i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 8 9 8 5 e8 9 9 6

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier.com/locate/he

Co-Ni supported yttrium oxide material

as a catalyst for ammonia decomposition
to COx-free hydrogen

Honghai Li a, Lexin Guo a, Jianing Qu a, Xianxin Fang a, Yaqian Fu a,

Jihai Duan a, Weiwen Wang a, Chaojie Li b,*
a
Key Laboratory of Multiphase Fluid Reaction and Separation Engineering of Shandong Province, College of Chemical
Engineering Qingdao University of Science and Technology, Qingdao 266042, China
b
College of Environment and Safety Engineering, Qingdao University of Science and Technology, Qingdao 266042,
China

highlights graphical abstract

Bimetallic catalysts offer better

catalytic properties than mono-
metallic catalysts.

The ammonia decomposition rate
of 20Coe10Ni/Y2O3 was 85.02% at
550  C.

The mesoporous CoeNi/Y2O3 has a
high specific surface area and
metal dispersion.

The catalysts have the benefits of
simple synthesis, good stability
and low price.

article info abstract

Article history: The preparation of CO2-free hydrogen by decomposing ammonia using non-precious metal
Received 11 October 2022 catalysts has received widespread attention. Co-based catalysts have good catalytic ac-
Received in revised form tivity and low synthesis costs. Therefore, we prepared a series of Y2O3-loaded Co and Co
23 November 2022 eNi bimetallic catalysts by the co-precipitation method. The structure-activity relation-
Accepted 29 November 2022 ship of the catalysts during NH3 decomposition was investigated by X-ray diffraction, N2
Available online 21 December 2022 adsorption-desorption, transmission electron microscopy, and Temperature-programmed
reduction by hydrogen. The ammonia decomposition performance of the bimetallic CoeNi
Keywords: was examined to elucidate the effect of the relative amount of the active metal of the
NH3 decomposition catalyst on the catalytic activity and compared it with the monometallic Co/Y2O3. Among
Hydrogen production the tested catalysts, 20Coe10Ni/Y2O3 showed the best catalytic performance as well as
Co-precipitation good stability. The ammonia decomposition rate of the 20Coe10Ni/Y2O3 catalyst was
CoeNi bimetal 85.02% at 550  C and GHSV of 9000 mL h1$gcat1, which was 28.5% higher than that of the

* Corresponding author.
E-mail address: lichaojie@qust.edu.cn (C. Li).
https://doi.org/10.1016/j.ijhydene.2022.11.338
0360-3199/© 2022 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
8986 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 8 9 8 5 e8 9 9 6

Synergistic effect 20Co/Y2O3 catalyst. The high activity of 20Coe10Ni/Y2O3 catalyst was attributed to the
Mesoporous catalysts uniform loading of Co and Ni, high specific surface area, and a large number of mesopores.
At the same time, Co and Ni exhibited synergistic effects to produce suitable adsorption
energy of nitrogen atoms, which further enhanced the catalytic activity of the catalyst.
© 2022 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

the chemical and physical properties of the carrier, such as

Introduction the degree of alkalinity of the carrier and the electrical con-
ductivity, and also to the promoter, so the activity of non-
The use of fossil fuels is involved in all aspects of life. precious metal catalysts can be promoted by changing the
Hydrogen has gained widespread attention due to the envi- catalyst carrier, active component, and adding promoter [31].
ronmental pollution caused by burning fossil fuels and the In ammonia decomposition, the main problem for non-
development of fuel cells in recent years [1,2]. Hydrogen has precious metals is the inappropriate M(metal)-N bond en-
the advantages of high calorific value, comprehensive source ergy [32]. Stronger M  N bonds favor NeH bonds dissociation
and high energy utilization, and no pollution. Still, it also has but not N atoms' desorption from the catalyst surface. While
the disadvantages of low density, high transportation and the weaker M  N favors the desorption of N atoms but not the
storage costs, and insecurity [3]. The search for chemical dissociation of NeH [33]. And the medium strength M  N
hydrogen storage in hydrides or organics has become a critical bond will combine both of these advantages. Developing
approach. Among all hydrogen carriers, ammonia has the bimetallic systems that combine bimetallic catalysts with
following advantages: high mass fraction (17.6 wt%), energy high M  N bond energy (>134 kcal mol1) with low M  N
density exceeding that of fuels such as methanol and gasoline bond energy (<134 kcal mol1) becomes an important way to
[4], ammonia (NH3) also has a much higher hydrogen storage obtain the best catalytic activity.
capacity than other traditional metal oxides, making it a very Yttrium is a rare earth element with a high proportion of
ideal fuel for hydrogen storage; ammonia is easily liquefied rare earth elements, strong alkalinity, and high dehydroge-
and can be stored and transported easily compared to nation activity compared to other rare earth elements [34,35].
hydrogen [5], ammonia gas has a strong irritating odor, Y2O3 is widely used in refractory materials, sensors, micro-
reducing the danger posed by its leakage; the only products of wave magnetic materials, optical glass, etc., because of its
ammonia decomposition are hydrogen and nitrogen, blocking high chemical stability, thermal conductivity, porosity, oxy-
the production of the toxic by-product COx at the source [6]. gen content capacity, and specific surface area [36]. Recent
The intensive research on proton exchange membrane fuel studies have shown that Y2O3-loaded catalysts have good
cells (PEMFC) and their rapid development has led to the catalytic activity and are widely used in the reforming of
widespread interest in the decomposition of ammonia to ethanol-water vapor [34,35], NOx oxidation, methanol oxida-
generate COx hydrogen [7e17]. tion [37] and CO oxidation, organic combustion [38], among
In recent years, significant progress has been made in the other reactions. Y2O3 is used as a carrier to immobilize the
research of ammonia decomposition catalysts, and various active center and to maintain the high dispersion of the metal
catalyst systems have been developed. For example, metal, at high temperatures. The loading of CoeNi bimetal on Y2O3 as
metal carbide, metal nitride [18e21], etc. Among all catalysts, an ammonia decomposition catalyst is less used for ammonia
ruthenium catalysts exhibit better catalytic activity than other decomposition. Therefore, the development is more
noble metals [22]. The carriers for loading ruthenium are promising.
diverse, such as carbon nanotubes (CNTs) [23,24], MgO, Al2O3, This paper prepared a series of catalysts with different
TiO2, etc. The results show that carbon nanotubes have better amounts of cobalt and nickel using Y2O3 as the carrier. XRD
catalytic activity than the catalysts loaded on other carriers. analyzed the structures of the catalysts and the ammonia
However, carbon-containing catalysts are prone to para- decomposition performance, BET, TEM, SEM, TCP, XPS, H2-
methanation in a hydrogen-rich atmosphere, which affects TPR, and other characterization means. The results showed
the stability of the catalyst. Although precious metals have that the structure of the catalyst had a significant influence on
shown excellent catalytic performance in the decomposition the ammonia decomposition performance. Using the hydro-
of ammonia, the low natural reserves of precious metals and thermal method, mesoporous Y2O3 was prepared, effectively
their relatively high prices have limited the large-scale use of protecting cobalt-nickel metal particles' agglomeration at high
catalysts [25]. This makes the development of inexpensive, temperatures. Meanwhile, a series of CoeNi/Y2O3 mesoporous
non-precious metals particularly important. Transition catalysts with good dispersion were successfully prepared by
metals (Fe [26], Co [27], Ni [9,18]) have the advantage of the co-precipitation method, and the ammonia decomposi-
abundant reserves and low prices and can be substituted for tion performance of each catalyst was compared. The excel-
precious metals. Among the non-precious metals, cobalt has lent catalytic performance of the 20Coe10Ni/Y2O3 catalyst
the most excellent catalytic potential compared to Fe and Ni was mainly attributed to the high specific surface area of the
[28,29]. Because it has a weak nitrogen binding energy, the catalyst and the high dispersion of Co and Ni nanoparticles,
activation energy is low at higher ammonia concentrations which facilitated the diffusion of NH3 molecules and the
[30]. In general, cobalt-based catalysts are closely related to active substances in contact, as in Scheme 1.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 8 9 8 5 e8 9 9 6 8987

Scheme 1 e Schematic diagram of the catalyst used for the catalytic decomposition of NH3 to produce hydrogen.

Stir for 30 min, and under vigorous stirring, configure the

Experimental appropriate amount of Co(NO3)2$6H2O and Ni(NO3)2$6H2O into
solution one. Then a mixture of 0.453 mol of K2CO3 and
Material 0.1225 mol of KOH is added dropwise to solution one, adjust the
pH to 11, and stir for 30 min at room temperature. The mixture
All materials are analytically pure and can be used without was reacted at 80  C for 12 h. The precipitate was separated by
further modification or purification. Nickel nitrate hexahy- filtration and washed several times with deionized water until
drate Ni(NO3)2$6H2O (AR, 99%), ammonia (NH3$H2O, 25e28%), pH ¼ 7. The resulting product was dried at 80  C for 8 h. Finally,
potassium hydroxide (KOH, 85%) were purchased from it was reduced to 500  C in a 10% H2/Ar atmosphere for 2 h.
China Shanghai Sinopharm Chemical Reagent Co. Cobalt ni- These samples were named xCo-yNi/Y2O3, where x and y are
trate hexahydrate Co(NO3)2$6H2O (AR, 99%), yttrium nitrate the loadings of Co (wt%) and Ni (wt%), respectively.
hexahydrate (Y(NO3)3$6H2O, 95%) were purchased from
Shanghai Maclean Biochemical Technology Co. Potassium Sample characterization
carbonate (K2CO3, 99%) was purchased from Tianjin Beilian
Fine Chemicals Development Co. Ammonia (NH3, 99.999%), Powder X-ray diffraction (XRD) was performed on an X-ray
hydrogen-argon mixture (H2/Ar, 10%) and nitrogen (N2, diffractometer (Rigaku, D-MAX, 2500/PC). Microstructural
99.99%) were supplied by Qingdao Deyi Gas Co. images of the catalysts were obtained by transmission elec-
tron microscopy (TEM, JEM-2100, Japan) and high-resolution
Catalyst preparation projection electron microscopy (HRTEM), and the catalyst
morphology was characterized using scanning electron mi-
Preparation of Y2O3 carrier croscopy (SEM, SU8010). X-ray photoelectron spectroscopy
Dissolve Y(NO3)3$6H2O in 50 mL of deionized water and adjust (XPS, Thermo 250XI) determined the chemical states of the
the pH value of the solution to 12 with 25e28% of NH3$H2O. A elements in the samples. The cobalt and nickel content of the
white turbid liquid is immediately formed at this point. After catalyst was characterized using inductively coupled plasma
stirring, the mixture is poured into a stainless-steel autoclave (ICP, Agilent ICPMS 7700).
lined with Teflon-lined. The autoclave was sealed and heated The physisorption of nitrogen was performed on an ASAP
at 180  C for 12 h. The product was collected by filtration and 2020 physisorption instrument and BET data were obtained
then rinsed with deionized water until pH ¼ 7. The product for all samples at 77 K The degassing temperature was 200  C
was then dried at 80  C for 12 h. Finally, the Y2O3 carrier was and the degassing time was 6 h. Meanwhile, the reduction of
calcined at 700  C for 2 h under a nitrogen atmosphere [36]. the samples was characterized by hydrogen programmed
temperature reduction (H2-TPR, AutoChemi II 2920) with the
Preparation of CoeNi/Y2O3 catalyst following procedure. First, the sample (50 mg) was purged at
Co(NO3)2$6H2O and Ni(NO3)2$6H2O were used as metal pre- room temperature for 2 h under 10% H2/Ar flow, and the TPR
cursors and loaded on a Y2O3 carrier by the co-precipitation curve was obtained by ramping up from room temperature to
method. Disperse 0.3 g of Y2O3 in 30 mL of deionized water. 800  Cat a ramp-up rate of 10  C/min.
8988 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 8 9 8 5 e8 9 9 6

where [NH3]in, [NH3]out, VNH3 and mcat are the amounts of NH3
in the feed stream, the amount of NH3 in the discharge stream,
the flow rate of NH3, and the mass of the catalyst.

Results and discussion

Structure and structural properties of catalysts

An X-ray diffractometer analyzed the structure of the calcined

catalysts. As shown in Fig. 1, all the catalysts have diffraction
peaks located at 29.4 , 34.0 , 48.8 , and 58.0 , corresponding to
the (111), (200), (220), and (311) crystallographic planes of the
Scheme 2 e Schematic illustration of the reaction setup.
cubic phase of Y2O3 (JCPDS No.43-0661). The diffraction peak
located at 44.2 in the XRD of the 20Co/Y2O3 catalyst can be
Catalytic activity evaluation matched with and the (111) crystallographic plane of face-
centered cubic Co (JCPDS No.15-0806), indicating that the
The NH3 decomposition performance of the catalyst was synthesized 20Co/Y2O3 catalyst was reduced to the metallic
evaluated in an atmospheric pressure stainless steel fixed-bed state before contact with NH3. By comparing the XRD of the
reactor (10 mm inner diameter). 200 mg of catalyst was loaded 20Co/Y2O3 catalyst, only the broadening of the diffraction
into the reactor and treated with N2 (30 mL/min) for 30 min peak of Co was observed in the XRD pattern of the 20Coe10Ni/
before the reaction to expel air from the device. Then the Y2O3 catalyst, and no obvious peaks of the metal Ni or addi-
catalyst was heated from room temperature to 300  C in the N2 tives were observed, which may be due to the conversion of Co
stream. Finally, the gas was switched to pure NH3 to CoeNi alloy after the addition of Ni [39]. The catalysts were
(30e100 mL
min1, corresponding to a gas per hour air velocity also characterized after 27 h of reaction at high temperature.
(GHSV) of 9000e30000 mL
h1$gcat1), and after reaching the From Fig. 1(b), it can be seen that the characteristic peaks of
steady state, the activity of the catalyst was evaluated for the the 20Coe10Ni/Y2O3 fresh catalyst did not change signifi-
decomposition of NH3 at 300e600  C (as shown in Scheme 2). cantly compared with the 20Coe10Ni/Y2O3 catalyst after use,
The ammonia conversion in the carrier gas was analyzed using which indicates that the structure of the catalyst did not
indophenol blue spectrophotometry. The conversion of change before and after the reaction and the catalyst has good
ammonia by NH3 decomposition (XNH3) and H2 formation (rH2) stability.
at different temperatures at a GHSV of 9000 mL$h1gcat1 were The morphology and structure of the catalysts were
calculated as follows: analyzed by scanning electron microscopy (SEM). Among
them, the pictures of scanning electron microscopy of the
½NH3 in  ½NH3 out carrier Y2O3 are shown in Fig. 2(a, b, c). The Y2O3 prepared by
XNH3 ð%Þ ¼  100 (1)
½NH3 in
the hydrothermal method is relatively uniform in size, shows
a rod-like structure, and has a relatively smooth surface. After
VNH3
 
,XNH3 ,1:5 co-precipitation, the pictures of scanning electron microscopy
rH2 mmol gcat min ¼ 22:4
(2)
mcat of the 20Coe10Ni/Y2O3 catalyst are shown in Fig. 2(d, e, f). It

Fig. 1 e XRD patterns of (a) 20Co/Y2O3, 20Coe10Ni/Y2O3;(b) XRD patterns of fresh and used 20Coe10Ni/Y2O3 catalyst.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 8 9 8 5 e8 9 9 6 8989

Fig. 2 e SEM image of 20Co/Y2O3(a, b, c) and 20Coe10Ni/Y2O3(d, e, f).

Fig. 3 e TEM (a, b) and HRTEM (c) images of 20Coe10Ni/Y2O3 catalyst.

can be seen from the figures that the surface of the Y2O3 Co (0.206 nm, JCPDS No.15-0806) and metallic Ni (0.205 nm,
carrier changed significantly with the co-precipitation, and JCPDS No.04-0850), corresponding to the CoeNi alloy (111).
the surface of the catalyst became rough and showed the In order to determine the distribution of Co and Ni metals
accumulation of nanosheets, indicating that the addition of on the catalyst surface, the catalyst was analyzed by energy
cobalt and nickel changed the original morphology of the dispersive X-ray (EDX). From Fig. 4, it can be seen that Co and
catalyst. The same results as obtained by XRD also proved that Ni are uniformly dispersed on the surface of the catalyst,
Co and Ni were successfully loaded on the surface of the while Y and O elements can be detected. This is consistent
catalyst. with the results of XPS, indicating that Co and Ni are uni-
Considering that 20Coe10Ni/Y2O3 has the best NH3 formly dispersed in Y2O3. Also, it can be seen that Co and Ni
decomposition properties, transmission electron microscopy are dispersed at approximately the same position, indicating
(TEM) and high-resolution electron microscopy (HRTEM) that there is some kind of force between Co and Ni and
studies were performed to examine the morphology of Co/Ni proving the generation of CoeNi alloy.
particles on yttrium oxide and their dispersion. From the Fig. 5 shows the samples' N2 adsorption-desorption iso-
transmission electron microscopy images (Fig. 3(a, b)), it can therms and Barrett-Joyner-Halenda (BJH) pore size distribu-
be seen that the Co/Ni particles are uniformly dispersed on the tion. The Brunauer-Emmett-Teller (BET) specific surface area,
Y2O3 carrier. Also, clear lattice stripes were observed from the pore volume, and pore size of the samples are shown in Table 1.
high-resolution electron microscopy images (Fig. 3(c)) with a According to the IUPAC classification, the N2 adsorption-
lattice spacing of 0.205 nm, which is close to that of metallic desorption isotherm curves of Y2O3 and 20Coe10Ni/Y2O3
8990 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 8 9 8 5 e8 9 9 6

Fig. 4 e EDS elemental mapping of 20Coe10Ni/Y2O3.

Fig. 5 e N2 adsorption-desorption isothermal curves (a) and BJH pore size distribution (b) for 20Coe10Ni/Y2O3 and Y2O3
catalysts.

Table 1 e Physical properties of Y2O3 and 20Coe10Ni/Y2O3.

Samples N2 adsorption-desorptiona ICP-OESb
2 3
SBET (m /g) Pore volume (cm /g) Pore diameter (nm) Co content (wt%) Ni content (wt%)
Y2O3 26 0.17 33.0 e e
20Coe10Ni/Y2O3 58 0.46 29.5 12.91 6.53
a
The specific surface area (SBET), pore volume, and average pore size of the catalysts were determined by N2 adsorption-desorption analysis.
b
The content of Ni and Co in the samples was determined using ICP-OES.

(Fig. 5(a)) belong to type IV and H3 hysteresis loops, indicating contact area for reactant molecules and promote the adsorp-
that all samples have slit-like pores with wide openings. The tion and desorption of reactant molecules, as well as promote
hysteresis curves were observed in the relative pressure P/P0 the reaction between reactants and catalysts and the dissoci-
range of 0.8e1.0, indicating that 20Coe10Ni/Y2O3 formed a ation of products, increasing the mass transfer efficiency
distinct mesoporous structure. The pore size distribution dia- [33,40,41]. As shown in Table 1, the specific surface area of the
gram (Fig. 5(b)) also proves that the pore size of the catalyst is 20Coe10Ni/Y2O3 catalyst prepared using the co-precipitation
mainly composed of microporous and mesoporous. The mes- method was 58 m2/g, and the pore volume was 0.46 cm3/g,
oporous catalysts have a large specific surface area, relatively and the average pore size was 29.5 nm. After comparison with
large pore size, and regular pore structure, which provide more the carrier Y2O3, the specific surface area and pore volume
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 8 9 8 5 e8 9 9 6 8991

Fig. 6 e XPS analysis data of 20Coe10Ni/Y2O3 catalyst - (a) XPS full spectrum, (b) Co 2p, (c) Ni 2p, (d) O 1s and (e) Y 3d.

increased significantly, which was attributed to the tendency attributed to the spin-orbit peaks of the spinel of Co3O4 [43],
of the metal CoeNi to adhere to the catalyst at high tempera- generated as a result of the oxidation of Co0 nanoparticles in
tures and thus form pores on the catalyst surface. air. Ni with the metallic state of nickel oxide [42]. The peaks
Fig. 6 shows the XPS spectrum of the calcined 20Coe10Ni/ are located at 861.2 eV and 879.3 eV are attributed to satellite
Y2O3 catalyst, and it can be seen from Fig. 6(a) that the catalyst peaks from the oscillation process. The formation of nickel
contains Co, Ni, O, and Y. The Co 2p1/2 peak at 794.7 eV in oxide should be similar to that of O, as a result of rapid
Fig. 6(b) is indicated as Co0 [42], the peaks at 803 eV and oxidation during sample transfer. O 1s spectrum (Fig. 6(d)) is
786.2 eV are attributed to the satellite peaks (labeled as “sat”) fitted to three peaks, 532.3 eV, 531.5 eV, and 529.2 eV. The peak
from the oscillation process. The peaks at 780.2 eV and at 529.2 eV is attributed to lattice oxygen in Y2O3 [44], 531.5 eV
796.5 eV for Co 2p are attributed to Co 2p3/2 and Co 2p1/2 peaks, is attributed to oxygen adsorbed on the catalyst surface, and
corresponding to Co3þ and Co2þ. These seals may be 532.3 eV is attributed to oxygen ions in carbonate or
8992 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 8 9 8 5 e8 9 9 6

carboxylate groups. Fig. 6(e) shows the Y 3d XPS spectra, ammonia decomposition performance of the CoeNi bime-
where 156.4 eV and 158.2 eV belong to the binding energies of tallic catalyst was significantly higher than that of the single-
Y 3d5/2 and Y 3d3/2 of Y 3d [45]. metal catalyst compared to the catalyst without Ni, indicating
a synergistic effect between Co and Ni. The high activity of the
Catalytic performance of the catalyst for ammonia bimetal is attributed to the simultaneous reduction of Co and
decomposition Ni to form a CoeNi alloy [46], which results in a synergistic
effect between CoeNi, resulting in a suitable metal (M)-N
The catalytic performance of monometallic catalysts with binding energy that facilitates the dissociation and desorption
bimetallic catalysts for NH3 decomposition was investigated of N2.
at a vacancy rate of 9000 mL
h1$gcat1. The test results are Fig. 7(b) examines the variation of hydrogen yield of each
shown in Fig. 7(a), and all the results were obtained in a steady catalyst at GHSV ¼ 9000 mL
h1$gcat1, from which it can be
state. From Fig. 7(a), it can be seen that for all synthesized seen that the hydrogen yield gradually increases with the in-
catalysts, the conversion of NH3 increases with the increase of crease of reaction temperature, where the catalyst with the
reaction temperature, which is consistent with the ammonia highest hydrogen yield is 20Coe10Ni/Y2O3.It can be seen from
decomposition reaction being a heat-absorbing reaction. It the figure that the ammonia decomposition performance of
can be seen from the figure that the ammonia decomposition the catalysts is affected by the CoeNi The order of hydrogen
activity is in the order of: 20Co/Y2O3 < 30Coe10Ni/Y2O3 < production was 20Coe10Ni/Y2O3 > 30Coe10Ni/Y2O3 > 10-
10Coe10Ni/Y2O3 < 20Coe5Ni/Y2O3 < 20Coe10Ni/Y2O3. with Coe10Ni/Y2O3 > 20Co/Y2O3. at 550  C, the hydrogen yield of
the increase of cobalt content, the catalytic effect of the 20Coe10Ni/Y2O3 could reach 51.2 mmol
g1
h1.
catalyst increases, and the 20Coe10Ni/Y2O3 catalyst has the Ammonia decomposition is a primary reaction, and also
highest catalytic activity with 85.02% conversion at 550  C. ammonia decomposition is a function of reaction tempera-
However, the catalyst activity decreased a further increase of ture. As the reaction temperature increases, the rate of
cobalt, so the appropriate amount of Co with Ni was beneficial ammonia decomposition increases. According to the Arrhe-
to the catalyst performance. 20Co/Y2O3 had the lowest cata- nius equation, the plot of log rate against 1/T can produce a
lytic activity with 56.55% ammonia conversion at 550  C. The straight line as shown in equation (3). Thus the apparent

Fig. 7 e (a) NH3 decomposition activity versus hydrogen yield of individual catalysts in the temperature range of
300e600  C (b) and (c) Arrhenius plots showing activation energy of catalysts and (d) durability experiments of 20Coe10Ni/
Y2O3 at 550  C for 30 h at GHSV ¼ mL·h¡1·gcat¡1.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 8 9 8 5 e8 9 9 6 8993

Fig. 9 e H2-TPR curves of 20Co/Y2O3 and 20Coe10Ni/Y2O3

Fig. 8 e NH3 decomposition properties on 20Coe10Ni/Y2O3 catalysts.
under different GHSV conditions.

activation energy can be determined by fitting this equation to conversion of the catalyst at 550  C and 530  C with 27 h
the rate of hydrogen decomposition as a function of reaction exposure to ammonia, which proved the good stability of
temperature. 20Coe10Ni/Y2O3 at high temperatures. The comparison of the
used catalyst with the calcined catalyst by XRD (Fig. 7(b)) also
Ea 1 proved the good stability of the catalyst.
lnðkÞ ¼  þ constant (3)
R T The 20Coe10Ni/Y2O3 catalyst has the best catalytic perfor-
The kinetics of CoeNi catalysts with different loading mance, and its decomposition performance on the 20Coe10Ni/
amounts were investigated at different reaction tempera- Y2O3 catalyst was investigated at different GHSVs. From Fig. 8,
tures. Fig. 7(c) shows the Arrhenius plot of hydrogen genera- it can be seen that the 20Coe10Ni/Y2O3 catalyst can reach
tion rate with temperature, and the apparent activation 96.61% NH3 conversion at a GHSV ¼ 9000 mL
h1$g1 cat and
energies were calculated to be 67 kJ/mol, 75 kJ/mol, 76 kJ/mol, temperature of 600  C. The NH3 conversion decreases with the
and 80 kJ/mol for 20Coe10Ni/Y2O3, 30Coe10Ni/Y2O3, increase of GHSV at the same temperature, which means that
10Coe10Ni/Y2O3, and 20Co/Y2O3. The Ea value of 80 kJ/mol for the low air velocity is favorable to the decomposition of NH3.
20Co/Y2O3 is similar to that of Co/CNTs reported in the liter- Meanwhile, Table 2 presented the 20Coe10Ni/Y2O3 catalyst
ature [47], so the active site of 20Co/Y2O3 may be similar to activity in comparison with other catalysts published in the
that of Co/CNTs from metallic cobalt nanoparticles, At the literature. In this work, the 20Coe10Ni/Y2O3 catalyst has a
same time, Y2O3 has the same support as carbon nanotubes. better conversion rate at 550  C compared to other catalysts.
The large Ea values observed on the 20Co/Y2O3 catalysts
suggest that the activation energy of NH3 on pure Co catalysts Redox properties of the catalyst
is more difficult, while the addition of Ni reduces the activa-
tion energy of the catalysts. H2-TPR is commonly used to detect the redox ability of the
In addition to the activity of the catalyst, the stability of the active component and the interaction between the metal and
operation at high temperatures is of great importance for the the carrier. Fig. 9 shows the H2-TPR curves for different cata-
practical application of the catalyst. Therefore, 20Coe10Ni/ lysts. These catalysts were not subjected to a reduction step in
Y2O3 was selected for testing in stability tests. As Fig. 7(d) a tube furnace. According to previous literature [55], the TPR
shows the stability tests of 20Coe10Ni/Y2O3 catalyst at 550  C spectra of Co-based catalysts consist of two main processes:
and 530  C. There was no significant decrease in NH3 (1) reduction peaks at low temperatures (255e343  C),

Table 2 e Catalytic activities of different catalysts for NH3 decomposition.

Catalysts Metal loadin5g (wt%) NH3 GHSV (ml/gcat/h) Temperature ( C) XNH3 (%) Ref
Ni/Al2O3 8.9 9000 500 27.0 [48]
Ni/MWCNT 6 6000 500 57.6 [49]
Ni_MgAl(6:1) 15 30000 550 48 [50]
5Ni/5 MgLa 10 30000 550 54 [51]
Ni-10/ATP 8.2 30000 650 64.3 [52]
15-Ni/MRM-600 15 30000 600 55 [53]
Fe/MWCNTs 5 36000 600 ~45 [54]
20Coe10Ni/Y2O3 30 9000 550 85.02 This study
20Coe10Ni/Y2O3 30 30000 550 71.24 This study
8994 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 8 9 8 5 e8 9 9 6

attributed to the reduction of Co3O4 to CoO. (2) reduction and a series of CoeNi catalysts with different amounts of co-
peaks at high temperatures (343e650  C), attributed to the balt and nickel were prepared by co-precipitation. The cata-
reduction of CoO to the metal Co. In the TPR plots of 20Co/Y2O3 lysts were investigated for NH3 decomposition experiments,
and 20Coe10Ni/Y2O3, the shoulder peaks observed at 311 in which the 20Coe10Ni/Y2O3 catalyst had the best catalytic

Cand 281  C correspond to the reduction of Co3þ to Co2þ, performance as well as stability. The ammonia decomposition
while the main peaks at 379  Cand 390  C can be attributed to rate of 20Coe10Ni/Y2O3 was 85.02% at 550  C and GHSV of
the reduction of Co2þ to Co0. With the addition of Ni, the 9000 mL
h1$gcat1, which was 28.5% higher than that of the
reduction peaks that can be observed at 449  C and 506  C can 20Co/Y2O3 catalyst. The high catalytic activity of the
be attributed to the process of reduction of Ni2þ to Nidþ and 20Coe10Ni/Y2O3 catalyst was attributed to the large specific
Nidþ to Ni0. Meanwhile, the high temperature reduction peaks surface area and a large number of mesopores of the catalyst.
of the 20Coe10Ni/Y2O3 catalyst move to higher temperatures, It was also found that there was a strong interaction between
which indicates a strong interaction between mesoporous Co and Ni as well as a strong synergistic effect between CoeNi
yttrium oxide and CoeNi, and it is this strong metal-carrier and the carrier, which improved the catalyst activity. The re-
interaction that effectively prevents the self-aggregation of sults obtained from the experimental data help to extend the
the active species and thus improves the catalytic perfor- application of highly active and thermally stable catalysts in
mance of the catalyst. the ammonia decomposition to COx-free hydrogen.
The mechanism of ammonia decomposition can be divided
into the following steps [33,56]: (1) ammonia adsorption at the
active center of the catalyst; (2) dehydrogenation of adsorbed Declaration of competing interest
ammonia to produce nitrogen and hydrogen atoms; (3)
desorption and recombination of nitrogen and hydrogen The authors declare that they have no known competing
atoms to produce hydrogen and nitrogen, as shown below: financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
NH3 þ * /NH3;ad (4)

NH3;ad þ * / NH2;ad * þ H* (5)

Acknowledgements
* * * *
NH2;ad þ / NH;ad þ H (6)
This work is supported by Natural Science Foundation of
* * * *
Shandong Province [grant number ZR2021QB144].
NH;ad þ / Nad þ H (7)

Nad * þ Nad * / N2 þ 2* (8) references

H* þ H* / H2 þ 2* (9)
[1] Bohn CD, Müller CR, Cleeton JP, Hayhurst AN, Davidson JF,
Since N is present in the interstitial position after the NH3
Scott SA, Dennis JS. Production of very pure hydrogen with
dehydrogenation step, the magnitude of the association simultaneous capture of carbon dioxide using the redox
desorption energy barrier of N is critical and is the key step in reactions of iron oxides in packed beds. Ind Eng Chem Res
determining the rate of the ammonia decomposition reaction. 2008;47:7623e30.
And the bond strength between the active metal (M) and N is [2] Qiu Y, Ma L, Li M, Cui D, Zhang S, Zeng D, Xiao R. Copper and
one of the key factors affecting ammonia decomposition. If the cobalt co-doped ferrites as effective agents for chemical
looping CO2 splitting. Chem Eng J 2020;387.
bond energy of M  N is too strong, it is favorable for the
[3] Ma Y, Wang XR, Li T, Zhang J, Gao J, Sun ZY. Hydrogen and
adsorption of ammonia but not for the dissociation of N. On the ethanol: production, storage, and transportation. Int J
contrary, if the bond energy of M  N is too weak, it is unfa- Hydrogen Energy 2021;46:27330e48.
vorable for the adsorption of ammonia and the full reaction of [4] El-Shafie M, Kambara S, Hayakawa Y. Plasma-enhanced
the catalyst with ammonia. Therefore, it is particularly catalytic ammonia decomposition over ruthenium (Ru/Al2O3)
important to find medium M  N bond energy. The generation and soda glass (SiO2) materials. J Energy Inst 2021;99:145e53.
[5] Cechetto V, Di Felice L, Medrano JA, Makhloufi C, Zuniga J,
of CoeNi in the catalyst was seen from XRD, XPS, and TEM
Gallucci F. H2 production via ammonia decomposition in a
characterization. Wu et al. [39] characterized using NH3-TPD,
catalytic membrane reactor. Fuel Process Technol 2021:216.
N2-TPD, etc., and concluded that CoeNi/SiO2 has moderate [6] Yin SF, Xu BQ, Zhou XP, Au CT. A mini-review on ammonia
M  N bonding energy compared to Ni/SiO2 compared to Co/ decomposition catalysts for on-site generation of hydrogen
SiO2. Also, the results of ammonia decomposition activity for fuel cell applications. Appl Catal Gen 2004;277:1e9.
indicate that the moderately strong M  N bond energy of [7] Yu X, Cheng L, Liu Y, Manthiram A. A membraneless direct
CoeNi catapults the ammonia decomposition reaction. isopropanol fuel cell (DIPAFC) operated with a catalyst-
selective principle. J Phys Chem C 2018;122:13558e63.
[8] Idem RO, Bakhshi NN. Production of hydrogen from
methanol. Part 1. Catalyst characterization studies. Ind Eng
Conclusion Chem Res 1994;33:2047e55.
[9] Chellappa AS, Fischer CM, Thomson WJ. Ammonia
In conclusion, mesoporous Y2O3 with homogeneous decomposition kinetics over Ni-Pt/Al2O3 for PEM fuel cell
morphology was prepared by a simple hydrothermal method, applications. Appl Catal Gen 2002;227:231e40.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 8 9 8 5 e8 9 9 6 8995

[10] Choudhary TV, Goodman DW. Stepwise methane steam table: bimetallic ammonia synthesis catalysts. J Am Chem
reforming: a route to CO-free hydrogen. Catal Lett Soc 2001;123:8404e5.
1999;59:93e4. [30] Garcı́a-Bordeje E, Armenise S, Rolda  n L. Toward practical
[11] Choudhary VR, Uphade BS, Mamman AS. Oxidative application of H2 generation from ammonia decomposition
conversion of methane to syngas over nickel supported on guided by rational catalyst design. Catal Rev
commercial low surface area porous catalyst carriers 2014;56:220e37.
precoated with alkaline and rare earth oxides. J Catal [31] Morlane s N, Sayas S, Shterk G, Katikaneni SP, Harale A,
1997;172:281e93. Solami B, Gascon J. Development of a BaeCoCe catalyst for
[12] Armor JN. The multiple roles for catalysis in the production the efficient and stable decomposition of ammonia. Catal Sci
of H2. Appl Catal Gen 1999;176:159e76. Technol 2021;11:3014e24.
[13] Choudhary TV, Sivadinarayana C, Goodman DW. Catalytic [32] Yi Y, Wang L, Guo Y, Sun S, Guo H. Plasma-Assisted
ammonia decomposition: COx-free hydrogen production for ammonia decomposition over FeeNi alloy catalysts for COx-
fuel cell applications. Catal Lett 2001;72:197e201. Free hydrogen. AIChE J 2019;65:691-701..
[14] Abashar MEE, Al-Sughair YS, Al-Mutaz IS. Investigation of [33] Huo L, Liu B, Li H, Cao B, Hu X-c, Fu X-p, Jia C, Zhang J.
low temperature decomposition of ammonia using spatially Component synergy and armor protection induced superior
patterned catalytic membrane reactors. Appl Catal Gen catalytic activity and stability of ultrathin Co-Fe spinel
2002;236:35e53. nanosheets confined in mesoporous silica shells for
[15] Yin S-F, Xu B-Q, Ng C-F, Au C-T. Nano Ru/CNTs: a highly ammonia decomposition reaction. Appl Catal B Environ
active and stable catalyst for the generation of CO -free 2019;253:121e30.
hydrogen in ammonia decomposition. Appl Catal B Environ [34] Sun J, Qiu X, Wu F, Zhu W. From steam reforming of ethanol
2004;48:237e41. at low temperature over, and catalysts for fuel-cell
[16] Yin S-F, Zhang Q-H, Xu B-Q, Zhu W-X, Ng C-F, Au C-T. application. Int J Hydrogen Energy 2005;30:437e45.
Investigation on the catalysis of COx-free hydrogen [35] Sun J. Hydrogen from steam reforming of ethanol in low and
generation from ammonia. J Catal 2004;224:384e96. middle temperature range for fuel cell application. Int J
[17] Yin SF, Xu BQ, Wang SJ, Ng CF, Au CT. Magnesiaecarbon Hydrogen Energy 2004;29:1075e81.
nanotubes (MgOeCNTs) nanocomposite: novel support of Ru [36] Feng J, Liu L, Ju X, Wang J, Zhang X, He T, Chen P. Highly
catalyst for the generation of COx-free hydrogen from dispersed ruthenium nanoparticles on Y2O3 as superior
ammonia. Catal Lett 2004;96:113e6. catalyst for ammonia decomposition. ChemCatChem
[18] Liang C, Li W, Wei Z, Xin Q, Li C. Catalytic decomposition of 2021;13:1552e8.
ammonia over nitrided MoNx/a-Al2O3 and NiMoNy/a-Al2O3 [37] Chen X, Cai H, Tang Q, He B, Chen H. Eggshell membrane
catalysts. Ind Eng Chem Res 2000;39:3694e7. templated Y2O3@Pd catalyst for enhanced methanol
[19] Guo J, Wang P, Wu G, Wu A, Hu D, Xiong Z, Wang J, Yu P, oxidation and CO tolerance. Mater Res Innovat
Chang F, Chen Z, Chen P. Lithium imide synergy with 3d 2013;18:300e6.
transition-metal nitrides leading to unprecedented catalytic [38] Sreethawong T, Sitthiwechvijit N, Rattanachatchai A,
activities for ammonia decomposition. Angew Chem Int Ed Ouraipryvan P, Schwank JW, Chavadej S. Preparation of Au/
Engl 2015;54:2950e4. Y2O3 and Au/NiO catalysts by co-precipitation and their
[20] Guo J, Chang F, Wang P, Hu D, Yu P, Wu G, Xiong Z, Chen P. oxidation activities. Mater Chem Phys 2011;126:212e9.
Highly active MnN-Li2NH composite catalyst for producing [39] Wu Z-W, Li X, Qin Y-H, Deng L, Wang C-W, Jiang X. Ammonia
COx-free hydrogen. ACS Catal 2015;5:2708e13. decomposition over SiO2-supported NieCo bimetallic
[21] Lorenzut B, Montini T, Bevilacqua M, Fornasiero P. FeMo- catalyst for COx-free hydrogen generation. Int J Hydrogen
based catalysts for H2 production by NH3 decomposition. Energy 2020;45:15263e9.
Appl Catal B Environ 2012;125:409e17. [40] Li G, Zhang H, Yu X, Lei Z, Yin F, He X. Highly efficient Co/NC
[22] Huang C, Yu Y, Yang J, Yan Y, Wang D, Hu F, Wang X, catalyst derived from ZIF-67 for hydrogen generation
Zhang R, Feng G. Ru/La2O3 catalyst for ammonia through ammonia decomposition. Int J Hydrogen Energy
decomposition to hydrogen. Appl Surf Sci 2019;476:928e36. 2022;47:12882e92.
[23] Wang S. A smooth surface interpolation to 3D triangulations. [41] Ju X, Liu L, Yu P, Guo J, Zhang X, He T, Wu G, Chen P.
J Comput Appl Math 2004;163:287e93. Mesoporous Ru/MgO prepared by a deposition-precipitation
[24] Hu Z, Mahin J, Datta S, Bell TE, Torrente-Murciano L. Ru- method as highly active catalyst for producing COx-free
based catalysts for H2 production from ammonia: effect of 1D hydrogen from ammonia decomposition. Appl Catal B
support. Top Catal 2018;62:1169e77. Environ 2017;211:167e75.
[25] Chen C, Wu K, Ren H, Zhou C, Luo Y, Lin L, Au C, Jiang L. Ru- [42] Zhao L, Wei Q, Zhang L, Zhao Y, Zhang B. NiCo alloy
based catalysts for ammonia decomposition: a mini-review. decorated on porous N-doped carbon derived from ZnCo-ZIF
Energy Fuels 2021;35:11693e706. as highly efficient and magnetically recyclable catalyst for
[26] Zhang H, Gong Q, Ren S, Arshid MA, Chu W, Chen C. hydrogen evolution from ammonia borane. Renew Energy
Implication of iron nitride species to enhance the catalytic 2021;173:273e82.
activity and stability of carbon nanotubes supported Fe [43] Razzaq R, Zhu H, Jiang L, Muhammad U, Li C, Zhang S.
catalysts for carbon-free hydrogen production via low- Catalytic methanation of CO and CO2 in coke oven gas over
temperature ammonia decomposition. Catal Sci Technol NieCo/ZrO2eCeO2. Ind Eng Chem Res 2013;52:2247e56.
2018;8:907e15. [44] Le TA, Kim Y, Kim HW, Lee S-U, Kim J-R, Kim T-W, Lee Y-J,
[27] Huang C, Yu Y, Tang X, Liu Z, Zhang J, Ye C, Ye Y, Zhang R. Chae H-J. Ru-supported lanthania-ceria composite as an
Hydrogen generation by ammonia decomposition over Co/ efficient catalyst for COx-free H2 production from ammonia
CeO2 catalyst: influence of support morphologies. Appl Surf decomposition. Appl Catal B Environ 2021;285.
Sci 2020:532. [45] Huang X, Xue G, Wang C, Zhao N, Sun N, Wei W, Sun Y.
[28] Boisen A, Dahl S, Norskov J, Christensen C. Why the optimal Highly stable mesoporous NiOeY2O3eAl2O3 catalysts for CO2
ammonia synthesis catalyst is not the optimal ammonia reforming of methane: effect of Ni embedding and Y2O3
decomposition catalyst. J Catal 2005;230:309e12. promotion. Catal Sci Technol 2016;6:449e59.
[29] Jacobsen CJH, Dahl S, Clausen BS, Bahn S, Logadottir A, [46] Xiang J, Wen X, Zhang F. Supported nickelecobalt bimetallic
Nørskov JK. Catalyst design by interpolation in the periodic catalysts derived from layered double hydroxide precursors
8996 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 8 ( 2 0 2 3 ) 8 9 8 5 e8 9 9 6

for selective hydrogenation of pyrolysis gasoline. Ind Eng [52] Li L, Chen F, Shao J, Dai Y, Ding J, Tang Z. Attapulgite clay
Chem Res 2014;53:15600e10. supported Ni nanoparticles encapsulated by porous silica:
[47] Qiu Y, Fu E, Gong F, Xiao R. Catalyst support effect on thermally stable catalysts for ammonia decomposition to
ammonia decomposition over Ni/MgAl2O4 towards hydrogen COx free hydrogen. Int J Hydrogen Energy 2016;41:21157e65.
production. Int J Hydrogen Energy 2022;47:5044e52. [53] Cao J-L, Yan Z-L, Deng Q-F, Yuan Z-Y, Wang Y, Sun G,
[48] Sima D, Wu H, Tian K, Xie S, Foo JJ, Li S, Wang D, Ye Y, Wang X-D, Hari B, Zhang Z-Y. Homogeneous precipitation
Zheng Z, Liu Y-Q. Enhanced low temperature catalytic activity method preparation of modified red mud supported Ni
of Ni/AleCe0.8Zr0.2O2 for hydrogen production from ammonia mesoporous catalysts for ammonia decomposition. Catal Sci
decomposition. Int J Hydrogen Energy 2020;45:9342e52. Technol 2014;4:361e8.
[49] Zhang H, Alhamed YA, Kojima Y, Al-Zahrani AA, Miyaoka H, [54] Zhang J, Müller J-O, Zheng W, Wang D, Su D, Schlӧgl R.
Petrov LA. Structure and catalytic properties of Ni/MWCNTs Individual Fe-Co alloy nanoparticles on carbon nanotubes:
and Ni/AC catalysts for hydrogen production via ammonia structural and catalytic properties. Nano Lett
decomposition. Int J Hydrogen Energy 2014;39:277e87. 2008;8:2738e43.
[50] Sato K, Abe N, Kawagoe T, Miyahara S-i, Honda K, Nagaoka K. [55] Zhang H, Alhamed YA, Chu W, Ye Z, AlZahrani A, Petrov L.
Supported Ni catalysts prepared from hydrotalcite-like Controlling Co-support interaction in Co/MWCNTs catalysts
compounds for the production of hydrogen by ammonia and catalytic performance for hydrogen production via NH3
decomposition. Int J Hydrogen Energy 2017;42:6610e7. decomposition. Appl Catal Gen 2013;464e465:156e64.
[51] Yu Y, Gan Y-M, Huang C, Lu Z-H, Wang X, Zhang R, Feng G. [56] Zhong JQ, Zhou X, Yuan K, Wright CA, Tadich A, Qi D, Li HX,
Ni/La2O3 and Ni/MgOeLa2O3 catalysts for the decomposition Wu K, Xu GQ, Chen W. Probing the effect of the Pt-Ni-Pt(111)
of NH3 into hydrogen. Int J Hydrogen Energy bimetallic surface electronic structures on the ammonia
2020;45:16528e39. decomposition reaction. Nanoscale 2017;9:666e72.