Enhanced Removal of Acid Orange 7 onto Layered Interleaved Symmetrical 3D Flower-like CeO₂ with Y(III) Doping

Abstract

CeO₂ has a potential application in the purification of organic dye wastewater because of the abundant oxygen vacancy (V_O) defects in its crystals. In this study, a cubic CeO₂ microsphere with layered interleaved symmetrical 3D flower-like morphology was synthesized, and its adsorption capacity for acid orange 7 (AO7) was further enhanced by Y doping. The impact of varying amounts of Y ions on the phase composition, lattice parameters, and morphology of the product was investigated, revealing that 4 mol.% was determined as the doping level limit of Y ions in CeO₂ crystals. XPS, Raman, and H₂−TPR techniques were employed to compare surface species changes before and after 4 mol.% Y doping in the CeO₂ crystals, including O−Ce(III), O−Ce(IV), O−Y(III), and V_O correlation, yielding a rough quantitative assessment of these species. The 4 mol.% Y-doped CeO₂ (2.0 g/L) demonstrated the highest removal rate for 20 mg/L of AO7 dye within just 20 min to reach adsorption–desorption equilibrium, half the time required by undoped CeO₂, achieving an impressive adsorption rate of 94.6%, compared to only 69.5% for undoped CeO₂ at 20 min. The adsorption capacity of undoped CeO₂ was enhanced by 19.05% through the doping of 4 mol.% Y, achieving a value of 16.56 mg/L. The feasibility of enhancing the adsorption capacity of CeO₂ by Y doping provides a reference for the application of CeO₂ and other metal oxides.

1. Introduction

The emergence and rapid development of synthetic dyes have had a profound impact on the textile printing and dyeing industry, not only enriching the choice of color, but also improving dyeing efficiency, reducing production costs, and, importantly, promoting the innovation of chemical technology [1]. The application of synthetic dyes is not only limited to the printing and dyeing of textiles, but has also widely penetrated into many fields such as paper making, plastics, leather, rubber, paint, ink, cosmetics, photosensitive materials, and so on, which has profoundly changed industrial production patterns [2,3]. However, it also brings about certain environmental problems, as a large amount of wastewater is produced in the production and use of synthetic dyes, causing pollution to the environment. Therefore, dye removal from textile wastewater is a significant environmental concern, particularly regarding azo dyes, which contain one or more azo bonds. These dyes make up approximately 60~70% of the dyes used in the textile industry. Some of these azo dyes are known to be aquatic toxic or allergenic, and can produce carcinogenic aromatic amines under reducing conditions [4]. In addition to being non-biodegradable, their toxic and potentially carcinogenic properties give rise to a serious threat to aquatic organisms, as well as animals and humans [5].

Traditional physico-chemical processes such as filtration, coagulation, flocculation, ion exchange, and photocatalytic degradation are not suitable for treating large areas of sewage [6,7]. In addition to this limitation, these methods may incur higher costs and can potentially lead to the generation of additional toxic byproducts. Moreover, azo dyes also pose a significant challenge to biological treatment processes due to their high toxicity and non-biodegradability [8]. Among the various water treatment technologies available, the liquid-phase adsorption process, using a solid adsorbent as the stationary phase to adsorb the contaminants in a liquid, has been proven to be one of the most reliable and effective dye removal technologies [9,10]. Although adsorbents usually require regeneration, which increases the cost of the process and can be time-consuming, adsorption remains the most widely used wastewater treatment process today [11]. This is due to its ability to handle considerable flow rates, produce high-quality wastewater, and avoid the formation of toxic and harmful intermediates such as free radicals, ozone, and dye degradants, etc. Currently, activated carbon particles [12] and fibers [13], zeolite [14], and humic-acid- [15] and coal-based [16] adsorbents are widely utilized due to their high specific surface area and excellent adsorption capacity [17]. However, their application is constrained by production costs and the challenges associated with waste regeneration. Consequently, there has been a growing interest in exploring the potential of low-cost agricultural waste as biosorbents [18]. Nevertheless, these cost-effective alternatives either have a limited adsorption capacity or require carbonization, which escalates treatment expenses [19]. Therefore, there remains a significant demand for affordable adsorbent materials with high adsorption capabilities.

There is an urgent necessity to develop cost-effective, uncomplicated, and efficient micro/nano adsorbent materials for the effective removal of various pollutants. Among the various metal oxide semiconductors, cerium dioxide (CeO₂) is widely utilized in polishing materials, catalysts, catalyst carriers or auxiliaries, automobile exhaust absorbers, ultraviolet absorbers, solid oxide fuel cell electrolytes, electronic ceramics, and many other fields due to its exceptional physical and chemical properties and crystallography characteristics. However, there are limited studies on the application of CeO₂ as an adsorbent [20]. A few reports have demonstrated that CeO₂ is effective in removing highly toxic pollutants such as arsenic (V) and chromium (VI), as well as azo dyes like Congo red and acid orange from water. CeO₂ crystals contain abundant oxygen vacancy defects, which can effectively bind with acidic dye molecules (R–SO₃Na) through the electron-rich sulfonic acid group (SO₃⁻) [21]. Therefore, theoretically, CeO₂ is suitable for adsorbing acidic dye molecules, and the interaction between CeO₂ and dye molecules is more stable than physical adsorption.

The doping of various metallic elements into the CeO₂ lattice has been demonstrated to be an effective strategy for enhancing its oxygen storage capacity, catalytic performance, transport properties, and room-temperature ferromagnetism. In this study, yttrium (Y), a rare earth element, was selected as a dopant for incorporation into the CeO₂ lattice based on the similarity–intermiscibility theory. A series of samples with varying ratios of Ce and Y were synthesized using a solvothermal method combined with a calcination process, employing only Ce(NO₃)₃∙6H₂O and Y(NO₃)₃∙6H₂O as raw materials, along with a solvent mixture of ethylene glycol and distilled water. The introduction of trace amounts of Y (4 mol.%) not only reduced the adsorption time required for CeO₂ to reach adsorption–desorption equilibrium with AO7 dye by half-achieving equilibrium in just 20 min, but also enhanced its adsorption capacity by 19.05%. This innovative approach achieved the optimal adsorption properties with minimal levels of impurity elements, offering a promising alternative for merchants seeking advanced solutions for treating dye wastewater.

2. Materials and Methods

2.1. Starting Materials

Ce(NO₃)₃∙6H₂O (99.95%), Y(NO₃)₃∙6H₂O (99.9%), ethylene glycol (99.8%), and acid orange 7 (AO7, 98.0%) were supplied by Aladdin Co., Ltd. (Shanghai, China). The water used in the experiment was distilled water.

2.2. Synthesis of Y-Doped CeO₂

Firstly, the desired amounts of Ce(NO₃)₃∙6H₂O and Y(NO₃)₃∙6H₂O with different Y/(Y + Ce) (mol.%) were dissolved in a solution consisting of 25 mL of ethylene glycol and 5 mL of distilled water, resulting in a total amount of 4.0 mmol of Ce³⁺ and Y³⁺ ions. Subsequently, the mixed solution was transferred into a Teflon-lined stainless-steel autoclave and sealed for the solvothermal process, lasting for 24 h at a temperature of 200 °C. Following the reaction, the resulting precipitates were collected by centrifugation and washed alternately with distilled water and ethanol, then dried in air at 80 °C for 12 h. Finally, a series of Y-doped CeO₂ powders were obtained through calcination in air at 500 °C for 2 h.

For comparison purposes, two reference samples were also synthesized using an identical procedure as controls, but without Ce(NO₃)₃∙6H₂O or Y(NO₃)₃∙6H₂O.

2.3. Characterization

The phases of the samples were determined by X-ray diffraction analysis (XRD, 40 mA, 40 kV; Rigaku D/MAX 2200 PC, Rigaku, Tokyo, Japan). The morphologies of the samples were observed by field-emission scanning electron microscopy (SEM, 20 kV; JEOL–7500F, Tokyo, Japan). The binding energy and surface composition of undoped and 4 mol.% Y-doped CeO₂ were determined by X-ray photoelectron spectroscopy (XPS, Monochromated Al Kalph 150W; ESCALAB 250, Thermo Scientific, Waltham, MA, USA). The natures of the surface V_O defects were identified using Raman spectroscopy with a He–Cd laser of 325 nm (LabRAM Aramis, Horiba Jobin Yvon, Paris, France). The V_O defects in the undoped and 4 mol.% Y-doped CeO₂ were quantified by hydrogen temperature-programmed reduction (H₂–TPR) measurements (TP–5080, Tianjin Xianquan industry and Tradde Development Co., Ltd., Tianjin, China). Furthermore, 0.05 g CeO₂ powders were initially treated in a 5% O₂/N₂ stream at 500 °C for 1 h, then cooled and purged with N₂ to remove excess O₂. Subsequently, a flow of 5% H₂/N₂ was introduced into the reactor at a flow rate of 30 mL/min, and the temperature was raised to ~960 °C with a heating rate of 10 °C/min. The specific surface areas of the samples were determined using the Brunauer–Emmett–Teller method, based on nitrogen (N₂) adsorption–desorption isotherms obtained from a QuadraSorb SI (Quantachrome, Boynton Beach, FL, USA).

2.4. Removal of AO7 Dye

AO7 azo dye was selected as the model target to evaluate the removal capacity of the CeO₂ product. The adsorption experiments were conducted without pH pre-adjustments in the dark at a constant ambient temperature (26 °C). Briefly, the resulting 0.2 g powders were dispersed into 100 mL of AO7 solution (20~40 mg/L) under continuous stirring (200 rpm) until the suspension was withdrawn at designated time (10~60 min). After filtering out about 5 mL of the clear liquid, it was immediately taken to test its absorbance using a U−3900 Uv–vis spectrophotometer (Hitachi, Tokyo, Japan). The removal rate at the set time, t, labeled as η_t (%), was calculated using Equation (1). Significantly, the adsorption experiments were conducted in triplicate to ensure credibility, and the resultant average was utilized for the subsequent analysis.

$${\eta }_{\mathrm{t}}(\%)={\displaystyle \frac{A_0-A_{\mathrm{t}}}{A_0}}\times 100$$

(1)

where A₀ represents the absorbance of a 20 mg/L AO7 solution at 485 nm and A_t represents the absorbance of the AO7 filtrate at 485 nm at a specific time t.

3. Results and Discussion

XRD was utilized to investigate the impact of the incorporation of Y elements on the phase composition and crystallographic structure of the samples. Figure 1a presents the XRD patterns of the samples synthesized without Y. All discernible diffraction peaks were consistent with the standard CeO₂ pattern (Cubic, Fm–3m(225); JCPDS no. 34−0394), and no impurity phases, such as CeCO₃ or CeCO₃OH, were detected. For the sample synthesized with a 10 mol.% Y content (Figure 1b), the XRD diffraction peak was similar to that of the undoped CeO₂ in Figure 1a, however, there was a slight but not negligible shift towards lower 2θ values. This shift might be attributed to the partial substitution of host Ce⁴⁺ (0.97 Å) ions by larger Y³⁺ (1.02 Å) ions and the resulting local lattice expansion of the CeO₂ crystal structure. For the samples synthesized with 30 mol.% (Figure 1c) and 50 mol.% (Figure 1d) contents, their XRD diffraction peaks exhibited similar peak positions and shapes to those of the undoped CeO₂ in Figure 1a, without any noticeable shifts. However, when the amount of Y increased to 70 mol.%, only four smooth diffraction peaks appeared in the XRD pattern of the obtained sample (Figure 1e). In contrast, when no Ce was added, only three smooth diffraction peaks with a lower intensity were observed in the XRD pattern of Figure 1f, and these could be matched with the three strong peaks characteristic of standard Y₂O₃ (Cubic, Fm–3m(225); JCPDS no.43–0661). This suggested that either a low crystallinity or amorphous substances were present in these obtained Y₂O₃ samples.

When impurity ions entered the lattice, they induced lattice distortion, which could be discerned by alterations in the lattice parameters to determine whether the lattice had expanded or contracted. In this work, the lattice parameters of the resulting cubic samples were determined based on Bragg’s equation, and the findings are presented in Figure 2. Specifically, the lattice parameter of undoped CeO₂ (i.e., 0 mol.%) was measured at 5.4117 Å, while that of CeO₂ doped with 10 mol.% Y was found to be slightly larger at 5.4227 Å. This observation aligns with the phenomenon depicted in Figure 1b, wherein a shift in the XRD diffraction peak towards lower 2θ values was evident for the doped sample compared to its undoped counterpart. The introduction of Y was shown to cause an expansion of the CeO₂ lattice, indicating a significant influence on its structural properties. With the gradual addition of Y, the lattice parameters of the cubic samples decreased. When 70 mol.% of Y was added, the resulting sample showed a lattice parameter of 5.3891 Å, which closely approximated the theoretical lattice parameter of cubic Y₂O₃ (5.2644 Å).

SEM analysis was undertaken to elucidate the morphological characteristics of the samples synthesized with varying dosages of Y. In the sample synthesized without Y, namely, the undoped cubic CeO₂ (Figure 3a), the morphology exhibited a multilayered structure consisting of interwoven nanodisk sheets (20 nm) with a diameter of approximately 5 μm. Upon introducing 10 mol.% of Y (Figure 3b), the multilayered structure remained unchanged. However, with an increase to 30 mol.% of Y (Figure 3c), a new morphology emerged in the form of coral-like aggregate particles alongside the original multilayered structure. Subsequently, at a 50 mol.% addition of Y (Figure 3d), the multilayered structure was completely replaced by coral-like aggregate particles. Notably, as the Y content increased to 70 mol.% (Figure 3e), there was a reduction in the size observed for these coral-like aggregate particles. Conversely, in the sample synthesized without Ce, namely, cubic Y₂O₃ (Figure 3f), two distinct types of coral-like aggregates were identified, one large and one small in size. Based on the shifts of 2θ in Figure 1, the variation in lattice parameters in Figure 2, and the evolution of morphology in Figure 3, it can be concluded that the doping levels for Y within CeO₂ were controlled within a range not exceeding 10 mol%.

Figure 4a illustrates the time-dependent adsorption profiles of AO7 dye in the presence of undoped CeO₂ (i.e., without Y). The removal rate within 10 min of contact reached 60.2% at an initial AO7 concentration of 20 mg/L. As the adsorption reaction progressed, the process was mostly complete within 40 min, indicating that adsorption–desorption equilibrium on the undoped CeO₂ surface was established within this timeframe. The removal rate within 60 min was found to be 83.6%. The rapid and efficient removal of AO7 within 10 min might be attributed to the presence of special oxygen vacancy defects (V_O), which could be enhanced by the introduction of Y ions, while the removal rate at adsorption–desorption equilibrium within 60 min represents the adsorption capacity of CeO₂. Therefore, we provide a comparison of the removal rates for all samples at both 10 and 60 min intervals, as shown in Figure 4b. For samples with 10 and 15 mol.% of Y additions, the removal rates were nearly identical at 10 min, reaching 76.6% and 76.3%, respectively, which surpassed the 60.2% removal rate observed for undoped CeO₂. Furthermore, at 60 min, the removal rates increased to 88.2% and 87.2%, slightly exceeding the value of 83.6% for undoped CeO₂. In contrast, for the sample with an Y addition of 30 mol.%, there was a significant decrease in the removal rate at both time points, dropping to only 41.6% at 10 min and further declining to just 60.2% at 60 min. As the Y addition increased to 50 mol.%, there was a sharp decline in the removal rate of the obtained sample, plummeting to only 3.0% at 10 min and 5. 4% at 60 min. Notably, for Y additions of 70 mol.% and 100 mol.% (i.e., without Ce, or Y₂O₃), their respective removal rates for AO7 were negligible. It was evident that Y₂O₃ exhibited no adsorption capacity for AO7.

Figure 5a illustrates the lattice parameters of undoped CeO₂ (namely, 0 mol.%), as well as doped CeO₂ with 1~9 mol.% Y, which provides a comprehensive view of the structural changes induced by doping in the CeO₂ lattice. As observed, the lattice parameters of the resulting CeO₂ sample exhibited a gradual increase with the addition of Y, ultimately reaching a maximum value of 5.4242 Å at a 4 mol.% Y addition. Subsequently, there was a slight decrease in the lattice constant with further increases in Y addition. The inflection point of the lattice constant curve can be regarded as the doping limit of Y ions in the CeO₂ lattice, indicating that the solid solubility of Y in CeO₂ was 4 mol%. The inset in Figure 5a presents the SEM image of the 4 mol.% Y-doped CeO₂ sample. In comparison to the undoped CeO₂ (Figure 3a) and the 10 mol.% Y-doped CeO₂ (Figure 3b), this sample retained a multilayered structure. Figure 5b displays the XRD pattern for the 4 mol.% Y-doped CeO₂ sample, which features sharp peaks, with all diffraction peaks aligning with those of standard cubic CeO₂ (Fluorite, JCPDS no. 34–0394). The indexed planes include (111), (200), (220), (311), (222), (400), (331), and (420). These findings were consistent with the XRD analysis results presented in Figure 1. Notably, no signals corresponding to Y₂O₃ or other Y compounds were detected in the 4 mol.% Y-doped CeO₂ sample.

XPS analysis was utilized to examine the chemical composition and oxidation states of the CeO₂ surface before and after Y doping, with the aim of gaining insight into any potential alterations. Figure 6a depicts the wide-scan XPS spectra of undoped and 4 mol.% Y-doped CeO₂. All wide-scan XPS spectra exhibited distinct CeO₂ features, as evidenced by the signals of Ce 3p3, Ce 3d, Ce Auger, Ce 4p3, Ce 4d, and O 1s, which were in excellent agreement with previously reported XPS patterns of both undoped and Y-doped CeO₂ [22,23,24]. Notably, the prominent C 1s peaks observed at approximately 284.8 eV were attributed to adventitious carbon present for sample calibration purposes. Furthermore, a faint Y 3d signal was visible in the vertical ellipse, and the corresponding Y 3d XPS region is shown in the inset of Figure 6a. The presence of an Y 3d characteristic peak in the XPS spectrum strongly suggested the existence of Y(III) species within the lattice structure of CeO₂. This finding provides compelling evidence for the incorporation of Y ions into this crystal framework, thereby enriching our understanding of the complex interplay between Y and Ce oxides at a molecular level. In order to investigate the chemical states of Ce in CeO₂, the Ce 3d XPS core-level regions of undoped and 4 mol.% Y-doped CeO₂ were recorded, as depicted in Figure 6b. Upon comparison with the Ce 3d XPS region of undoped CeO₂, no significant alterations in peak shape and binding energy were discerned for the 4 mol.% Y-doped CeO₂, suggesting that the Y doping had minimal impact on the Ce species in CeO₂. To delve into the chemical states of O in CeO₂, the O 1s XPS core-level regions of the undoped and 4 mol.% Y-doped CeO₂ were recorded and fitted, as illustrated in Figure 6c and d. For undoped CeO₂, its O 1s XPS region could be curve-fitted into three peaks. The peaks with binding energies of approximately 528.3 and 529.8 eV could be attributed to the lattice oxygen of O−Ce(III) and O−Ce(IV) species, respectively, while that at approximately 531.6 eV could be assigned to chemisorbed oxygen or/and weakly bonded oxygen species associated with V_O defects. As for the 4 mol.% Y-doped CeO₂, alongside the aforementioned three peaks (representing O−Ce(III), O−Ce(IV), and V_O correlation), a new curve was fitted at around 527.9 eV, which might be ascribed to corresponding O−Y(III) species. Furthermore, the relative V_O content could be approximated by determining the ratio of the integrated area of the peak associated with the V_O defect to that of all peaks. The calculated value for 4 mol.% Y-doped CeO₂ was found to be 27.2%, which was notably higher than that of undoped CeO₂ at 13.4%. This implied that the incorporation of Y doping had a positive impact on the generation of V_O in CeO₂, suggesting an advantageous influence on its overall performance.

Raman scattering was utilized to explore the structural characteristics of the undoped and 4 mol.% Y-doped CeO₂, owing to its sensitivity towards V_O defects. This advanced technique allowed for a comprehensive investigation into the composition and arrangement of these materials, shedding light on their intricate properties. In Figure 7a, the Raman spectrum of undoped CeO₂ is elegantly fitted into three distinct peaks. The peaks observed at Raman shifts of approximately 419.9 cm⁻¹ and 463.7 cm⁻¹ were confidently attributed to the lattice oxygen of O−Ce(III) and O−Ce(IV) bonds, respectively, while the peak at approximately 583.5 cm⁻¹ was unequivocally assigned to the oxygen species associated with V_O defects. In contrast, for the 4 mol.% Y-doped CeO₂, in addition to three peaks corresponding to O−Ce(III), O−Ce(IV), and V_O correlations, a novel curve emerged at around 346.4 cm^–1, which was ascribed to the O–Y(III) bond. Upon comparison of the Raman spectra of the undoped and 4 mol.% Y-doped CeO₂, it became evident that a significant transformation occurred in two peaks at 580 cm⁻¹ and 460 cm⁻¹. In the case of undoped CeO₂, the dominant Raman peak belonged to the Ce species, whereas after Y doping, the Vo-related peak took precedence. This radical shift in peak dominance underscores the impact of Y doping on the Raman spectra of CeO₂. Moreover, akin to the XPS analysis, the relative V_O and O−Ce(III) contents were also estimated by determining the ratio of the integrated area of the peak associated with the VO defect or O−Ce(III) to that of all peaks, respectively. The calculated values for V_O and O−Ce(III) in the 4 mol.% Y-doped CeO₂ were determined to be 64.9% and 66.6%, which significantly exceeded those of undoped CeO₂ at 25.0% and 32.1%. Additionally, the relative O−Y(III) content was approximated to be 4.6%. This suggests that undoped CeO₂ inherently possesses a certain number of V_O defects and Ce(III) species, while Y doping promotes the presence of more species.

To more precisely ascertain the quantity of V_O defects, H₂ was employed as a reducing agent in H₂−TPR apparatus to facilitate the reduction of the CeO₂ sample while undergoing a continuous temperature elevation. Figure 8 illustrates the H₂–TPR profile of the 4 mol.% Y-doped CeO₂ alongside that of the undoped CeO₂ for comparison. In Figure 8a, the H₂–TPR spectrum of undoped CeO₂ reveals a reduction process occurring at approximately 200 °C, with two distinct peaks in H₂ consumption observed at around 500 °C and 770 °C. The reduction peak around 500 °C was attributed to the direct reduction in surface oxygen species by H₂, while the peak around 600 °C was associated with a reduction in bulk oxygen that migrated to the CeO₂ surface through V_O defects and reacted with H₂. In contrast to undoped CeO₂, the 4 mol.% Y-doped CeO₂, as shown in Figure 8b, exhibited a certain ability to be reduced at room temperature up to 200 °C, with a shift in the low-temperature reduction peak towards higher temperatures by approximately 14 °C. Additionally, an observable shoulder from 360 °C to 440 °C appeared. Furthermore, the reduction band at about 600 °C was significantly elevated compared to baseline levels. These phenomena suggested that Y doping optimized the surface states of CeO₂, thereby enhancing its redox properties. For the solid adsorbents and catalysts, physicochemical reactions mainly occurred on the surface of materials, so it is crucial to discuss their surface properties. The reduction peak at 500 °C in Figure 8 corresponds to the adsorption of oxygen by V_O on the surface/sub-surface of CeO₂. Therefore, the V_O content on the surface and sub-surface of CeO₂ was quantified using the amount of H₂ consumption per gram of CeO₂ powders (mmol H₂/g CeO₂) by measuring the corresponding peak areas of the H₂–TPR profiles. For the undoped CeO₂, the value of H₂ consumption per gram from 20 °C to 593 °C was 0.23 mmol/g, while for the 4 mol.% Y-doped CeO₂ from 20 °C to 612 °C, it was measured at 0.39 mmol/g, with an increase of 70.6%.

According to the analyses conducted via XPS in Figure 6, Raman scattering in Figure 7, and H₂−TPR in Figure 8, it was evident that undoped CeO₂ contained a significant number of intrinsic V_O defects. The formation and filling of these vacancies were accompanied by the release and storage of oxygen atoms within the CeO₂ lattice, leading to a substantial deviation from stoichiometry in the atmosphere. This resulted in the formation of nonstoichiometric oxide CeO_2-δ, as expressed by Equation (2). During the synthesis of Y-doped CeO₂, the primary phase component identified was cerium carbonate hydroxide (CeCO₃OH) (JCPDS no. 52−0352), which was initially obtained through a solvothermal process. The Y compound either adhered to the surface of CeCO₃OH or integrated into its lattice structure. Subsequently, calcination in the air at 500 °C facilitated not only the phase transformation from CeCO₃OH to CeO₂, but also enabled doping modification of the CeO₂ lattice with Y ions, corroborated by our previous report [25]. Following doping with Y³⁺ cations, CeO₂ maintained its fluorite crystal structure (refer to Figure 1 and Figure 5b), while concurrently resulting in the considerable generation of extrinsic V_O defects, as indicated by the Kroger and Vink notations in Equation (3).

$$\mathrm{C}\mathrm{e}{\mathrm{O}}_2\underset{\mathrm{O}\mathrm{x}\mathrm{y}\mathrm{g}\mathrm{e}\mathrm{n} \mathrm{s}\mathrm{t}\mathrm{o}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}{\overset{\mathrm{O}\mathrm{x}\mathrm{y}\mathrm{g}\mathrm{e}\mathrm{n} \mathrm{r}\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{a}\mathrm{s}\mathrm{e}}{\rightleftarrows }}\mathrm{C}{\mathrm{e}}_{1-x}^{4+}\mathrm{C}{\mathrm{e}}_x^{3+}{\mathrm{O}}_{2-{\displaystyle \frac{x}{2}}}{(V_{\mathrm{O}})}_{{\displaystyle \frac{x}{2}}}+{\displaystyle \frac{x}{4}}{\mathrm{O}}_2$$

(2)

$${\mathrm{Y}}_2{\mathrm{O}}_3\stackrel{{2\mathrm{CeO}}_2}{\rightleftarrows }{2\mathrm{Y}}_{\mathrm{Ce}}^{\prime }+3{\mathrm{O}}_{\mathrm{O}}^{\times }+V_{\mathrm{O}}^{••}$$

(3)

where ${\mathrm{Y}}_{\mathrm{Ce}}^{\prime }$ denotes a Y³⁺ cation that occupies the site of a Ce⁴⁺ cation. ${\mathrm{O}}_{\mathrm{O}}^{\times }$ refers to a lattice oxygen atom. $V_{\mathrm{O}}^{••}$ represents an V_O defect with two positive charges, respectively, which is produced through the vacancy compensation mechanism.

The temporal variations in the AO7 removal rates in the presence of undoped and doped CeO₂ with varying Y concentrations are illustrated in Figure 9a. As observed, it was noted that all CeO₂ samples exhibited the most rapid rates of AO7 removal within the initial 10 min of the adsorption process. In the case of Y-doped CeO₂ samples with a solid solubility lower than 4 mol.%, the sequence of removal rates at 10 min unfolded as follows: 93.7% (4 mol.% Y) > 85.0% (3 mol.% Y) > 82.0% (2 mol.% Y) > 81.5% (1 mol.% Y) > 60.2% (undoped). The removal rates of the CeO₂ samples at this juncture experienced a decline with an increase in the quantity of Y, specifically only 71.5% for the CeO₂ with a higher concentration of Y at 9 mol.%. Most importantly, for 1~8 mol.% Y-doped CeO₂, the adsorption–desorption equilibrium was basically reached at 20 min of the adsorption reaction, and the time to reach adsorption–desorption equilibrium was reduced by half compared with 40 min for undoped CeO₂. The order of adsorption rate at 20 min was as follows: 94.6% (4 mol.% Y) > 92.3% (3 mol.% Y) > 91.3% (2 mol.% Y) > 91.0% (1 mol.% Y) > 69.5% (undoped). For CeO₂ doped with 5~8 mol.% of Y, the removal rates at 20 min remained consistently high, ranging between 91 and 93%. However, the adsorption rate for CeO₂ doped with 9 mol.% of Y was notably lower at only 77.7%. Considering the varying removal rates of CeO₂ with different Y addition amounts for AO7 dye, businesses are advised to integrate their specific requirements into the selection process for suitable products in the purification treatment of dye wastewater. Factors such as the cost of Y element addition, the wastewater treatment cycle, and other practical considerations should be taken into account in practical applications. Moreover, the specific surface areas of the undoped, 2 mol.%, 4 mol.%, and 9 mol.% Y-doped CeO₂ were estimated based on N₂ adsorption–desorption experiments conducted using the Brunauer–Emmett–Teller method. The results are presented in Figure 9b as a histogram. The specific surface areas for the undoped, 2 mol.%, 4 mol.%, and 9 mol.% Y-doped CeO₂ were found to be 96.0, 102.0, 98.1, and 98.4 m²/g, respectively. These findings indicate that low-concentration Y doping had a minimal impact on the specific surface area of CeO₂ products. In conjunction with the analyses of the morphology and specific surface area of both undoped and Y-doped CeO₂ samples, it can be concluded that the enhanced adsorption capacity might be attributed to the incorporation of Y³⁺ ions into the CeO₂ lattice, which partially substituted for host Ce⁴⁺ ions and promoted the formation of additional V_O defects.

The effects of the initial concentration of AO7 on its removal rate within 60 min of the reaction are illustrated in Figure 10a. As observed, the removal rates for both undoped and 4 mol.% Y-doped CeO₂ decreased with increasing initial concentrations of AO7. This decline could be attributed to the limited number of adsorption sites available on the adsorbent. Moreover, the saturated adsorption capacity of AO7 was determined using the Langmuir isotherm model (Equations (4) and (5)) [25]. The Langmuir linear fits for the experimental data regarding the adsorption of AO7 dye onto both undoped and 4 mol.% Y-doped CeO₂ are presented in Figure 10b, and the corresponding Langmuir parameters calculated are listed in

Table 1. Estimated parameters of Langmuir linear fitting for AO7 molecule adsorption onto either undoped or 4 mol.% Y-doped CeO₂.

Model	Langmuir Isotherm Model
Parameters	qm (mg/g)	KL	R2
Undoped CeO2	13.91	0.4174	0.9971
4 mol.% Y-doped CeO2	16.56	1.4724	0.9945

. The saturated adsorption capacities (q_m) of AO7 were found to be 13.91 mg/g for undoped CeO₂ and 16.56 mg/g for 4 mol.% Y-doped CeO₂, as determined through Langmuir linear fitting. The high correlation coefficients (R²) of 0.9971 and 0.9945 indicated that the Langmuir isotherm model provided an excellent fit for modeling the adsorption of AO7 onto both undoped and 4 mol.% Y-doped CeO₂.

$$q_e={\displaystyle \frac{(C_0-C_{\mathrm{e}})V}{m}}$$

(4)

$${\displaystyle \frac{C_{\mathrm{e}}}{q_{\mathrm{e}}}}={\displaystyle \frac{1}{q_{\mathrm{m}}}}{C}_{\mathrm{e}}+{\displaystyle \frac{1}{K_{\mathrm{L}}{q}_{\mathrm{m}}}}$$

(5)

where C₀ (mg/L) and C_e (mg/L) are the initial concentration and the concentration of AO7 solution at equilibrium, respectively. q_e (mg/g) is the amount of AO7 adsorbed per gram of CeO₂ at equilibrium, corresponding to a specific initial concentration of AO7. q_m (mg/g) is the maximum amount of AO7 molecules adsorbed per gram of CeO₂. K_L is the Langmuir constant related to the energy of adsorption.

The possible adsorption mechanisms between AO7 molecules and the Y-doped CeO₂ adsorbent are illustrated in Figure 11. In an aqueous solution, the SO₃^– groups within the AO7 molecule dissociated, resulting in the formation of anionic dye ions. The hydroxyl groups on the surface of CeO₂ (Ce–OH) could be protonated to form Ce–OH₂⁺ species. The adsorption process was further facilitated by electrostatic attraction between these oppositely charged ions, as depicted in Figure 11a. Additionally, bidentate structures formed between SO₃^– groups and Ce⁴⁺ cations, representing another significant pathway for CeO₂ to effectively capture AO7 molecules, as shown in Figure 11b,c. Furthermore, V_O defects might have induced reactive oxygen species, enhancing the potential of CeO₂ as a photocatalyst for organic matter degradation.

4. Conclusions

Three-dimensional flower-like CeO₂ with a layered interleaved structure was synthesized, and its removal rate for AO7 dye was optimized by incorporating 10 mol.% Y while maintaining its original morphology. The doping limit of Y ions in CeO₂ was established at 4 mol.% with a positive trivalent oxidation state. The introduction of 4 mol.% Y ions resulted in an increased abundance of O−Ce(III) and V_O defects in the CeO₂ crystals. The quantity of surface/sub-surface V_O was measured at 0.23 mmol/g for undoped CeO₂, rising to 0.39 mmol/g following 4 mol.% Y doping based on H₂–TPR. The incorporation of 1~8 mol.% Y ions could shorten the adsorption equilibrium of CeO₂ by half, bringing it down to just 20 min, and the maximum removal rate of AO7 was achieved with 4 mol.% Y-doped CeO₂, reaching an impressive value of 94.6%, in contrast to only 69.5% for undoped CeO₂. The adsorption of AO7 onto both undoped and 4 mol.% Y-doped CeO₂ could be well described by the Langmuir isotherm model. The saturated adsorption capacity of AO7 dye on 4 mol.% Y-doped CeO₂ was 16.56 mg/g, as determined through Langmuir linear fitting, representing an enhancement of 19.05% compared to the undoped CeO₂. The strategy of doping CeO₂ with trace amounts of rare earth elements to enhance its adsorption capacity for AO7 dye is a viable approach, offering an innovative alternative method for improving the properties of CeO₂ and other metal oxides. In subsequent experiments, we will utilize 4 mol.% Y-doped CeO₂ as a catalyst to investigate its photocatalytic degradation characteristics concerning AO7 dyes under simulated ultraviolet light or natural sunlight.