Comparative Studies on Effects of Metal Cation (La) and Non-Metal Anion (N) Doping on CeO₂ Nanoparticles for Regenerative Scavenging of Reactive Oxygen Radicals

Abstract

The intrinsic effects of metal cation (La) and non-metallic anion (N) doping of CeO₂ nanoparticles (NPs) for regenerative scavenging of reactive oxygen radicals were studied. La-doped CeO₂ NPs were prepared by the conventional impregnation method at various La doping levels. N-doped CeO₂ NPs were prepared by urea thermolysis with two different methods: (i) direct thermolysis of urea after physical mixing with CeO₂ NPs and (ii) wet impregnation of CeO₂ NPs with urea followed by thermolysis under inert N₂ atmosphere. Physicochemical properties of samples were characterized by X-ray diffraction, X-ray photoelectron spectroscopy, Raman spectroscopy, and N₂ sorption measurement. Radical scavenging properties of the samples were characterized by applying Fenton’s reaction. Results indicated that atomic N doping on CeO₂ NPs significantly enhanced radical scavenging properties of CeO₂ NPs, resulting in an activity of N-doped CeO₂ about 3.6 times greater than the pristine CeO₂ NPs and 1.6 times higher than the La-doped CeO₂ NPs. This result suggests that anionic N doping of CeO₂ NPs is highly effective in enhancing radical scavenging properties of CeO₂ NPs, whereas such modifications have been typically practiced by hetero-metal doping with rare earth metal elements. A collective structure–property correlation analysis suggested that enhancement of radical scavenging properties of heteroatom-doped CeO₂ NPs was largely attributed to an increase in surface oxygen vacancies on CeO₂ NPs due to heteroatom doping.

1. Introduction

Cerium oxides are known to have regenerative scavenging properties for reactive oxygen radicals. Extensive studies have been conducted on their applications in various areas such as polymer electrolyte membrane fuel cells (PEMFCs) and biomedical antioxidants. It has been reported that adding cerium oxide nanoparticles (NPs) into the perfluorosulfonated acid (PFSA) membrane of PEMFCs can mitigate degradation of the polymer membrane caused by reactive hydroxyl and hydroperoxyl radicals generated during electrochemical cell reactions [1,2,3,4,5,6,7,8,9]. In this context, various metal oxides such as ZrO₂, YSZ, TiO₂, MnO₂, and silica-supported Mn, Cr, and Co oxides [5,9,10,11,12] have been explored for radical scavenging applications, with cerium oxides being explored mainly due to their superior properties [13]. In parallel, there have been significant efforts to enhance the radical scavenging properties of cerium oxides. Previous studies have reported that regenerative radical scavenging properties of cerium oxides are attributed to redox cycles between Ce³⁺ and Ce⁴⁺ species involving surface lattice oxygen vacancies [14,15]. Effects of the size and exposed crystal facets of CeO₂ NP have been emphasized, in which a decrease in its particle size can lead to an increase in surface oxygen vacancies [13].

The chemical compositions of cerium oxides can also vary by hetero-atom doping, which can modify the intrinsic properties of metal oxides. Various studies have reported promotional effects of hetero-metal doping as metal cations, such as with La [16,17,18], Eu [17,19], Nd [17], Gd [20], Pr [17,20], W [21], and Zr [20,22,23,24]. In these studies, enhancements of radical scavenging activities of heterometal-doped CeO₂ are generally rationalized by increases in surface oxygen vacancies and Ce³⁺ contents. In contrast, only a few recent studies have reported the enhancement of radical scavenging properties of CeO₂ NPs by non-metallic anion doping, such as nitrogen doping on the CeO₂ surface [25]. Atomic N doping has been applied previously for band gap modification of TiO₂ and CeO₂ for photocatalytic applications [26,27,28]. Enhancement of catalytic activity of CeO₂ NPs by N doping has also been reported for catalytic reduction of nitrogen oxides in automotive applications [29].

The aim of this work is to compare the intrinsic effects of metallic and non-metallic heteroatom doping of CeO₂ NPs for radical scavenging properties. Various previous studies have reported enhancements in the radical scavenging properties of CeO₂ by heteroatom doping. However, the materials were prepared and their properties were characterized under very different conditions. In addition, comparative studies on the effects of metal cation and non-metal anion dopants on the intrinsic radical scavenging activities of CeO₂ NPs have not been reported yet. Herein, La-doped CeO₂ and N-doped CeO₂ NPs were prepared with various methods at controlled heteroatom doping levels and their intrinsic radical scavenging properties were studied considering doping species, concentration, and surface oxygen vacancy concentrations.

2. Results and Discussion

2.1. Physicochemical Properties of La-Doped CeO₂ NPs

Figure 1a shows photographs of pristine CeO₂ and La-doped CeO₂ NPs obtained at various La loading amounts (2, 5, and 10 wt.%). Pristine CeO₂ NPs showed a typical bright yellow color. The La-doped CeO₂ NPs also displayed a yellow color. However, the color became darker when La loading amount increased. Figure 1b displays XRD patterns of pristine CeO₂ and La-doped CeO₂ NPs at various La loading levels. All these samples exhibited a typical diffraction pattern of a face-centered cubic (FCC) crystal structure of CeO₂ (JCPDS #82-0792) without any additional diffraction peaks. Formation of any hetero-crystalline phases by La doping treatment was not found.

Table 1. Cubic lattice parameter and BET surface area of the pristine CeO₂ and the La-doped CeO₂ NPs.

Sample	Cubic Lattice Parameter, a (Å)	BET Surface Area (m2 g−1)
Pristine CeO2	5.408	28.5
2 wt% La-CeO2	5.408	25.9
5 wt% La-CeO2	5.409	19.8
10 wt% La-CeO2	5.405	13.5

shows the cubic lattice parameter (a, Å) of samples calculated from the (111) diffraction peak of the spectra. The pristine CeO₂ showed a lattice parameter value of 5.408 Å, which was close to the average value of a stoichiometric CeO₂ (5.4112 Å). The La-doped CeO₂ samples showed cubic lattice parameter values (5.405–5.408 Å) similar to the pristine CeO₂. BET surface areas of the pristine CeO₂ and La-doped CeO₂ samples are summarized in Table 1. The pristine CeO₂ NPs showed a moderate surface area of 31.8 m² g⁻¹. La-doped CeO₂ samples exhibited a gradual decrease in their surface areas with an increase in La loading amount. The average surface area of 5 wt.% La-CeO₂ NPs was 19.8 m² g⁻¹, which was about 34% lower than the initial surface area of the pristine CeO₂ (28.5 m² g⁻¹).

Figure 2a shows XPS spectra of the pristine CeO₂ and La-doped CeO₂ NPs at different La loading levels. La-doped CeO₂ samples showed typical peaks of lanthanum oxides at 3d binding energy levels [30].

Table 2. Atomic content and the relative Ce concentration on the pristine CeO₂ and the La-doped CeO₂ NPs.

Sample	Absolute Atomic %				Relative Ce Concentration (%)
	Ce	C	O	La	Ce3+	Ce4+
Pristine CeO2	18.23	30.04	51.73	-	25.4	74.6
2 wt.% La-CeO2	19.25	17.07	59.25	4.42	21.6	78.4
5 wt.% La-CeO2	15.17	15.83	60.99	8.00	22.0	78.0
10 wt.% La-CeO2	13.61	14.74	62.48	9.18	20.5	79.5

summarizes surface atomic contents of La on samples measured by XPS analysis. The results showed incorporation of La on the sample surface at various loading levels as prepared accordingly. Figure 2b shows deconvoluted XPS spectra of the samples at Ce 3d binding energy level characterized for quantifying Ce³⁺ and Ce⁴⁺ concentrations on the pristine CeO₂ and La−CeO₂ NPs. The relative Ce³⁺ concentration was defined by the Ce³⁺/(Ce⁴⁺ + Ce³⁺) atomic ratio on the sample surface. As shown in Table 2, the relative Ce³⁺ concentration slightly decreased as the La loading on the CeO₂ NPs increased.

Figure 3a shows Raman spectra of the pristine CeO₂ and La-doped CeO₂ NPs at various La loading levels. Samples exhibited the typical characteristic Raman peak of CeO₂ at 460 cm⁻¹, corresponding to the symmetrical stretching mode of the Ce–O₈ vibrational unit [23,31]. Oxygen vacancy concentrations of samples were estimated from broadening of the peak at 460 cm⁻¹ [25,32]. As shown in Figure 3b), oxygen vacancy concentrations of the pristine CeO₂ increased by approximately 143–271% with the La doping treatment.

2.2. Radical Scavenging Properties of Pristine CeO₂ and La-Doped CeO₂ NPs

Radical scavenging properties of the pristine CeO₂ and La-doped CeO₂ NPs were characterized by conducting Fenton’s reaction with Rhodamine-B (RhB) in a solution. Briefly, reactive hydroxyl radicals generated by Fenton’s reactions (Equation (1)) degraded RhB molecules (Equation (3)) in the solution:

$${\mathrm{H}}_2{\mathrm{O}}_2+{\mathrm{Fe}}^{2+}\to {\mathrm{HO}}^{•}+{\mathrm{HO}}^{-}+{\mathrm{Fe}}^{3+}$$

(1)

$${\mathrm{Fe}}^{3+}+{\mathrm{H}}_2{\mathrm{O}}_2\to {\mathrm{Fe}}^{2+}+{\mathrm{HOO}}^{•}+{\mathrm{H}}^{+}$$

(2)

$${\mathrm{HO}}^{•}+\mathrm{RhB}\to \mathrm{Degradation} \mathrm{of} \mathrm{RhB}$$

(3)

Therefore, scavenging of reactive radicals by a radical scavenger could reduce the decay of RhB concentration in the solution, which could be used to compare radical scavenging properties of materials under the same reaction condition [33].

Figure 4a shows typical UV–Vis spectra of Fenton’s solution taken every 10 min for 1 h, with the addition of pristine CeO₂ or La-doped CeO₂ NPs (25 mg of the radical scavenger sample per 1 mL Fenton’s solution). All spectra displayed a peak at 550 nm attributed to light absorption by RhB in the solution. However, the peak intensity attenuated over time, indicating a decrease in RhB concentration due to degradation caused by hydroxyl radicals generated by Fenton’s reaction. Figure 4b shows relative RhB concentration (C_t/C_i) in the Fenton’s solution over time, where C_t is the RhB concentration at the given time and C_i is the initial RhB concentration in the solution. The results showed a significant and rapid decrease of RhB concentration in the blank solution. The relative RhB concentration in the solution was less than 10% after 60 min. In comparison, the decay of RhB concentration in the presence of pristine CeO₂ NPs was much less, retaining about 40% of the initial RhB concentration after completion of the Fenton’s reaction. Radical scavenging properties of La-doped CeO₂ NPs depended significantly on the La loading amount. The 5 wt.% La-CeO₂ NPs exhibited a higher radical scavenging property than the pristine CeO₂, retaining about 50% of the initial RhB concentration after completion of the reaction. The 2 wt.% La-CeO₂ NPs showed similar radical scavenging properties to the pristine CeO₂ NPs. The 10 wt.% La-CeO₂ NPs displayed much smaller radical scavenging properties than the pristine CeO₂, suggesting detrimental effects of excessive La loading on the radical scavenging properties of CeO₂ NPs. Figure 4c displays photographs of the solution taken before and after completion of the Fenton’s reaction. The results clearly showed that the color of the blank solution significantly changed from a thick red to a pale pink due to degradation of RhB in the solution. Comparatively, the solution containing the 5 wt.% La-CeO₂ NPs exhibited a thick red color similar to that of the initial solution, demonstrating the significant radical scavenging properties of La-CeO₂ NPs.

2.3. Physicochemical Properties of N-Doped CeO₂ NPs

Atomic N doping of the pristine CeO₂ was conducted with the urea thermolysis method. Scheme 1 briefly shows the procedure for atomic N doping on the pristine CeO₂ NPs. As described in the experimental section, N doping was conducted by urea thermolysis via two different approaches: (i) mixing pristine CeO₂ NPs with urea powder (the dry method); and (ii) impregnating pristine CeO₂ NPs with aqueous solution of urea (the wet method). These samples were subsequently treated at 873 K under two different conditions: inert N₂ or air atmosphere. Figure 5a shows photographs of the pristine CeO₂, N-CeO₂-D-I, N-CeO₂-W-I, and N-CeO₂-W-A samples. The pristine CeO₂ NPs showed a typical bright yellow color. The N-CeO₂-D-I and N-CeO₂-W-I samples prepared by the heat treatment in an inert N₂ atmosphere exhibited dark brown colors. However, samples prepared in an air atmosphere (N-CeO₂-W-A) showed a bright yellow color similar to the pristine CeO₂. Figure 5b shows the XRD patterns of the pristine CeO₂ and N-doped CeO₂ samples. As discussed previously, the pristine CeO₂ NPs exhibited the typical diffraction pattern of FCC crystal structure of CeO₂. The N-doped CeO₂ samples also exhibited the typical diffraction pattern of CeO₂, retaining the bulk FCC crystal structure of the pristine CeO₂.

Table 3. Cubic lattice parameter and BET surface area of the pristine CeO₂ and the N-doped CeO₂ NPs.

Sample	Cubic Lattice Parameter, a (Å)	BET Surface Area (m2 g−1)
Pristine CeO2	5.408	28.5
N-CeO2-D-I	5.406	12.2
N-CeO2-W-I	5.408	22.3
N-CeO2-W-A	5.408	30.0

shows cubic lattice parameter values and BET surface areas of the pristine CeO₂ and N-doped CeO₂ NPs prepared by various methods. The N-doped CeO₂ samples exhibited similar cubic lattice parameter values (5.406–5.408 Å) to the pristine CeO₂ NPs (5.4112 Å). Surface areas of the N-doped CeO₂ samples showed considerable differences depending on the N doping method. The N-CeO₂-W-I NPs prepared by the urea impregnation followed by thermolysis in N₂ atmosphere showed a moderate surface area (22.3 m² g⁻¹), which was about 21% smaller than that of the pristine CeO₂ (28.5 m² g⁻¹). The surface area of the N-CeO₂-D-I prepared by the dry method was far smaller (12.2 m² g⁻¹), which was about 57% lower than that of the pristine CeO₂. Differently, the surface area of the N-CeO₂-W-A NPs (30.0 m² g⁻¹) prepared in the air atmosphere was similar to that of the pristine CeO₂ NPs.

Figure 6a shows the deconvoluted XPS spectra of the pristine CeO₂ and N-doped CeO₂ samples at the N 1s binding energy level. It was notable that N doping did not occur on the N-CeO₂-W-A NPs prepared by heat treatment in an air atmosphere. Differently, N-CeO₂-D-I and the N-CeO₂-W-I samples prepared in an inert N₂ atmosphere exhibited two deconvoluted peaks centered at binding energies of 396 and 399 eV, respectively, that could be attributed to Ce-N and C-O bonding configurations.

Table 4. Atomic content and the relative Ce concentration on the pristine CeO₂ and the N-doped CeO₂ NPs.

Sample	Absolute Atomic %				Relative Ce Concentration (%)
	Ce	C	O	N	Ce3+	Ce4+
Pristine CeO2	18.23	30.04	51.73	-	25.4	74.6
N-CeO2-D-I	22.04	16.73	53.72	7.51	28.0	72.0
N-CeO2-W-A	22.33	20.44	57.22	-	25.2	74.8
N-CeO2-W-I	24.7	10.83	60.15	4.45	24.7	75.3

shows surface N atomic contents and relative bonding configuration obtained from XPS results. The N-CeO₂-D-I NPs exhibited a higher N atomic content (7.51%) than the CeO₂-W-I NPs (4.45%), with a high Ce-N bonding configuration (51.6%). Figure 6b shows the deconvoluted XPS spectra of samples at Ce 3d binding energy level for quantifying Ce³⁺ and Ce⁴⁺ concentrations. The results indicated that Ce³⁺ concentrations on N-doped CeO₂ samples (24.7–28%) were similar to or slightly higher than that on the pristine CeO₂ (25.4%).

Figure 7a shows Raman spectra of the pristine CeO₂ and N-doped CeO₂ samples. Oxygen vacancy concentrations on samples were estimated from peak broadening [25,32]. The results are summarized in Figure 7b. Surface oxygen vacancy concentrations on the N-CeO₂-D-I and the N-CeO₂-W-I samples prepared by urea thermolysis in an inert N₂ atmosphere were about 4-fold and 8-fold higher than that on the pristine CeO₂. However, the N-CeO₂-W-A sample prepared in an air atmosphere showed practically the same oxygen vacancy concentration as the pristine CeO₂. Similar oxygen vacancy concentrations of the pristine CeO₂ and the N-CeO₂-W-A were reasonable considering that N doping did not occur on the N-CeO₂-W-A NPs as shown in XPS results. Collectively, these results clearly indicate that N doping of CeO₂ NPs could lead to a significant increase in surface oxygen vacancy concentration on the surface.

2.4. Radical Scavenging Properties of N-Doped CeO₂ NPs

Radical scavenging properties of the N-doped CeO₂ NPs are characterized by applying the Fenton’s reaction as previously described. Figure 8a compares the UV–Vis spectra of Fenton’s solution taken every 10 min for 1 h after addition of a radical scavenger (pristine CeO₂, N-CeO₂-D-I, N-CeO₂-W-I, or N-CeO₂-W-A NPs, 25 mg of radical scavenger per 1 mL Fenton’s solution). Figure 8b shows changes of RhB concentration (C_t) normalized by its initial concentration (C_i) in the solution measured over time. The results showed that normalized RhB concentration (C_t/C_i) decreased over time due to degradation of RhB caused by hydroxyl radicals generated by Fenton’s reaction. These results clearly showed that RhB was retained at much higher concentrations in solutions in the presence of N-doped CeO₂ samples (N-CeO₂-D-I and the N-CeO₂-W-I) than the pristine CeO₂ NPs, reflecting significant enhancements to the radical scavenging properties of CeO₂ NPs by N doping on the surface. In comparison, N-CeO₂-W-A NPs prepared by heat treatment in air atmosphere showed practically the same radical scavenging property as pristine CeO₂ NPs. As discussed previously, N doping did not occur on the N-CeO₂-W-A sample and its surface area was also practically the same as the pristine CeO₂. These results support that N doping of the CeO₂ NPs is the primary reason for the enhancement of the radical scavenging properties of N-doped CeO₂ NPs. These results collectively suggest that the doping method has significant effects on radical scavenging properties of N-doped CeO₂ NPs. The wet method conducted with urea impregnation and subsequent heat treatment in an inert atmosphere gave rise to higher radical scavenging properties than the dry method conducted by heat treatment of a mixture of urea powder and CeO₂ NPs. Figure 8c compares photographs of the initial solution and the solutions after completion of Fenton’s reaction. The results clearly show that the solutions containing N-doped CeO₂ samples maintained the initial thick red color, confirming the high radical scavenging properties of the N-doped CeO₂ NPs.

2.5. Comparison of Radical Scavenging Properties of La- and N-Doped CeO₂ NPs

Radical scavenging properties of pristine CeO₂, La-doped CeO₂, and N-doped CeO₂ NPs were compared based on the RhB concentration remained in the solution after Fenton’s reaction for 60 min. Figure 9a shows relative RhB concentrations in the blank and the radical scavenger containing solutions after completion of Fenton’s reaction. For comparison, the 5 wt.% La-CeO₂ and the N-CeO₂-W-I NPs were chosen among the La-doped and the N-doped CeO₂ samples because these two samples exhibited the best radical scavenging properties in each group of samples. The results showed that RhB concentration fell rapidly to 10% in the blank solution after 60 min. In the presence of pristine CeO₂ NPs, about 38% of the initial RhB concentration was retained after the test. Comparatively, the RhB concentration was retained at significantly higher concentrations with La-CeO₂ NPs (50%) and N-CeO₂ NPs (80%) than with pristine CeO₂ NPs, indicating a considerable enhancement to radical scavenging activities by La and the N atomic doping on CeO₂ NPs.

Intrinsic radical scavenging properties of pristine CeO₂, La-CeO₂, and N-CeO₂ NPs were estimated by normalization of radical scavenging properties with their BET surface areas. Figure 9b shows the surface area-normalized radical scavenging properties of samples. The results revealed that intrinsic radical scavenging properties of samples had the following decreasing order: N-CeO₂ > La-CeO₂ > pristine CeO₂ NPs. The intrinsic radical scavenging property of the N-CeO₂ NPs was about 3.6 times greater than that of the pristine CeO₂ and 1.6 times greater than that of the La-CeO₂ NPs. These results suggest that atomic N doping is highly effective for enhancing the radical scavenging properties of CeO₂ NPs.

The radical scavenging properties of pristine CeO₂, La-doped CeO₂, and N-doped CeO₂ NPs were correlated with surface oxygen vacancy concentrations on these samples. Figure 9c shows the RhB concentrations retained after 60 min of Fenton’s reaction in the solution with respect to surface oxygen vacancy concentrations of the radical scavenger samples. The collective structure-property correlation analysis showed that RhB concentration retained in the solution increased with surface oxygen vacancy concentration of the radical scavengers. These results suggest that oxygen vacancies play a primary role in regenerative radical scavenging activities of these samples. Comparative studies revealed that atomic N doping of CeO₂ NPs was highly effective for increasing surface oxygen vacancies on CeO₂ NPs, which resulted in large enhancements in the intrinsic radical scavenging properties of CeO₂ NPs.

3. Experimental

3.1. Atomic Hetero-Metal (La) Doping on CeO₂ NPs

Scheme 1 describes the atomic doping of metal cation (La) and non-mental anion (N) on pristine CeO₂ NPs prepared in this work. La doping of CeO₂ NPs was conducted with the incipient wetness impregnation method. Briefly, an aqueous solution of La(NO₃)₃ was prepared by dissolving 2.6 g of La(NO₃)₃ (Alfa Aesar, 99%) in 5 mL of D.I. H₂O. This 1.58 mL La(NO₃)₃ solution was then mixed dropwise with 5 g of pristine CeO₂ powder (Sigma-Aldrich, 99.9%) for uniform impregnation (5 wt.% La loading). The resultant was then dried at 353 K in a convection oven for 12 h. The sample was placed in a ceramic boat and placed inside a horizontal tubular quartz reactor (I.D. = 50 mm, L = 1,200 mm) equipped with a temperature-controlled electric furnace. Both ends of the reactor were capped gas tight, and air (99.999%, Deokyang, Republic of Korea) was fed to the reactor (100 mL min⁻¹). While maintaining the air flow through the reactor, the temperature was raised (ramp = 10 K min⁻¹) and held at 873 K for 4 h. The reactor was then cooled to room temperature under the air flow. The loading amount of La on the sample was varied (2, 5, and 10 wt.%) by changing the concentration of La(NO₃)₃ in the impregnation solution.

3.2. Atomic Non-Metallic Heteroatom (N) Doping on CeO₂ NPs

Atomic N doping on CeO₂ NPs was conducted by urea thermolysis with two different methods: (i) dry urea thermolysis and ii) wet urea impregnation followed by thermolysis. In the dry method, CeO₂ NPs were physically mixed with solid urea powder (98%, Alfa Aesar) at a fixed weight ratio (CeO₂/urea = 4) and placed in a ceramic boat. The sample boat was placed inside a horizontal tubular quartz reactor. Both ends of the reactor were capped gas tight. The inside reactor was purged by flowing an inert N₂ gas (100 mL min⁻¹, 99.999%, Deokyang, Republic of Korea) at 298 K for 1 h. The reactor temperature was raised (ramp = 10 K min⁻¹) and held at 873 K for 2 h, flowing N₂ gas through the reactor. Finally, the reactor was cooled to room temperature under the N₂ gas flow. The resulting sample is denoted as N−CeO₂-D-I, where D and I indicate the dry method and inert atmosphere, respectively.

In the wet impregnation method, CeO₂ NPs were impregnated with an aqueous solution of urea. The impregnation amount of urea on the CeO₂ was fixed at 3 wt.%. Urea solution was prepared by dissolving 2.1 g of urea powder in 5 mL of D.I. H₂O. Then, 1.58 mL of urea solution was dropwise mixed with 5 g CeO₂ NPs for uniform impregnation. The urea impregnated CeO₂ was then dried in a convection oven at 353 K for 12 h. The sample was placed in a ceramic boat and loaded in the tubular reactor. Subsequent heat treatment procedures were the same as those described above in the dry method except that it was conducted in two different atmospheres: inert N₂ and air. The sample was treated flowing N₂ gas (99.999%, Deokyang Korea) or air (99.999%, Deokyang Korea) at 873 K for 4 h. Resulting samples are denoted as N-CeO₂-W-I and N-CeO₂-W-A, respectively, where W and A indicated wet method and air atmosphere, respectively.

3.3. Characterizations

Surface atomic compositions of the pristine CeO₂, La-doped CeO₂, and N-doped CeO₂ NPs were characterized by X-ray photoelectron spectroscopy (XPS, Theta Probe AR-XPS, Thermo Fisher Scientific, Waltham, MA, U.S.A.). The atomic contents of Ce³⁺ and Ce⁴⁺ on the samples were obtained by deconvolution of corresponding characteristic peaks at the Ce 3d binding energy level according to previously reported methods in the literature [25,34]. Doping amount and structural configuration of La and the N species on La-doped and N-doped CeO₂ NPs were obtained by deconvolution of the La 3d and the N 1s peak of the XPS spectra. The crystal structure of each sample was characterized by X-ray diffraction (XRD, SmartLab, Rigaku, Tokyo, Japan) with monochromic Cu–Kα radiation operated at 3 kW (scan rate = 0.05° min⁻¹). Defects on the sample surfaces were characterized by Raman spectroscopy (Vertex 80, Bruker, Billerica, MA, U.S.A.) conducted at 532 nm laser wavelength. The concentration of surface oxygen vacancies was calculated from the full width at half maximum (FWHM) of the characteristic Raman peak according to the method reported in the literature [25,32]. The specific surface area of each sample was obtained with the Brunauer–Emmett–Teller (BET) method using an N₂ adsorption–desorption isotherm obtained in a volumetric unit (Tristar, Micromeritics, Norcross, GA, USA).

3.4. Measurement of Radical Scavenging Properties

Radical scavenging properties of the pristine CeO₂, La-doped CeO₂, and N-doped CeO₂ NPs were characterized by conducting Fenton’s reaction after adding Rhodamine B (RhB) and radical scavenger into the solution. Added RhB was used as a marker to estimate radical scavenging properties of the scavenger materials added into the solution. First, an aqueous solution was prepared by dissolving 0.6 mg RhB (Alfa Aesar) and 3 mg FeSO₄ (98.5%, Junsei Chem., Tokyo, Japan) in 100 mL D.I. H₂O. Then, 50 mg of the radical scavenger was added to 2 mL of the solution with stirring. Separately, an aqueous solution of H₂O₂ (0.08 vol.%) was prepared by mixing H₂O₂ (30 wt.%, DaeJung Chem., Siheung, Republic of Korea) and D.I. H₂O. Finally, a fixed amount (3 μL) of the aqueous H₂O₂ solution was added dropwise to the first solution containing FeSO₄ and stirred for 10 min. Subsequently, light adsorption spectra of RhB in the mixed solution were obtained using a UV–Vis spectrometer (Mega-800, Sinco, Daejeon, Republic of Korea) at a wavelength of 300 to 700 nm. Dropwise H₂O₂ addition, stirring, and UV–Vis measurements were repeated at 10 min intervals for 1 h to measure changes of RhB concentration in the solution over time. Radical scavenging properties of materials added into the solution were estimated from the extent of RhB concentration decay in the solution in comparison with that in the blank Fenton’s solution measured without adding radical scavengers.

4. Conclusions

The effects of metal cation and non-metal anion doping on CeO₂ NPs for radical scavenging properties were studied. La-doped CeO₂ NPs (8 atomic % La) showed about 2.0 times greater intrinsic radical scavenging activities than the pristine CeO₂ NPs. In comparison, N-doped CeO₂ NPs (4.5 atomic-% N) prepared by a wet urea impregnation with thermolysis method exhibited about 2.8 times higher intrinsic radical scavenging activities than the pristine CeO₂, suggesting the high effectiveness of atomic N doping for enhancing radical scavenging properties. A collective structure–property correlation analysis conducted with pristine CeO₂, La-doped CeO₂, and N-doped CeO₂ NPs suggested that surface oxygen vacancies provide relevant active sites for regenerative radical scavenging. Surface oxygen vacancy concentrations could be effectively introduced on CeO₂ NPs by heteroatom doping, which gave rise to significant enhancements of radical scavenging properties of CeO₂ NPs.