1. Introduction
In recent years, cerium dioxide (CeO
2), as a typical rare earth metal oxide, has received widespread attention due to its excellent performance in photocatalysis [
1,
2,
3], fuel cells [
4,
5,
6,
7], sensors [
8,
9,
10,
11], CO oxidation [
12,
13,
14,
15,
16], water gas shift reaction [
17,
18,
19,
20], and other fields [
21,
22,
23]. This is mainly attributed to its two important characteristics. Firstly, in the CeO
2 lattice, rapid conversion of two valence states (Ce
4+/Ce
3+) is achieved through the formation/loss of oxygen vacancies, thus CeO
2 has excellent redox ability [
24]. Secondly, CeO
2 has a cubic fluorite structure containing many oxygen vacancies, which are beneficial for improving oxygen mobility [
25,
26].
CeO
2 is an n-type semiconductor material with oxygen vacancies sites. The Kroger- Vink formula is expressed as 2CeO
2=2Ce'
Ce+V¨
O+3O
Ox+1/2O
2 ↑, where Ce'
Ce represents the presence of one-unit negative charge at the Ce
4+ position, V¨
O represents an oxygen vacancy with a two-units positive charge, and O
Ox represents the oxygen atom on the CeO
2 lattice site. The presence of oxygen vacancies generates Ce
3+, therefore CeO
2 has a high lattice ion mobility and excellent oxygen storage and release ability and is used as a catalyst in various fields [
27].
CeO
2 nanocrystals typically expose low index crystal planes (111), (110), and (100) [
28]. Theoretical calculations indicate that the (110) plane has the lowest vacancy formation energy of 1.99 eV, the (100) plane has a vacancy formation energy of 2.27 eV, and the most stable (111) plane has a maximum vacancy formation energy of 2.60 eV. Consequently, the order of the formation energy of oxygen vacancies on different crystal planes of CeO
2 is: (110)<(100)<(111) [
29]. Therefore, the formation of oxygen vacancies on the CeO
2 (110) crystal plane is easier. CeO
2 has different exposed crystal planes based on its morphology. Polyhedral CeO
2 mainly exposes (111) crystal planes, while cubic CeO
2 mainly exposes (100) crystal planes and rod-shaped CeO
2 mainly simultaneously exposes (110) and (100) crystal planes [
28]. It is possible to regulate the morphology of CeO
2 to alter oxygen vacancies. Yuan investigated that the effect of CeO
2 morphology on the catalytic activity of nitrobenzene hydrogen transfer reduction reaction. Oxygen vacancies and basic sites can selectively activate ethanol molecules to reduce nitro groups. The catalytic activity is sorted in the order: CeO
2 nanorods, CeO
2 nanopolyhedrons, and CeO
2 nanocubes [
30].
Doping different metals, such as precious metals, transition metals, alkali metals, and rare earth metals, into the CeO
2 lattice can improve its catalytic activity and stability [
31,
32]. Through metal doping, lattice distortion can be induced, resulting in abundant oxygen vacancies and Ce
3+, which are widely used in catalysis [
33]. Transition metal catalysts have received considerable attention due to their low cost and excellent activity. Researchers have found that the substitution of Ce
4+ with transition metals can significantly alter the geometric and electronic structures of CeO
2 systems, leading to the reappearance of enriched electronic regions in CeO
2 and weakening of Ce-O or M-O bonds [
34,
35]. The results indicate that the atomic radius has a significant impact on the structure of doped CeO
2, and radii larger or smaller than those of Ce
4+ ions usually produce significant geometric distortions. La doped CeO
2 nanorod shows significantly higher H
2 production compared to CeO
2 in photocatalytic reaction [
36]. We selected the fourth period transition metal element Ni, which has the advantage of low cost, as the doping element to investigate its effect on oxygen vacancies in CeO
2.
In this article, we control the concentration of oxygen vacancies in CeO2 by combining morphology control and metal ion doping, thereby affecting its catalytic activity. This study prepared a series of Ni-doped CeO2 nanorods using the hydrothermal method. In addition, the structural, thermal stability, changes in oxygen vacancy concentration between CeO2 and the doped CeO2 nanorods were thoroughly analyzed using a combination of X-ray power diffusion (XRD), transmission electron microscopy (TEM), thermogravimetry analysis (TGA), and Raman spectroscopy.
3. Results and Discussion
The phase structure of the products was characterized by X-ray powder diffraction (XRD), as shown in
Figure 1. The characteristic diffraction peaks of pure CeO
2 samples appear at 28.5°, 33.1°, 47.5°, 56.3°, 59.1°, 69.4°, 76.7°, and 79.1°, corresponding to the cubic fluorite crystal planes of CeO
2 for (111), (200), (220), (311), (222), (400), (331), and (420), respectively. As the Ni doping amount increases, the width of the diffraction peaks also increases, indicating a decrease in crystallite size with increasing Ni doping amount (
Table 1.). Additionally, upon comparison with the (111) peak of pure CeO
2, the peaks of the doped samples are observed to shift to lower angles. This phenomenon occurs due to the replacement of cerium sites in the cerium oxide lattice by Ni
2+, with a radius of 0.63 Å, resulting in the observed shift of the cerium oxide peaks [
37,
38]. This shift indicates the successful doping of metallic nickel into the cerium oxide lattice. According to the Bragg equation, 2
dsin
θ =
nλ, after doping with nickel metal, the decrease in
θ value leads to a slight increase in the interplanar spacing of the crystal cell, causing slight lattice expansion [
39]. However, this does not significantly affect the crystal cell structure of ceria itself. When the doping amount is increased to 15%, a new diffraction peak appears near 38°, corresponding to the standard diffraction peak of Ni
2O
3. This indicates the generation of Ni
2O
3 from some undoped metal Ni. However, no diffraction peak of Ni
2O
3 is observed at doping levels of 1% and 5%, suggesting that nickel can be completely doped into CeO
2 at low doping levels.
Figure 2(a–d) presents TEM images of CeO
2 and Ni-doped CeO
2. It is evident that rod-shaped CeO
2 nanomaterials have been successfully synthesized. As depicted in
Figure 2(b), even with a 1% doping amount, the rod-shaped morphology of CeO
2 nanomaterials remains unchanged. With a doping amount of 5% (
Figure 2(c)), the rod-shaped structure becomes wider. In
Figure 2(d), with a doping amount of 15%, although the structure remains rod-shaped, the length of the rods significantly decreases while the width increases.
Figure 3 shows the thermogravimetric analysis of undoped CeO
2 and Ni-doped CeO
2 nanorods. Up to about 150 °C, the weight loss due to the loss of surface moisture can be ignored. Therefore, the weight loss above 150 °C can be attributed to the loss of surface or lattice oxygen. It can be observed that the pattern of weight loss changes around 320 °C, indicating different energy release mechanisms between surface and lattice oxygen [
40,
41]. By analyzing the weight loss in each temperature range, the weight loss due to surface oxygen release was approximately 2 wt% for pure CeO
2, 5 wt% for 1% Ni-CeO
2, 4 wt% for 5% Ni-CeO
2, and 3 wt% for 15% Ni-CeO
2. On the other hand, the weight loss due to lattice oxygen release was approximately 1 wt% for pure CeO
2, 5 wt% for 1% Ni-CeO
2, 2 wt% for 5% Ni-CeO
2, and 1 wt% for 15% Ni-CeO
2. The weight loss rate after Ni doping was higher than that without doping, except for 1% Ni-CeO
2, and generally decreased with increasing doping level (
Table 2.). This may be attributed to the non-equilibrium reactions leading to the formation of NiO
x when Ni is added at high concentrations [
42]. Furthermore, as the doping effect increases, the activation of surface oxygen also increases, leading to a greater impact on weight loss at lower temperatures. In other words, in the temperature range below 800 °C, the influence of surface oxygen release becomes more significant with increasing doping. Calculating the ratio of surface oxygen to lattice oxygen, it increased to 1 for 1% Ni-CeO
2, 2 for 5% Ni-CeO
2, and 3 for 15% Ni-CeO
2.
Figure 4, the Raman spectra provide the following insights: both undoped and Ni-doped CeO
2 samples exhibit characteristic peaks around 462 cm
–1, originating from the F
2g vibration of CeO
2 due to the symmetric arrangement of oxygen atoms in the CeO
2 lattice [
43]. Additionally, a weak absorption peak is observed near 595 cm
–1, corresponding to Frenkel-type oxygen vacancies [
44]. Notably, the characteristic peak of undoped CeO
2 at 462 cm
–1 appears strongest among the four images. As the doping level of Ni increases, the intensity of absorption peaks at this location gradually diminishes due to the substitutional reaction
, leading to the generation of oxygen vacancies and thus symmetry degradation [
45,
46].
To gauge the relative concentration of surface oxygen vacancies in the samples, the intensity of the absorption peak at 595 cm
–1 (
A595) was compared to that at 462 cm
–1 (
A462) [
47]. A higher ratio indicates a greater presence of oxygen vacancies on the sample surface [
48,
49]. Specific ratios are provided in
Table 3. As evident from these findings, increasing Ni doping correlates with a rise in the concentration of oxygen vacancies on the surface of cerium dioxide, peaking at a 5% doping level. However, at a 15% doping level, the peak intensity slightly decreases, attributed to incomplete Ni doping and the formation of Ni
2O
3, consistent with previous discussions.