Nitrate reduction on the surface of bimetallic catalysts supported by nano-crystalline beta-zeolite (NBeta)

Article abstract of DOI:10.1039/c6gc02349e

Nitrate contamination of surface and groundwater is one of the most serious environmental issues due to its adverse effects on human health and the ecosystem. In this study, novel bimetallic catalysts supported by nanocrystalline beta zeolite (NBeta) were firstly developed for the complete and selective reduction of nitrate to environment-friendly nitrogen. The synthesis of a NBeta support and bimetallic catalysts was confirmed by X-ray diffraction and Fourier transformation infra-red spectroscopy. The morphology of the NBeta support and metal dispersion was confirmed by transmission electron microscopy and scanning electron microscopy with energy dispersive X-ray, which revealed that newly synthesized NBeta consisted of individually separated cubical shaped crystals and that the metals were well dispersed in uniform and closely packed ensembles. The nitrate reduction mechanism on NBeta-supported bimetallic catalysts was also identified by inductively coupled plasma mass spectroscopy, X-ray photoelectron spectroscopy, temperature-programmed reduction, and H2 pulse chemisorption. All catalysts successfully achieved 100% NO3 ? and NO2 ? removal. Sn-Pd-NBeta showed the highest reduction kinetics (k = 19.09 × 10?2 min?1) followed by In-Pd-NBeta (k = 2.89 × 10?2 min?1) and Cu-Pd-NBeta (k = 2.31 × 10?2 min?1). However, N2 selectivity was in the order of Cu-Pd-NBeta (92.68%) > In-Pd-NBeta (82.93%) > Sn-Pd-NBeta (80.80%). Characterization of the catalysts revealed that the highest N2 selectivity of Cu-Pd-NBeta was due to the suppression of a deep hydrogenation capacity of Pd by Cu. The NBeta-supported Cu-Pd, In-Pd, and Sn-Pd catalysts also showed a complete NO3 ? reduction with consistent N2 selectivity in successive and regenerated cycles. The results suggest an eco-friendly catalytic system which can be applied to the optimal ex situ remediation of water and wastewater contaminated with nitrate.

Full text of DOI:10.1039/c6gc02349e

View Article Online
View Journal
Green
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: S. Hamid, M. A.
Kumar, J. Han, H. Kim and W. Lee, Green Chem., 2016, DOI: 10.1039/C6GC02349E.
This is an Accepted Manuscript, which has been through the
Volume 18 Number
7 7 April 2016 Pages 1821–2242
Royal Society of Chemistry peer review process and has been
accepted for publication.
Green
Chemistry
Accepted Manuscripts are published online shortly after
Cutting-edge research for a greener sustainable future
www.rsc.org/greenchem
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
author guidelines.
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, outlined
in our author and reviewer resource centre, 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
CRITICAL REVIEW
G. Chatel et al.
Heterogeneous catalytic oxidation for lignin valorization into valuable
chemicals: what results? What limitations? What trends?
rsc.li/green-chem

Page 1 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
Nitrate Reduction on the Surface of Bimetallic Catalysts Supported by Nano-crystalline
Beta-zeolite (NBeta)
Shanawar Hamida, Macharla Arun Kumara, Jong-In Hana, Hyungjun Kimb, Woojin Leec,*
a Department of Civil and Environmental Engineering, Korea Advanced Institute of Science and
Technology, 291 Daehakꢀro, Yuseongꢀgu, Daejeon 34141, Republic of Korea
b Graduate School of Energy, Environment, Water, and Sustainability (EEWS), Korea Advanced
Institute of Science and Technology, 291 Daehakꢀro, YuseongꢀGu, Daejeon 305ꢀ701, Korea
c School of Environmental Science and Engineering, Pohang University of Science and
Technology, 77 Cheongamꢀro, Namꢀgu, Pohang 37673, Republic of Korea
*Corresponding author: phone: 82ꢀ10ꢀ8835ꢀ4292
Eꢀmail: woojin_lee@kaist.edu
A manuscript submitted to Green Chemistry
November, 2016
1

Green Chemistry
Page 2 of 45
View Article Online
DOI: 10.1039/C6GC02349E
Abstract
Nitrate contamination of surface and groundwater is one of the most serious environmental
issues due to its adverse effects on human health and ecosystem. In this study, novel bimetallic
catalysts supported by nanocrystalline beta zeolite (NBeta) were firstly developed for the
complete and selective reduction of nitrate to environmentꢀfriendly nitrogen. The synthesis of
NBeta support and bimetallic catalysts was confirmed by XꢀRay diffraction and Fourier
transformation infraꢀred spectroscopy. Morphology of NBeta support and metal dispersion was
confirmed by transmission electron microscopy and scanning electron microscopy with energy
dispersive Xꢀray, which revealed that newly synthesized NBeta consisted of individually
separated cubical shaped crystals and that the metals were well dispersed in uniform and closely
packed ensembles. The nitrate reduction mechanism on NBetaꢀsupported bimetallic catalysts
was also identified by inductively coupled plasma mass spectroscopy, Xꢀray photoelectron
spectroscopy, temperatureꢀprogrammed reduction, and H2 pulse chemisorption. All catalysts
ꢀ
ꢀ
successfully achieved 100% NO3 and NO2 removal. SnꢀPdꢀNBeta showed highest reduction
kinetics (k = 19.09×10ꢀ2×minꢀ1) followed by InꢀPdꢀNBeta (k = 2.89×10ꢀ2×minꢀ1) and CuꢀPdꢀ
NBeta (k = 2.31×10ꢀ2×minꢀ1). However, the N2 selectivity was in the order of CuꢀPdꢀNBeta
(92.68%) > InꢀPdꢀNBeta (82.93%) > Snꢀ PdꢀNBeta (80.80%). Characterization of catalysts
revealed that highest N2 selectivity by CuꢀPdꢀNBeta was due to the suppression of deep
hydrogenation capacity of Pd by Cu. The NBetaꢀsupported CuꢀPd, InꢀPd, and SnꢀPd catalysts
ꢀ
also showed a complete NO3 reduction with consistent N2 selectivity in successive and
regenerated cycles. The results suggest an ecoꢀfriendly catalytic system which can be applied to
optimal exꢀsitu remediation of water and wastewater contaminated with nitrate.
2

Page 3 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
1. Introduction
ꢀ
In recent decades, nitrate (NO3 ) input to surface and groundwater bodies have been rapidly
increasing due to excessive use of agricultural fertilizers and insufficient treatment of domestic
ꢀ
and industrial effluents. NO3 pollution is one of the most serious water pollution issues affecting
both potable and nonꢀpotable water sources. Consuming the drinking water contaminated by
ꢀ
NO3 is detrimental to human health because it causes methaemoglobinanemia and blue baby
ꢀ
syndrome.1 NO3 is also toxic to ecological cycles as it enhances algal blooms in large water
bodies which deplete dissolved oxygen.1ꢀ3 Many conventional technologies, e.g. biological
ꢀ
denitrification, reverse osmosis, and ion exchange are used for NO3 removal, however, they
suffer from many challenges, e.g. production of undesirable solid byproducts and concentrated
ꢀ
NO3 streams.1ꢀ5
ꢀ
Recently, catalytic NO3 hydrogenation by bimetallic catalysts has emerged as a viable
solution due to its inherent higher removal capacity and environmentally benign end product,
nitrogen (N2). Bimetallic catalysts are generally prepared by dispersing trace amounts of a
promoter metal (e.g. Cu, Sn, In, Ni, Zn) and a noble metal (e.g. Pd, Pt, Au) on a support surface.
A superior role of Sn, Cu, and In as promoters and Pd as a noble metal have been extensively
studied and well established. However, the selection of support material is still a prime area of
research because the reactivity and stability of bimetallic catalyst depend mainly on the
properties of the support surface.1,6ꢀ8 In addition, the inherent properties of a support material
ꢀ
(e.g., ecoꢀtoxicity) and its behavior in NO3 hydrogenation process (e.g., dissolution and stability)
are also important in view of field applications.7
Zeolites are crystalline solid structures consisting of a three dimensional network of
microporous aluminosilicates linked by corners sharing oxygen ions.9ꢀ11 Recent studies show that
3

Green Chemistry
Page 4 of 45
View Article Online
DOI: 10.1039/C6GC02349E
bimetallic catalysts supported by microcrystalline commercial zeolites (e.g. NaY, ZSMꢀ5, and
mordenite) are very effective in the hydrogenation of oxyanions due to their unique physcoꢀ
chemical properties.12ꢀ15 The physicoꢀchemical properties of a zeolite can be further enhanced by
its particle size shift from micro to nanoꢀscale. For example, a decrease in particle size of the
zeolite results in high external surface area, more exposed active sites, reduced diffusion pathꢀ
length, and increase in pore opening density.7,16 An enhancement of these properties of a zeolite
support results in a highly reactive and selective bimetallic catalyst.7,16
Nanoꢀcrystalline Betaꢀzeolite (NBeta) is a three dimensional network of 12ꢀmembered
ring which has shown intriguing results for hydrocarbon catalysis but has not been used as a
ꢀ
support material for NO3 reduction to date. NBeta could prove an excellent support material for
bimetallic catalysts because 1) it has higher surface area, unique shape for selectivity, and
surficial silanol groups thus an optimal metal dispersion could be achieved, 2) it is stable in
extreme thermal and chemical environments, thereby synthesis of a robust catalyst could be
achievable, 3) it is resistant to aggregation thus more active sites could be exposed and available,
and 4) it is insoluble and ecoꢀfriendly in aqueous solution thus an environmentally safe
remediation process could be developed. Due to these unique properties, NBeta was prepared
and used as a support material in this study to develop novel bimetallic catalysts for highly
ꢀ
reactive and selective removal of aqueous NO3 .
The main objective of this research is the synthesis and evaluation of NBeta as a novel
ꢀ
support surface of bimetallic catalysts for reactive and selective catalytic NO3 reduction. Cu, Sn,
and In (as promoter metals) and Pd (as noble metal) were used to prepare CuꢀPdꢀNBeta, SnꢀPdꢀ
NBeta, and InꢀPdꢀNBeta bimetallic catalysts. The specific objectives were 1) Characterization of
newly synthesized NBeta and CuꢀPdꢀNBeta, SnꢀPdꢀNBeta, and InꢀPdꢀNBeta bimetallic catalysts
4

Page 5 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
by XꢀRay diffraction (XRD), Fourier transformation infraꢀred spectroscopy (FTꢀIR),
transmission electron microscopy (TEM) and scanning electron microscopy (SEM), Ammonia
temperatureꢀ programmed desorption (NH3ꢀTPD), and BrunauerꢀEmmettꢀTeller (BET) specific
ꢀ
ꢀ
surface area, and 2) the effect of promoter metal type on NO3 and NO2 reduction kinetics and
ꢀ
product selectivity, 3) the systematic elucidation of reaction mechanism of NO3 reduction by
characterization experiments, i.e. inductively coupled plasma mass spectroscopy (ICPꢀMS), Xꢀ
ray photoelectron spectroscopy (XPS), temperatureꢀprogrammed reduction (TPR) and H2 pulse
chemisorption, and 4) evaluation of the stability of bimetallic catalysts over multiple cycles.
2. Materials and methods
2.1 Synthesis of NBeta
A 15.5 mL solution of 37 mmol tetraethyl ammonium hydroxide (TEAOH, 35% Aldrich), was
dropwise added to 30 mmol of NaOH (98%, Aldrich) and subsequently stirred for 10 min.17 This
stirred mixture was then added to a 22.3 ml solution of 100 mmol tetraethyl orthosilicate (99%,
Aldrich) at room temperature with stirring. After 30 min of stirring, a 10.2 mL solution
(3.33mmol prepared in warm water) of aluminum sulphate (≥95%, Fluka Japan) was added to
the mixture and stirred for 1 h more. Then 1.4 g (3.8 mmol) of cetyltrimethylammonium bromide
(≥99%, Sigma) dissolved in 40 mL of ethanol was added drop wise and the resultant gel was
finally stirred for 2 h. It was then dried by heating at 90 °C on a hotplate under continuous
stirring. The dried matrix was then ground into fine powder and shifted to a Teflon cup. The
Teflon cup was then transferred to a Teflonꢀlined autoclave for crystallization. A minute quantity
of water (0.2 mL water: 1 g dry gel) was also added into the autoclave. The autoclave was
completely sealed and placed in an oven at 175 °C for 24 h to achieve hydrothermal
5

Green Chemistry
Page 6 of 45
View Article Online
DOI: 10.1039/C6GC02349E
crystallization of the dry gel. After hydrothermal treatment, the solid crystalline powder was
removed from autoclave and washed with deionized water (DIW) via vacuum filtration until
neutral pH was reached and then dried in air. The asꢀsynthesized nanoꢀcrystalline beta zeolite
was then calcined at 550 °C for 8 h to remove surfactant.
2.2 Synthesis of bimetallic catalysts supported by NBeta
Bimetallic catalysts supported by NBeta were synthesized by successive impregnation method as
reported in our earlier study.7 The measured amount of NBeta was put into respective volume of
deꢀaerated deionized water (DDIW) (ratios were fixed as 1.5 g NBeta: 200 mL DDIW), stirred
10 minutes to homogenize the mixture and then sonicated 6 min (ultraꢀsonication at 40%
amplitude and pulse). Promoter metal solution was prepared by dissolving 2.2% (wt.% of NBeta)
promoter metal precursor i.e. Sn (tin(IV) chloride pentahydrate, 98%, Junsei Chemical Co.), Cu
(copper(II) chloride dehydrate, 97.5%, Samchun Pure Chemicals Co.), and In (In(III) chloride
(99.8%, Samchun Pure Chemicals Co.), in DDIW (ratio were fixed at 25 mL DDIW for
precursor solution: 200 mL NBeta aqueous NBeta solution). Similarly, 1.6% Pd precursor
solution was prepared by adding Pd precursor salt (Palladium (II) chloride, 99.9%, Sigmaꢀ
Aldrich) with ratio fixed at 50 mL DDIW for Pd precursor solution: 200 mL NBeta aqueous
NBeta solution and then sonicated for 2 h to completely disperse PdCl2. Respective promoter
metal solution was added dropwise into NBeta suspension and continuous stirred for 2 h. Then
Pd solution was dropwise added to promoterꢀ NBeta suspension and again continuous stirred for
2 h at room temperature. The mixture including respective promoter metal and Pd precursors was
then dried at 90 °C under continuous stirring on a hot plate. Finally the dried catalysts (SnꢀPdꢀ
6

Page 7 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
NBeta, CuꢀPdꢀNBeta and InꢀPdꢀNBeta) were calcined at 350 °C for 2 h to stabilize the metals on
the surface of NBeta.
Next, the calcined bimetallic catalyst was reduced by adding 0.01 mM NaBH4 (98.0%,
Samchun Pure Chemical Co., Korea) dropwise to bimetallic catalyst suspension.1,7,19 The
reduced bimetallic catalysts were vacuumꢀfiltered using a 0.2 ꢁm membrane filter and washed
with DDIW twice to remove residual chemicals. The catalyst thus reduced was immediately
ꢀ
shifted to batch reactor for NO3 reduction experiments.
2.3 Experimental procedure
ꢀ
ꢀ
The reactivity test for the catalytic NO3 and nitrite (NO2 ) reduction was performed in a 500 mL
glass reactor as reported in our previous studies.1,7,18 In general, a 100 mL DDIW was taken into
the reactor and H2 was purged through a glass diffuser in order to remove the dissolved oxygen.
After 30 min, a 0.25 g of reduced bimetallic catalyst was added to the reactor and the catalytic
ꢀ
suspension was continuously stirred at 300 rpm till the end of the experiment. The catalytic NO3
ꢀ
reduction was initiated by the injecting NO3 (potassium nitrate, 99.0%, Duksan Pure Chemical
ꢀ
Co.) stock solution (20,000 mg/L as NO3 ꢀN) in order to obtain an initial concentration of 30
mg/L as NO3ꢀN at room temperature (25 °C). H2 and CO2 gas were continuously supplied
through flow regulators as reducing and buffering agents respectively.
ꢀ
ꢀ
The catalysts were also evaluated for successive and repeated reduction of NO3 and NO2
according to the procedures described in our recent research works.7,18,19 In successive reactions,
a membrane filter (0.2 µm mixed cellulose ester, Advantech Japan) was attached on the sampling
point inside the reactor to avoid loss of catalyst.7,18 The treated water samples were obtained by
opening the sampling port and transferring into a falcon tube.7 After each successive reaction, the
7

Green Chemistry
Page 8 of 45
View Article Online
DOI: 10.1039/C6GC02349E
supply of H2 and CO2 gas was turned off and supply of argon gas was initiated to keep the
system at anaerobic conditions. Finally, the treated water remaining was drained though the filter
on the sampling port. The catalyst inside the reactor was rinsed with DDIW and it was drained
several times to wash out residual chemical species on the catalyst surface.7 To start next
successive reaction, a 100 mL of DDIW was injected to the reactor and purged with H2 for 30
ꢀ
ꢀ
min. Then respective stock solution (i.e. NO3 or NO2 ) was added into the reactor and CO2
supply restarted to initiate nitrate reduction as described above. This procedure was repeated for
3 times to complete successive reaction tests. For repeated reaction tests, we followed the same
procedure as described above and regenerated the catalyst at the end of each repeated test. For
the regeneration, the catalyst was filtered and washed with a vacuum filtration assembly at the
end of each experiment. The collected catalyst was dried overnight at 105 °C to remove all
moisture, and then calcined at 350 °C. The new experiment was started by reducing the catalyst
with NaBH4 as described above. The loss of catalyst during each regeneration process was
compensated by adjusting the volume to mass ratio of catalyst.7,19 Similarly, the volume of NO3
ꢀ
and NO2 stock solutions were also adjusted to keep constant concentration at 30 mg/L. 7,19
ꢀ
Unless stated otherwise, the experimental conditions were fixed at 2.2 wt% promoter
metal (i.e., Sn: 0.046 mmol; Cu: 0.087 mmol; and In: 0.048 mmol), 1.6 wt% noble metal (i.e. Pd:
0.0376 mmol), 30 mL/min H2, and 60 mL/min CO2. Batch experiments were conducted to
investigate the effect of different promoter metals (Sn, Cu, and In). All experiments were
conducted in duplicate.
2.4 Characterization of Catalysts
8

Page 9 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
The NBetaꢀsupported bimetallic catalysts were characterized using XRD, FTꢀIR, SEM/EDX,
TEM/EDX, ICPꢀMS, XPS, TPR, NH3ꢀTPD, BET, and H2 pulse chemisorption. More details on
instrumentation, sample preparation, and analytical procedures are provided in the SI.
2.5 Analytical Methods
We used all the analytical methods as reported in our previous studies.1,7,18 Briefly, an aliquot
amount of sample (4 mL) was collected from the reactor using a 5 mL disposable syringe (Korea
Vaccine Corp., Korea) at respective sampling time. The sample was immediately filtered using a
0.2 ꢁm membrane syringe filter and these filtered samples were used to determine the
ꢀ
ꢀ
+
concentrations of NO3 , NO2 , and ammonium (NH4 ), using ion chromatography (IC) (883 basic
IC plus, Metrohm, U.K.). The IC was equipped with a compact autoꢀsampler (863 Compact IC,
Metrohm, USA), anion column (Metrosep A Supp 4ꢀ250/4.0), and cation column (Metrosep C4ꢀ
ꢀ
ꢀ
150/4.0). The concentrations of NO3 and NO2 were also determined by high performance liquid
chromatograph (HPLC) (ProStar, Varian, USA), equipped with RP C18 column (Shiseido,
Japan) and photodiode array (PDA, 335, Varian, USA) detector at 205 nm, as described
ꢀ
elsewhere.1 The NO3 removal (RNO3−) and byproduct selectivity (Sbyproduct), were calculated
using the following equations.
ꢄ
ꢄ
ꢅ
ꢇ
ꢅ
ꢇ
ꢁꢂ ꢆ ꢁ ꢆ ꢁꢂ ꢆ ꢁ
ꢉ
ꢃ
ꢈ
ꢃ
ꢄ
ꢀꢁꢂ (%) =
ꢄ
ꢃ
ꢅꢁꢂ ꢆꢁꢇ
ꢃ
ꢄ
ꢅꢁꢂ ꢆ ꢁꢇ
ꢋ
ꢉ
ꢄ
ꢊꢁꢂ (%) = ꢅ
ꢄ
ꢄ
ꢋ
ꢇ
ꢅ
ꢇ
ꢁꢂ ꢆ ꢁ ꢆ ꢁꢂ ꢆꢁ
ꢃ
ꢃ
ꢉ
ꢈ
ꢎ
ꢅꢁꢌ ꢆꢁꢇ
ꢍ
ꢉ
ꢎ
ꢊꢁꢌ (%) = ꢅ
ꢄ
ꢄ
ꢍ
ꢇ
ꢅ
ꢇ
ꢁꢂ ꢆꢁ ꢆ ꢁꢂ ꢆ ꢁ
ꢉ
ꢃ
ꢃ
ꢈ
ꢄ
ꢄ
ꢄ
ꢎ
ꢅ
ꢇ
ꢅ
ꢇ
ꢅ
ꢇ
ꢉ
ꢁꢂ ꢆꢁ ꢆ ꢁꢂ ꢆ ꢁ ꢆ ꢁꢂ ꢆꢁ ꢆ ꢏꢁꢌ ꢆꢁꢐ
ꢃ
ꢃ
ꢉ
ꢋ
ꢍ
ꢈ
ꢉ
ꢄ
( )
ꢊꢁ % =
ꢄ
ꢄ
ꢋ
ꢅ
ꢇ
ꢅ
ꢇ
ꢁꢂ ꢆ ꢁ ꢆ ꢁꢂ ꢆꢁ
ꢉ
ꢃ
ꢃ
ꢈ
9

Green Chemistry
Page 10 of 45
View Article Online
DOI: 10.1039/C6GC02349E
The subscripts i and f represent initial and final concentrations, respectively. The byꢀproducts
+
ꢀ
other than NH4 and NO2 (e.g., NO and N2O), were negligible in this study.1,4,7,19
3. Results and discussions
3.1 Characterization of catalysts
XRD analysis was performed to check the crystallinity of NBetaꢀsupported bimetallic catalysts
and their successful synthesis and stability by the preparation procedure (Fig. 1). Fig. 1(a) shows
the diffraction peaks at 6–9o and 21–23o assigned to fully crystalline beta zeolite structure.20 The
average size of the newly produced NBeta crystals was also calculated by Scherrer’s equation
which was ~ 80 nm indicating that NBeta was successfully synthesized. Figure 1(bꢀd) shows that
the diffractograms of SnꢀPdꢀNBeta, CuꢀPdꢀNBeta and InꢀPdꢀNBeta are almost same to that of
NBeta, indicating that no significant changes occurred in the crystallinity of NBeta support
during the catalyst synthesis procedure. The diffraction peaks of promoter (i.e. Sn, Cu and In)
and noble (i.e. Pd) metals were not observed due to their lower contents (i.e. 2.2% for promoter
and 1.6% of Pd metal).1,7,19
FTꢀIR analysis was performed to confirm the structural information of the synthesized
NBeta zeolite (Fig. 2). The bands at 463 cmꢀ1 and 520ꢀ570 cmꢀ1 are assigned to a high
crystallinity of beta zeolite.21 Peak appeared at 947 cmꢀ1 is related to terminal SiꢀOꢀ groups on the
external surface of the crystallite. The large peak appearing in the range of 1060ꢀ1090 cmꢀ1 is
due to asymmetric stretching vibration. Peak appearing in 3200ꢀ3500 cmꢀ1 is characteristic of OH
of the silanol groups attached through strong hydrogen bonds with neighboring silanols and
hence corresponds essentially to internal SiꢀOH defect groups. No remarkable changes in the FTꢀ
IR spectra of NBetaꢀsupported bimetallic catalyst was observed compared to that of NBeta. The
10

Page 11 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
results from FTꢀIR analyses of NBeta and NBetaꢀsupported bimetallic catalyst indicate that
NBeta was successfully synthesized and it was stable during bimetallic catalyst preparation.
SEM/EDX analysis was conducted to examine the morphological evidence of NBeta
synthesis and dispersion of Sn, Cu, In and Pd on NBeta surface (Fig. 3). Fig. 3a(1ꢀ3) shows that
sharp edged and uniformly shaped cubical crystals of NBeta are present in three different
samples. The cubic crystals of NBeta are also well separated from each other. The measured size
of individual crystal is ~70ꢀ80 nm which is in well agreement to that calculated from XRD data.
EDX mapping of bimetallic catalysts confirmed that SnꢀPd (Fig. 3b(1ꢀ3)), CuꢀPd (Fig. 3c(1ꢀ3))
and InꢀPd (Fig.3 d(1ꢀ3)) ensembles are present on the surface of NBeta in closely packed and
evenly distributed patterns. In addition, EDX analysis also revealed that the concentration of
each promoter metal (~2.15%) and Pd (~1.56) was in close agreement to initial input
concentrations of Sn, Cu, In (i.e. 2.2% each) and Pd (i.e. 1.6%). The results of SEM/EDX
analysis confirmed the successful synthesis of NBeta support and NBetaꢀsupported bimetallic
catalysts.
TEM analysis was performed to acquire a more detailed morphological information of
fresh NBeta and bimetallic catalyst supported by NBeta (Fig. 4). The TEM image of freshly
synthesized NBeta shows square shaped NBeta crystals with particle size of ~70ꢀ80 nm (Fig.4a).
The TEM images of SnꢀPdꢀNBeta (Fig. 4b(1ꢀ3)), CuꢀPdꢀNBeta (Fig. 4c(1ꢀ2), and InꢀPdꢀNBeta
(Fig. 4d(1ꢀ2)) show a dense and uniform dispersion of nanoꢀsized particles (2–8 nm) on the
surface of NBeta. The EDX spectra confirmed that these were respective promoter (Sn, Cu, In)
and noble (Pd) metal particles (Fig. 4(b3), (c2) and (d2). The size of bimetallic ensembles
(typically < 10 nm) and their dispersion in close proximity have been known as a vital factor for
ꢀ
the reactive and selective NO3 reduction.6,19 Characterization of NBetaꢀsupported bimetallic
11

Green Chemistry
Page 12 of 45
View Article Online
DOI: 10.1039/C6GC02349E
catalysts by SEM and TEM analysis showed that NBeta is a proper support surface for the dense
and uniform dispersion of different bimetallic ensembles of the catalysts developed in this study.
To determine the strength of the acidity of NBeta zeolite, NH3ꢀTPD has been performed on
parent NBeta zeolite. Fig. S1 shows the typical NH3ꢀTPD profile obtained for NBeta zeolite.
Total acidity and acid strength distribution can be obtained from the total peak area and relative
area of NH3ꢀTPD peaks at lower and higher temperature. NBeta zeolite exhibits two stages of
NH3 desorption peaks appeared at ~140 oC and ~270 oC, and are similar to the conventional Beta
zeolite.22 The low temperature peak is due to the desorption of NH3 from weak acid sites and the
high temperature peak is ascribed to NH3 desorption from the strong acid sites. The total amount
of NH3 desorbed from weak and strong acid sites for NBeta zeolite was 0.93 mmol/g.
The surface area and pore size distribution of NBeta and NBeta supported bimetallic
catalysts are presented in Table S1. The specific surface area and external surface area of the
parent NBeta zeolite was reduced after impregnation with bimetals. The pore volume and pore
size of the NBeta zeolite was also reduced after the impregnation, which is due to the occupancy
of the pores by the impregnated metal particles.
3.2 Nitrate reduction by NBeta-supported mono metallic and bimetallic catalysts
ꢀ
Fig. 5 shows the NO3 removal kinetics by bare NBeta, monometallic PdꢀNBeta, and bimetallic
catalysts, i.e. CuꢀPdꢀNBeta, InꢀPdꢀNBeta, and SnꢀPdꢀNBeta. The bare NBeta could not remove
ꢀ
ꢀ
NO3 indicating that it does not possess a reductive capacity for NO3 . Its removal by PdꢀNBeta in
the absence of both H2 and CO2 was also negligible indicating that Pd alone is not capable of
ꢀ
NO3 reduction. PdꢀNBeta in the presence of CO2 (no H2) was also negligible showing that CO2
ꢀ
ꢀ
cannot induce NO3 reduction. However, PdꢀNBeta slowly removed ~8% of NO3 in the presence
12

Page 13 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
ꢀ
of H2 (no CO2) and ~40% NO3 in the presence of both H2 and CO2 showing that NBetaꢀ
ꢀ
supported Pd monometallic catalyst have a limited reductive capacity of NO3 due to PdꢀH2
ꢀ
ꢀ
activation system (Pd(0) + H2 → Pdꢀ2Hads and NO3 (aq) + 2Hads → NO2 (aq) + H2O). The results
from these control experiments indicate that H2 (more precisely Hads) is the only reducing agent
ꢀ
ꢀ
for NO3 in a PdꢀNBeta + H2+ CO2 system implying that a proper contact time between NO3 and
ꢀ
ꢀ
Hads should be provided for an enhanced NO3 removal. However, NO3 reduction on Pd surface
by Hads is not ideal due to poor affinity between Pd surface and NO3ꢀ.1,7 It implies that Hads must
ꢀ
desorb from Pd surface to bulk aqueous phase to come in contact with NO3 for its reduction.
ꢀ
However, the contact between Hads and NO3 could be irregular in the aqueous solution thus
resulting in the slow reaction kinetics. The slow and incomplete removal by PdꢀNBeta indicates
ꢀ
that it is not feasible for effective denitrification. However, the NO3 reduction kinetics were
dramatically enhanced by ~9 to 73 times when promoter metals (Cu, In, and Sn) were added to
ꢀ
PdꢀNBeta (Fig. 5). CuꢀPdꢀNBeta and InꢀPdꢀNBeta catalysts achieved more than ~95% NO3
ꢀ
removal in 120 min, while SnꢀPdꢀNBeta removed ~98% of NO3 in 30 min. This is due to the
higher affinity of NO3 towards promoters (i.e., Cu, In, and Sn) than that towards noble metal.1,7
ꢀ
The experimental results also show that the reactivity of SnꢀPdꢀNBeta (k = 19.09×10ꢀ
2×minꢀ1) was highest followed by InꢀPdꢀNBeta (k = 2.89×10ꢀ2×minꢀ1) and CuꢀPdꢀNBeta (k =
2.31×10ꢀ2×minꢀ1) at the same experimental conditions (Table 1). This variation could be due to
1) leaching of promoter and noble metals, 2) different rejuvenation potentials of promoter metals
ꢀ
by activated H2 (Hads) (MpromoterOx + Hads ꢀ Mpromoter(0))+ H2O) and 3) different NO3 affinity
towards the metals at controlled reaction conditions. We evaluated the leaching of Cu, In, Sn and
Pd by ICPꢀMS, which showed that leaching of each metal was negligible (i.e., <0.1%) indicating
that variation of reactivity was not due to metal leaching. To pursue understanding on the
13

Green Chemistry
Page 14 of 45
View Article Online
DOI: 10.1039/C6GC02349E
different reaction kinetics by examining the change of metal oxidation state, we performed XPS
analysis for samples under three conditions, i.e. synthesize asꢀis catalyst, catalyst after reduction
ꢀ
with NaBH4, and catalyst after one cycle of NO3 reduction reaction (Figs 6ꢀ8). Narrow scan of
Pd(3d) confirmed the presence of Pd in metallic form in all three catalysts (Fig. 6ꢀ8), indicating
that it was readily available for H2 activation. Narrow scan of Cu (3d5/2) on synthesize asꢀ Cuꢀ
PdꢀNBeta shows one main peak at 933.9 eV corresponding to Cu2+ (100%), while two main
peaks after reduction with NaBH4 at 932 and 933.9 eV corresponding to Cu0 (96.15%) and Cu2+
(3.85%), respectively (Fig.6a,b & Table 2). The narrow scan of In before reduction shows one
main peak at 445.10 eV corresponding to (100%) In3+, while two peaks were observed after
reduction at 443.8 and 445.10 eV corresponding to In0 (75.76%) and In3+ (24.25%), respectively
(Fig. 7a,b & Table 2). Similarly, narrow scan of Sn(3d5/2) before reduction shows two main
peaks at 486 and 486.7 eV corresponding to Sn2+ (77%) and Sn4+ (29%), while three peaks were
observed after reduction at 485, 486, and 486.7 eV corresponding to Sn0 (91%), Sn2+ (5%) and
Sn4+ (4%), respectively (Fig.8a,b & Table 2). Using experimental oxide formation free energies
of CuO, In2O3, and SnO/SnO2,23 we calculated the free energetic tendency to reduce metal oxide
into zeroꢀvalent metal (ꢀGred°), which is defined by the reaction free energy of MxO + 1/2 H2 (g)
→ xM + H2O (l). ꢀGred° values were calculated as ꢀ108.3 kJ/mol for CuO, +40.0 kJ/mol for
In2O3, and +19.7 kJ/mol and +23.2 kJ/mol for SnO and SnO2, respectively. This explains the
XPS results showing the trend of reduction propensity toward the formation of zeroꢀvalent metal
using NaBH4, which is in the order of In < Sn < Cu (Table 2).
ꢀ
A comparison of oxidation states among Cu, In, and Sn before and after the NO3 reduction
(Table 2) shows that a considerable amount of promoter metals was successfully rejuvenated and
indicates that all the catalysts could be sustainable during the reaction. The amount of
14

Page 15 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
rejuvenation for each promoter metal was in the order of Cu < Sn ≈ In; 96.15 % → 63.67 % of
Cu0 after the reaction (66% rejuvenation); 75.76 % → 69.93 % of In0 after the reaction (92%
rejuvenation); and 91.01 % → 80.71 % of Sn0 after the reaction (89% rejuvenation). We note that
the observed rejuvenation tendency cannot be rationalized using the simple thermodynamic
argument based on ꢀGred°, predicting the inverse order. On the other hand, using DFT calculated
H binding energy difference of each promoter metal with respect to Pd, we evaluated the
energetic cost for the diffusion of Hads from Pd to each promoter metal. The calculated values
were 36.7 kJ/mol for Cu, 110.0 kJ/mol for In, and 48.2 kJ/mol for Sn, which still predict the
easiest rejuvenation of Cu compared with other metals in contrast to the XPS results. This
implies that the variation in the rejuvenation of the promoter metals may not be due to H
binding/diffusion on the promoter surface.
We therefore conceive that the different rejuvenation tendency of promoter metals by
ꢀ
activated H on Pd surface during NO3 reduction is probably due to 1) loss of active phase (zero
valent metal) via permanent oxidation and/or 2) ) different rejuvenation kinetics of each metal
oxide. The permanent oxidation was less likely to occur under a continuous H2 supplying
ꢀ
environment, which could be also supported by the fact that a complete and sustainable NO3
reduction was achieved by all three catalysts. The variation of active phase could be related to
rejuvenation kinetics, i.e., kinetic consideration is important for understanding the rejuvenation.
CuO reduction by H2 occurs after an induction period of several minutes.23ꢀ26 It implies that the
CuO rejuvenation to Cu0 could be slower due to induction period as compared to that of SnO,
SnO2 and In2O3. The slow rejuvenation of CuO means the presence of a lower active phase of
Cu0 and ultimately a slower reaction kinetics by CuꢀPdꢀNBeta as compared to InꢀPdꢀNBeta and
SnꢀPdꢀNBeta.
15

Green Chemistry
Page 16 of 45
View Article Online
DOI: 10.1039/C6GC02349E
It is also interesting to note that the reactivity of In (2.89×10ꢀ2 minꢀ1) is slightly higher than
that
of Cu (2.31×10ꢀ2 minꢀ1) and far less than that of Sn (19.09×10ꢀ2 minꢀ1), although In have highest
ꢀ
rejuvenation during the NO3 reduction. This indicates that rejuvenation of promoter metals is not
ꢀ
a unique factor governing the NO3 reduction on the surface of bimetallic catalysts. Another
ꢀ
potential factor could be the adsorption of NO3 on the promoter metal of bimetallic catalyst. We
measured the zeta potential of CuꢀPdꢀNBeta, InꢀPdꢀNBeta, and SnꢀPdꢀNBeta to characterize the
ꢀ
adsorption behavior of NO3 on their surfaces (Fig. 9). The zeta potential of NBeta is also shown
for the purpose of comparison.27 The zeta potential values of the bimetallic catalysts are higher
than that of NBeta indicating the influence of metal doping on the change of surface charge of
NBeta. The zeta potential of SnꢀPdꢀNBeta (~ ꢀ17 mV) at pH 4.5 ꢀ 5 is higher than that of CuꢀPdꢀ
NBeta (~ ꢀ27 mV) and InꢀPdꢀNBeta (~ ꢀ26 mV) indicating that the Sn doping on NBeta surface
provided higher surface charge change than In and Cu doping. Furthermore, the activated H on
Pd surface can also increase the positive surface charge due to the polar HꢀMetal bonding.28 Due
to the continuous activation of H2 by Pd, the negatively SnꢀPdꢀNBeta surface could be easily
shifted to positively charged one at the beginning of the experiment. This is because the surface
charge of SnꢀPdꢀNBeta is much closer to zero than that of CuꢀPdꢀNBeta and InꢀPdꢀNBeta. Once
the bimetallic SnꢀPdꢀNBeta surface is positively charged, the adsorption of target contaminant,
ꢀ
ꢀ
NO3 , its reduction could be intensified on the catalytic surface leading to accelerated NO3
reduction kinetics. It is important to note that this is the first study which found that affinity of a
catalyst towards contaminants could also be enhanced or suppressed by the dopant metals and
reaction conditions. Above discussion indicates that SnꢀPdꢀNBeta showed one of the highest
16

Page 17 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
reduction kinetics due to its 1) higher and faster rejuvenation kinetics and 2) higher surface
ꢀ
affinity for NO3 as compared to that of CuꢀPdꢀNBeta and InꢀPdꢀNBeta.
-
3.3 Reaction mechanism of NO3 reduction by NBeta-supported bimetallic catalysts
Table 3 shows the results of control and reduction experiments performed to elucidate the
ꢀ
ꢀ
reaction mechanism of NO3 reduction by NBetaꢀsupported bimetallic catalysts. NO3 removal on
NBeta and monometallic CuꢀNBeta, InꢀNBeta, and SnꢀNBeta catalysts was negligible even in
the presence of H2 and CO2 indicating that the NBeta and NBetaꢀsupported monometallic
ꢀ
catalysts are inefficient to remove NO3 through their surface adsorption and reduction. SnꢀPdꢀ
ꢀ
NBeta bimetallic catalyst could reduce only 6.64% NO3 in the absence of H2, while 39.61%
ꢀ
NO3 in the absence of CO2. A mixture of SnꢀNBeta and PdꢀNBeta monometallic catalysts could
ꢀ
remove only 14.33% NO3 , although H2 and CO2 were supplied together. In contrast to the result,
ꢀ
SnꢀPdꢀNBeta, CuꢀPdꢀ NBeta, and InꢀPdꢀ NBeta achieved 100% NO3 reduction with high N2
+
selectivities (>80%) and relatively lower NH4 selectivities (<20%) when they were supplied
with both H2 and CO2. The results from the control experiments show that 1) the presence of
ꢀ
closely packed promoterꢀnoble ensembles enhances the NO3 reduction and N2 selectivity, and 2)
a simultaneous supply of H2 as reducing agent and CO2 as buffering agent is essential for
ꢀ
enhanced NO3 removal.
ꢀ
The stable decrease in NO3 concentration (Fig. 5), and a subsequent increase in Cu2+, Sn2+,
ꢀ
Sn4+, and In3+ after NO3 reduction reaction (Table 2), during batch experiments indicate that
ꢀ
NO3 reduction preferentially took place on promoter metal sites. Furthermore, the production of
ꢀ
+
ꢀ
NO2 (in the absence of CO2 buffer, Table 3, exp. 6), NH4 and N2 indicates that NO3 removal
ꢀ
on NBetaꢀsupported bimetallic catalysts is a twoꢀstep reduction process. In the first step NO3 is
17

Green Chemistry
Page 18 of 45
View Article Online
DOI: 10.1039/C6GC02349E
ꢀ
ꢀ
adsorbed on zero valent promoter metal site (i.e. Sn0, In0, and Cu0) and is reduced to NO2 (NO3
ꢀ
ꢀ
ꢀ
+ Mpromoter(0) ꢀ NO2 + MpromoterOx). In the second step, NO2 is reduced to N2 (2NO2 + 6Hadsꢀ
+
ꢀ
+
Pd ꢀ N2 + 2OHꢀ + 2H2O+ Pd0 and/or NH4 (2NO2 + 12HadsꢀPd0 ꢀ 2NH4 + 4OHꢀ + Pd0) by
ꢀ
Hads spillover Pd surface (Pd0 + H2 → Pdꢀ2Hads). The detailed reaction mechanism of NO3
reduction on NBetaꢀsupported bimetallic catalysts is depicted in Scheme 1, which shows that
ꢀ
promoter metals favor NO3 adsorption and reduction, while Pd sites favor H2 adsorption and
ꢀ
activation. The scheme 1 also shows that NO2 migration to PdꢀHads and Hads spillover to
promoter oxides (i.e CuO, In2O3, and SnO/SnO2) could occur simultaneously due to close
ꢀ
proximity of promoter and Pd particles. The Hads then reduced both NO2 and promoter oxide by
abstraction of oxygen and producing OHꢀ. The OHꢀ buildup is effectively neutralized by CO2
ꢀ
buffer to maintain a constant pH. The simultaneous consumption of Hads by NO2 and promoter
oxides could create a limiting reaction environment due to the limited supply of H2, i.e. 30 mg/L.
The limited availability of Hads could enhance the probability of NꢀN recombination to produce
+ 18
ꢀ
N2, while it could diminish that of NꢀH to produce NH4 . This leads to a highly selective NO3
reduction to N2 by CuꢀPdꢀNBeta, InꢀPdꢀNBeta and SnꢀPdꢀNBeta bimetallic catalysts (Table 3).
3.4 Effect of promoter metals on products’ selectivity during nitrate and nitrite reduction
ꢀ
The reaction mechanism of catalytic NO3 reduction indicates that the selectivity of products
+
ꢀ
(i.e., NH4 and N2) is governed by the noble metal assisted reduction of NO2 and N:Hads ratio in
ꢀ
+
the system (NO2 + PdꢀHadsꢀ Pd(0)+ N2/NH4 ). Higher N:Hads (e.g. 1:3) ratio lead to a high N2,
+
while lower N:Hads (e.g. 1:6) result in a high NH4 selectivity. The N:Hads ratio could be easily
influenced by several factors, i.e. 1) initial N concentration, 2) H2 supply, 3) Pd content, 4)
buffering, and 5) promoterꢀPd interaction. However, initial N (30mg/L NO3ꢀN), H2 supply (30
18

Page 19 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
mL/min), CO2 supply (60 mL/min), and Pd contents (1.6%) were fixed in this study to evaluate
ꢀ
ꢀ
the effect of promoters on the product selectivity for NO2 and NO3 reduction. Firstly, we
ꢀ
conducted an experiment to observe the product’s selectivity for the reduction of NO2 (30 mg/L
NO2ꢀN) by monometallic PdꢀNBeta and bimetallic CuꢀPdꢀNBeta, InꢀPdꢀNBeta, and SnꢀPdꢀ
ꢀ
NBeta catalysts (Table1). The results showed that a complete NO2 reduction was achieved
within 15 mins by all four catalysts and the reaction kinetics were in the order of PdꢀNBeta (k=
45.44×10ꢀ2 minꢀ1) > SnꢀPdꢀNBeta (43.45×10ꢀ2 minꢀ1) > InꢀPdꢀNBeta (38.97×10ꢀ2 minꢀ1) > CuꢀPdꢀ
NBeta (33.14×10ꢀ2 minꢀ1). The lower reactivity of bimetallic catalysts than that of monometallic
ꢀ
catalyst shows that promoter metal addition to monometallic PdꢀNBeta affected the NO2
reduction kinetics (Table 1). However, all bimetallic catalysts showed a very high N2 selectivity
ꢀ
(>98%) similar to that of the monometallic catalyst (Table 1) indicating that NO2 reduction to
N2 is a main reaction pathway for NBetaꢀsupported monometallic and bimetallic catalysts. This
ꢀ 7
is because PdꢀHads has a very high affinity towards NO2 , implying that PdꢀHads would have been
ꢀ
ꢀ
instantaneously occupied by NO2 when a high dose of NO2 (e.g. 30 mg/L NO2ꢀN) was supplied.
ꢀ
As a result of this, a higher rate of both NO2 reduction and Hads consumption could be expected.
The faster consumption of Hads could create an instantaneous condition where H is deficient (due
ꢀ
to fixed and limited H2 supply), while N is abundant (due to faster NO2 reduction). This implies
+
that NꢀN recombination to form N2 could be promoted, while NꢀH recombination to form NH4
ꢀ
is suppressed.1,18,19 This is the reason why all the catalysts have highest N2 selectivity for NO2
reduction in this study.
ꢀ
ꢀ
In contrast to the results of NO2 reduction, N2 selectivity from NO3 (30 mg/L NO3ꢀN)
reduction by monometallic and bimetallic catalysts varied over a wide range (Fig. 10 and Table
1). CuꢀPdꢀNBeta showed the highest N2 selectivity (92.68%) followed by InꢀPdꢀNBeta (82.93%)
19

Green Chemistry
Page 20 of 45
View Article Online
DOI: 10.1039/C6GC02349E
and SnꢀPdꢀNBeta (80.80%) indicating that promoter metals could influence the product
ꢀ
ꢀ
selectivity during the NO3 reduction. This variation of product selectivity in bimetallic NO3
reduction could be due to change in hydrogenation capacity of Pd as a result of second metal
doping.
To verify this, we analyzed PdꢀNBeta, CuꢀPdꢀNBeta, InꢀPdꢀNBeta and SnꢀPdꢀNBeta by H2ꢀ
TPR (Fig. 11). A wellꢀformed strong peak by PdꢀNBeta, SnꢀPdꢀNBeta and InꢀPdꢀNBeta while a
o
relatively weaker signal by CuꢀPdꢀNBeta was observed at 76.9 C (Fig. 11). This peak is
assigned to H2 adsorption by PdOx.29,30 The low intensity peak by CuꢀPdꢀNBeta as compared to
other samples indicates that H2 sorption capacity of Pd was suppressed due to addition of Cu. To
further verify this we also performed H2 pulse chemisorption which confirmed that PdꢀNBeta
have highest H2 cumulative uptake quantity (0.661 cm3/g) followed by SnꢀPdꢀNBeta (0.408
cm3/g), InꢀPdꢀNBeta (0.324 cm3/g) and lowest by CuꢀPdꢀNBeta (0.036 cm3/g). This trend of H2
activation by monometallic and bimetallic catalysts can be explained by the intermetallic
interaction (also referred as an ensemble effect) between promoters (Cu, Sn, In) and Pd.31ꢀ33 Pd
crystals possess multiple sites for hydrogenation, i.e. edge or corner having higher hydrogenation
capacity, and terrace site having medium hydrogenation capacity.34 It means that hydrogenation
capacity (and Hads production) of bimetallic catalyst with promoter metal doped towards terrace
site would be higher (e.g. SnꢀPdꢀNBeta and InꢀPdꢀNBeta) as compared to that of metal doped at
edge or corner site (e.g. CuꢀPdꢀNBeta). The lowest H2 uptake capacity of CuꢀPdꢀNBeta could be
due to the selective deposition of Cu towards corner sites.34,35 Thus a lower Hads could have been
produced in CuꢀPdꢀNBeta which resulted in a high N:Hads, suppression of excessive
ꢀ
+
ꢀ
hydrogenation of NO2 to NH4 (2NO2 + 6HadsꢀPdterrace ꢀ N2 + 2OHꢀ + 2H2O + Pd(0)) and
ultimately a highest N2 selectivity (92.68%). On the contrary, a higher H2 uptake capacity of Inꢀ
20

Page 21 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
PdꢀNBeta and SnꢀPdꢀNBeta could be due to the selective deposition of In and Sn towards terrace
sites which lead to relatively higher Hads production (as confirmed by H2 chemisorption results),
+
ꢀ
+
a lower N:Hads and consequently a higher NH4 selectivity (2NO2 + 12HadsꢀPdedge ꢀ 2NH4
+
4OHꢀ + Pd(0)). A detailed comparison of selectivity by different bimetallic catalysts is provided
in SI Table S2. The N2 selectivities by the bimetallic catalysts developed in this study are among
the highest in the comparison. This comparison indicates that NBeta support provided an optimal
ꢀ
ꢀ
surface environment for the selective NO2 and NO3 reduction.
ꢀ
ꢀ
Surprisingly, NO2 was not detected during the NO3 reduction by CuꢀPdꢀNBeta, InꢀPdꢀ
NBeta
and SnꢀPdꢀNBeta unlike the CuꢀPd, InꢀPd and SnꢀPd bimetallic catalyst supported by Hematite,
ꢀ
NZVI, maghemite, ZrO2, TiO2, SiO2, AC, Al2O3, and resins.1,6,18,19,36ꢀ40 The detection of NO2
ꢀ
during bimetallic NO3 reduction is attributed to its slower or incomplete reduction due to 1)
change in catalyst and support structure, 2) lack of close proximity between promoter and noble
metal, 3) overlapping of Pd by promoters due to low surface area of support, 3) insufficient H2
activation due to low H2 or Pd contents and 4) lack of proper buffering.1,18,19 In contrast, the
ꢀ
results from this study show that NO2 reduction by NBetaꢀsupported monometallic and
ꢀ
ꢀ
bimetallic catalysts is 2 to 14 times faster than that of NO3 (Table 1) implying that NO2 was
+
instantaneously and completely reduced to N2 or NH4 before it could be detected. The results
also indicate that NBeta support enhanced the efficiency of monometallic and bimetallic
ꢀ
catalysts for the reduction of NO2 by facilitating the uniform dispersion of proximate promoterꢀ
noble ensembles due to its unique surface feature, high crystallinity and stability. Thus it can
ꢀ
make NO2 migration from promoter to PdꢀHads easier and enhancing its reduction by more
21

Green Chemistry
Page 22 of 45
View Article Online
DOI: 10.1039/C6GC02349E
optimal utilization of PdꢀHads and CO2 buffer as compared to other supports previously reported
in the literature.
3.5 Stability of NBeta-supported bimetallic catalysts
ꢀ
In order to investigate the stability of CuꢀPdꢀNBeta, InꢀPdꢀNBeta and SnꢀPdꢀNBeta for NO3
reduction, the catalysts were first tested for four successive reactions (without regeneration) and
then for four repeated reactions (by regenerating the same catalyst after every use) (Fig. 12). All
ꢀ
bimetallic catalysts successfully achieved 100% NO3 reduction with consistent N2 selectivities
both in successive as well as in repeated cycles (Fig. 12), indicating that NBetaꢀsupported
ꢀ
bimetallic catalysts possess very stable NO3 reduction capacities. To further confirm the stability
of catalytic reaction steps, CuꢀPdꢀNBeta, InꢀPdꢀNBeta and SnꢀPdꢀNBeta were also evaluated for
ꢀ
four successive and repeated reductions of NO2 (Fig. 13). The results show that all three
ꢀ
catalysts successfully reduced 100% NO2 to more than 98% N2 which is consistent to that of
first reaction. The results in Fig. 12 and 13 indicate that NBetaꢀsupported bimetallic catalysts are
ꢀ
ꢀ
ꢀ
stable for 1) NO3 reduction to NO2 , 2) NO2 reduction to N2 and 3) rejuvenation of promoter
metal oxide to zeroꢀvalent by PꢀHads system as described in reaction mechanism (Scheme 1).
The leaching of bimetals is another crucial parameter for reactivity and durability of
bimetallic catalysts and metal associated toxicity in treated effluents. We analyzed the leaching
ꢀ
of Pd, Cu, In and Sn by ICPꢀMS to investigate bimetallic stability during recycling test of NO3
reduction. The results revealed that dissolution of Pd (<0.01%), Cu(<0.8%), In (<0.3%) and Sn
(0.1%) was negligible during stability test indicating that bimetallic ensembles were stable
during catalytic reduction. Remarkable durability of bimetallic catalysts, their fast and steady
ꢀ
ꢀ
NO3 and NO2 reduction capacities with consistent and high N2 selectivities achieved in this
22

Page 23 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
study imply that NBeta is an excellent support for bimetallic catalysts. The results also show that
bimetallic catalysts reported in this study are much suitable to develop an ecofriendly, stable,
efficient and economical denitrification process for water treatment.
4. Conclusions
In this study we report the successful synthesis and use of NBeta zeolite as support surface to
develop novel SnꢀPdꢀNBeta, CuꢀPdꢀNBeta and InꢀPdꢀNBeta bimetallic catalysts for highly
ꢀ
ꢀ
reactive and selective NO3 reduction. All three catalysts achieved 100% NO3 removal with Snꢀ
PdꢀNBeta (k = 19.09×10ꢀ2×minꢀ1, K' = 76.36×10ꢀ2 minꢀ1gcatꢀ1, K”=477.25×10ꢀ2 L.minꢀ1gPdꢀ1, TOF
= 11.13×10ꢀ3 Sꢀ1) as the most reactive catalyst followed by InꢀPdꢀNBeta (k = 2.89×10ꢀ2×minꢀ1, K'
= 11.56×10ꢀ2 minꢀ1gcatꢀ1, K”=72.25×10ꢀ2L.minꢀ1gPdꢀ1; TOF = 0.74×10ꢀ3 Sꢀ1) and CuꢀPdꢀNBeta (k =
ꢀ
2.31×10ꢀ2×minꢀ1, K' = 9.24×10ꢀ2 minꢀ1gcat 1, K”=57.75×10ꢀ2L.minꢀ1gPdꢀ1; TOF = 0.41×10ꢀ3 Sꢀ1).
The characterization results confirmed that the higher reactivity of SnꢀPdꢀNBeta was due to its
ꢀ
higher NO3 affinity and rejuvenation capacity as compared to that of InꢀPdꢀNBeta and CuꢀPdꢀ
NBeta. In this study, all the catalysts also achieved high N2 selectivity which was in the order of
CuꢀPdꢀNBeta (92.68%)>InꢀPdꢀNBeta (82.93%)>SnꢀPdꢀNBeta (80.80%). By characterization of
ꢀ
catalysts, it was firstly confirmed that N2 selectivity during NO3 reduction is a function of
hydrogenation capacity of Pd which depends upon the promoterꢀPd ensemble type. NBetaꢀ
supported CuꢀPd ensembles could achieve highest N2 selectivity because Cu could suppress over
ꢀ
+
hydrogenation of NO2 to NH4 by Pd more than that by In and Sn. Nevertheless, the reactivity
and selectivity by NBetaꢀsupported SnꢀPdꢀNBeta, InꢀPdꢀNBeta and CuꢀPdꢀNBeta is one of the
highest among those reported in literature (SI Fig. S4 and SI Table S2). To our best knowledge
this also the first report which explain that the addition of promoter metal can also enhance the
23

Green Chemistry
Page 24 of 45
View Article Online
DOI: 10.1039/C6GC02349E
ꢀ
surface charge of the support in such a way that NO3 reduction is enhanced. The NBetaꢀ
ꢀ
supported CuꢀPd, InꢀPd and SnꢀPd catalysts also showed remarkable NO3 reduction with
consistent N2 selectivity in four successive and then four repeated reactions. It is therefore
concluded that NBeta is an excellent support surface to synthesize efficient, stable, durable and
robust catalyst for highly reactive, selective and sustainable denitrification processes. The
procedures and results explained in this study could help to improve fundamental knowledge and
ꢀ
to develop novel and viable solutions useful in solving global challenges of NO3 contamination.
Acknowledgements
The author is sincerely thankful to all EGRL heroes who have made its wonderful scientific
legacy possible at KAIST since 2005. Special thanks should be given to Dr. Shanawar Hamid
and Prof. Hyungjun Kim, the main contributors of this masterpiece. This research work was
supported by the Korean Ministry of Environment (project no. RE201402059) and the Korean
National Research Foundation (project no. 2012M1A2A2026587).
References
1 S. Jung, S. Bae and W. Lee, Environ. Sci. Technol., 2014, 48, 9651ꢀ9658.
2 A. Pintar, Catal. Today, 2003, 77, 451ꢀ465.
3 H. Shin, S. Jung, S. Bae, W. Lee and H. Kim, Environ. Sci. Technol., 2014, 48, 12768ꢀ12774.
4 Y. H. Liou, C. J. Lin, S. C. Weng, H. H. Ou and S. L. Lo, Environ. Sci. Technol., 2009, 43,
2482ꢀ2488.
24

Page 25 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
5 O. S. G. P. Soares, J. J. M. Orfa, E. GallegosꢀSuarez, E. Castillejos, I. RodríguezꢀRamos and
M. F. R. Pereira, Environ. Technol., 2012, 33, 2353ꢀ2358.
6 Z. Gao, Y. Zhang, D. Lia, C. J. Werth, Y. Zhang and X. Zhou, J. Hazard. Mater., 2015, 286,
425ꢀ431.
7 S. Hamid, A. K. Macharla, and W. Lee, Appl. Catal. B: Environ., 2016, 187, 37ꢀ46.
8 J. K. Chinthaginjala and L. Lefferts, Appl. Catal. B: Environ., 2010, 101, 144ꢀ149.
9 A. Dyer, An Introduction to Zeolite Molecular Sieves, Wiley, Chichester, 1988.
10 N. Narender, K. S. K. Reddy, A. K. Macharla, C. N. Rohitha and S. J. Kulkarni, Catal. Lett.,
2010, 134, 175ꢀ178.
11 A. K. Macharla. M. Reddy, P. Swamy, M. Naresh, K. Srujana, K. V. V. K. Mohan and N.
Narender, Catal. Lett., 2015, 145, 1922ꢀ1930.
12 O. S. G. P. Soares, L. Marques, C. M. A. S. Freitas, A. M. Fonseca, P. Parpot, J. J. M. Orfao,
M. F. R. Pereira and I. C. Neves, Chem. Eng. J., 2015, 281, 411ꢀ417.
13 C. M. A. S. Freitas, O. S. G. P. Soares, J. J. M. Orfao, A. M. Fonseca, M. F. R. Pereira and I.
C. Neves, Green Chem., 2015, 17, 4247ꢀ4254.
14 K. Nakamura, Y. Yoshida, I. Mikami and T. Okuhara, Appl. Catal. B: Environ., 2006, 65,
31ꢀ36.
15 K. Nakamura, Y. Yoshida, I. Mikami and T. Okuhara, Chem. Lett., 2005, 34, 678ꢀ679.
16 L. Tosheva and V.P. Valtchev, Chem. Mater., 2005, 17, 2494ꢀ2513.
17 A. Sakthivel, K. Komura and Y. Sugi, Chem. Lett., 2007, 36, 894ꢀ895.
18 S. Hamid, S. Bae, W. Lee, M. T. Amin and A. A. Alazba, Ind. Eng. Chem. Res., 2015, 54,
6247ꢀ 6257.
19 J. Jung, S. Bae and W. Lee, Appl. Catal. B: Environ., 2012, 127,148ꢀ158.
25

Green Chemistry
Page 26 of 45
View Article Online
DOI: 10.1039/C6GC02349E
20 M. J. Eapen, K. S. N. Reddy and V. P. Shiralkar, Zeolites, 1994, 14, 295ꢀ302.
21 B. Modhera, M. Chakraborty, P. A. Parikh and R. V. Jasra, Cryst. Res. Technol., 2009, 44,
379ꢀ385.
22 C. Li and K. Fujimoto, Catal. Sci. Technol., 2015, 5, 4501ꢀ4510.
23 T. J. Singler, S. J. Meschter, J. Spalik in Handbook of Lead-Free Solder Technology for
Microelectronic Assemblies, eds. K. J. Puttlitz and K.A. Stalter, Marcel Dekker, New York,
2004, 11, 409ꢀ410.
24 J. A. Rodriguez, J. Y. Kim, J. C. Hanson, M. Perez and A. I. Frenkel, Catal. Lett., 2003, 85,
247ꢀ254.
25 B. Delmon in: Handbook of Heterogeneous Catalysis, eds. G. Ertl, H. Knozinger and J.
Weitkamp, VCHꢀWiley, New York, 1997, 264ꢀ265.
26 H. H. Kung, Transition Metal Oxides: Surface Chemistry and Catalysis, Elsevier, New York,
1989.
27 W. Yang, X. Wang, Y. Tang, Y. Wang, C. Ke, and S. Fu, J. Macromol. Sci., A, 2002, 39,
509ꢀ526.
28 H. A. Aleksandrov, S. M. Kozlov, S. Schauermann, G. N. Vayssilov and K. M. Neyman,
Angew. Chem. Int. Ed., 2014, 53, 13371ꢀ13375.
29 D. Fang, W. Li, J. Zhao, S. Liu, X. Ma, J. Xu and C. Xia, RSC Adv., 2014, 4, 59204ꢀ59210.
30 L. Yue, C. He, X. Zhang, P. Li, Z. Wang, H. Wang, Z. Hao, J. Hazard. Mater., 2013, 244-
245, 613ꢀ620.
31 S. Sakong, C. Moscha and A. Grob, Phys. Chem. Chem. Phys., 2007, 9, 2216ꢀ2225.
32 I. Witonska, S. Karski, J. Rogowski and N. Krawczyk, J. Mol. Catal. A: Chem., 2008, 287,
87ꢀ94.
26

Page 27 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
33 Y. Yoshinaga, T. Akita, I. Mikami and T. Okuhara, J. Catal., 2002, 207, 37ꢀ45.
34 R. Melendrez, G. Del Angel, V. Bertin, M. A. Valenzuela and J. Barbier, J. Mol. Catal. A,
2000, 157, 143ꢀ149.
35 C. L. Constantinou, C. N. Costa and A. M. Efstathiou, Environ. Sci. Technol., 2007, 41, 950ꢀ
956.
36 D. Gasparovicova, M. Kralik, M. Hronec, Z. Vallusova, H. Vinek and B. Corain, J. Mol.
Catal. A: Chem., 2007, 264, 93ꢀ102.
37 I. Dodouche, D. P. Barbosa, M. D. Rangel, and F. Epron, Appl. Catal. B: Environ., 2009, 93,
50ꢀ55.
38 J. Sa, D. Gasparovicova, K. Hayek, E. Halwax, J. A. Anderson and H. Venek, Catal. Lett.,
2005, 105, 209ꢀ217.
39 H. Berndt, I. Monnich, B. Lucke, M. Menzel, Appl. Catal. B: Environ., 2001, 30, 111ꢀ122.
40 N. Krawczyk, S. Karski and I. Witonska, Reac. Kinet. Mech. Cat., 2011, 103, 311ꢀ323.
27

Green Chemistry
Page 28 of 45
View Article Online
DOI: 10.1039/C6GC02349E
(d) InꢀPdꢀNBeta
(c) CuꢀPdꢀNBeta
(b) SnꢀPdꢀNBeta
(a) NBeta
Reference pattern of NBeta
0
20
40
60
80
2 Theta (degrees)
Fig. 1 XRD patterns of (a) NBeta, (b) SnꢀPdꢀNBeta, (c) CuꢀPdꢀNBeta, and (d) InꢀPdꢀNBeta.
28

Page 29 of 45
Green Chemistry
View Article Online
DOI: 10.1039/C6GC02349E
(d) InꢀPdꢀNBeta
(c) CuꢀPdꢀNBeta
(b) SnꢀPdꢀNBeta
(a) NBeta
4000
3000
2000
1000
ꢀ1
Wave number (cm )
Fig. 2 FTꢀIR spectra of (a) NBeta, (b) SnꢀPdꢀNBeta, (c)CuꢀPdꢀNBeta, and (d) InꢀPdꢀNBeta.
29