Y.-S. Jeong, et al.
Catalysis Today xxx (xxxx) xxx–xxx
ethanol [18,25], selective reduction of NO [26], and carbon dioxide
2.2. Catalyst characterization
reforming of methane [15,27–29]. In addition, Y
2
O
3
has been re-
portedly used as a support and promoter because of its ability to resist
carbon deposition [20,30,31]. Furthermore, it is highly effective in
dehydrogenation reactions [18] and has excellent optical properties
The phase structures of the powder samples were confirmed by X-
ray diffraction (XRD, Rigaku, SmartLab) measurements. The XRD
measurements were carried out with Cu Kα radiation
(λ = 0.15428 nm) at a scanning rate of 0.02°/s from 10° to 90° in 2θ.
High-angle annular dark-field scanning transmission electron mi-
croscopy (HADDF-STEM) imaging using a JEOL JEM-2100 F micro-
scope was operated at 200 kV.
[
32], high thermal stability [33,34], and advantageous chemical
properties such as high surface oxygen mobility [15,35]. Among the
various properties, we have focused on its ability to resist carbon de-
position and its effectiveness for dehydrogenation because the hydro-
genation of ethanol to acetaldehyde is the rate determining step for the
reductive amination of ethanol [36,37]. In particular, the ability to
resist carbon deposition can reduce the catalytic deactivation. The
The specific surface area, total pore volume, and average pore size
of the Ni(X)/SY catalysts were determined from N -sorption isotherms
2
at -196 ℃ using a Micromeritics ASAP2020 apparatus. Before N -
2
surface properties of Ni/Y
2
O
3
catalysts and the relationship between
sorption analysis, the samples were degassed at 200 ℃ for 4 h. The
the physicochemical properties of the catalysts and the catalytic re-
ductive amination performance have not been studied yet. In addition,
to date, researchers have focused on the amination of epoxides [38],
ketones [39], and alcohols such as ethanol [10,40–42], allyl alcohol
specific surface areas of the samples were then calculated using the
Brunauer–Emmet–Teller (BET) equation in the relative pressure (P/P )
0
range of 0.05–0.2. The total pore volumes were obtained from the vo-
lumes of N adsorbed at a relative pressure 0.995, and the average pore
2
[
43], dodecanol [44], and myrtenol [45]. A few papers have discussed
diameter was calculated as 4 × (Vpore/SBET), where Vpore is the total
pore volume and SBET is the BET surface area. The calculation assumes
that the pores are cylindrical. The pore size distribution was calculated
from the desorption branch using the Barrett–Joyner–Halenda (BJH)
equation.
the reaction rates for the amination reaction, and power-law type re-
action rate equations have been developed by considering the forma-
tion of acetonitrile over Co-Ni/γ-Al
2
O catalysts [41]. In addition,
3
Ibanez et al. showed the promoting effects of the hydrogen pressure on
the catalytic amination of 1-octanol over Ag-Co/ Al
2
O
3
catalysts using a
The metal dispersion, particle sizes, and metallic surface areas of the
kinetic simulation approach [46]. Kinetic modeling for the direct ami-
different Ni(X)/SY catalysts were determined by volumetric H che-
2
nation of dodecanol over carbon-based catalysts has also been reported
misorption (Micromeritics ASAP2020C). Before adsorption experi-
[
44]. However, there are scant reports concerning the amination of
ments, each sample was reduced at 500 ℃ for 3 h under flowing H
2
3
−1
ethanol with NH
3
and H
2
based on a detailed reaction mechanism.
catalysts having dif-
(50 cm min ). The metal dispersion was calculated for each sample
assuming an H/Ni stoichiometry of 1.0.
Herein, we report the preparation of Ni/Y
O
2 3
ferent Ni loadings for application in the reductive amination of ethanol
to EAs. In addition, the relationship between the physicochemical
properties of the catalysts and their catalytic performance was carefully
investigated. Furthermore, we propose a reaction mechanism for the
reductive amination of ethanol by changing the reaction parameters
such as reaction temperature (T), and the partial pressures of ethanol,
H
2
-Temperature programmed reduction (H -TPR) was performed
2
using a quadrupole mass spectrometer (QMS, Balzers QMS 200) as the
detector. Before measurement, each 0.1 g sample was pretreated under
3
−1
flowing Ar (50 cm min ) at 450 ℃ and held for 2 h to remove ad-
sorbed water and other contaminants before being cooled to 30 ℃
under flowing Ar. The reducing gas was
a
5%
H /Ar mixture
2
3
−1
NH
3
, and H
2
. Finally, the cause of deactivation affecting the catalyst
(50 cm min ), and this was passed over the sample at a heating rate
−1
stability are discussed.
of 10 ℃ min
until 900 ℃. The mass signal of H (m/z = 2) was de-
2
tected by the QMS.
To investigate the catalytic deactivation behavior, X-ray photo-
electron spectroscopy (XPS) and temperature programmed oxidation
2
. Experimental
.1. Catalyst preparation
The Y support was synthesized via a precipitation method using
(
TPO) were performed. The surface chemical states of the calcined Ni
2
(10)/SY catalysts were identified by XPS using an PHI Quantera II
photoelectron spectrometer (Al Kα radiation; hν = 1486.6 eV), and the
XPS data were calibrated using the binding energy of adventitious
carbon (i.e., C 1s = 284.5 eV) as a standard. To determine the peak
areas and positions, the C 1s, N 1s, Ni 2p, Y 3d, O 1s, and Si 2p spectra
were deconvoluted using Gaussian–Lorentzian curve fitting after
background subtraction using the Shirley method. In the TPO experi-
O
2 3
yttrium nitrate hexahydrate (Y(NO
3
)
2
·6H O, SAMCHUN) as a precursor.
2
The yttrium precursor was dissolved in deionized water at room tem-
perature with vigorous stirring, and the final concentration of the
precursor was 0.5 M. Subsequently, 2 M NH OH solution (SK Chemical)
4
was added to the yttrium solution until the pH reached 10. The re-
sulting solution was aged at 100 ℃ for 10 days in a 1-L round-bottomed
Pyrex reactor fitted with a condensation reflux system. The solid aged
product was washed with distilled water. During the aging process, we
found that Si was dissolved from the Pyrex reactor at pH 10 and the
reflux temperature (100 ℃). The dried Si-Y hydroxide was calcined at
ment, the used catalyst was exposed to a 5% O
2
/Ar mixed gas
3
−1
−1
(30 cm min ) and heated to 800 ℃ at a rate of 10 ℃ min . The
effluent gases were analyzed continuously using the QMS. The mass
signals at m/z = 46, 44, 30, and 28 were assigned to NO
2
, CO , NO, and
2
CO or N
2
, respectively.
6
00 ℃ for 6 h in an flowing air. The Si and Y contents in the prepared
2.3. Catalytic reaction test
support SiO
2
-Y
2
O were analyzed by inductively coupled plasma spec-
3
troscopy (ICP, iCAP 6300 Duo). The resulting SiO
hereafter, denoted SY.
2
-Y
2
O
3
sample is,
The reductive amination of ethanol was carried out in a fixed-bed
quartz reactor. Before the reaction, the Ni(X)/SY catalyst was reduced
3
−1
The Ni loading in the Ni/SY catalysts ranged from 5 to 25 wt%, and
the catalysts were synthesized on the SY support by an incipient wet-
ness impregnation method using a nickel nitrate hexahydrate solution
at 500 ℃ for 3 h under H flow (50 cm min ). The standard reaction
2
conditions were as follows: 0.05 g catalyst, T = 200 ℃, weight hourly
−
1
3
−1
space velocity (WHSV) = 3.65 h , total flow rate = 50 cm min , a
(
Ni(NO
3
)
2
2
·6H O, > 98%, SAMCHUN). The impregnated samples were
molar feed composition of EtOH/NH
3
/H
2
= 1/3/6 balanced with N ,
2
then dried at 60 ℃ for 24 h and, subsequently, calcined at 500 ℃ for 2 h
under flowing air. The prepared Ni/SY catalysts are denoted Ni(X)/SY,
where X (X = 5, 10, 15, 20, and 25) represents the Ni loading in weight
percent.
and a fixed EtOH partial pressure of 3 kPa. In addition, the effect of the
reaction parameters such as the partial pressure of ammonia
(3–27 kPa), the partial pressure of H (0–50 kPa), reaction temperature
2
−1
(160–230 ℃), and WHSV (1.82–5.11 h ) controlled by variations in
3
−1
the total flow rate (30–70 cm min ), and catalyst quantity (0.05 and
2