Enhanced catalytic reduction of N-nitrosodimethylamine over bimetallic Pd-Ni catalysts

Article abstract of DOI:10.1016/j.molcata.2016.05.026

Catalytic reduction of N-nitrosodimethylamine (NDMA) was investigated over γ-Al₂O₃ supported bimetallic Pd-Ni catalysts (3%(Pd_xNi_1-x)). NDMA could be reduced to dimethylamine over 3%(Pd_0.8Ni_0.2) with a metal-loading-normalized pseudo-first-order rate constant of 836 ± 21 L g_me^-1 h^-1. Characterization results showed that the reducibility of PdO (or NiO) was improved by addition of Ni (or Pd) in 3%(Pd_xNi_1-x); Pd-Ni ensembles were formed in 3%(Pd_xNi_1-x) and there was an electronic transfer from Pd to Ni; metal dispersion was affected by the formation of Pd-Ni ensembles and a volcano curve of metal dispersion via Pd/Ni ratio was observed. The activity profiles demonstrated that TOF had a significant positive relationship with metal dispersion of 3%(Pd_xNi_1-x), indicating high metal dispersion favor NDMA reduction. NDMA reduction over 3%(Pd_0.8Ni_0.2) catalyst could be described by Langmuir-Hinshelwood model, reflecting an adsorption controlled reduction mechanism. The reduction mechanism of NDMA over 3%(Pd_0.8Ni_0.2) catalyst was proposed to be that H₂ was activated by Pd, and H spillover from Pd to Ni/NiO reduced NDMA though N-N cleavage.

Full text of DOI:10.1016/j.molcata.2016.05.026

Journal of Molecular Catalysis A: Chemical 421 (2016) 167–177
Contents lists available at ScienceDirect
Journal of Molecular Catalysis A: Chemical
journal homepage: www.elsevier.com/locate/molcata
Enhanced catalytic reduction of N-nitrosodimethylamine over
bimetallic Pd-Ni catalysts
Huan Chen , Ting Li , Fang Jiang , Zhe Wangb
a
a
a,∗
a
Key Laboratory of Jiangsu Province for Chemical Pollution Control and Resources Reuse, School of Environmental and Biological Engineering, Nanjing
University of Science and Technology, Nanjing 210094, China
b
Department of Civil and Environmental Engineering, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Catalytic reduction of N-nitrosodimethylamine (NDMA) was investigated over ␥-Al O3 supported
2
Received 29 March 2016
Received in revised form 25 May 2016
Accepted 26 May 2016
bimetallic Pd-Ni catalysts (3%(PdxNi1−x)). NDMA could be reduced to dimethylamine over 3%(Pd0.8Ni0.2)
−1 −1
.
with a metal-loading-normalized pseudo-first-order rate constant of 836 ± 21 L gme
h
Characterization results showed that the reducibility of PdO (or NiO) was improved by addition of
Ni (or Pd) in 3%(PdxNi1−x); Pd-Ni ensembles were formed in 3%(PdxNi1−x) and there was an electronic
transfer from Pd to Ni; metal dispersion was affected by the formation of Pd-Ni ensembles and a volcano
curve of metal dispersion via Pd/Ni ratio was observed. The activity profiles demonstrated that TOF had a
significant positive relationship with metal dispersion of 3%(PdxNi1−x), indicating high metal dispersion
favor NDMA reduction.
Available online 27 May 2016
Keywords:
N-Nitrosodimethylamine (NDMA)
␥
-Al2O3 supported bimetallic Pd-Ni
catalysts
Catalytic reduction
Metal dispersion
NDMA reduction over 3%(Pd0.8Ni0.2) catalyst could be described by Langmuir-Hinshelwood model,
reflecting an adsorption controlled reduction mechanism. The reduction mechanism of NDMA over
3
%(Pd0.8Ni0.2) catalyst was proposed to be that H2 was activated by Pd, and H spillover from Pd to Ni/NiO
reduced NDMA though N-N cleavage.
©
2016 Published by Elsevier B.V.
1
. Introduction
N-nitrosamine moiety in NDMA is susceptible to be reduced. For
example, zero-valent iron (Fe ) and improved iron catalyst (Fe-Ni)
0
N-Nitrosodimethylamine (NDMA) has been widely detected as
disinfection byproducts in drinking water and wastewater effluent
1,2]. Results of toxicity tests have shown that NDMA can cause liver
can transform NDMA to dimethylamine (DMA) and ammonium
(NH4+)/N O [13,14]. And Zn(0) is also found efficient to reduce
2
[
NDMA and lower pH promotes the process [15]. However, using
zero-valent metal as reductant has a limitation in NDMA treat-
ment for the exhaustion of zero-valent metal and requirement of
additional treatment of oxidized metal.
cancer, lung cancer and nervous system damage to living organ-
isms, and hence NDMA is defined as “Probable Human Carcinogen”
by the U.S. Environmental Protection Agency (EPA) [3]. The frequent
detection of NDMA in source water raises serious concerns for its
potential threat to public health. The cleanup level in groundwater
established by U.S. EPA is 0.7 ng/L [4]. Therefore, there is a consid-
erable need for developing rapid and effective methods to remove
NDMA from drinking water.
Catalysts, especially Ni-based (such as Raney Ni [12], porous Ni
[16] and NiB [11]) and Pd-based catalysts (such as Pd [17], Pd-Cu
[18] and Pd-In [19]), have the ability to promote NDMA reduction
using hydrogen as a reducing agent. As reported, Ni-based cata-
lysts presented a high ability to reduce NDMA to dimethylamine
Traditional water treatment technologies, such as bio-reduction
(DMA) and inorganic nitrogen species (NH4+ or N ) via the N
N
2
[
5], ion exchange and adsorption [6], are inefficient and/or expen-
bond and/or N O bond cleavage [20,21]. Rapid reduction of NDMA
could be achieved over Raney Ni, and further investigations showed
that porous nickel/nickel-boron had much higher catalytic activ-
ity/stability [16,11]. As for Pd-based catalysts, bimetallic catalysts
were found more effective to reduce NDMA than monometallic cat-
alyst; one reason was that each active metal in bimetallic catalysts
had a respective role. For example, Pd was used to active hydrogen
and Cu was proposed for activation of NDMA in Pd-Cu bimetallic
catalyst [18]. Similar results were observed in Pd-In catalyst that
sive to remove NDMA. Hence, other chemical approaches, including
advanced oxidation processes (AOPs) [7], photocatalysis [8,9],
electrochemical oxidation [10] and catalytic/non-catalytic reduc-
tion [11,12] have been explored recently. Non-catalytic/catalytic
reduction method shows potential to eliminate NDMA for the
∗ Corresponding author.
E-mail address: fjiang@njust.edu.cn (F. Jiang).
http://dx.doi.org/10.1016/j.molcata.2016.05.026
1
381-1169/© 2016 Published by Elsevier B.V.

1
68
H. Chen et al. / Journal of Molecular Catalysis A: Chemical 421 (2016) 167–177
Pd could activate H2 to H* and In could activate NDMA to NDMA*,
and subsequently NDMA* was reduced by H* [19].
3%Ni/Al O3 were also prepared by the same method as bimetallic
catalysts.
2
Generally, electronic effects, geometric effects, the occurrence
of mixed sites and the disappearance of ␤-PdH have been devel-
oped to elucidate the improved activity and stability of bimetallic
catalysts [22]. Nutt et al. [23] studied Pd-on-Au bimetallic cata-
lysts for liquid phase hydrodechlorination of trichloroethene and
an improved activity and selectivity was observed due to electronic
2
.3. Characterization of catalysts
The Brunauer-Emmett-Teller (BET) surface area of the cat-
alysts was measured by N2 adsorption-desorption method on
Micromeritics ASAP 2200. The catalysts were dried at 200 C under
◦
effects. Pd-Ni supported on Al O3 catalysts were studied in gas
2
vacuum (1.33 Pa) for 4 h before measurement. The Pd and Ni con-
tents in the catalysts were determined by ICP (OPTIMA7000DV,
Perkin Elmer, US). X-ray diffraction (XRD) patterns were obtained
by a Rigaku D/max-RA powder diffraction-meter.
The transmission electron microscopy (TEM) images were
recorded by JEOL 2100. Before measurement, catalysts were dis-
persed in water by ultrasonic method, and then the catalyst
particles were deposited on a holey-carbon/Cu grid (300 Mesh).
Surface analyses were also conducted by TEM with a field emis-
sion gun using a FEI Tecnai G2 F20 TEM/STEM, with transient EDX
phase hydrodechlorination of chlorobenzene and the formation of
x+
active Pd-Ni interfaces embedded with Pd favored the reaction
[
24]. Besides, the promoted activity of Pd by Ni in PdNi/NaY cata-
lysts in the gas phase hydrogenation of butyronitrile was attributed
to a higher reduction degree of Ni and the presence of mixed
ensembles [25]. The mixed sites could be described as Pd-Mx+ and
synergistic effects in such sites were found. In the hydrogenation of
2
,4-dinitrotoluene on PdFe/SiO , the positive effect of Fe addition
2
n+
could be invoke to the activation of N O bonds by Fe in mixed
site Pd-Fe [26]. However, no paper concerning these effects on
n+
−
1
mapping (1 nm s ).
X-ray photoelectron spectroscopy (XPS) was evaluated on
a PHI 5000 VersaProbe XPS instrument using Al K␣ radiation
NDMA reduction has been reported.
In this study, we prepared ␥-Al O3 supported bimetallic Pd-Ni
2
catalysts (3%(PdxNi1 )) and investigated their catalytic perfor-
−x
(hv = 1486.6 eV). The calcined catalysts were reduced in a H2 flow
mance on NDMA reduction. The objectives of this paper were to
investigate the mechanism of NDMA reduction by catalytic hydro-
◦
at 300 C for 2 h and then embedded in i-octane in order to avoid
exposure to air. All catalysts were outgassed and then trans-
ferred to the ion-pumped analysis chamber with ultra-high vacuum
genation over 3%(PdxNi1 ); research the metal–metal interaction
−x
in 3%(PdxNi1 ) and assess the relationship between metal disper-
−x
−8
(<10 Torr).
sion and catalytic activity; quantify the effect of NDMA adsorption
and solution pH on NDMA reduction. The results showed that
Temperature programmed reduction (TPR) of the catalysts was
performed on a home-made apparatus consisting of gas chro-
matogram (GC) analyzer equipped with a TCD detector. The
catalysts were packed into a U-shape quartz tube and pretreated
under Ar flow (40 mL min ) at 300 C for 1 h. After cooling down
to room temperature, the samples were treated with H /Ar mix-
3
%(PdxNi1 ) exhibited a promising catalytic activity towards
−x
NDMA reduction, highlighting its potential as an effective catalyst
to eliminate NDMA contamination.
−
1
◦
2
−
1
◦
2
. Experimental
ture gas (40 mL min ) and the temperature was raised to 700 C
◦
−1
with a heating rate of 10 C min . The thermal conductivities of the
samples were subsequently monitored by a thermal conductivity
detector (TCD) and the hydrogen consumption was calculated from
the results of thermal conductivity.
2.1. Materials
2
−1
The ␥-Al O with BET surface area of 137.8 m g was obtained
2
3
from Aladdin (Shanghai, China). NDMA was purchased as analytical
grade solutions in isopropyl from o2si smart solutions (American,
CO chemisorption was carried out on the same apparatus as TPR.
Typically, the catalysts were pretreated in a H2 at 300 C for 1 h and
◦
1
mg/L), and DMA (1 g, >98%) was purchased as a solution from CNW
subsequently purged with He at the same temperature. Then, at
−
1
Technologies GmbH (Düsseldorf, Germany). PdCl , Ni(NO ) ·6H O,
room temperature, CO flow (30 mL min ) was injected until the
outlet peaks were constant. The CO adsorption capacity was then
recorded using the pulse titration model by a thermal conductivity
detector (TCD). The metal dispersion was calculated on the assump-
tion of adsorption stoichiometry of CO to either Pd or Ni equaled to
2
3
2
2
Na CO , HCl, AgNO and methanol were acquired from Aladdin
2
3
3
(
Shanghai, China). Hydrogen (H , 99.99%) and nitrogen (N , 99.99%)
2
2
gases were supplied by Nanjing Specialty Gases Co. All chemicals
were used as received without further purification.
1
.
The zeta potentials were performed on a Zeta Potential Ana-
lyzer (Zeta PALS, Brookhaven Instruments Co.). Briefly, 30 mg of
the catalysts were dispersed in 1 L KCl solution (10 M) and the
pH was adjusted by 0.1 M HCl or 0.1 M NaOH. The samples were
equilibrated for 24 h in an incubator before measurement.
2
.2. Catalyst preparation
−
3
The ␥-Al O supported Pd-Ni bimetallic catalysts were pre-
2
3
pared by a co-deposition-precipitation method. In brief, PdCl2 was
dissolved in 1 M HCl solution and mixed with a certain amount
of Ni(NO3)2 solution. The ␥-Al O3 support was dispersed in the
2
bimetallic precursor solution and Pd and Ni metal particles were
2.4. Catalytic NDMA reduction
then precipitated by adding the mixed solution with a 1 M Na CO3
2
solution slowly until the pH of the mixture reached 10.5. The mix-
ture was maintained for 1 h and the obtained precipitates were
The catalytic reduction of NDMA was conducted at atmospheric
pressure in a 250 mL three-necked flask reactor with a magnetic
stirrer. Prior to reaction, catalysts were ground to pass through
a 400-mesh sieve (<37 ␮m) to ensure negligible intraparticular
washed with deionized water until no chlorine was detected. The
◦
resultant catalysts were dried at 100 C for 6 h, calcined under N
2
◦
◦
◦
at 300 C for 2 h, and followed by reduction under H2 at 300 C
for 4 h. The loading amount of total metals (Pd and Ni) was
adjusted to be 3 wt.% for bimetallic catalysts. The catalysts with
diffusion limitation [27,28]. The reaction temperature was 25 C
controlled by a water bath. Briefly, 30 mg of the catalyst was
added into the reactor containing 200 mL of NDMA solution and
−
1
different ratios of Pd/Ni were designated as 3%(PdxNi1 ), where
the solution was purged with N2 (40 mL min ) under stirring
−x
x referred to the mass percent of Pd. The as-prepared bimetal-
lic catalysts included 3%(Pd0.9Ni0.1), 3%(Pd0.8Ni0.2), 3%(Pd0.5Ni0.5),
rate of 1200 rpm for 30 min. Afterwards, the N switched to H2
2
−
1
(40 mL min ) and the samples were taken at the desired contact
time.
3
%(Pd0.2Ni0.8) and 3%(Pd0.1Ni0.9). In addition, 3%Pd/Al O3 and
2

H. Chen et al. / Journal of Molecular Catalysis A: Chemical 421 (2016) 167–177
169
The obtained samples were filtered with a 0.45 ␮m membrane
to remove catalyst particles and the concentrations of NDMA and
DMA in the filtrate were determined by high performance liquid
chromatography (HPLC Waters e2695) equipped with a UV–vis
detector. The absorbance of NDMA and DMA was measured at a
wavelength of 230 nm and a flow rate of 1.0 mL min . The mobile
phase was water/methanol (v/v) = 85/15. Ammonium was deter-
mined using IC method with a DX 500 ion chromatographic system
increased metal dispersion [31]. In addition, both CO uptake
and the metal dispersion showed volcanic-type curves via the
ratio of Pd/Ni. The decrease of CO uptake and metal dispersion
decreased due to that when Ni became the main metal in bimetal-
lic catalyst, Ni particle is prone to aggregate. Furthermore, the
decrease of metal dispersion may be also correlated with the
decreased number of active metal sites with increase of Ni content.
Thus, 3%(Pd0.5Ni0.5), 3%(Pd0.2Ni0.8) and 3%(Pd0.1Ni0.9) had lower
metal dispersion, although they showed higher CO uptake than
3%Pd/Al O .
−
1
(
Dionex, Sunnyvale, CA, USA) consisting of a quaternary gradient
pump with automated membrane eluent degassing. The eluent was
0 mM HCl and 4 mM 1,2-diaminopropionic acid monohydrochlo-
2
3
4
−
1
ride, and the flow-rate was 1.0 mL min with the injection volume
3.1.2. TEM
of 10 ␮L.
Figs.
1 and S3 shows the TEM images of 3%Pd/Al O ,
2 3
3
%(PdxNi1 ) and 3%Ni/Al O3 catalysts. Uniform Pd particles could
−x
2
be clearly seen as dark spheres in 3%Pd/Al O , while the Ni/NiO par-
2
3
3
3
3
. Results and discussion
ticle is large and obscure in 3%Ni/Al O . As for 3%(PdxNi ), Pd-Ni
2
3
1−x
ensembles were observed. The particle distributions and average
particle sizes of the tested catalysts are shown in the insert figures
of Fig. 1. The average particle sizes could be calculated as follows:
.1. Catalyst characterization
.1.1. BET surface area and pulse CO chemisorption
The BET specific surface areas, average pore diameters and pore
ꢀ
ꢀ
nidi3/
nidi2
(1)
d¯ s =
volumes of Al O , 3%Pd/Al O , 3%(PdxNi
evaluated by N2 adsorption-desorption isotherms and the results
are shown in Table S1 and Fig. S1. The Al O supported monometal-
lic and bimetallic catalysts had similar BET surface areas pore
volumes and pore size distribution, reflecting that the textural
) and 3%Ni/Al O3 were
2
3
2
3
1−x
2
where ni is the number of counted particles with diame-
ter of di, and the total number of particles (ꢀni) is larger
than 100. The average metal particle size decreased in the
2
3
order:
3%Ni/Al2O3 > 3%(Pd0.1Ni0.9) > 3%(Pd0.2Ni0.8) > 3%Pd/Al2O3
characteristics of the Al O3 support were not significantly affected
either by Pd-loading or by Ni-loading. The elemental analyses of
> 3%(Pd0.5Ni0.5) > 3%(Pd0.9Ni0.1) > 3%(Pd0.8Ni0.2). This result, com-
bined with the results of CO chemisorption, indicated that catalyst
with high metal dispersion had small metal size. The 3%(Pd0.8Ni0.2)
catalyst exhibited the narrowest size distribution with a smallest
particle size of 2.55 nm, indicating of well metal dispersion in
3%(Pd0.8Ni0.2).
Dark field TEM images and elemental maps obtained via EDX
analysis for 3%(Pd0.8Ni0.2) are shown in Fig. 2. It could be observed
Pd and Ni coincided in this region, reflecting of the formation of
Pd-Ni ensembles.
2
Pd and Ni in 3%Pd/Al O , 3%(PdxNi
were determined by ICP technique and the results are also sum-
marized in Table 1. The experimental values of Pd and Ni contents
) and 3%Ni/Al O3 catalysts
2
3
1−x
2
in 3%Pd/Al O , 3%(PdxNi
deviations from the theoretic values, indicating of effective loading
of Pd and Ni.
) and 3%Ni/Al O3 catalysts had minor
2
3
1−x
2
Furthermore, CO chemisorption was used to estimate the dis-
persion of active metals and the CO chemisorption results of
3
%Pd/Al O , 3%(PdxNi
) and 3%Ni/Al O3 are shown in Table 1.
2
3
1−x 2
In the case of monometallic catalysts, 3%Pd/Al O had higher CO
3.1.3. XRD
2
3
uptake and metal dispersion than 3%Ni/Al O , due to the different
The XRD patterns of ␥-Al O , 3%Pd/Al O , 3%(PdxNi
) and
2
3
2
3
2
3
1−x
adsorption ability of metal precursor on Al O3 surface during the
3%Ni/Al O3 samples are shown in Fig. S4. All samples displayed
2
2
◦ ◦ ◦ ◦
characteristic diffraction peaks at 2␪ of 32.9 , 39.3 , 37.5 , 45.9 ,
synthesis process of the catalyst [24]. As shown in Fig. S2, ␥-Al O
2
3
◦
◦
◦
has an isoelectric point (IEP) of 7.5. At acid metal precursor solution,
the ␥-Al O surface was positively charged. Thus, the anionic Pd
60.5 , 67.3 and 84.9 , corresponding to the (022), (222), (025),
(220), (511), (042) and (444) lattice planes of the ␥-Al O (JCPDS
2
3
2
3
◦ ◦
precursor (HPdCl ) had a higher adsorption ability than the cationic
10-0425), respectively. New diffraction peaks at 40.2 and 46.8
4
Ni precursor (Ni(NO ) ). The higher adsorption ability resisted sin-
appeared in 3%Pd/Al O3 (Fig. S4) were assigned to (111) and
3
2
2
tering of metal particles, leading to higher dispersion of Pd particles
29,30].
In comparison to the 3%Pd/Al O3 catalysts, an enhancement in
(200) planes of cubic Pd (JCPDS 46-1043), respectively. Weak char-
acteristic reflections of Pd could be observed in 3%(Pd0.9Ni0.1),
3%(Pd0.8Ni0.2) and 3%(Pd0.5Ni0.5) catalysts, and no obvious Pd peak
was found in 3%(Pd0.2Ni0.8) and 3%(Pd0.1Ni0.9). The weakened peaks
of cubic Pd could be attributed to reduced Pd loading and introduc-
tion of Ni. On the one hand, decreasing Pd loading amount led to
smaller Pd crystal size, leading to weak peaks of cubic Pd [32,33].
On the other hand, addition of Ni led to a hindrance of crystalliza-
[
2
CO uptake was observed in 3%(PdxNi1 ) bimetallic catalysts. The
−x
reason is assumed to be that adding Ni can improve the reducibil-
ity of PdO, consequently hinder the sintering of Pd nanoparticles
during H2 reduction. Besides, the formation of mixed Pd-Ni oxides
prevented the surface mobility of the active metals, leading to
Table 1
Physico-chemical properties of 3%Pd/Al2O3, 3%(PdxNi1−x) and 3%Ni/Al2O3.
Catalyst
Pd contenta (wt%) Ni contenta (wt%) Pd/Ni (mol/mol) CO uptakeb (ml gcat−1
)
Metal dispersionb (%) Particle sizeb (nm) Particle sizec (nm)
3
3
3
3
3
3
3
%Pd/Al2O3
3.20
2.88
2.74
1.87
0.73
0.34
–
–
–
2.15
2.42
3.40
2.67
2.43
2.15
1.89
31.8
32.5
43.7
26.0
22.0
19.4
17.1
3.5
3.2
2.4
3.4
3.6
4.0
4.4
3.4
3.1
2.6
3.4
3.9
4.1
4.6
%(Pd0.9Ni0.1)
%(Pd0.8Ni0.2)
%(Pd0.5Ni0.5)
%(Pd0.2Ni0.8)
%(Pd0.1Ni0.9)
%Ni/Al2O3
0.36
0.53
1.65
2.49
2.72
2.89
4.45
2.84
0.62
0.16
0.07
–
a
b
c
Determined by ICP-OES chemical analysis.
Calculated from CO chemisorption.
Calculated from TEM.

1
70
H. Chen et al. / Journal of Molecular Catalysis A: Chemical 421 (2016) 167–177
Fig 1. TEM images and particle size distributions of (a) 3%Pd/Al2O3, (b) 3%(Pd0.9Ni0.1), (c) 3%(Pd0.8Ni0.2), (d) 3%(Pd0.5Ni0.5), (e) 3%(Pd0.2Ni0.8), (f) 3%(Pd0.1Ni0.9) and (g)
%Ni/Al2O3.
3
tion and crystal growth of Pd particles in 3%(PdxNi1 ) bimetallic
3%Ni/Al O3 catalysts, due to the formation of amorphous or well
dispersed Ni/NiO, and the similar phenomenon has been recorded
in the Pd-NiO@SiO2 catalyst [34].
−x
2
catalysts. The presence of characteristic reflection peaks of Pd indi-
cated the Pd-Ni might not form. If the alloy is formed, characteristic
reflections of Pd will not be seen and a shift of Pd standard peaks to
Pd-Ni intermetallic ones will be observed in bimetallic catalysts. In
3.1.4. H -TPR
2
0
addition, no Ni or NiO peak could be detected in 3%(PdxNi1 ) and
Hydrogen temperature programmed reduction (H2-TPR) can be
−x
used to explore the reducibility of the catalysts [35,36] and the H2-

H. Chen et al. / Journal of Molecular Catalysis A: Chemical 421 (2016) 167–177
171
Fig. 2. TEM dark field images of 3%(Pd0.8Ni0.2). Elemental EDX maps were taken from area 1. Area 2 was used to correct for any drift of the samples. EDX maps illustrate the
distribution of (a) Al, (b) O, (c) Pd, (d) Ni.
TPR profiles of 3%Pd/Al O , 3%(PdxNi
) and 3%Ni/Al O3 catalysts
of ␤-PdH phase [24,37]. The presence of ␤-PdH is associated with
2
3
1−x
2
0
are shown in Fig. 3. For 3%Pd/Al O3 catalyst in Fig. 3a, a nega-
large Pd particles, which are easily to be reduced to Pd . In the
2
◦
◦
tive H2 consumption peak at about 60 C (region I) and a broad
region II, the broad H2 consumption peaks at 350–550 C could be
◦
H2 consumption peak at 350–550 C were observed (region II). The
assigned to the reduction of subsurface PdO species [38] and the
peaks shifted to lower temperature as Ni contents increased. It indi-
cated that increasing of Ni content could improve the reducibility of
negative H2 consumption peak was also detected in 3%(Pd0.9Ni0.1),
3
%(Pd0.8Ni0.2) and 3%(Pd0.5Ni0.5), and the intensity of the peak
◦
decreased with the decrease of Pd/Ni ratio. The reduction peak
PdO species. In the region III, H2 consumption peaks at 100 C and
◦
◦
at 350–550 C in 3%(Pd0.9Ni0.1) and 3%(Pd0.8Ni0.2) shifted toward
70 C are attributed to the reduction of PdO species with strong
lower temperatures as compared to 3%Pd/Al O catalyst. More-
metal-support interaction [39]. It demonstrated that small Pd par-
2
3
◦
◦
over, H2 consumption peak at 100 C and 70 C were found for
%(Pd0.2Ni0.8) and 3%(Pd0.1Ni0.9), respectively (region III). There
ticles were generated and metal (Pd)-support (␥-Al O ) interaction
2 3
3
existed [24,39].
are two factors associated with the Pd reduction behaviors, which
are Pd particle size and Ni addition. In the region I, the nega-
tive H2 consumption peaks are characteristics of decomposition
The TPR profile of 3%Ni/Al O3 presents a broad temperature
reduction signal around 300–660 C (region IV), corresponding to
the reduction of fixed NiO species and NiO interacting with the
2
◦
(
a)
(b)
Fig. 3. Temperature programmed reduction patterns of (a) 3%Pd/Al2O3, 3%(Pd0.9Ni0.1), 3%(Pd0.8Ni0.2) and 3%(Pd0.5Ni0.5) and (b) 3%(Pd0.2Ni0.8), 3%(Pd0.1Ni0.9) and 3%Ni/Al2O3.

1
72
H. Chen et al. / Journal of Molecular Catalysis A: Chemical 421 (2016) 167–177
Table 2
alumina support [40]. Hilli et al. [41] reported that NiO peaks had
three typical reduction species, including bulk NiO reduction at
XPS data of 3%Pd/Al2O3, 3%(PdxNi1−x) and 3%Ni/Al2O3 catalysts.
◦
◦
4
00 C, well dispersed NiO reduction at 650 C and pseudo-spinel
Catalyst
Binding energy (eV)
Pd/Ni (mol/mol)
◦
◦
or spinel NiAl O reduction at 820 C or above 900 C. In this paper,
2
4
Pd 3d5/2
Ni 2p3/2
the NiO peaks in 3%Ni/Al O3 could be divided into two reduction
2
Pdn+
Ni0
Ni2+
peaks, which were denoted as ␣ peak (low temperature region at
◦
◦
3
00–500 C) and ␤ peak (high temperature region at 500–660 C).
3%Pd/Al2O3
335.1
335.1
335.3
335.5
336.1
–
–
–
–
–
–
–
–
3
3
3
%(Pd0.9Ni0.1)
%(Pd0.8Ni0.2)
%(Pd0.5Ni0.5)
1.32
0.54
0.19
0.07
–
The two peaks corresponded to NiO with a weak interaction with
852.8
853.0
852.9
853.5
853.7
861.5
861.6
861.3
861.7
861.7
855.7
855.9
855.7
855.8
855.9
the Al O support (␣ peak) and NiO with a strong interaction with
2
3
the Al O support (␤ peak), respectively [41]. As shown in Fig. 3b,
2
3
3%(Pd0.2Ni0.8
3%(Pd0.1Ni0.9
3%Ni/Al2O3
)
)
the appearance of strong ␣ peak and weak ␤ peak in 3%Ni/Al O3
2
indicated that the reduction process was mainly occurred in the
–
–
bulk NiO with weak interaction with ␥-Al O support. As for
2
3
3
%(Pd0.2Ni0.8) and 3%(Pd0.1Ni0.9) catalysts, ␣ peak was weakened
leading to the highest B.E of Pd 3d5/2 in 3%(Pd0.2Ni0.8). Moreover,
surface Pd species were not detected in 3%(Pd0.1Ni0.9) because of
low Pd loading amount. Similarly, no Ni species could be detected
in 3%(Pd0.9Ni0.1) catalyst.
and ␤ peak was strengthened, showing bulk NiO decreased and
well dispersed NiO increased with the increase of the ratio of Pd/Ni.
Meanwhile, the ␣ and ␤ peaks in 3%(Pd0.2Ni0.8) and 3%(Pd0.1Ni0.9
catalysts are shifted to lower temperature compared to 3%Ni/Al O ,
indicating the introduction of Pd could improve the reducibility of
NiO species.
)
2
3
Fig. 4b shows the Ni 2p region of XPS for 3%(Pd Ni ) and
1−x
x
3%Ni/Al O3 catalysts. The B.E of Ni 2p
in 3% Ni/Al O3 at 853.7 eV
2
3/2
2
0
was assigned to Ni , and the satellite peaks at 855.9 eV and 861.7 eV
were attributed to two forms of NiO phases [46,47]. It could be seen
from Fig. 4b that most of Ni species in 3%Ni/Al O were NiO, while
3.1.5. XPS
2
3
In order to determine the electronic properties of metal species,
few Ni0 specie was observed, reflecting of low reduction degree of
Ni species. The result could be speculated from the result of TPR
XPS analysis was conducted over all catalysts and the spectra of
Pd 3d and Ni 2p are presented in Fig. 4. The Pd 3d region of XPS
for 3%(PdxNi1 ) and 3%Pd/Al O3 catalysts are shown in Fig. 4a
and the binding energies (B.Es) of Pd 3d5/2 for all catalysts are
listed in Table 2. For 3%Pd/Al O , the B.E of Pd 3d was recorded
◦
that NiO could not be completely reduced by H2 at 300 C. As for
−x
2
3%(Pd Ni
), the Ni 2p regions of XPS are also shown in Fig. 4b
x
1−x
0
and the B.Es of Ni and NiO are listed in Table 2. The B.Es of Ni
2
3
5/2
2p3/2 (855.7–855.9 eV) in 3%(Pd Ni
) were slightly lower than
x
1−x
0
as 335.1 eV, higher than the reported B.E of Pd , due to the exis-
that in 3%Ni/Al O3 (855.9 eV), probably due to the transfer of elec-
2
tence of electron deficient Pd species (Pdn+) generated as a result
tron density from Pd to Ni species. Similar negative shift of Ni 2p3/2
in bimetallic Pd-Ni catalyst compared to the monometallic Ni sam-
ple was also observed by Singh [48]. As a result, it confirmed the
interaction and electronic transfer between Pd and Ni species in
of metal-support interaction [42–44]. In the case of 3%(PdxNi1
catalysts, the B.Es of Pd 3d5/2 (335.1–336.1 eV) were higher than
that of 3%Pd/Al O , due to the interaction between Pd and Ni
)
−x
2
3
species in Pd-Ni ensembles. Similar B.E enhancement of Pd 3d5/2
was observed in previous reports and the reason was found to be
the electronic relaxation energy of Pd levels in the presence of Ni
3%(Pd Ni
) bimetallic catalysts as mentioned above.
x
1−x
In addition, the surface Pd/Ni molar ratios in bimetallic
3%(Pd Ni ) catalysts were calculated and the results are listed
x
1−x
[
45]. It was also proposed by Fang et al. [46] that there was an
electronic transfer from PdOx to NiOx, leading to higher B.E of Pd
d5/2 in Pd-Ni/Al O catalysts compared to Pd/Al O3 catalysts. In
in Table 2. The ratios analyzed by XPS are higher than that from ICP
analysis, indicating the surface enrichment of Ni on these bimetallic
catalysts.
3
2
3
2
this work, it could also be concluded that the enhancement of B.Es
of Pd 3d5/2 in 3%(PdxNi ) may be attributed to electron transfer
3.2. Catalytic investigations
1−x
from Pd species to Ni species. In addition, it was also found from TPR
that there was an metal (Pd)-support interaction in 3%(PdxNi1
3
.2.1. NDMA reduction over 3%(Pd Ni
)
0
.8 0.2
)
,
Catalytic reduction process of NDMA via reaction time over
3%(Pd0.8Ni0.2) is shown in Fig. 5a. The NDMA solution was rapidly
−x
with low ratio of Pd/Ni, which could give rise to B.Es of Pd 3d5/2
3
% Pd/Al O
3%Ni/Al O
2 3
(b)
2
3
(a)
3
%(Pd Ni )
3
%(Pd Ni
)
0.1 0.9
0
.9 0.1
3
%(Pd Ni )
3
%(Pd Ni
)
0.2 0.8
0
.8 0.2
3
%(Pd Ni
)
3%(Pd Ni )
0.5 0.5
0
.5 0.5
3
%(Pd Ni
)
3%(Pd Ni )
0.8 0.2
0
.2 0.8
346
344
342
340
338
336
334 332
330 885
880
875
870
865
860
855
850
Binding Energy (eV)
Binding Energy (eV)
Fig. 4. (a) XPS spectra of Pd 3d for 3%Pd/Al2O3 and 3%(PdxNi1−x) catalysts; (b) XPS spectra of Ni 2p for 3%Ni/Al2O3 and 3%(PdxNi1−x) catalysts.

H. Chen et al. / Journal of Molecular Catalysis A: Chemical 421 (2016) 167–177
173
(
a)
(b)
Fig. 5. (a) Catalytic reduction process of NDMA versus reaction time over 3%(Pd0.8Ni0.2); insert figure show pseudo-first-order model fit of NDMA reduction; (b) comparative
−
1
tests over 3%Pd/Al2O3 and 3%Ni/Al2O3. Reaction conditions: 2.02 ␮M NDMA, pH 7.2, 0.15 g L catalyst.
Table 3
reduced by 3%(Pd0.8Ni0.2) in the presence of H and NDMA could not
2
Comparison of NDMA reduction rates over various catalysts.
be detected after reaction for 60 min. In addition, catalytic reduc-
tion of NDMA has no significant effect on pH variation (from 7.0
to 7.3) during the reaction process. For comparison, the NDMA
reduction processes without catalyst and with ␥-Al O were con-
Catalyst
Kobs (h−1)
Metal weight loading
Ref.
normalized pseudo-first-order
−
1
−1
h ]
rate [L gme
2
3
ducted and the results are shown in Fig.S5. No NDMA reduction was
observed under H2 without catalyst, indicating that NDMA could
not be reduced by H2 only. The concentration of NDMA slightly
Pd
Pd-Cu
Ni
Fe
Fe-Ni
Mn
Porous Ni
Fe/Ni
Pd/In
NiB
ZnO
3%(Pd0.8Ni0.2)
–
–
–
–
–
–
11.5
66.5
8.30
0.13
0.65
0.07
61.4
–
41.7
29.5
0.12
836 ± 21
[18]
[18]
[18]
[18]
[18]
[18]
[16]
[13]
[19]
[11]
[15]
This study
decreased in the presence of ␥-Al O , due to weak adsorption of
2
3
NDMA on ␥-Al O . Considering of the saturation of H under the
2
3
2
30.7
21.0
0.25
18.9
1.22
4.1 ± 0.1
reaction condition, the effect of mass transfer limitation was eval-
uated by changing the dosage of 3%(Pd0.8Ni0.2) and stirring speed.
The dosage of catalyst varied from 10–50 mg and the stirring speed
changed from 900–1500 rpm. As showed in Fig. S6, increase of the
catalyst dosage and stirring rate had no significant effect on the
initial catalytic activities, indicating no mass transfer limitation
existed in the reduction reaction.
In the absence of any external and internal diffusion limitation,
the Langmuir-Hinshelwood kinetic model [49,50] could be used to
describe the process of NDMA reduction.
rate is lower than those of porous Ni, Fe/Ni and NiB catalysts,
%(Pd0.8Ni0.2) catalyst in this study owned the highest metal weight
3
loading normalized pseudo-first-order rate among these catalysts.
It indicated that small quantity of active metal in 3%(Pd0.8Ni0.2) was
needed to reduce NDMA as compared to other catalysts, and the
results emphasized that the 3%(Pd0.8Ni0.2) catalyst was a potential
catalyst for NDMA reduction.
dC
bC
r0 = − dt = kꢁs = k
(2)
(3)
1 + bC
ꢁ
ꢂ
C0
C
ln
+ b(C − C) = kbt
0
Comparative tests of NDMA reduction over 3%Pd/Al O3 and
2
where r is the initial NDMA reduction rate at concentration C, ꢁs
3%Ni/Al2O3 were also carried out under the same conditions as
3%(Pd0.8Ni0.2), and the results are shown in Fig. 5b. NDMA had no
observable reduction by 3%Pd/Al2O3 or 3%Ni/Al2O3, reflecting that
sole Pd or Ni/NiO had poor catalytic activity on NDMA reduction. As
reported, activation of H2 and break of N N bond are two main pro-
cesses on NDMA reduction [19]. In the case of 3%Pd/Al2O3, it was
hard to break the N N bond for the N N bond cleavage energy
over Pd was higher than that over Ni [20], leading to a low cat-
alytic activity. Other researchers using monometallic Pd catalyst in
NDMA reduction also gained low catalytic activities [17]. Adding
another metal to Pd could improve the catalytic activity as sug-
gested in other catalysts, such as Pd-Cu [18] and Pd-In [19]. In the
case of 3%Ni/Al O , NiO (the main Ni specie) could not effectively
0
is the NDMA coverage on the catalyst surface, k is the reaction rate
constant and b is the equilibrium constant for NDMA adsorption on
the catalyst.
When the adsorption of NDMA is relatively weak and/or the
concentration of NDMA is low in the reduction reaction, Eq. (3)
can be simplified to a pseudo-first order equation with an apparent
first-order rate constant (Kobs), as expressed in Eq. (4).
ꢁ
ꢂ
C0
C
ln
= kbt = Kobs
t
(4)
The insert figure in Fig. 5a shows the relationship between
ln(C /C) and the reaction time. The reaction process could be
0
2
3
2
described by a pseudo-first-order model with R = 0.955, although
activate H , causing low catalytic activity of this catalyst.
2
there is a minor deviation during the first 7 min. The dupli-
cate tests were operated for three times and the data showed
good reproducibility. The rate constant of NDMA reduction over
According to previous reports, the expected reduction pro-
ductions of NDMA were DMA and ammonium [13,18]. During
the catalytic reduction process, the expected product DMA was
detected and the total concentration of NDMA and DMA were
nearly constant except for deviation of measurement. However,
the expected product ammonium was not detected, possibly due
to the detection method. The final concentration of ammonium
was supposed to be 36.5 ␮g/L, which was below the detection lim-
−
1
3
%(Pd0.8Ni0.2) was calculated to be 4.1 ± 0.1 h , and the corre-
sponding metal weight loading normalized pseudo-first-order rate
was 836 ± 21 L gme
−1
−1
h (Table 3). The metal weight loading nor-
malized pseudo-first-order rates as reported in other literatures
were also summarized in Table 3. Although the pseudo-first-order

1
74
H. Chen et al. / Journal of Molecular Catalysis A: Chemical 421 (2016) 167–177
(
a)
(b)
Fig. 6. Catalytic reduction of NDMA over 3%(PdxNi1−x) catalysts: (a) Initial NDMA reduction rates versus Pd/Ni molar ratio and (b) linear plot of TOF versus metal dispersion.
−1
Reaction conditions: 2.02 ␮M NDMA, pH 7.2, 0.15 g L catalyst.
(
a)
(b)
Fig. 7. (a) Initial activity of 3%(Pd0.8Ni0.2) as a function of initial NDMA concentration and (b) linear plot of 1/r0 versus 1/C. Reaction conditions: pH 7.2, 0.15 g L−1 catalyst.
itation of IC measurement. In addition, no intermediate, such as
dimethylhydrazine was found during the reduction process.
understand the effect of NDMA adsorption. As shown in Fig. 7a, the
initial NDMA reduction rate on 3%(Pd0.8Ni0.2) was increased from
−
1
−1
1
04.28 to 232.37 ␮M gcat
h with the initial NDMA concentra-
tion increasing from 1.35 to 6.75 ␮M, and then remained nearly
constant with further increase of initial NDMA concentration.
As discussed in Section 3.2.1, the Langmuir-Hinshelwood kinetic
model [45,46] could be used to describe the mechanism based on
the adsorption of NDMA. As for NDMA reduction with different
initial concentration, Eq. (2) can be transferred to Eq. (5):
3
.2.2. Effect of Pd/Ni ratio
Catalytic reduction of NDMA over 3%(PdxNi1 ) catalysts were
−x
operated to investigate the effect of Pd/Ni ratio and the results
are shown in Fig. 6a. The 3%(PdxNi1 ) catalysts with different
Pd/Ni ratio presented different reduction efficiencies on NDMA
reduction. The initial catalytic activities via the moral ratio of
Pd/Ni are shown in the insert figure of Fig. 6a. The initial catalytic
activities of 3%(PdxNi1 ) catalysts were decreased in the order:
3
>
−x
1
r0
1 1
kb C
1
=
+ k
(5)
−x
%(Pd0.8Ni0.2) > 3%(Pd0.5Ni0.5) > 3%(Pd0.9Ni0.1) > 3%(Pd0.2Ni0.8
)
3%(Pd0.1Ni0.9). There appeared to be an optimal Pd/Ni ratio, and
However, the concentration C here refers to the different ini-
beyond which further increase of the ratio would lead to decrease
in the catalytic activity. The influence of the moral ratio of Pd/Ni
was further evaluated by TOF, for which the number of exposed
metal amount was determined by CO chemisorption measure-
ment. TOF values via metal dispersion are shown in Fig. 6b and
there is an obvious increase in TOF with the metal dispersion of the
tial concentration of 2,4-D. As shown in Fig. 7b, the plot of 1/r0
versus 1/C was fitted to a liner (R2 higher than 0.96), implying
that the NDMA reduction could be well described by Langmuir-
Hinshelwood model and controlled by the pre-adsorption behavior
of NDMA. In detail, the adsorption of NDMA on the surface of
3%(Pd0.8Ni0.2) increased with the increase of initial concentration of
NDMA, and the reduction rate consequently increased; with further
increase of NDMA concentration, the initial NDMA reduction rate
showed no longer increase for the NDMA adsorption was saturated.
3
%(PdxNi1 ) bimetallic catalysts. And the plot of the TOF values
−x
2
versus metal dispersion was fitted to a liner model (R = 0.9366),
indicating the catalytic activities of 3%(Pd0.8Ni0.2) was related to
the metal dispersion and the catalyst with high metal dispersion
favored the NDMA reduction.
3.2.4. Effect of pH
Solution pH has an impact on adsorption of the reactants onto
3
.2.3. Effect of initial concentration of NDMA
As discussed in Section 3.2.1, NDMA adsorption is an initial and
the catalyst surface [32,51], thus influence the catalytic reduction of
NDMA. The NDMA reduction process over 3%(Pd0.8Ni0.2) at varied
pH is shown in Fig. S7. The reduction efficiency of NDMA reduction
was 13.0% at pH = 2.0, while 100% at pH = 11.0 after reaction for
essential step for catalytic reduction, and hence we studied the cat-
alytic performance of NDMA with different initial concentration to

H. Chen et al. / Journal of Molecular Catalysis A: Chemical 421 (2016) 167–177
175
Fig. 8. (a) Dependence of initial NDMA reduction rates over 3%(Pd0.8Ni0.2) and existent species at different pH. (b) Surface charge densities (␨) of ␥-Al2O3 and 3%(Pd0.8Ni0.2)
−
1
catalysts as a function of solution pH. Reaction conditions: 2.02 ␮M NDMA, 0.15 g L catalyst. (c) Recycling runs of NDMA reduction over 3%(Pd0.8Ni0.2). Reaction conditions:
−1
2
.02 ␮M NDMA, pH 11.0, 0.15 g L catalyst.
6
0 min. The initial catalytic activity of 3%(Pd0.8Ni0.2) versus pH is
charged, causing an electrostatic attractive interaction between
+
presented in Fig. 8a. The initial catalytic activity of 3%(Pd0.8Ni0.2
)
3%(Pd0.8Ni0.2) and (CH ) N
N
OH , further elevated the adsorp-
3
2
was increased slowly from pH 2.0 to 5.6, while a significant increase
was observed when the pH increased from 5.6 to 7.2, and then a
slow increase occurred when the pH increased from 7.2 to 11.
The impact of solution pH could be explored by both alteration
tion and reduction of NDMA on catalyst. At 7.2 < pH < 11(region IV),
+
the content of (CH ) N
N
OH was gradually decreased and the
3
2
negative charge of 3%(Pd0.8Ni0.2) was increased, the electrostatic
attractive interaction between NDMA and catalyst slowly increased
the initial catalytic activity. Recycling tests under pH 11 was con-
ducted to evaluate the stability of 3%(Pd0.8Ni0.2) catalyst and the
results are shown in Fig. 8c. After reused for three times, the ini-
tial activity presented slight variation, reflecting of the excellent
stability of 3%(Pd0.8Ni0.2).
of NDMA speciation and the adsorption properties of 3%(Pd0.8Ni0.2).
+
On the one hand, NDMA existed in two forms as (CH ) N
N OH
3
2
and (CH ) N
N O in aqueous solution and the forms were pH
3
2
dependent as shown in the insert figure of Fig. 8a. At pH < 6,
+
only (CH ) N
N
OH existed in the solution, and the amount of
3
2
(
CH ) N
N
O increased when the pH increased from 6 to 11.
3
2
On the other hand, surface zeta potentials (␨) of 3%(Pd0.8Ni0.2
)
changed with the alteration of pH. As shown in Fig. 8b, the isoelec-
tric point (IEP) was 6.5 for 3%(Pd0.8Ni0.2). At pH < 5.6 (region I), the
3.2.5. Hypothesized reaction mechanism
3
%(Pd0.8Ni0.2)/Al O catalyst surface was positively charged and
The reaction mechanism of NDMA reduction over 3%(Pd0.8Ni0.2)
was proposed in Scheme 1. Firstly, NDMA and H2 were adsorbed
onto the surface of the catalyst. There are two species of
2
3
+
NDMA existed mainly as (CH ) N
N OH , which caused an elec-
3
2
trostatic repulsive interaction between NDMA and 3%(Pd0.8Ni0.2).
When the pH increased from 2 to 5.6, the decreased charge
of 3%(Pd0.8Ni0.2) weakened the electrostatic repulsive interac-
tion, increasing the initial catalytic activity due to enhanced
NDMA in different pH solution, which are (CH3)2N
N O and
+
(CH3)2N
N
OH , respectively. The species of NDMA could influ-
ence the adsorption interaction between NDMA and catalyst, and
consequently affect the catalytic activity. NDMA molecule could
adsorb onto Pd or Ni/NiO by upright adsorption modes via NO frag-
ment or flat adsorption modes via ONN plane [20,21], and then be
activated. Meanwhile, the adsorbed H2 was activated by Pd and
“spillover” to the surface of Ni/NiO, and then the N N bond in
NDMA was cleaved by active H. Finally, DMA and nitrosyl radical
were formed and the latter was further reduced to ammonium.
NDMA adsorption. At 5.6 < pH < 6.5 (region II), the content of
+
(
CH ) N
N
OH gradually decreased while that of (CH ) N
N
O
3
2
3 2
+
increased. The lower content of (CH ) N
N OH weakened the
3
2
electrostatic repulsive interaction between NDMA and the cata-
lyst, and further improved the NDMA reduction. At 6.5 < pH < 7.2
(
region III), the 3%(Pd0.8Ni0.2) catalyst surface was negatively

1
76
H. Chen et al. / Journal of Molecular Catalysis A: Chemical 421 (2016) 167–177
Scheme 1. The reaction mechanism of NDMA reduction over 3%(Pd0.8Ni0.2).
Similar mechanism in NDMA reduction over Pd-In was reported
Appendix A. Supplementary data
to be activation of hydrogen by Pd, activation of NDMA by In and
then NDMA reduction by active H [19]. In this work, it could be
concluded that both Pd and Ni/NiO played key roles on NDMA
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.molcata.2016.05.
026.
reduction reaction: Pd was used to activate H and Ni/NiO provided
2
the active site to cleave N-N in NDMA. Moreover, the Pd-Ni/NiO
ensembles, involved in metal–metal interaction, was found bene-
ficial for hydrogenation reactions [52], resulting in high catalytic
activity in 3%(Pd0.8Ni0.2).
References
[
[
[
[
1] A.C. Gerecke, D.L. Sedlak, Environ. Sci. Technol. 37 (2015) 1331–1336.
2] W.A. Mitch, D.L. Sedlak, Environ. Sci. Technol. 36 (2002) 588–595.
3] U.S. EPA, Fed. Reg. 63 (1998) 10273–10287.
4] U.S. EPA, Record of Decision for the Western Groundwater Operable Unit
OU-3, Aerojet Sacramento Site, (2001) July 20.
4
. Conclusions
In summary, ␥-Al O supported Pd-Ni bimetallic catalysts were
[5] J. Chung, C.H. Ahn, Z. Chen, B.E. Rittmann, Chemosphere 70 (2008) 516–520.
[6] C. Chen, W. Zhou, D. Lin, Bioresour. Technol. 179 (2015) 359–366.
[7] C. Lee, J. Yoon, U. Von Gunten, Water Res. 41 (2007) 581–590.
2
3
prepared via co-deposition-precipitation methods. Well dispersed
Pd-Ni/NiO ensembles on ␥-Al O3 were observed in 3%(PdxNi
)
1−x
[8] J. Lee, W. Choi, J. Yoon, Environ. Sci. Technol. 39 (2005) 6800–6807.
[9] X. Guo, Q. Li, M. Zhang, M. Long, L. Kong, Q. Zhou, H. Shao, W. Hu, T. Wei,
Chemosphere 120 (2015) 521–526.
2
bimetallic catalysts. Increasing of Ni in 3%(PdxNi1 ) could improve
−x
the reducibility of PdO, and introduce an electronic transfer from
Pd to Ni. The formation of mixed Pd-Ni metal oxides could prevent
the migration of active metals, consequently enhanced the metal
[
[
10] B.P. Chaplin, G. Schrader, J. Farrell, Environ. Sci. Technol. 43 (2009) 8302–8307.
11] A.J. Frierdich, C.E. Joseph, T.J. Strathmann, Appl. Catal. B: Environ. 90 (2009)
175–183.
[
[
12] G. Lunn, E.B. Sansone, L.K. Keefer, Environ. Sci. Technol. 17 (1983) 240–243.
13] L. Gui, R.W. Gillham, M.S. Odziemkowski, Environ. Sci. Technol. 34 (2000)
dispersion in 3%(PdxNi1 ). However, further increase of Ni make a
−x
decrease of metal dispersion due to the aggregation of Ni particles.
Results of catalytic activity evaluation showed that
3
489–3494.
[14] M.S. Odziemkowski, L. Gui, R.W. Gillham, Environ. Sci. Technol. 34 (2000)
495–3500.
3
3
%(Pd0.8Ni0.2) catalyst could reducte NDMA to dimethylamine
DMA) with a much higher metal-loading-normalized pseudo-
first-order rate constant in comparison with other catalysts
[
15] Y. Han, Z.L. Chen, L.N. Tong, L. Yang, J.M. Shen, B.Y. Wang, Y. Liu, Q. Chen,
Water Res. 47 (2013) 216–224.
(
[16] A.J. Frierdich, J.R. Shapley, T.J. Strathmann, Environ. Sci. Technol. 42 (2007)
62–269.
2
reported before. The catalytic activity of 3%(PdxNi1 ) was posi-
−x
[
17] D. Shuai, D.C. McCalman, J.K. Choe, J.R. Shapley, W.F. Schneider, C.J. Werth,
ACS Catal. 3 (2013) 453–463.
tively relative to metal dispersion. The NDMA reduction process
could be described by Langmuir-Hinshelwood model, reflecting
of a pre-adsorption controlled reaction. The trend of catalyst
activity of 3%(Pd0.8Ni0.2) versus pH confirmed the importance
of NDMA adsorption. With pH changed from 2.0 to 11.0, both
surface charge of 3%(Pd0.8Ni0.2) and species of NDMA changed;
[18] M.G. Davie, M. Reinhard, J.R. Shapley, Environ. Sci. Technol. 40 (2006)
329–7335.
7
[
19] M.G. Davie, K. Shih, F.A. Pacheco, J.O. Leckie, M. Reinhard, Environ. Sci.
Technol. 42 (2008) 3040–3046.
[20] V.A. Ranea, T.J. Strathmann, J.R. Shapley, W.F. Schneider, ChemCatChem 3
(2011) 898–903.
[
[
[
21] V.A. Ranea, J. Mol. Catal. A: Chem. 392 (2014) 157–163.
22] B. Coq, F. Figueras, J. Mol. Catal. A: Chem. 173 (2001) 117–134.
23] M.O. Nutt, K.N. Heck, P. Alvarez, M.S. Wong, Appl. Catal. B: Environ. 69 (2006)
115–125.
high adsorption interaction between NDMA and 3%(Pd0.8Ni0.2
)
catalyst was beneficial to the reduction process. The reduction
mechanism was proposed to be that H2 was activated by Pd and
transport to Ni/NiO, and then reduce NDMA to DMA.
[
24] N.S. Babu, N. Lingaiah, P.S. Sai Prasad, Appl. Catal. B: Environ. 111–112 (2012)
309–316.
[
[
25] Y. Huang, W.M.H. Sachtler, J. Catal. 188 (1999) 215–225.
26] A. Benedetti, G. Fagherazzi, F. Pinna, G. Rampazzo, M. Selva, G. Strukul, Catal.
Lett. 10 (1991) 215–223.
Acknowledgements
[
[
27] H. Chen, P. Zhang, W. Tan, F. Jiang, R. Tang, RSC Adv. 4 (2014) 38743–38749.
28] M.A. Aramendía, V. Borau, I.M. García, C. Jimenez, F. Lafont, A. Marinas, F.J.
Urbano, J. Mol. Catal. A: Chem. 184 (2002) 237–245.
The financial supports from the Natural Science Foundation
of China (Nos. 51208257, 51178223 and 51478223) and the Nat-
ural Science Foundation of Jiangsu Province (BK2012405), China
Postdoctoral Science Foundation (No. 2013M541677), the Jiangsu
Planned Projects for Postdoctoral Research Funds (1202007B),
the Fundamental Research Funds for the Central University (Nos.
[
29] S.E. Wanke, P.C. Flynn, Catal. Rev. 12 (1975) 93–135.
[30] Y.H. Wang, H.M. Liu, B.Q. Xu, J. Mol. Catal. A: Chem. 299 (2009) 44–52.
[
[
[
31] L. Guczi, I. Kiricsi, Appl. Catal. A: Gen. 186 (1999) 375–394.
32] P. Zhang, F. Jiang, H. Chen, Chem. Eng. J. 234 (2013) 195–202.
33] J. Zhou, K. Wu, W. Wang, Y. Han, Z. Xu, H. Wan, D. Zhu, Appl. Catal. B: Environ.
162 (2015) 85–92.
34] H. Liu, K. Tao, C. Xiong, S. Zhou, Catal. Sci. Technol. 5 (2015) 405–414.
35] P. Casta n˜ o, B. Pawelec, J.L.G. Fierro, J.M. Arandes, J. Bilbao, Fuel 86 (2007)
[
[
3
0915011308 and 30920140112010) and the Priority Academic
Program Development of Jiangsu Higher Education Institution are
gratefully acknowledged.
2262–2274.
[36] M. Mihet, M.D. Lazar, Chem. Eng. J. 251 (2014) 310–318.

H. Chen et al. / Journal of Molecular Catalysis A: Chemical 421 (2016) 167–177
177
[
[
[
37] H. Chen, Z.Y. Xu, H.Q. Wan, J.Z. Zheng, D.Q. Yin, S.R. Zheng, Appl. Catal. B:
Environ. 96 (2010) 307–313.
38] N.S. Babu, N. Lingaiah, J.V. Kumar, P.S. Prasad, Appl. Catal. A: Gen. 367 (2009)
[45] T. Jiang, Q. Huai, T. Geng, W. Ying, T. Xiao, F. Cao, Biomass Bioenergy 78 (2015)
71–79.
[46] R. Fang, Y. Cui, Z. Shi, M. Gong, Y. Chen, Chin. J. Catal. 36 (2015) 994–1000.
[47] P. Lu, T. Teranishi, K. Asakura, M. Miyake, N. Toshima, J. Phys. Chem. B 103
(1999) 9673–9682.
[48] S.K. Singh, Y. Iizuka, Q. Xu, Int. J. Hydrogen Energy 36 (2011) 11794–11801.
[49] A. Pintar, J. Batista, J. Levec, T. Kajiuchi, Appl. Catal. B: Environ. 11 (1996)
81–98.
[50] Z.M. De Pedro, J.A. Casas, L.M. Gomez-Sainero, J.J. Rodriguez, Appl. Catal. B:
Environ. 98 (2010) 79–85.
7
0–76.
39] N.S. Babu, N. Lingaiah, R. Gopinath, P.S. Sankar Reddy, P.S. Sai Prasad, J. Phys.
Chem. C 111 (2007) 6447–6453.
[
[
40] U. Oemar, K. Hidajat, S. Kawi, Appl. Catal. A: Gen. 402 (2011) 176–187.
41] Y. Hilli, N.M. Kinnunen, M. Suvanto, A. Savimäki, K. Kallinen, T.A. Pakkanen,
Appl. Catal. A: Gen. 497 (2015) 85–95.
[
[
[
42] Y. Yazawa, H. Yoshida, N. Takagi, S. Komai, A. Satsuma, T. Hattori, J. Catal. 187
(
1999) 15–23.
43] A. Barrera, M. Viniegra, S. Fuentes, D. Gabriela, Appl. Catal. B: Environ. 56
2005) 279–288.
44] F. Cárdenas-Lizana, S. Gómez-Quero, C. Amorim, M.A. Keane, Appl. Catal. A:
Gen. 473 (2014) 41–50.
[51] G. Yuan, M.A. Keane, Appl. Catal. B: Environ. 52 (2004) 301–314.
[52] A. Corma, P. Serna, J.J. Calvino, J. Am. Chem. Soc. 130 (2008) 8748–8753.
(