Construction of Oxygen Vacancies of Zr-Doped CeO 2 with Enhanced Dye Adsorption Performance

: Congo red (CR), a highly pigmented anionic dye, is highly toxic and resistant to degradation. The discharge of CR wastewater into the natural environment can lead to ecological destruction and harm to human health. CeO 2 as an adsorbent possesses the advantages of excellent acid and alkali resistance, biocompatibility, stable physical and chemical properties, and nontoxic by-products. The impact of Zr doping on the adsorption performance of nano-CeO 2 was investigated. XPS and Raman characterisation revealed that Zr doping effectively enhanced the oxygen vacancy ratio at the active sites for CR adsorption on the surface of nano-CeO 2 . When the doping amount of Zr was 3%, the nanoparticles with the best adsorption properties were obtained, and the adsorption amount of CR at room temperature was as high as 3642.05 mg/g, which was approximately three times the adsorption amount of undoped CeO 2 . This excellent adsorption property shows good prospects for the removal of anionic dyes from wastewater.


Introduction
The dyeing, textile, leather, plastic, and other industries generate significant amounts of dye-containing wastewater [1][2][3][4].Anionic dyes such as Congo Red (CR) exhibit intense colouration and chromaticity, lead to a substantial pollution load, and have limited degradation under natural conditions [5,6].Discharging untreated CR wastewater into the natural environment poses a threat to ecological integrity and carcinogenic and mutagenic risks to living organisms [7,8].Therefore, the removal of these substances is crucial to safeguard the environment and public health against their deleterious effects.A number of methods have been used to remove CR dye from wastewater, including photocatalytic (redox) reactions, flocculation, coagulation, ion exchange, reverse osmosis, membrane filtration, and adsorption [9][10][11][12][13][14].Compared to other techniques, the adsorption technique is best suited for practical applications, since it offers advantages such as operational simplicity, cost-effectiveness, and the generation of non-toxic byproducts, making it widely employed in dye wastewater treatment [15][16][17].Metal oxides serve as advanced adsorbents because of their unique characteristics, including their large surface areas, diverse forms, stable chemical properties, and abundant surface functional groups [18][19][20][21].Among the various metal oxides, cerium oxide has garnered special attention for wastewater treatment owing to its nontoxic nature, economic viability, and environmental friendliness [15][16][17].Compared to traditional adsorbents, such as activated carbon [22], hydrogels [23], zinc oxide [24], and alumina [25], nano-CeO 2 exhibits superior acid-alkali resistance, biocompatibility, and stable physical/chemical properties [26].
It has been reported that oxygen vacancies play a crucial role as active sites for the adsorption of CR dyes [27,28].Zhao [27] discovered that increasing the concentration of oxygen vacancies on the surface of Al 2 O 3 enhances the adsorption of CR.Joshy [29] used the First, by adjusting the ratio of zirconium oxychloride to cerium nitrate and then adding deionised water as solvent, a 0.05 mol/L cerium nitrate solution with 50 mL Zr/(Zr + Ce) and molar ratio of 0% (1, 2, 3, 4, and 5%) was obtained.Subsequently, 0.005 mol of anionic surfactant sodium dodecyl benzene sulfonate (SDBS) was added to the cerium nitrate solution, which was magnetically stirred at 298 K until the surfactant and cerium nitrate were evenly mixed.We prepared 100 mL of 2 mol/L ammonia solution and used a peristaltic pump to extract cerium nitrate precursor solution at a constant speed and dropped it into the ammonia solution stirred at a high-speed using magnetic force.After the cerium nitrate drops were added, an appropriate amount of hydrogen peroxide was added to change the solution from purple to light yellow.The mixture was stirred for 30 min to complete the reaction.The samples were then centrifuged and washed alternately with deionised water and anhydrous ethanol until the washing liquid was neutral, dried in a vacuum oven at 333 K for 12 h, and then calcined at 773 K for 2 h.The dried powder was then grounded using an agate mortar.ZC0, ZC1, ZC2, ZC3, ZC4, and ZC5, representing Zr, were used to add CeO 2 nanoparticles in the proportions of 0, 1, 2, 3, 4, and 5%, respectively.

Characterization
The morphologies of the surfaces and cross-section of various Zr doping ratios of CeO 2 nanoparticles were observed via field-emission scanning electron microscopy (FE-SEM, Hitachi SU-70, Tokyo, Japan).Transmission electron microscope analysis (TEM, Hitachi H9500, Almelo, The Netherlands) for nano-CeO 2 was also performed.The surface elemental composition with various Zr doping ratios of nano-CeO 2 was examined using X-ray photoelectron spectroscopy (XPS, Thermo Scientific ESCALAB Xi+, Waltham, MA, USA).
Al-Kα radiation was used as the X-ray source, and the binding energies were referenced to the C 1s peak (284.8 eV).X-ray diffraction (XRD; Bruker-axs, San Jose, MA, USA) was conducted to characterise the particle size of the nanoadsorbent.CR adsorption capacity tests were performed to measure the absorption spectra using a UV-Vis spectrometer (UV-VIS, Puli TU-1901, Beijing, China).N 2 adsorption isotherms were measured at 77.3 K by using an automatic specific surface and micropore analyser (BET, BSD Instrument 3H-2000PM2, Beijing, China).Before the adsorption measurements, the nano-CeO 2 powders were degassed under flowing N 2 for 5 h at 393 K to remove any residual water and surface species.Raman spectra were used to analysis the Zr-doped CeO 2 nanocomposite material at room temperature using Horiba Raman system (Raman, HR Evolution, Montpellier, France) with a continuous wave laser operating at 532 nm.

Batch Adsorption Study
An aqueous stock solution of CR (500 mg/L) was prepared by dissolving CR in deionized water.The required concentration (100, 200, 300, and 400 mg/L) was obtained by continuous dilution with deionized water.A total of 100 mL each of CR dye solution with different concentrations were taken and 10 mg nano-CeO 2 adsorbent was added.To adjust the pH levels, solutions of 0.1 mol/L HCl and 0.1 mol/L NaOH were utilized.Then the solutions were stirred sufficiently with magnetic stirrer at 500 rpm in a neutral environment with a temperature of 298 K and a pH of 7. At different time intervals, 5 mL of suspension samples was taken for centrifugation separation, and the supernatant was collected to measure its absorbance at the maximum absorption wavelength of CR at 496 nm via UV-Vis spectrophotometry.All studies were conducted at pH of 7 except for the pH effect study.Equilibrium adsorption isotherms were studied at different initial concentrations (pH of 7).At different pH values (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12), the CR removal efficiency was investigated at an initial concentration of 100 mg/L and an adsorption time of 120 min.Finally, all batch adsorption measurements were performed three times, and the average was used for the experiments.The adsorption capacity Qe (mg/g) and removal rate (%) by the CeO 2 nanoparticles were calculated using the formulae (1) and (2), respectively.
where Qe is the equilibrium adsorption capacity (mg/g); C 0 and C f are the initial and final concentrations of CR (mg/L), respectively; and V (L) and M (g) are the volume of the CR solution and adsorbent dosage, respectively.

Phase Analysis
The X-ray diffraction patterns of cerium oxide and Zr 4+ -doped cerium oxide revealed a crystal structure that aligns with the cubic fluorite structure of cerium oxide (PDF.34-0394), as shown in Figure 1.The main characteristic peaks at 2θ = 28.6 • , 33.0 • , 47.5 • , and 56.3 • correspond to the crystal faces of (111), (200), (220), and (311) planes [35], respectively.The increase in the Zr doping content did not affect the positions of the four prominent peaks of CeO 2 in the XRD pattern.However, this resulted in a less-pronounced peak deformation and weakened peak intensity for CeO 2 .This observation suggests that the characteristic peaks tend to weaken with increasing Zr doping concentration, indicating a decrease in the degree of sample crystallisation.The crystallite sizes of the ZC0, ZC1, ZC2, ZC3, ZC4, and ZC5 nanoparticles were 9.8, 9.0, 8.3, 7.1, 6.6, and 6.7 nm, respectively, as determined by the Scherrer formula (3) [36].This analysis revealed a decreasing trend in grain size with increasing Zr doping ratio.These phenomena may be attributed to the doping of Zr into the CeO 2 lattice, which hindered the growth of the CeO 2 crystals.The reduction in particle size suggests an increase in the surface adsorption of active sites of CeO 2 , thereby enhancing its adsorption performance.
respectively.The increase in the Zr doping content did not affect the positions of the four prominent peaks of CeO2 in the XRD pattern.However, this resulted in a less-pronounced peak deformation and weakened peak intensity for CeO2.This observation suggests that the characteristic peaks tend to weaken with increasing Zr doping concentration, indicating a decrease in the degree of sample crystallisation.The crystallite sizes of the ZC0, ZC1, ZC2, ZC3, ZC4, and ZC5 nanoparticles were 9.8, 9.0, 8.3, 7.1, 6.6, and 6.7 nm, respectively, as determined by the Scherrer formula (3) [36].This analysis revealed a decreasing trend in grain size with increasing Zr doping ratio.These phenomena may be attributed to the doping of Zr into the CeO2 lattice, which hindered the growth of the CeO2 crystals.The reduction in particle size suggests an increase in the surface adsorption of active sites of CeO2, thereby enhancing its adsorption performance.

SEM Analysis
Scanning electron microscopy (SEM) was employed to investigate the surface morphology of the pure cerium oxide ZC0 (a) and Zr-doped cerium oxides ZC1, ZC2, ZC3, ZC4, and ZC5 (b, c, d, e, and f, respectively), as shown in Figure 2. The results demonstrate that CeO2 nanoparticles prepared by high-temperature calcination and doped with Zr exhibit a uniform particle size distribution.Compared to the undoped samples, the morphology of the CeO2 nanoparticles remained largely unchanged, with a granular particle size distribution.Additionally, all samples exhibited a regular crystal structure which enhanced the adsorption process.

SEM Analysis
Scanning electron microscopy (SEM) was employed to investigate the surface morphology of the pure cerium oxide ZC0 (a) and Zr-doped cerium oxides ZC1, ZC2, ZC3, ZC4, and ZC5 (b, c, d, e, and f, respectively), as shown in Figure 2. The results demonstrate that CeO 2 nanoparticles prepared by high-temperature calcination and doped with Zr exhibit a uniform particle size distribution.Compared to the undoped samples, the morphology of the CeO 2 nanoparticles remained largely unchanged, with a granular particle size distribution.Additionally, all samples exhibited a regular crystal structure which enhanced the adsorption process.

TEM Analysis
Figure 3 displays the TEM images and elemental distribution images of the nano-ZC3 particles doped with 3 atm% Zr at varying magnifications.As shown in Figure 3a,

TEM Analysis
Figure 3 displays the TEM images and elemental distribution images of the nano-ZC3 particles doped with 3 atm% Zr at varying magnifications.As shown in Figure 3a, the Zr-doped nano-CeO 2 , prepared through the co-precipitation of cerium nitrate and zirconium oxychloride with ammonia water, exhibited a distinct particle structure, indicating that the structure of the ZC3 nanoparticles remained unchanged after Zr doping while maintaining their nanoparticle form.At low magnification, the good dispersion of the ZC3 nanoparticles was observed using TEM.The average particle size was calculated to be 11.05 nm by the TEM image (a) statistics.Additionally, the elemental distribution diagram in Figure 3c demonstrates the uniform distribution of Zr, Ce, and O on the surface of the ZC3 nanoparticles.

TEM Analysis
Figure 3 displays the TEM images and elemental distribution images of the nano-ZC3 particles doped with 3 atm% Zr at varying magnifications.As shown in Figure 3a, the Zr-doped nano-CeO2, prepared through the co-precipitation of cerium nitrate and zirconium oxychloride with ammonia water, exhibited a distinct particle structure, indicating that the structure of the ZC3 nanoparticles remained unchanged after Zr doping while maintaining their nanoparticle form.At low magnification, the good dispersion of the ZC3 nanoparticles was observed using TEM.The average particle size was calculated to be 11.05 nm by the TEM image (a) statistics.Additionally, the elemental distribution diagram in Figure 3c demonstrates the uniform distribution of Zr, Ce, and O on the surface of the ZC3 nanoparticles.

BET Analysis
The effect of the doping amount on the surface area was evaluated using nitrogen adsorption-desorption isotherms.As depicted in Figure 4, the ZC0, ZC1, ZC2, ZC3, ZC4, and ZC5 nanoparticles exhibit typical class IV isotherms and H2-type [37,38] hysteresis with a distinct jump observed at a relative pressure of P/P 0 = 0.4.These results indicated the presence of a mesoporous structure with a relatively uniform pore size distribution.The SEM and TEM analyses revealed that the formation of pore structures in the material primarily resulted from the aggregation of numerous nanoparticles, whereas interparticle pores facilitated adsorption.The specific surface areas of the ZC0, ZC1, ZC2, ZC3, ZC4, and ZC5 nanoparticles were 65.31, 73.90, 88.07, 82.38, 68.23 and 62.47 m 2 /g, respectively.The specific surface areas and average pore diameters of the samples are listed in Table 1.
and ZC5 nanoparticles exhibit typical class IV isotherms and H2-type [37,38] hysteresis with a distinct jump observed at a relative pressure of P/P0 = 0.4.These results indicated the presence of a mesoporous structure with a relatively uniform pore size distribution.The SEM and TEM analyses revealed that the formation of pore structures in the material primarily resulted from the aggregation of numerous nanoparticles, whereas interparticle pores facilitated adsorption.The specific surface areas of the ZC0, ZC1, ZC2, ZC3, ZC4, and ZC5 nanoparticles were 65.31, 73.90, 88.07, 82.38, 68.23 and 62.47 m 2 /g, respectively.The specific surface areas and average pore diameters of the samples are listed in Table 1.

X-ray Photoelectron Spectroscopy
To further study the surface composition and chemical valence of the Zr-doped CeO2 nanoparticles and determine whether Zr was successfully introduced into the CeO2 nanoparticles, an XPS analysis was performed on the ZC0, ZC1, ZC2, ZC3, ZC4, and ZC5 nanoparticles.The full XPS spectrum of the Zr-doped CeO2 nanoparticles, as depicted in Figure 5a, confirms the presence of characteristic peaks for Ce 3d, O 1s, C1s, and Zr 3d.This observation is consistent with the TEM mapping results and indicates the successful introduction of Zr into the CeO2 nanoparticles. Figure 5b

X-ray Photoelectron Spectroscopy
To further study the surface composition and chemical valence of the Zr-doped CeO 2 nanoparticles and determine whether Zr was successfully introduced into the CeO 2 nanoparticles, an XPS analysis was performed on the ZC0, ZC1, ZC2, ZC3, ZC4, and ZC5 nanoparticles.The full XPS spectrum of the Zr-doped CeO 2 nanoparticles, as depicted in Figure 5a, confirms the presence of characteristic peaks for Ce 3d, O 1s, C1s, and Zr 3d.This observation is consistent with the TEM mapping results and indicates the successful introduction of Zr into the CeO 2 nanoparticles.Figure 5b shows the Zr 3d spectra of the sample, where peg-split fitting was applied to the Zr 3d energy spectrum of ZC3 doped with 3% Zr.The binding energies at 182.4 and 185.6 eV [39] correspond to Zr 3d5/2 and Zr 3d3/2, respectively, suggesting the presence of surface-bound Zr 4+ components in the ZC3 material [40,41].
The O 1s energy spectrum was fitted using Advantage software (5.918).The O 1s energy spectra of the undoped pure CeO 2 sample ZC0 and doped nanoparticles with 1~5% Zr, as depicted in Figure 6, were analysed using peak fitting to quantify the surface oxygen content of different types on the surface of CeO 2 particles with varying levels of Zr doping.The peak at approximately 530 eV [42] in the figure corresponds to the lattice oxygen (O L ) within the CeO 2 structure, whereas the peak at approximately 532 eV primarily represents surface vacancies (O V ) [43], and the peak near 533 eV indicates adsorbed oxygen on the surface (O S ).By calculating the ratio of peak areas, it can be determined that ZC3 nanoparticles possess an approximate oxygen vacancy content of 60.69%, which is significantly higher than that of undoped ZC0 (%).This demonstrates that doping with Zr 4+ effectively reduces the lattice oxygen on the CeO 2 surface, resulting in a greater number of oxygen vacancies on ZC3 than on pure CeO 2 .Based on these findings, it can be inferred that calcined samples with a 3% Zr doping content exhibit numerous adsorption sites for CR.
sample, where peg-split fitting was applied to the Zr 3d energy spectrum of ZC3 doped with 3% Zr.The binding energies at 182.4 and 185.6 eV [39] correspond to Zr 3d5/2 and Zr 3d3/2, respectively, suggesting the presence of surface-bound Zr 4+ components in the ZC3 material [40,41].The O 1s energy spectrum was fitted using Advantage software (5.918).The O 1s energy spectra of the undoped pure CeO2 sample ZC0 and doped nanoparticles with 1~5% Zr, as depicted in Figure 6, were analysed using peak fitting to quantify the surface oxygen content of different types on the surface of CeO2 particles with varying levels of Zr doping The peak at approximately 530 eV [42] in the figure corresponds to the lattice oxygen (OL) within the CeO2 structure, whereas the peak at approximately 532 eV primarily represents surface vacancies (OV) [43], and the peak near 533 eV indicates adsorbed oxygen on the surface (OS).By calculating the ratio of peak areas, it can be determined that ZC3 nanoparticles possess an approximate oxygen vacancy content of 60.69%, which is significantly higher than that of undoped ZC0 (%).This demonstrates that doping with Zr 4+ effectively reduces the lattice oxygen on the CeO2 surface, resulting in a greater number of oxygen vacancies on ZC3 than on pure CeO2.Based on these findings, it can be inferred that calcined samples with a 3% Zr doping content exhibit numerous adsorption sites for CR.From Table 2, it is evident that Zr doping induces a plethora of oxygen vacancies tha are dependent on the degree of doping.If the doping level is extremely low, the lattic distortion will not be conspicuous, and the concentration of oxygen vacancies will remai low.Conversely, excessive doping impedes the diffusion of oxygen vacancies onto th sample surface, resulting in a decrease in their concentration.Therefore, an optima threshold exists for CeO2 to achieve the maximum oxygen vacancy content.In this system CeO2 attained its highest oxygen vacancy content (60.69%) when the Zr doping reache 3%.From Table 2, it is evident that Zr doping induces a plethora of oxygen vacancies that are dependent on the degree of doping.If the doping level is extremely low, the lattice distortion will not be conspicuous, and the concentration of oxygen vacancies will remain low.Conversely, excessive doping impedes the diffusion of oxygen vacancies onto the sample surface, resulting in a decrease in their concentration.Therefore, an optimal threshold exists for CeO 2 to achieve the maximum oxygen vacancy content.In this system, CeO 2 attained its highest oxygen vacancy content (60.69%) when the Zr doping reached 3%.

Raman Analysis
The Raman spectra of the Zr-doped CeO 2 are presented in Figure 7.All spectra exhibit a main peak at n = 464 cm −1 , which could be assigned to the F 2g Raman-active mode of the fluorite-structured CeO 2 .An additional small band at around n = 610 cm −1 is caused by the defects related to oxygen vacancies (O v ) [44][45][46][47].No Raman bands that correspond to dopant oxides are detected, which is in agreement with the XRD results.Notably, the F 2g band of the Zr-doped CeO 2 is shifted to a lower wavenumber for relative to pure CeO 2 .The shift in the F 2g mode reveals the variation in the M (metal)-O vibration frequency, which is attributed to the difference in the radii of Ce 4+ and the dopants [44][45][46].The difference in the ionic radius of the dopants enables changes in the lattice parameter caused by cell contraction or expansion.Hence, vibrations are stiffer in a contracted lattice and more relaxed in an expanded lattice so that the bands shift to higher and lower wavenumbers [48], respectively.Further, the incorporation of dopant ions into the CeO 2 changes the relative amount of oxygen vacancies (Figure 7, inset).The Zr-doped CeO 2 and pure CeO 2 samples exhibited only one type of O v band at around n = 610 cm −1 .Generally, the extrinsic vacancies could be formed because of the oxygen vacancies introduced into the CeO 2 to maintain the charge neutrality if Ce 4+ ions are replaced with trivalent cations [46,48], whereas the intrinsic vacancies are generated by the presence of Ce 3+ ions.Therefore, the Zr-doped CeO 2 and pure CeO 2 promote only intrinsic vacancies attributed to the presence of Ce 3+ ions in the samples.
Crystals 2023, 13, x FOR PEER REVIEW 9 of 18 pure CeO2 samples exhibited only one type of Ov band at around n = 610 cm −1 .Generally, the extrinsic vacancies could be formed because of the oxygen vacancies introduced into the CeO2 to maintain the charge neutrality if Ce 4+ ions are replaced with trivalent cations [46,48], whereas the intrinsic vacancies are generated by the presence of Ce 3+ ions.Therefore, the Zr-doped CeO2 and pure CeO2 promote only intrinsic vacancies attributed to the presence of Ce 3+ ions in the samples.
As the bands at n = 610 cm −1 are related to oxygen vacancies and the band at n =464 cm −1 is the main band of the CeO2-based samples, the area ratio of these bands A(Ov)/A(F2g) (Figure 7, inset) reflects the relative concentration of oxygen vacancies.As a result, A(Ov)/A(F2g) follows the sequence ZC3 > ZC2 > ZC4 > ZC5 > ZC1 > ZC0, which indicates that the ZC3 sample had the most defect sites and the ZC0 sample had the least, which is consistent with the conclusions drawn from the results of the XPS.Notably, Raman spectroscopy is essentially a bulk technique.Therefore, the detected oxygen vacancies in the ceria and doped ceria samples are a sum of both the surface and bulk defects.As the particle sizes of all the samples here are small (~11 nm) and similar, it is reasonable to compare the relative amount of defect sites via Raman spectroscopy.

Effect of Zr Doping Amount on Adsorption Properties
To compare the effect of the Zr doping ratio on the adsorption properties of CeO2 nanoparticles, the adsorption properties of samples with different Zr doping levels were compared at room temperature with the other conditions remaining unchanged (the amount of adsorbent added was 10 mg, and the CR solution was 100 mg/L of 100 mL), as shown in Figure 8.The adsorption properties of CeO2 prepared with different Zr doping ratios were significantly improved compared with those of undoped CeO2 nanoparticles, As the bands at n = 610 cm −1 are related to oxygen vacancies and the band at n = 464 cm −1 is the main band of the CeO 2 -based samples, the area ratio of these bands A(O v )/A(F 2g ) (Figure 7, inset) reflects the relative concentration of oxygen vacancies.As a result, A(O v )/A(F 2g ) follows the sequence ZC3 > ZC2 > ZC4 > ZC5 > ZC1 > ZC0, which indicates that the ZC3 sample had the most defect sites and the ZC0 sample had the least, which is consistent with the conclusions drawn from the results of the XPS.Notably, Raman spectroscopy is essentially a bulk technique.Therefore, the detected oxygen vacancies in the ceria and doped ceria samples are a sum of both the surface and bulk defects.As the particle sizes of all the samples here are small (~11 nm) and similar, it is reasonable to compare the relative amount of defect sites via Raman spectroscopy.

Effect of Zr Doping Amount on Adsorption Properties
To compare the effect of the Zr doping ratio on the adsorption properties of CeO 2 nanoparticles, the adsorption properties of samples with different Zr doping levels were compared at room temperature with the other conditions remaining unchanged (the amount of adsorbent added was 10 mg, and the CR solution was 100 mg/L of 100 mL), as shown in Figure 8.The adsorption properties of CeO 2 prepared with different Zr doping ratios were significantly improved compared with those of undoped CeO 2 nanoparticles, which may be due to doping changes in its physical and chemical properties compared with those of pure CeO 2 .Doping has a significant influence on the surface O V of an adsorbent, and the surface O V of vacant oxygen plays an important role in the adsorption process [27,49].The interaction between the ammonium radical (NH 3 + ) in CR and oxygen vacancy O V determines the adsorption efficiency of the adsorbent [43,50,51].The amount of vacant oxygen O V may be affected by the amount of Zr doping [52].The adsorption capacity of the nanoparticles with 3% Zr doping exhibited the best adsorption capacity, reaching 974 mg/g, whereas the adsorption performance of the undoped pure CeO 2 was only 787 mg/g.

Effect of pH
The influence of pH on the CR adsorption performance was evaluated by changing the pH of the solution (pH = 1, 3, 5, 7, 9, 10, 11, and 12) at different initial pH values [50,53].In this study, 0.01 g ZC3 particles were used to adsorb 100 mL of a 100 mg/L CR solution at different initial pH values, as shown in Figure 9.The influence of the pH on ZC0 is shown in Figure S1a.At pHs between 1 and 10, the removal of dyes was high, which can be attributed to the strong electrostatic interaction between the positively charged ZC3 nanoparticles and anionic CR (pH = 1-10) and the hydrogen bonding between the oxygen vacancy and the amino group of the CR molecules.At a higher pH (pH > 10), the significant decrease in the adsorption performance may have been attributed to the high pH neutralising the positive charge on the surface of the ZC3 nanoparticles, resulting in a negatively charged surface.Therefore, the anion of CR has a stronger repulsive effect on the negatively charged ZC3 nanoparticles, which leads to a reduction in the adsorption efficiency.However, because there are a large number of oxygen vacancies on the surface

Effect of pH
The influence of pH on the CR adsorption performance was evaluated by changing the pH of the solution (pH = 1, 3, 5, 7, 9, 10, 11, and 12) at different initial pH values [50,53].In this study, 0.01 g ZC3 particles were used to adsorb 100 mL of a 100 mg/L CR solution at different initial pH values, as shown in Figure 9.The influence of the pH on ZC0 is shown in Figure S1a.At pHs between 1 and 10, the removal of dyes was high, which can be attributed to the strong electrostatic interaction between the positively charged ZC3 nanoparticles and anionic CR (pH = 1-10) and the hydrogen bonding between the oxygen vacancy and the amino group of the CR molecules.At a higher pH (pH > 10), the significant decrease in the adsorption performance may have been attributed to the high pH neutralising the positive charge on the surface of the ZC3 nanoparticles, resulting in a negatively charged surface.Therefore, the anion of CR has a stronger repulsive effect on the negatively charged ZC3 nanoparticles, which leads to a reduction in the adsorption efficiency.However, because there are a large number of oxygen vacancies on the surface of the doped nanoparticles, which is also an important adsorption site for CR molecules, the increased pH can still maintain a certain adsorption performance.The pH (pH > 10) has a considerable effect on the adsorption efficiency, confirming that electrostatic interactions play an important role in the adsorption mechanism; the optimal pH value for adsorption studies was 7, which was used throughout the experiment.As the initial pH value increased from 7 to 12, the removal of CR gradually decreased, and the removal of ZC3 nanoparticles from the CR solution decreased from 98.7% to 44.6%.At a pH of >7, the adsorption performance of ZC3 was significantly better than that of ZC0, indicating that doping effectively improves the acid-base resistance of the adsorbent.

Effect of Contact Time
The change in the adsorption capacity of ZC3 nanoparticles on the CR solution over time is shown in Figure 10 and the effect of the contact time of ZC0 is shown in Figure S1b.There were a large number of vacancies on the surface of the nanoparticle adsorbent, which was conducive to improving the adsorption rate in the initial stage.BET data also confirmed the existence of a large number of mesoporous pores on the surface of ZC3 nanoparticles.Moreover, a high specific surface area of the ZC3 nanoparticle adsorbent ensured that there were sufficient vacancies for small dye molecules to reside in and interact with the adsorbent.Over time, the number of vacancies gradually decreases, owing to increased competition for active adsorption sites.Thus, during dye adsorption, a preliminary process involved the formation of a thin monolayer from the dispersed dye solution on the outer surface of the adsorbent.The CR molecules then diffused from the monolayer to the pores formed by the stacking of the adsorbent nanoparticles.Over time, the mesoporous pores in the adsorbent were filled.Ultimately, this provided resistance to the diffusion of the accumulated dye molecules onto the adsorbent.In this study, the amount of adsorbent was kept constant at 0.01 g/L.Therefore, for lower initial dye concentrations, the removal rate of the CR dye was higher and the time to reach the adsorption equilibrium was shorter.The CR adsorption dynamics included two phases.The first stage was the rapid adsorption stage, in which the adsorption (which may be instantaneous) was mainly on the outer surface.In the second stage, the adsorption gradually slowed.Finally, the adsorption of the two dyes reached an equilibrium.Considering these facts, for a lower initial concentration (100 mg/L), the time to reach an equilibrium was relatively less than that for a higher initial concentration (400 mg/L).Therefore, a constant adsorption dose indicates a high adsorption rate.In the initial stage of the adsorption reaction, the adsorption capacity of the CR molecules on the ZC3 nanoparticles increased with increasing reaction time, and the rate of increase was

Effect of Contact Time
The change in the adsorption capacity of ZC3 nanoparticles on the CR solution over time is shown in Figure 10 and the effect of the contact time of ZC0 is shown in Figure S1b.There were a large number of vacancies on the surface of the nanoparticle adsorbent, which was conducive to improving the adsorption rate in the initial stage.BET data also confirmed the existence of a large number of mesoporous pores on the surface of ZC3 nanoparticles.Moreover, a high specific surface area of the ZC3 nanoparticle adsorbent ensured that there were sufficient vacancies for small dye molecules to reside in and interact with the adsorbent.Over time, the number of vacancies gradually decreases, owing to increased competition for active adsorption sites.Thus, during dye adsorption, a preliminary process involved the formation of a thin monolayer from the dispersed dye solution on the outer surface of the adsorbent.The CR molecules then diffused from the monolayer to the pores formed by the stacking of the adsorbent nanoparticles.Over time, the mesoporous pores in the adsorbent were filled.Ultimately, this provided resistance to the diffusion of the accumulated dye molecules onto the adsorbent.In this study, the amount of adsorbent was kept constant at 0.01 g/L.Therefore, for lower initial dye concentrations, the removal rate of the CR dye was higher and the time to reach the adsorption equilibrium was shorter.The CR adsorption dynamics included two phases.The first stage was the rapid adsorption stage, in which the adsorption (which may be instantaneous) was mainly on the outer surface.In the second stage, the adsorption gradually slowed.Finally, the adsorption of the two dyes reached an equilibrium.Considering these facts, for a lower initial concentration (100 mg/L), the time to reach an equilibrium was relatively less than that for a higher initial concentration (400 mg/L).Therefore, a constant adsorption dose indicates a high adsorption rate.In the initial stage of the adsorption reaction, the adsorption capacity of the CR molecules on the ZC3 nanoparticles increased with increasing reaction time, and the rate of increase was significantly fast.After 30 min, the absorption rate stabilised.

Adsorption Kinetics
Pseudo-first-and second-order kinetic models were used to fit the adsorption data.The linear expressions of the quasi-first-and second-order kinetic models are detailed in (4) and ( 5) [54][55][56], respectively.The adsorption data of the CR molecules by the ZC0 and ZC3 nanoparticles were fitted using pseudo-first-and second-order kinetic models, and the results are shown in Figure S1c,d and Figure 11.The pseudo-second-order kinetic models' t/qt with t exhibited a good linear relationship, as shown in Figures 11b and S1d.By calculating the fitting results, the parameters of the quasi-first-and second-order kinetic models were obtained [57], as presented in Table 3.The fitting results showed that the correlation coefficients R 2 of the quasi-second-order kinetic model for the molecular adsorption of CR by the ZC3 nanoparticles were 0.9932, 0.9966, 0.9967, and 0.9845.The equilibrium adsorption capacity fitted by the pseudo-second-order kinetic model was closer to the experimental results, indicating that the adsorption process of ZC3 nanoparticles on the CR molecules can be fitted by the quasi-second-order kinetic model.As listed in Table S1, the undoped ZC0 also conformed to pseudo-second-order kinetics, indicating chemisorption.

𝑙𝑛 𝑞 𝑞
(4) Here, qe is the CR adsorption capacity at the time of equilibrium, while qt is related to the CR adsorption capacity at time t; K1 is the kinetic rate constant when applying the pseudo-first-order reaction (min −1 ) using the prepared materials of ZC3; and K2 is the rate constant of the pseudo-second-order reaction (g/mg −1 •min −1 ).

Adsorption Kinetics
Pseudo-first-and second-order kinetic models were used to fit the adsorption data.The linear expressions of the quasi-first-and second-order kinetic models are detailed in (4) and ( 5) [54][55][56], respectively.The adsorption data of the CR molecules by the ZC0 and ZC3 nanoparticles were fitted using pseudo-first-and second-order kinetic models, and the results are shown in Figures S1c,d and 11.The pseudo-second-order kinetic models' t/q t with t exhibited a good linear relationship, as shown in Figures 11b and S1d.By calculating the fitting results, the parameters of the quasi-first-and second-order kinetic models were obtained [57], as presented in Table 3.The fitting results showed that the correlation coefficients R 2 of the quasi-second-order kinetic model for the molecular adsorption of CR by the ZC3 nanoparticles were 0.9932, 0.9966, 0.9967, and 0.9845.The equilibrium adsorption capacity fitted by the pseudo-second-order kinetic model was closer to the experimental results, indicating that the adsorption process of ZC3 nanoparticles on the CR molecules can be fitted by the quasi-second-order kinetic model.As listed in Table S1, the undoped ZC0 also conformed to pseudo-second-order kinetics, indicating chemisorption.ln(q e − q t ) = lnq e − K 1 t (4) Here, qe is the CR adsorption capacity at the time of equilibrium, while q t is related to the CR adsorption capacity at time t; K 1 is the kinetic rate constant when applying the pseudo-first-order reaction (min −1 ) using the prepared materials of ZC3; and K 2 is the rate constant of the pseudo-second-order reaction (g/mg −1 •min −1 ).The adsorption properties of the ZC0 and ZC3 nanoparticles for the CR at 298 K were studied.The Langmuir (6) and Freundlich (7) isothermal absorption models [58,59] were used to fit the experimental data for the adsorption of CR on the Zc0 and ZC3 particles, and the results are shown in Figures S1e,f and 12.As summarized in Tables 4 and S2, the correlation coefficients R 2 of the Langmuir and Freundlich models were both greater than 0.9; however, the fitting effect of the Freundlich model was better, indicating that the adsorption process of Zr-doped CeO2 nanoparticles on CR may be a combination of physical adsorption and chemisorption.Physical adsorption is caused by electrostatic attraction, and chemisorption is caused by the interaction between surface oxygen vacancies and dye molecules.A zeta potential analysis was performed on ZC0, ZC3, and CR dyes, as shown in Figure S2.The CR dyes were negatively charged, whereas ZC0 and ZC3 were positively charged.The ZC3 potential increased after doping.A comparison of the adsorption properties with those of other materials is presented in Table S3. 
where qm was the theoretical maximum adsorption capacity of the adsorbent (mg/g); b is one of the important parameters of the Langmuir isotherm adsorption model (L/mg); Ce is the adsorption equilibrium of the solution concentration (mg/L); qe is the unit mass of the adsorbent adsorption capacity (mg/g); and Kf and 1/n are related to the adsorption capacity and adsorption intensity of the Freundlich isothermal adsorption model parameters.The adsorption properties of the ZC0 and ZC3 nanoparticles for the CR at 298 K were studied.The Langmuir (6) and Freundlich (7) isothermal absorption models [58,59] were used to fit the experimental data for the adsorption of CR on the Zc0 and ZC3 particles, and the results are shown in Figure S1e,f and 12.As summarized in Tables 4 and S2, the correlation coefficients R 2 of the Langmuir and Freundlich models were both greater than 0.9; however, the fitting effect of the Freundlich model was better, indicating that the adsorption process of Zr-doped CeO 2 nanoparticles on CR may be a combination of physical adsorption and chemisorption.Physical adsorption is caused by electrostatic attraction, and chemisorption is caused by the interaction between surface oxygen vacancies and dye molecules.A zeta potential analysis was performed on ZC0, ZC3, and CR dyes, as shown in Figure S2.The CR dyes were negatively charged, whereas ZC0 and ZC3 were positively charged.The ZC3 potential increased after doping.A comparison of the adsorption properties with those of other materials is presented in Table S3. ) where q m was the theoretical maximum adsorption capacity of the adsorbent (mg/g); b is one of the important parameters of the Langmuir isotherm adsorption model (L/mg); Ce is the adsorption equilibrium of the solution concentration (mg/L); q e is the unit mass of the adsorbent adsorption capacity (mg/g); and K f and 1/n are related to the adsorption capacity and adsorption intensity of the Freundlich isothermal adsorption model parameters.From Table 5, we can see that the adsorption capacity and removal rate of adsorbent are affected by the amount of adsorbent added and the concentration of the dye.The ZC3 anionic dye prepared by this study still has an excellent adsorption performance when there is a high concentration of dye wastewater, and only a small amount of adsorbent is needed to remove the Congo red anionic dye.

Conclusions
Based on a study on the adsorption process of CR dye, CeO2 nanoparticles synthesised with a 3% Zr doping ratio demonstrated the highest adsorption capacity for CR.These nanoparticles exhibit an excellent adsorption performance across a wide pH range and achieved a maximum adsorption capacity of 3642.05 mg/g at room temperature.The adsorption process followed a quasi-second-order kinetic model, and the isothermal data were consistent with both the Langmuir and Freundlich models.The XPS analysis of different doping ratios of nano-CeO2 combined with CR in the adsorption experiments revealed that oxygen vacancies on the CeO2 surface significantly influenced the adsorption performance of CR and may serve as primary active sites for adsorption.In conclusion, Zr-doped CeO2 nanoparticles have significant potential for the development of and applications for the effective removal of anionic dyes from wastewater because of their straightforward preparation methods and exceptional adsorptive properties.From Table 5, we can see that the adsorption capacity and removal rate of adsorbent are affected by the amount of adsorbent added and the concentration of the dye.The ZC3 anionic dye prepared by this study still has an excellent adsorption performance when there is a high concentration of dye wastewater, and only a small amount of adsorbent is needed to remove the Congo red anionic dye.

Conclusions
Based on a study on the adsorption process of CR dye, CeO 2 nanoparticles synthesised with a 3% Zr doping ratio demonstrated the highest adsorption capacity for CR.These nanoparticles exhibit an excellent adsorption performance across a wide pH range and achieved a maximum adsorption capacity of 3642.05 mg/g at room temperature.The adsorption process followed a quasi-second-order kinetic model, and the isothermal data were consistent with both the Langmuir and Freundlich models.The XPS analysis of different doping ratios of nano-CeO 2 combined with CR in the adsorption experiments revealed that oxygen vacancies on the CeO 2 surface significantly influenced the adsorption performance of CR and may serve as primary active sites for adsorption.In conclusion, Zr-doped CeO 2 nanoparticles have significant potential for the development of and applications for the effective removal of anionic dyes from wastewater because of their straightforward preparation methods and exceptional adsorptive properties.

Figure 1 .
Figure 1.X-ray XRD diffraction pattern of CeO2 doped with varying amounts of Zr.

Figure 1 .
Figure 1.X-ray XRD diffraction pattern of CeO 2 doped with varying amounts of Zr.

Figure 3 .
Figure 3. (a) Low-magnification TEM images of ZC3 nanoparticles; (b) High-magnification TEM images of ZC3 nanoparticles; (c) Image of element distribution of ZC3 nanoparticles; (d) The elemental distribution of Ce; (e) The elemental distribution of O; (f) The elemental distribution of Zr.

Crystals 2023 , 18 Figure 8 .
Figure 8. Influence of Zr doping levels on the adsorption properties of CeO2 nanoparticles.

Figure 8 .
Figure 8. Influence of Zr doping levels on the adsorption properties of CeO 2 nanoparticles.

Figure 9 .
Figure 9. Influence of pH on the adsorption properties.

Table 1 .
Specific surface area and average pore size of Zr-doped CeO2 nanoparticles.

Table 1 .
Specific surface area and average pore size of Zr-doped CeO 2 nanoparticles.

Table 2 .
Content ratio of OL, OV, and OS of Zr-doped CeO2.

Table 2 .
Content ratio of O L , O V, and O S of Zr-doped CeO 2 .

Table 3 .
Kinetic parameters for the adsorption of the CR onto ZC3.

Table 3 .
Kinetic parameters for the adsorption of the CR onto ZC3.

Table 4 .
Isothermal adsorption model of CR molecules on ZC3.

Table 5 .
Various recent reports on the CeO2-and Zr-doped CeO2 for pigmented anionic dye adsorption.

Table 4 .
Isothermal adsorption model of CR molecules on ZC3.

Table 5 .
Various recent reports on the CeO 2 -and Zr-doped CeO 2 for pigmented anionic dye adsorption.