images were recorded in a JEOL JEM 2010 transmission electron
microscope with operating voltage 200 kV equipped with a FEG. Field emission scanning
electron microscopic images of samples were obtained using a JEOL JEM 6700 field
emission scanning electron microscope (FE-SEM). Nitrogen sorption isotherms were
obtained using a Quantachrome Autosorb 1C surface area analyzer at 77 K. Prior to the
measurement, the samples were degassed at 393 K for approximately 6 h in high vacuum.
Surface areas were calculated from the adsorption data using Brunauer-Emmett-Teller (BET)
in the relative pressure (P/P ) range of 0.01-0.1. The total pore volumes and pore size
0
distribution curves were obtained from the adsorption branches using nonlocal density
functional theory (NLDFT) method. FT IR spectra of the samples were recorded using a
Nicolet MAGNA-FT IR 750 Spectrometer Series II. Thermogravimetry (TGA) analyses of
the all samples were carried out using a TGA Instruments thermal analyzer TA-SDT Q-600
-1
under nitrogen atmosphere with a heating rate of 10°C min from 25 to 600°C. Solid-state
1
3
C CP-MAS NMR studies were performed using a Bruker Avance III HD 400 MHz NMR
spectrometer. Transmission electron microscope high-annular dark-field scanning
HAADFSTEM) and energy-dispersive X-ray mapping images were obtained with a
(
TECNAI G2 F20 equipped with an EDX detector.X-ray photoelectron spectroscopy (XPS)
was performed on an Omicron nanotech operated at 15 kV and 20 mA with a monochromatic
Al Kα X-ray source. Quadrupole ion trap Mass Spectrometer equipped with Thermo Accela
6
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
LC and Agilent 6890 GC system equipped with a flame ionization detector were used for
analysis of catalytic reactions. The loading amounts of Pd were determined using an
inductively coupled plasma optical emission spectrometer (ICP-OES, iCAP-6500DUO,
Thermo Fisher Scientific).
Synthesis of Porous Organic Polymer TRIA:
Porous Organic Polymer (TRIA) has been synthesized based on the protocol previously
[14a]
reported by Zhao et al.
In a typical synthetic procedure, divinylbenzene (5.99 mmol, 781
mg), 2, 4, 6-triallyoxy-1, 3, 5-triazine (1.49 mmol, 373 mg) and AIBN (0.152 mmol, 25 mg)
were mixed together in acetone (15 mL) and the resulting mixture was stirred at room
2
temperature under N atmosphere for about 6 h. After that, the reaction mixture was
solvothermally treated in a teflon lined stainless steel autoclave at 120°C under static
condition for 24 h. After cooling, the resulting white color solid material TRIA was isolated
washed with acetone followed by methanol and dried in air.
Synthesis of Pd-A:
Typically, 400 mg TRIA polymer was suspended in MeOH (100 mL) and the mixture was
sonicated for 30 minutes until it become homogeneous. With constant vigorous stirring 40
mg of Pd(OAc)
Then the resulting reaction mixture was continuously stirred at room temperature for another
5 h. The as-obtained brown colour Pd(II)-POP composite was isolated from the reaction
2
in 10 mL of MeOH solution was added dropwise over a period of 15 min.
1
mixture by centrifugation and washed with MeOH for several times. After that the brown
solid was re-dispersed in 100 mL water by stirring at room temperature for 2 h. A freshly
prepared 1(M) NaBH (2 mL) solution was subsequently added and stirred for another 15
4
minutes. The black colour solid powder was isolated and designated as Pd-A.
Synthesis of Pd-B:
The brown colour Pd(II)-POP composite was synthesized following the same procedure as
described above. After that the brown color solid was re-dispersed in 100 mL water with
addition of 0.5 mL formalin (37 wt%) solution and heated the reaction mixture at 80°C for
3h. The mixture was allowed to cool at room temperature and the black colour solid was
recovered with careful centrifugation followed by washing with acetone and methanol which
was designated as Pd-B.
Synthesis of Pd-C:
7
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
For the synthesis of Pd-C material the as-synthesized brown colour solid was subjected for
heat treatment in a stream of H /N (10% H , 100 mL/min) at 200°C for 4 h. The black colour
2
2
2
solid was isolated and designated as Pd-C.
Catalytic Hydrodeoxygenation (HDO) of Vanillin:
The catalytic hydrodeoxygenation (HDO) of vanillin was performed in a stainless steel
reactor inbuilt with a pressure gauge setup. In a typical procedure under, vanillin (100 mg,
0
.657 mmol), Pd-POP catalyst (20 mg, 0.014 wt% of Pd) and isopropanol (30 mL) were
charged into the reactor. The reactor was evacuated using a vacuum pump for 15 min at room
temperature to remove dissolved O or air and then pressurized with H at a desired pressure
with continuous purging by hydrogen gas. The temperature of the autoclave was fixed at
40°C for the desired reaction time with continuous stirring at a speed of 800 rpm. After the
2
2
1
reaction, the autoclave was cooled at room temperature and catalyst was separated from the
reaction mixture by centrifugation prior to being analyzed by GC. Quantitative analysis for
the isolated liquid reaction mixture was performed using dodecane as an internal standard
followed by comparison with known standards with a standard deviation less than 2%.
Agilent 6980 Gas chromatograph equipped with a flame ionization detector and an SE-54
capillary column (30 m×0.32 mm×1.0 μm) with a stationary phase based on
poly(methylphenylsiloxane) has been utilized for the analysis and quantification of the
reaction products.
Recycling Test: In a typical recycling experiment of hydrodeoxygenation (HDO) of vanillin
with palladium nanoparticles supported POP catalyst, a mixture of vanillin (500 mg, 3.375
mmol), Pd-POP catalyst (100 mg, 0.07 wt% of Pd) and isopropanol (150 mL) was heated to
stir at 140°C in a 200 mL dry stainless steel reactor pressurised with hydrogen gas at 10 bar
for 15 h. After the reaction, the catalyst was recovered from the reaction mixture by
centrifugation, washed with methanol several times followed by acetone, finally dried
overnight in an oven at 80°C and then directly used for the next cycle reaction in succession.
No any additional specific approach including heat treatment and addition of acid or base is
necessary for the reactivation and regeneration of our catalyst.
Hot filtration test: We have conducted hot filtration test in order to establish that our
catalyst is unambiguously heterogeneous in nature and no considerable leaching of the
catalytic active sites took place into the solution. Typically, 50 mL dry stainless steel reactor
pressurised with hydrogen gas at 10 bar was charged with a mixture of vanillin (100 mg,
0.657 mmol), Pd-C catalyst (20 mg) and isopropanol (30 mL) and heated at 140°C. After 6 h
8
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
the catalyst was removed from the hot reaction mixture by centrifugation and mixture was
examined with GC analysis which provided 54.6% conversion of vanillin. Then we have
performed the catalytic reaction running the filtrate with same reaction mixture under
stainless steel reactor with identical conditions for further 8 h. After 15 h, no noticeable
increase in the vanillin conversion beyond 55% was attained as determined by GC analysis.
This investigation definitely illustrates that the palladium nanoparticles were strongly
encapsulated with the highly rigid porous organic framework making the system
heterogeneous in nature with no evident of metal leaching in the solution as confirmed by
AAS analysis measurement of the hot filtrate solution.
Hg(0) poisoning test: In order to examine active catalytic Pd-species is likely to be
heterogeneous in nature, a standard Hg-poisoning test was conducted for the successful
suppression of the catalytic reaction. Two sets of reactions for hydrodeoxygenation of
vanillin were carried charging the stainless steel reactor with the reaction mixtures of vanillin
(100 mg, 0.657 mmol), Pd-C catalyst (20 mg) and isopropanol (30 mL), one of them is
considered as a control experiment. Vanillin conversions for both set of reactions to 2-
methoxy-4-methylphenol were analysed at 140 °C after 6 h by GC-FID. At that time,
conversions of 54.6% and 55.3% have been attained for both the reactions, respectively.
Then, elemental mercury (100 mg) was introduced to one of the reactor system after 6 h and
both reactions were continued for about 18 h. After 18 h, it was discovered that the catalytic
reactor system containing Hg furnished only 57.9% vanillin conversion. But on the other
hand we have achieved 96.9 % vanillin conversion in the absence of Hg, demonstrating the
inactivation of the Pd-NPs surfaces owing to the formation of a Pd-Hg amalgam by the
introduction of Hg(0).
Results and Discussion:
Synthesis of POP and Pd-POP catalysts:
The route for stepwise structural evolution from TRIA-POP to three different Pd-POP
materials is schematically presented in the Scheme 1 and the synthetic details were provided
in the Experimental Section. Porous Organic Polymer (TRIA-POP) has been synthesized by
nonaqueous radical polymerization of 2,4,6-triallyoxy-1,3,5-triazine in the presence of
divinylbenzene as cross-linking agent under metal-free and template-free solvothermal
conditions using Azobisisobutyronitrile (AIBN) as a radical initiator following the method as
[14a]
previously reported by Zhao et al.
A solvothermal route in a sealed media has been
adopted in our present study where triazine monomers started to polymerize with divinyl
9
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
benzene under moderately high temperature (120°C) and high autoclave pressure in the
liquid-phase conditions. A highly cross-linked poly-divinylbenzene-co-triallyloxytriazine
organic framework was progressively developed in acetone solvent which is likely to be
considered to exhibit both as solvent and as well as template roles. White color solid TRIA-
POP sample with open disordered porosity were finally obtained with the slow removal of
solvent under the autoclave system during the course of gradual formation of framework
[14d]
material.
2
Facile treatment of as-synthesized POP with Pd(OAc) in MeOH afforded
yellow color Pd(II)-POP composites. Two weak signals appeared at 26.5 and 179.1 ppm in
13
the C CP MAS NMR spectrum of Pd(II)-POP composite (Figure S1, SI) could be assigned
to the carbonyl and methyl carbons of the incorporated Pd(OAc) , respectively. Additional
reaction of the composite following three different reduction routes in NaBH , HCHO and a
stream of H gave rise to blackish brown color materials which are designated as Pd-A, Pd-B
2
4
2
and Pd-C, respectively. The Pd contents in Pd-A, Pd-B and Pd-C are 0.213, 0.228 and 0.210
mmol/g, respectively, as measured by inductively coupled plasma optical emission
spectrometry (ICP-OES) analysis. Insignificant deviation in the experimental C and N
content (%) from the corresponding theoretical values were obtained in the Elemental (C, H
&
N) analyses (Table S1, SI) of the respective Pd-A, Pd-B and Pd-C materials which could
be correlated with the very common phenomenon of porous materials that is the inclusion of
guest molecules in the POPs. Similar finding regarding the deviation in elemental analysis
result has been observed and reported by Wang and co-workers in their triazol functionalized
[15]
POP decorated Pd-NPs.
was further confirmed by Thermogravimetric (TGA) analysis. An initial weight loss around
.8%, 3.5% and 5.6% have been observed (Figure S2A, SI) in the TGA curves of Pd-A, Pd-B
The existence of trapped guest molecules and thermal stability
2
and Pd-C, respectively, before 150°C which can be assigned to the removal of intercalated
and adsorbed water molecules from the POP. All the three respective Pd-POP materials are
thermally stable in nitrogen atmosphere up to 400°C temperature. Our respective POP based
material also show exceptional chemical stability after keeping the material in strongly acidic
solution circumventing the associated structural degradation, as experimentally evidenced by
13
solid state C MAS NMR studies (Figure S2B, SI).
10
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
13
Figure 1. a) Wide angle powder X-ray diffraction spectra, b) C CP Solid State MAS NMR
spectra, c) N adsorption/desorption isotherms, and d) the corresponding pore-size
2
distributions of Pd-A, Pd-B and Pd-C materials, respectively. Pore Size distributions
calculated by NLDFT method.
Porosity and Nanostructure:
Wide angle powder XRD patterns of Pd-A, Pd-B and Pd-C materials ( Figure-1a) exhibit
three types of characteristic diffractions at 2Ө values of 40.1°, 46.2°, and 67.2° which could
be indexed to the (111), (200), and (220) crystalline reflections, corresponding to the face
[16]
centered cubic lattice arrangement of Pd (0) NP (JCPDS no. 46-1043).
The additional
broad diffraction peak appeared at 2Ө=20.2° which could be attributed to the pi-stacking of
the aromatic building blocks in small domains. The size of the Pd NPs was measured to be
approximately 2.76-3.35±0.5 nm with the consideration of most highest intensity diffraction
peak of the (111) crystalline lattice plane by using Debye Scherrer equation, which is in good
agreement with the results of TEM and HAADF-STEM analysis of the respective Pd-A, Pd-B
and Pd-C materials. It is interesting to notice that the Pd (111) and Pd (200) crystalline
11
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
reflections in Pd-B material appeared with the stronger peak intensity than that of the
corresponding Pd-A and Pd-C. This observation could be recognized to the deposition of
most of the large Pd-NPs on the external surface of POP rather than inside of cavity, resulting
[17]
a great exposure of Pd-NP in surface.
The weak intensities of (111) and (200) Bragg
diffractions for Pd-C material confirm to the well distribution of small size Pd-NPs with the
coverage of POP framework, attributing to the minimum exposure of Pd nanocrystal in
surface. FT-IR spectra of the Pd-A, Pd-B, and Pd-C materials are displayed in Figure S3, SI.
-1
The absorption bands positioned at 1572, 1350, and 807 cm could be recognized for the
[18]
presence of characteristic triazine framework in all the metal-POP composite materials.
-1
The concomitant appearance of distinguishable peaks at 2927 and 2854 cm are suggestive to
[6c]
the stretching vibrations of the -CH
2
and -CH groups in the aliphatic moiety.
The
molecular connectivity, structural integrity of 1,3,5-triazolyl moiety and different chemical
environment of carbon nuclei of Pd-NP fabricated POP materials were assessed by
13
13
performing Solid state C CP MAS NMR analysis. C CP MAS NMR spectra of TRIA
POP) support, Pd-A, Pd-B and Pd-C materials were displayed in Figure-1b. Three broad
(
signals appeared at 112.0, 126.8, and 136.7 ppm are assigned to the characteristic aromatic
carbon atoms of the porous organic polymer (overlap of C7-C9) and the signals centered at
28.3 and 39.7 ppm correspond to the different aliphatic carbon atoms (overlap of C3-C6) of
the bridging unit. The concurrent emergence of characteristic peak at 144.1 ppm can be
attributed to the triazine carbon atom (C1) of the melamine unit, which is consistent with
[
14a]
previous analogous reports by Mondal et al.
Rest of these three peaks, the presence of
one small peak at 68.3 ppm (Figure-1b) likely refers to the carbon atom (C2) of the −OCH
2
1
3
bridging unit. According to FT-IR and solid state C CP MAS NMR spectral data, we could
be able to conclude that the structural integrity of the polymer back bone unit remains
undisturbed after Pd modification. The porous properties of Pd-A, Pd-B and Pd-C have been
investigated by N
isotherms of Pd-A, Pd-B and Pd-C materials, respectively, measured at 77 K exhibited a
rapid uptake at high pressure region of P/P > 0.4, (Figure 1c) which suggests the presence of
2
adsorption/desorption analysis at 77K (Figure-1c). Nitrogen-sorption
0
inter particle void space within the POP based materials. This sorption profile is described as
a type II isotherm, which is the characteristic of porous materials. The appearance of small
hysteresis loop in the isotherm could be ascribed to the deformation and swelling of the
polymeric network upon the gas adsorption or due to porosity originating from the
[17, 19]
interparticle void space.
A similar finding regarding this characteristic hysteresis loop in
12
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
the isotherm was previously reported by Bhaumik et al. in nitrogen rich porous organic
[20]
polymer.
The BET (Brunauer-Emmett-Teller) surface areas of the three materials Pd-A,
2
-1
2
-1
2 -1
Pd-B and Pd-C were found to be 542 m g , 491 m g and 456 m g , respectively.
Substantial decrease in the surface areas of Pd-POP materials compared with the parent POP
2
-1
material (The BET surface area of the TRIA material is 620 m g ) suggests to both partial
pore filling and mass increment after palladium loading. The preservation of N adsorption/
2
desorption isotherms shapes of Pd-A, Pd-B, and Pd-C (Figure 1c) materials signifies no
considerable alteration of the pore systems by the occupation of palladium NPs inside the
nanocage. The pore volumes of Pd-A, Pd-B and Pd-C materials were evaluated to be 0.46
3
-1
3
-1
3 -1
cm g , 0.39 cm g and 0.33 cm g , respectively. The pore size distributions of the respective
Pd-POP based materials have been calculated from the adsorption branches by employing
NLDFT (Non Local Density Functional Theory) method which is presented in the Figure 1d.
The corresponding pore-size distributions reveal the narrow pore size distribution with the
pores predominately in the microporous range (1.26 nm), which is in good agreement with
2
the shape of N isotherms. A contraction in the pore width in the Pd-POP materials compared
with the as-synthesized POP material (Calculated NLDFT pore size for TRIA POP is 2.9 nm)
was observed, thus indicating partial occupancy of nanocages with the introduction of Pd
nanoparticles in parent POP material. We also note that this observation is in full agreement
with the recent work by Shall and co-workers on incorporation of Palladium nanoparticles
[
10]
with the shrinkage of pore sizes within sulfonic acid-functionalized MIL-101(Cr).
13
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
Figure 2. a) XPS survey spectra, b) XPS spectra N 1s region, c) O 1s region and Pd 3d region
for d) Pd-A, e) Pd-B and f) Pd-C, respectively.
XPS analysis:
Further insights into the evaluation of oxidation state, metal support interaction and
change in the catalyst structure were revealed from XPS measurements. The X-ray
photoelectron spectroscopy (XPS) results of the three respective materials Pd-A, Pd-B and
Pd-C are provided in the Figure 2. The binding energy peaks in the survey spectra appeared
at 284.9, 397.8, 530.8 and around 338.4-339.3 eV could be assigned to the C 1s, N 1s, O 1s
and Pd 3d, respectively (Figure 2a). All the binding energies were calibrated via referencing
to C 1s binding energy (284.9 eV). The presence of N 1s binding energy peak at 397.8 eV of
the as-synthesized POP material corresponds to the single type symmetrical pyridinic N
atoms of the 1,3,5-triazine unit (Figure 2b). The XPS spectra of the N-1s core level of the
three Pd-POP based materials indeed showed a positive displacement by 0.6 and 1.5 eV,
respectively, in comparison with the as-prepared POP, demonstrating strong interaction
between the N-rich POP support and Pd-NPs (Figure 2b). It has also been noticed that the
interaction between support and Pd-NPs remarkably enhanced with the change of respective
synthesis procedure, referring to the strong co-ordination of small Pd-NPs in Pd-C material
inside the nanoporous channel. The O 1s spectral regions (Figure 2c) of the respective Pd-A,
Pd-B and Pd-C materials reveal a significant contribution in co-ordination with the Pd-NPs of
the allyloxy unit from the polymeric framework, as is evident from the O 1s binding energy
positive shift compared with the parent POP, which is further supported with previous report
[21]
in Urea-Based Porous Organic Frameworks. To have a clear insight in the oxidation state
of surface palladium NPs in the respective Pd-POP materials we have carried out XPS spectra
in Pd 3d binding energy core (Figures 2d, 2e & 2f). Two sets of spin-orbit doublet could be
recognized from Pd (3d5/2,3/2) binding energy signals with the intensity ratio at 3:2
considering spin-orbit separation of 5.2 eV. The low-binding-energy doublets of Pd 3d5/2 and
Pd 3d3/2 (Figures 2d, 2e & 2f) appeared at 335.35, 335.56 and 334.92 eV as well as 340.71,
340.75 and 340.23 eV for Pd-A, Pd-B and Pd-C, respectively, could be accepted to the
formation of Pd (0) species in the nanohybrids, while the characteristic peaks of Pd 3d5/2
binding energies at 336.61, 336.64 and 336.50 eV, respectively, pointed at the possibility for
the presence of Pd (II) species. In comparison with that in Pd-A and Pd-B, a lower binding
energy shift of about 0.44 eV in Pd-C in the Pd 3d XPS spectrum (Figure 2f) was observed
which may be assigned to the strong electron-donation effect of nitrogen atoms to small size
14
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
Pd(0) NPs distributed inside the nano cavities of the POP, resulting in more electron-rich
Pd(0) species in Pd-C. Very recently, Wang and co-workers have observed similar trend of
negative binding energy displacement in Pd 3d spectrum for Pd-POP based materials with
stronger interaction between nitrogen atoms and small size palladium NPs present in the
[22]
interior pores of the POP material. The ratios of Pd(0) to Pd(II) in Pd-A, Pd-B and Pd-C
are 1.96. 1.02 and 0.61, respectively. The occurrence of this Pd (II) species in Pd 3d XPS
spectra of the Pd-POP catalysts could be recognized to uniformly distributed single site
isolated Pd strongly interacting with the allyloxy triazine moiety through O or N atoms which
is in good agreement with the recent results of Bulusheva et al for their N-Doped Carbon
[23]
supported Pd catalyst. Very recently, Arrigo et al. have investigated variation in spectral
studies of N atoms after Pd deposition on support which could be attributed to N-Pd covalent
bond formation owing to strong interaction between pyridinic N atoms and Pd, with the
subsequent outcome of Pd 3d binding energy shift very close to divalent Pd and they have
also concluded that the coordination of Pd(II) species to nitrogen of the support makes it
[24]
difficult in reduction.
Electron microscopic analyses:
In order to investigate the distribution, shape, location of Pd-NPs on the polymer matrix
and their average particle sizes we have conducted Transmission Electron Microscopy (TEM)
analysis of the respective Pd-POP materials (Figure 3). TEM images illustrate black color Pd-
NPs in Pd-A (Figure 3A-C), Pd-B (Figure 3F, G) and Pd-C (Figure 3K, L) are strongly
encapsulated with their mean particle sizes 3.33±0.5, 8.0±0.33 and 2.71±0.25 nm,
respectively on the POP surface. A dual size and homogeneous distribution of Pd-NPs with
the average sizes 1.31±0.36 and 2.71±0.25 nm, respectively, has been observed from the
TEM images of Pd-C (Figure 3K & L), where relatively small Pd-NPs (1.31 nm) are
accommodated in the interior cavities and comparatively large Pd-NPs (2.71 nm) are
deposited on the external surface of POP. On the basis of Wang and co-workers previous
reports we can assume that selection of H
2
as the reducing agent may promote the particle
nucleation within the nanoporous channel under solid-phase conditions by nano-confinement
[22]
effect of POP, thereby restricting the unnecessary growth of Pd-NPs in Pd-C material. In
contrast, there is an obvious aggregation and worse dispersion of palladium in Pd-B (Figure
3G) leading to the formation of large Pd-NP (8-10 nm with mean particle size) than the Pd-A
and Pd-C which could be explained by stabilization of Pd-NPs on the basis of nitrogen
15
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
content variation as measured from C, H, N elemental analysis (Table S1, SI) and this
[25]
phenomena has already been shown and discussed by Zhang et al.
Figure 3. TEM images (A, B & C), (F & G), (K & L) and corresponding HR-TEM images
with clear lattice fringe (D), (H & I), (M & N) and SAED patterns with marked facets (E),
(J), (O) of Pd-A, Pd-B and Pd-C materials, respectively. Average particle size distributions
(P, Q and R) of Pd-A, Pd-B and Pd-C materials, respectively. High magnified TEM image
(S) of single Pd nanoparticle of Pd-B. Scale bars (A, 20 nm), (B, 50 nm), (C, 20 nm), (D, 2
nm), (F, 200 nm), (G, 50 nm), (H, 5 nm), (I, 5 nm), (K. 100 nm), (L, 20 nm), (M, 2 nm), (N, 2
nm).
Also Pd-A material exhibits a better homogeneous distribution (Figures 3B & C) but with the
larger particle size (3.33±0.5) than the Pd-C, which is probably ascribed to the fact that
4
aqueous solution of NaBH cannot access internal pores efficiently by capillary force due to
the hydrophobic nature of nanoporous polymer, causing diffusion of the Pd-NPs and their
precursors from interior cavity to the external surface of the polymer, resulting the formation
of large particle size.
16
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
Figure 4. HAADF-STEM images with the corresponding elemental mapping of C, N, O and
Pd for A) Pd-A, B) Pd-B and C) Pd-C materials, respectively.
The morphology and crystalline feature of the anchored Pd-NPs was further established at a
closer view from the High Resolution (HR) TEM images (Figures 3 D, H, I, M & N) of the
respective Pd-POP materials which demonstrate that the distinguished lattice fringes with the
d spacing about 2.14Å for Pd NPs, clearly assigning to the (111) lattice spacing of face
centered cubic (fcc) arrangement of the Pd(0) nanopartciles. Characteristic bright rings
associated with distinct spots are obviously observed in the Selected Area Electron
17
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
Diffraction (SAED) Patterns (Figure 3E, J & O), which could be readily indexed to the
corresponding (311), (400), (222), (200), (331), (220), (111) crystalline reflections planes of
[26]
fcc Pd. The average Pd NPs size distributions of the Pd-A, Pd-B and Pd-C materials are
displayed in the inset of Figure 3P, 3Q and 3R, respectively. From the above TEM images
analysis we can definitely conclude that the synthesis method by varying the reducing agents
4 2
from NaBH , HCHO to H exhibits a considerable effect on the size and dispersion of NPs.
High-angle annular dark-field scanning TEM (HAADF-STEM) images of the respective Pd-
A, Pd-B and Pd-C materials (Figure 4) reveal that the as-synthesized POP was decorated with
uniformly distributed palladium NPs (bright spots). The existence of evenly distributed Pd
NPs in the nitrogen rich carbon matrix can be further confirmed from their corresponding
EDX (Energy Dispersive X-ray) elemental mapping analysis (Figure 4). The particle size
distributions measured from the respective HAADF-STEM images clearly indicate to the
existence of similar average particle size with a very negligible difference as determined from
the conventional TEM images.
Catalytic Hydrodeoxygenation (HDO) of Vanillin with Pd-POP based catalysts:
We have investigated the catalytic performance of our newly designed Porous Organic
Polymer decorated Pd nano catalysts in biofuel upgrading considering liquid phase
hydrodeoxygenation of Vanillin into 2-methoxy-4-methylphenol, a typically model
component of pyrolysis oil derived from the lignin fraction. The catalytic transformation of
3
the C=O to -CH group is involved with the three different path ways which includes i)
hydrogenation/dehydration, (ii) hydrogenation/hydrogenolysis, and (iii) direct hydrogenolysis
of the C=O bond, respectively. Hydrogenation/dehydration pathway is not feasible due to
absence of α-hydrogen in the reduced product vanillin alcohol. So the progress of the
transformation from vanillin to 2-methoxy-4-methylphenol must be involved either by
hydrogenation/hydrogenolysis or direct hydrogenolysis. Our initial experiment began by
conducting hydrodeoxygenation (HDO) reaction of Vanillin (100 mg, 0.657 mmol) with
isopropanol (30 mL) over Pd-A nano catalyst in a stainless steel reactor inbuilt with a
°
pressure gauge setup under 10 bar hydrogen pressure at 140 C temperature. The
corresponding Pd-B and Pd-C catalysts were also tested under the identical reaction
conditions for comparison study. The contents of products and reactants were determined
using a GC-FID based on authentic samples. Among the diverse Pd-POP based catalysts
investigated, the Pd-C exhibits superior catalytic performance in HDO reaction of vanillin
providing conversion (96%) and selectivity (98%) to the desired product 2-methoxy-4-
18
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
-1
methylphenol, affording a magnitude TOF of 8.51 h (Figure 5A). On the other hand, we
have achieved a 91% and 84% conversion of vanillin with 98% and 86% selectivity of 2-
methoxy-4-methylphenol after 18 h reaction time for the corresponding Pd-A and Pd-B
-1
-1
catalysts, exhibiting TOF values of 7.18 h and 6.72 h , respectively. When we have
evaluated the catalytic activity of bare Pd-NPs in HDO reaction of vanillin under optimized
reaction conditions, the vanillin conversion was drastically decreased to 16% with a targeted
-1
product selectivity of 85%, displaying a TOF value of 1.39 h (Figure 5A). The superior
catalytic performance of Pd-C can be attributed not only due to its higher nitrogen content
which facilitates improved homogeneous dispersion of smaller mean size of Pd-NPs, but also
due to the existence of electron-rich Pd formed by the strong electronic interaction between
Pd and N atoms from 1,3,5-triazine unit, favouring preferential hydrogenation over the
[27]
electron-enriched metallic Pd, supported with the previous reports by Tsang et al.
The
existing electron transfer from nitrogen functional group of the support to the Pd was further
confirmed by the higher positive and negative displacement in XPS binding energies of N
and Pd core regions of Pd-C compared with the other catalysts. Similarly Arrigo and co-
workers have found that the N-Pd interaction in Pd nanoparticles immobilized on N-doped
carbon nanotubes was responsible for the enhanced catalytic activity in acetylene
[24]
hydrogenation. Small size Pd-NPs (1.31 nm) in the Pd-C catalyst may play a decisive role
for the improved catalytic performance compared with the other catalysts which possess large
size Pd-NPs (3.33-10 nm) on the external surface of the POP owing to their high surface-to-
volume ratio with huge number of available active sites per unit area for reaction substrates.
Shall and co-workers have investigated that large Pd-NPs (5.5 nm in size) distributed on the
outer surface of Pd/NH
2
-UiO-66 provided a lower catalytic activity and selectivity of desired
product in HDO of vanillin compared with the embedded Pd-NPs inside the cavity of NH
2
-
[9]
UiO-66 having mean size in the range of 1.5-2.5 nm. Similar conclusion has been drawn
regarding the higher catalytic activity of Pd@MIL-101 in HDO of vanillin where smaller Pd-
NPs with an average size of 1.8±0.2 nm are accommodated inside the mesoporous cavities of
[12]
MIL-101 than the large particles present on the external surface.
Following these
observations, we can emphasize that small Pd particle size become the determining factor for
such higher activity which resembles well for the poor activity in Pd-B. In order to validate
the feasibility of involved catalytic pathway for the hydrodeoxygenation of vanillin over the
newly designed Pd-POP catalysts, we have further examined the distribution of substrate and
product selectivity as a function of reaction time, influence of the reaction temperature and
19
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
effect of hydrogen pressure. Evolution in the reactant and product distributions with the
progress of the reaction as a function of reaction time over the Pd-A, Pd-B and Pd-C catalysts
are provided in the Figures 5B, 5C and 5D, respectively.
Figure 5. A) Comparison in catalytic activity with respect to TOF (turn over frequency) for
vanillin hydrodeoxygenation reaction with Pd based catalysts; Reaction conditions: vanillin
(
1
100 mg, 0.657 mmol), Pd catalyst (20 mg), isopropanol (30 mL), 10 bar H
8 h; Evolution of reactant and product distributions as a function of time over B) Pd-A, C)
Pd-B and D) Pd-C catalysts, respectively; Reaction conditions: vanillin (100 mg, 0.657
mmol), Pd catalyst (20 mg), isopropanol (30 mL), 10 bar H pressure, 140°C; E) Influence of
2
pressure, 140°C,
2
reaction temperature with diffrent Pd-POP catalysts in vanillin HDO reaction; Reaction
conditions: vanillin (100 mg, 0.657 mmol), Pd catalyst (20 mg), isopropanol (30 mL), 10 bar
H
2
pressure, 18 h; F) Effect of H
conditions: vanillin (100 mg, 0.657 mmol), Pd catalyst (20 mg), isopropanol (30 mL), 140°C,
8 h.
2
pressure on conversion and product selectivity; Reaction
1
The reaction was accompanied with a progressive increase in vanillin alcohol selectivity
~87-93%) and conversion of vanillin (~10-32%) at initial 3 h, which demonstrates that
(
vanillin, is mainly hydrogenated to vanillin alcohol in the first step. However, in the first 3 h
we have noticed appearance of little amount of 2-methoxy-4-methylphenol (~10%),
illustrating that vanillin alcohol underwent hydrogenolysis reaction to deliver the desired
product. Surprisingly, no peak for 2-methoxy-4-methylphenol at initial 3 h for Pd-B catalyst
has been observed from the Time-on-stream profile (Figure 5C), suggesting the poorer
catalytic activity in hydrogenolysis reaction. The catalytic conversion for this
20
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
hydrodeoxygenation reaction could be observed at after 3 h with the very les selectivity of 2-
methoxy-4-methylphenol but selectivity vanillin alcohol was found to be very high. But after
8 h, all the produced in situ vanillin alcohol had been completely transformed into 2-
1
methoxy-4-methylphenol via hydrogenolysis. These results consistently prove that
transformation of vanillin to 2-methoxy-4-methylphenol proceeded by hydrogenation to
[28]
vanillin alcohol followed by hydrogenolysis.
The decisive role of reaction temperature in hydrodeoxygenation of vanillin over our
newly developed Pd-POP catalysts has been thoroughly investigated by conducting the
catalytic reaction at different reaction temperatures (80°C-160°C) under 10 bar H
2
pressure
(Figure 5E). An enhancement in catalytic performance for Pd-A catalyst with a high vanillin
conversion (~95%) and 2-methoxy-4-methylphenol selectivity of >90% has been achieved,
upon raising the temperature from 100°C to 140°C. The examination results of the other two
respective Pd-B (from 44% to 84%) and Pd-C (from 52% to 96%) catalysts unambiguously
demonstrate the similar accelerating trends in catalytic activity with higher reaction
temperature, suggesting much faster reaction kinetics, which can be further enhanced to
1
9
00% at 160°C. However, a gradual drop in 2-methoxy-4-methylphenol selectivity (from
1% to 80%) on Pd-A, (from 90% to 83%) on Pd-B and (from 96% to 78%) on Pd-C
catalysts, respectively, at 160°C has been observed, which could be attributed to the
appearance of Guaiacol, p-cresol and Cyclohexanol by some undesired side reactions
including decarbonylation, deeper hydrogenation and hydrogenolysis thereby causing some
catalyst deactivation. Previous reports also indicated that elevated temperature offered an
acceleration in polycondensation, which was thought to be accountable for the loss in mass
[29]
balance. We have performed the catalytic activity of newly developed Pd-POP catalysts
for HDO of vanillin with various solvents including isopropanol, water, methanol and
isopropanol and water mixture (Table S2, SI). Among these solvents tested isopropanol-water
mixture proved to be the best one to provide better conversion under mild conditions (0.7
2
MPa H and 120 °C) but selectivity of the targeted product is less compared with the
isopropanol owing to the high solubility of vanillin alcohol in water. In contrast, we have
noticed very poor conversion of vanillin but with very high selectivity of 2-methoxy-4-
methylphenol in water. Most of the previous literature reports predict that the HDO of
vanillin was conducted very efficiently either in water or in water/oil emulsion medium. But
in our case the reverse trend is noticed. This new trend could be explained on the basis of
type of nitrogen atoms present inside the supported catalytic systems. It has been proposed by
21
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
Jiang and co-workers that polarization effect between graphitic-N and adjacent carbon atoms
significantly alter the charge distribution over N-atoms thereby favouring interaction with
nucleophilic H O molecules and as a result the superior catalytic performance for the carbon
nitride based catalyst in vanillin HDO reaction in aqueous medium is observed.
2
[11]
This
charge distribution of the N-doped graphitic material exhibits a crucial role in the
enhancement of catalytic activity. The absence of graphitic-N atoms rather than the presence
of pyridinic-N atoms in our as-synthesized Pd-POP based catalyst has been experimentally
evidenced by the appearance of XPS binding energy at around 398.4-399.3 eV in the N-1S
spectra (Figure 2b). Kim and co-workers proposed that lone-pair electrons over the pyridinic-
N atoms in the catalyst does not take part in favourable interaction with the water molecules,
[30]
resulting in poor contact with the substrates and suppression of the catalytic activity.
Different hydrogen concentration (3-15 bar) were subsequently screened on vanillin
hydrodeoxygenation at the optimized temperature of 140°C (Figure 5F). It is noticed from the
Figure 5F that, in the presence of hydrogen environment i.e., without any pressure showed
poor activity indicating hydrogen pressure is essential for the improvement of catalytic
activity. An increase in H pressure (from 5-10 bar) led to an enhancement in vanillin
2
conversion and selectivity of 2-methoxy-4-methylphenol. However, with further increase of
hydrogen pressure above 10 bar remarkably decreased the 2-methoxy-4-methylphenol
selectivity, with the subsequent enhancement in undesired side products (e.g. Guaiacol and
anisole) selectivity including over-hydrogenation and decarbonylation pathway owing to the
availability of high concentration of hydrogen. Encouraged with this promising result and
considering that Pd-based materials can serve as an effective catalysts in hydrodeoxygenation
of vanillin, we have synthesized a series of supported Pd catalysts following the solid phase
H
2
reduction method and investigated their catalytic activity under our standard reaction
conditions. The comparison study with various Pd based catalysts including Pd-C, Pd-SiO
Pd-CeO , Pd-TiO , are summarized in Figure 6A. The textural proprieties of the conventional
catalysts are provided in the Figure S4, SI. When Pd-C and Pd-SiO have been utilized, we
2
,
2
2
2
have achieved 84.7% and 82.1% conversion with 76.7% and 47.3% selectivity of desired
product, respectively. The observed poor selectivity of targeted product below 30% for the
2 2
catalysts of Pd-CeO and Pd-TiO could be delineated as the some extent of vanillin alcohol
are not fully converted to produce 2-methoxy-4-methylphenol. The observed relatively low
selectivity towards the targeted product (Figure 6A), probably due to their poor wettability,
low mass transfer, less surface area, poor dispersion of nanoparticles, less metal-support
22
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
interaction and large Pd NPs. Some aggregation of Pd-NPs was noticed from the TEM
images of the conventional catalysts (Figure S4, SI) which could be pointing out towards the
poor catalytic activity. Although the reference catalysts provided satisfactory product
conversions but didn’t show long-term durability in succession of several catalytic runs.
Leaching of Pd-NP from the reference catalysts has also been found out owing to the absence
of strongly anchoring sites causing to the diminishment in catalyst recyclability. Delightedly,
oxides supported Pd NPs were found to be reasonably effective for this catalytic bio-fuel
upgrading reaction, furnishing 2-methoxy-4-methylphenol in lower conversions than Pd-POP
(Figure 6A), inferring that the N dopant for the catalysts played a pivotal role for this
catalytic reaction. Compared with the highly conventional solid catalysts, doping with N
atoms enhanced the hydrophilic nature of Pd-POP, homogenous dispersion and stabilization
of Pd-NPs, which could improve the catalyst dispersion along with the maximum exposure of
catalytic active sites to the substrates (e.g., vanillin), thereby increasing the catalytic
performance significantly. Our explanation is in good agreement with the improved catalytic
activity of Pd@CN0.132 compared with the conventional catalysts in vanillin
[7]
hydrodeoxygenation reaction as reported by Wang et.al.
The superior catalytic performance of Pd-POP catalyst compared with the other
conventional catalysts can be attributed to the homogeneously distributed electron-rich
metallic Pd sites offering excellent hydrogenation activity and high surface area and
hierarchical pores facilitates hydrogenolysis of the intermediate to the desired product with
fast and easy diffusion. The high catalytic activity can also be explained by easy adsorption
of organic substrates onto the nanoporous matrix by hydrophobic interactions and from π-π
interactions between aromatic aldehyde and aromatic POPs framework. We have previously
reported this special advantage of the POP based catalysts in nitrobenzene and alkene
[14a,14g]
hydrogenation reactions with Ru-POP and Pd-POP based catalysts.
We have also
examined vanillin hydrodeoxygenation reaction of Pd/POP (Figure 6A), has been developed
by employing a physical mixture of triazine based polymer and Pd NPs, but surprisingly only
36.7% conversion with 26% and 70% selectivity of 2-methoxy-4-methylphenol and vanillin
alcohol, respectively, has been attained after 20 h. The greater catalytic efficiency of the
Pd@POP material than Pd/POP can be ascribed to the synergetic effects of nano-confinement
and electron-donation offered by the POP framework, making the Pd surface more electron
rich and active, thereby facilitating for the effective hydrogenation reaction. This finding
supports the previous assumption of the Xu research group associated with the efficient
23
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
Table 1: Comparison in catalytic performance for Hydrodeoxygenation of vanillin over
different catalysts
Entry
Catalyst
Con (%)
Selectivity (%)
Alcohol Cresol
Ref
8
7
45.8
5
4.2
Pd/MSMF
Pd/CN0.132
>
99.5
1
2
3
4
31
69
47
65
3
1
1
0
85
53
Pd/SWNT-SiO2
Pd/SO H-MIL-101 96.1
9.1
0
90.9
100
3
9
Pd@NH -UiO-66
100
2
5
6
7
11
Pd/NPC-ZIF-8
1
00
0
100
9
6.5
1.8
98.2
Our Work
Pd-POP
vanillin HDO reaction where higher quantity of vanillin alcohol has been observed for the
Pd-NPs with better accessibility but no nano-confinement effect in Pd/MIL-101 owing to a
[12]
weak interaction between the reactant and active centres. Furthermore, we have carried out
the comparison study in the catalytic performance, desired product selectivity over the newly
developed Pd-POP catalyst and as well as catalyst durability with those reported in the
literature (Table 1). The survey reports indicated that a 99.5% conversion of vanillin with
54.2% 2-methoxy-4-methylphenol selectivity could be achieved over 4.5 wt% Pd/MSMF
[8]
catalyst (entry 1 in Table 1) by Xiao et al. Additionally, superior catalytic performance for
our developed catalyst compared with the N-doped carbon-supported Pd catalyst and
palladium nanoparticles deposited onto the SWNT-inorganic oxide hybrid catalyst
[31]
(
Pd/SWNT-SiO ) has been observed (entries 2 and 3, Table 1). On the other hand, Shall
2
3
and co-workers have investigated hydrodeoxygenation of vanillin over Pd/SO H-MIL-101
24
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
catalyst and a 96.1% conversion of vanillin with lower selectivity (entry 4, Table 1) for 2-
[10]
methoxy-4-methylphenol has been achieved.
In contrast, Pd@NH -UiO-66 catalyst as
2
prepared by Shall and co-workers exhibits a remarkably high
Table 2: Use of different hydrogen source for hydrodeoxygenation of vanillin with Pd-POP
Nanocatalyst
a
Reaction conditions: vanillin (100 mg, 0.657 mmol), Pd catalyst (0.014 wt% of Pd),
b
HCOOH (3.28 mmol, 0.3 mL), EtOH, 120°C, 15 h, under sealed-tube conditions; vanillin
i
(
100 mg, 0.657 mmol), Pd catalyst (0.014 wt% of Pd), PrOH(30 mL), 150°C, 15 h, under
high pressure autoclave condition.
catalytic activity and selectivity for 2-methoxy-4-methylphenol (entry 5, Table 1) but the
[9]
durability of the catalyst in terms of recyclability does not show any impressive result. The
same phenomenon has been observed for the Pd/NPC-ZIF-8 catalyst (catalyst could be
[11]
recycled for only three successive runs) as reported by Jiang et al. (entry 6, Table 1). The
th
long term durability of our new catalyst in the 10 catalytic run for succession compared with
the previously reported catalysts could be explained on the basis of strong catalyst support
interaction which is strongly dependent on the configurations of the N-dopants inside the
25
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
highly crosslinked rigid polymer materials. It has been well documented that pyridinic-N
atoms bind more strongly with the transition metal than the other type of nitrogen atoms
owing to the larger overlap between nonbonding state of lone pair electrons and the d-orbital
[32]
of the transition metal. These newly designed Pd-POP catalysts hold promising potential
for biofuel upgrade processes, and additional work will be conducted toward the applications
to natural lignins and other model compounds. We have further examined the catalytic
activity of our newly designed Pd-POP catalysts in hydrodeoxygenation of vanillin (Table 2)
by employing various hydrogen sources including alcohol, formic acid, which can donate
hydrogen to reduce the substrate in a process called catalytic transfer hydrogenation (CTH).
i
HCOOH and PrOH both were considered as the poor hydrogen sources in HDO of vanillin
providing around 40-50% and 60-65% conversion, respectively, with higher selectivity for
vanillin alcohol. But the past decade has witnessed that the hydrodeoxygenation of vanillin to
2
-methoxy-4-methylphenol, a potential future biofuel could be executed in the presence of
Formic acid is a remarkable bio-derived source of hydrogen employing AgPd@g-C and
Pd/TiO @N-C) catalysts under visible light and UV light mediated reaction conditions.
Catalyst reusability:
3
N
4
[33]
(
2
The outstanding catalytic performance of Pd-A, Pd-B, and Pd-C encouraged us to investigate
their stability and recyclability, which are also important aspects in heterogeneous catalytic
systems. In order to evaluate the recyclability of the newly designed Pd-POP systems we
have conducted the HDO reaction of vanillin under the applied batch conditions and the
results are provided in the Figure 6B. The solid catalyst was recovered after the reaction by
centrifugation technique followed by washing with methanol 2-3 times, dried at 80°C in oven
overnight, and used for the next cycle reaction in succession. The detail about the
recyclability test has been provided in the Experimental Section. Under these conditions, our
Pd-C catalyst could be recycled in successive ten runs (Figure 6B) sustaining 2-methoxy-4-
methylphenol selectivity and yield. Only a little drop in vanillin conversion from 96.5 to
91.9% was observed upon the seventh reuse (Figure 6B). In contrast, Pd-B and Pd-A
catalysts could also be reused for fifth and seventh consecutive catalytic runs, affording a
significant decrease in conversion from 84.3 to 61.2% and 91.2 to 75.3%, in the fifth and
ninth runs, respectively. On the other hand from the above investigation we can conclude that
Pd-C is highly stable catalyst compared with the others providing a near-quantitative
26
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
Figure 6. A) Comparison study in catalytic performances with various supported Pd
nanohybrid catalysts, Reaction conditions: vanillin (100 mg, 0.657 mmol), Pd catalyst (20
mg), isopropanol (30 mL), 10 bar H pressure, 140°C, 18 h; B) Recycle potential diagram for
2
catalytic hydordeoxygenation of vanillin with Pd-A, Pd-B and Pd-C catalyts, respectively;
Reaction conditions: vanillin (500 mg, 3.375 mmol), Pd catalyst (100 mg), isopropanol (150
mL), 10 bar H pressure, 140°C, 18 h; C) Evolution of reactant and product distributions in
2
recyclability test for Pd-C catalyst; D) Hot-filtration experiments where Pd-C catalyst was
removed after 6 h and 18 h followed by addition of vanillin (100 mg, 0.657 mmol), Pd
catalyst (20 mg), isopropanol (30 mL), 10 bar H
2
pressure, 140°C, 18 h.
2-methoxy-4-methylphenol yield up to the tenth recycling run. Figure 6C suggested that
selectivity of the desired product and intermediate is nicely preserved in each catalytic run. A
slight drop in the vanillin conversion and desired product selectivity for Pd-C in the recycling
test at tenth cycle 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 the accessibility of active sites.
27
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
Reused Catalyst Characterization:
We have performed TEM analysis of the respective recycled catalysts after consecutive
catalytic reactions in order to check the location, distribution and stability of Pd-NPs of these
catalytic systems (Figure 7). It is quite evident from the Figures 7A & B, the mean sizes of
palladium particle uniformly dispersed on the external surface in Pd-A catalyst are slightly
th
increased from 3.33 to 3.54 nm after 7 catalytic run. In contrast, a severe agglomeration of
th
Pd-NPs for Pd-B catalyst has been observed after 5 catalytic run thereby increasing the
average particle sizes from 8.65 to 12.73 nm (Figures 7C & D). The increase in mean particle
size and severe aggregation of palladium nanoparticles signify to the drastic drop in
conversion (%) during the course of recyclability test for Pd-B catalyst. However, the average
palladium NPs size in Pd-C increases to (3.22±0.43) nm compared with the fresh one after
the tenth reaction cycle (Figures 7E & F), which may be attributed to the diffusion out of the
small palladium NPs from the interior cavity and deposition on the external surface during
the consecutive reaction resulting an increase in mean particle size. This observation nicely
[34]
resembles with the reported NPs encapsulated in the interior pores of POPs by Wang et al.
The impressive Pd-C catalyst stability during the consecutive reaction cycles could be
emphasized on the basis of the nano-confinement effect of the host framework and the strong
interaction of nitrogen rich 1,3,5-triazine linkage to encapsulated Pd-NPs thereby making
1
3
them leaching resistant. C CP solid state MAS NMR characterization study of the reused
catalysts (Figure S5, SI) revealed the mechanical and chemical stability, evading the
associated structural degradation of the catalyst.
28
This article is protected by copyright. All rights reserved.

ChemCatChem
10.1002/cctc.201700186
th
th
Figure 7: TEM images of reused Pd-A (A & B) after 7 catalytic run, Pd-B (C & D) after 5
th
catalytic run and Pd-C (E & F) after 10 catalytic run, respectively.
In order to access the robustness and heterogeneous nature of the designed Pd-POP
catalysts we have conducted Hot filtration and Hg (0) poisoning test under optimized reaction
conditions. Typically, in the hot-filtartion test, we have stopped the recation releasing the
hydrogen pressure from the stainless steel reactor after 6 h, followed by cooling down at
room temperature. The black solid was separated from the reaction mixture by centrifugation.
54.6% vanillin conversion after 6 h has been reached (Figure 6D). Then we have continoued
29
This article is protected by copyright. All rights reserved.