Nanostructured CeO₂-C Derived from Ce-BDC Precursors for Room-Temperature Ammonia Sensing

Abstract

The prompt and reliable detection of NH₃ leakage at room temperature (RT) is considered important for safety assurance and sustainable production. Although chemiresistive NH₃ sensors feature low cost and structural simplicity, their practical application is hindered by high operating temperatures and inadequate selectivity. Metal–organic frameworks (MOFs) and their derivatives offer a promising approach to address these limitations. In this work, Ce-BDC precursors with tunable particle sizes and crystallinity were synthesized by adjusting the raw material concentration. Controlled pyrolysis yielded a series of CeO₂-C-X (X = 0.5, 1, 1.5, 2) materials with nanosized particles. Among them, the CeO₂-C-1 sensor delivered a high response of 82% toward NH₃ under 40% relative humidity at RT. Moreover, it possessed excellent selectivity, repeatability, and rapid response-recovery behavior compared with the other samples. CeO₂-C-1 also remained stable under varying oxygen and humidity conditions, demonstrating high applicability. The superior sensing properties may be attributed to its high specific surface area and optimized mesoporous structure, which facilitated efficient gas adsorption and reaction. These findings demonstrated that precise control of MOF precursors and the structure in CeO₂ nanomaterials was critical for achieving high-performance gas sensing and established Ce-MOF-derived CeO₂ as a promising sensing material for NH₃ detection at RT.

1. Introduction

In recent years, gas-sensing materials and devices have attracted extensive attention in various fields such as environmental monitoring, industrial process control, and health management [1,2]. Ammonia (NH₃) is commonly employed in NH₃-selective catalytic reduction (NH₃-SCR) systems for the removal of NO_x from sintering flue gas and automobile exhaust [3]. However, excessive dosing can result in the release of unreacted NH₃ into the atmosphere [4]. Meanwhile, with NH₃ emerging as an efficient hydrogen carrier in the context of clean energy technologies, the risk of leakage has become a growing concern due to its high volatility [5,6]. NH₃ can participate in the formation of PM_2.5 by reacting with atmospheric SO₂ and NO_x to generate secondary inorganic aerosols [7]. Additionally, its strong irritant properties pose serious risks to the human respiratory system and eyes, and high concentrations (>50 ppm) can lead to irreversible damage [8,9]. Therefore, the prompt and accurate detection of its leakage is of paramount importance for both safety assurance and environmental sustainability.

Among various types of NH₃ gas sensors, chemiresistive sensors have attracted extensive attention due to their simple architecture, low cost, ease of miniaturization, and compatibility with online monitoring [10,11]. However, the high operating temperature and low selectivity of conventional NH₃ chemiresistive sensors have limited their widespread application [12]. Metal–organic frameworks (MOFs) have been regarded as promising materials for gas sensing applications due to their tunable porosity, high surface area, and structural diversity [13]. Most MOFs suffer from poor intrinsic conductivity due to the insulating nature of organic linkers and the lack of continuous electron pathways except for two-dimensional MOFs, limiting their direct use in chemiresistive sensing [14,15]. Pyrolyzing MOFs into metal oxide semiconductors (MOS) effectively overcomes these issues by enhancing conductivity, preserving high surface area and porosity, inheriting some intrinsic physical properties of the parent MOFs, and improving defect sites in the resulting metal oxides [16,17]. Song et al. [18] prepared ZIF-67-derived Co₃O₄ with a hollow structure, which exhibited abundant surface defects, leading to enhanced oxygen vacancy formation and improved NH₃ sensing performance. Ren et al. [19] prepared porous ZnO derived from ZIF-8, and the optimal sample (ZIF-8-500) exhibited excellent NO₂ sensing performance attributed to its mesoporous structure, high surface area, and abundant oxygen vacancies. MOF-derived materials have exhibited tremendous potential in sensing applications [20,21]. To the best of our knowledge, few studies have yet reported the use of Ce-MOF-derived CeO₂ for room-temperature gas sensing.

Herein, nano-MOF materials of Ce-BDC were successfully synthesized by adjusting the concentrations of Ce⁴⁺ and the organic ligand of BDC (terephthalic acid). A series of Ce-BDC-X (X = 0.5, 1, 1.5, 2) derivatives were subsequently obtained via controlled pyrolysis at appropriate temperatures and employed for NH₃ sensing at room temperature (RT). Notably, after thermal treatment, Ce-BDC precursors with varying average particle sizes and crystallinities yielded CeO₂ nanoparticles with residual carbon from organic linkers, which were labeled as CeO₂-C-X series. Among them, Ce-BDC-1 and its derivative CeO₂-C-1 maintained the closest particle size before and after pyrolysis and exhibited the best NH₃ gas-sensing performance. Furthermore, the selectivity and repeatability of the CeO₂-C-1 based sensing device were comprehensively evaluated. Under 40% relative humidity at RT, it demonstrated a high response of 82% toward NH₃ than others. In addition, CeO₂-C-1 sensor showed excellent response-recovery behavior (22 s/66 s) and selectivity at RT. The effects of oxygen and humidity on gas-sensing performance were investigated, indicating that the sensor maintained a relatively stable response even under increased humidity and oxygen-free conditions, thereby enhancing its applicability. Finally, the NH₃ sensing mechanism of CeO₂-C-1 at RT was proposed.

2. Experiment Section

2.1. Fabrication of CeO₂-C-X Gas Sensing Materials

Firstly, 0.2832 g (0.5 mmol) ammonium cerium nitrate (CAN, (NH₄)₃Ce(NO₃)₆) was dissolved in 8 mL of deionized water, while 0.0839 g terephthalic acid (H₂BDC, C₈H₆O₄) was dissolved in 24 mL of N,N-dimethylformamide (DMF). The molar ratio of CAN to H₂BDC was about 1. After complete dissolution under magnetic stirring, the CAN solution was slowly added to the H₂BDC solution and stirred for another 10 min to form a homogeneous mixture. The solution was then transferred to a Teflon-lined autoclave (XIUILAB X-SF-G100, Shanghai, China) and heated at 100 °C for 1 h. After natural cooling, the product was collected by centrifugation and washed with DMF and deionized water. The resulting precipitate was dried at 60 °C for approximately 12 h to obtain Ce-BDC-0.5. Similarly, the other Ce-BDC-X (X = 1, 1.5, 2) samples were synthesized by increasing the CAN and H₂BDC amounts (CAN amounts: 1, 1.5 and 2 mmol) proportionally. CeO₂-C-X (X = 0.5, 1, 1.5, 2) was further prepared by calcining at 500 °C for 2 h with a heating rate of 5 °C/min.

2.2. Characterization and Gas Sensing Properties

The morphology and structure of the Ce-BDC precursors and CeO₂-C-X gas sensing materials were characterized by SEM, TEM, TG, XRD, XPS, Raman spectroscopy, and N₂ adsorption–desorption measurements. The gas sensing tests were conducted using a setup with continuous gas flow, as shown in Figure S1. Moreover, the photograph of the sensor is shown in Figure S2. Detailed characterization information, sensor fabrication procedures, and gas-sensing test protocols can be found in the Supplementary Materials.

3. Results and Discussion

3.1. Morphology Characterization and Physical Properties

CeO₂-C-X semiconductor series were synthesized by the rapid solvothermal method and calcination of Ce-BDC-X precursors, and the schematic illustration of synthesis is shown in Figure 1a. In addition, according to the TG analysis, calcination at 500 °C could basically stabilize the mass of the derived CeO₂-C-1 (Figure S3). As the concentration of CAN and H₂BDC increased, the color of Ce-BDC-X gradually changed from yellow to white. The morphology of the CeO₂-C-X (X = 0.5, 1, 1.5, 2) was observed by SEM and presented in Figure 1(b1–e2). It was obvious that with the increasing reactant concentrations, the CeO₂-C-X nanoparticles became more uniformly dispersed. In CeO₂-C derived from Ce-BDC with low raw material concentration, more pronounced aggregation of particles was observed.

Further microstructural characterization of Ce-BDC-1 and its derived CeO₂-C-1 was carried out using TEM. As shown in Figure 2a,b, Ce-BDC-1 exhibited only a few lattice fringes corresponding to the (533), (862), and (600) planes of previously reported Ce-BDC, owing to its small particle size and low crystallinity [22,23]. Upon further examination of CeO₂-C-1 derived from Ce-BDC-1, as presented in Figure 2e, well-defined particles with an average size of approximately 7.73 nm were observed. Moreover, distinct lattice fringes corresponding mainly to the (111) and (200) planes of CeO₂ were evident. EDS was employed to analyze the presence of Ce, O, and C elements in both Ce-BDC-1 and CeO₂-C-1. As shown in Figure 2c,d,g,h, the content of C element was significantly reduced after calcination.

The phase structure and pore characteristics of the Ce-BDC precursor and the derived CeO₂-C were further characterized. As depicted in Figure 3a, the Ce-BDC-X (X = 0.5, 1, 1.5, 2) samples exhibited gradually enhanced diffraction peak intensities with the increased raw material concentration. By comparing the positions of the characteristic peaks, the formation of Ce-BDC materials could be further confirmed, with diffraction peaks located at 7.1° of (111) plane, 8.2° of (200) plane, and 24.9° of (600) plane, which match the reported structure of Ce₆O₄(OH)₄(BDC)₆ (CCDC-1036904) [24]. Ce-BDC consists of hexanuclear Ce₆O₄(OH)₄ clusters coordinated by twelve BDC linkers, forming a three-dimensional framework with fcu topology [25]. This structure was isostructural to the well-known Zr-based UiO-66 (University of Oslo-66), where Ce⁴⁺ replaces Zr⁴⁺ in the metal cluster [26]. The average crystallite size of the precursors was calculated using the Scherrer equation, revealing an increase from 2.50 to 19.21 nm as the reactants concentration increased, indicating that the crystallite size of Ce-BDC could be effectively tuned by adjusting the precursor concentration [27]. After calcination, CeO₂-C-X materials were obtained, and their XRD patterns are shown in Figure 3b. All Ce-BDC derived CeO₂ samples exhibited characteristic diffraction peaks corresponding to fluorite-structured CeO₂ (PDF 01-81-0792) [28]. The peaks at 28.5°, 33.1°, 47.5°, 56.3°, 59.1°, 69.4°, 76.7°, 79.1°, and 88.4° could be indexed to the (111), (200), (220), (311), (222), (400), (331), (420), and (422) crystal planes of CeO₂ (PDF 01-81-0792), respectively [29]. No additional peaks related to impurities were detected, suggesting that the sample was a single-phase CeO₂ [30]. Similarly, the crystallite sizes of CeO₂-C derived from Ce-BDC were estimated. The results showed nearly uniform crystallite sizes with only a slight increasing trend, ranging from 7.37 to 8.64 nm. The measured 7.96 nm particle size of CeO₂-C-1 closely corresponds to the average particle size observed by TEM. It could be inferred that after calcination, the CeO₂ derived from Ce-BDC underwent structural expansion or collapse, ultimately stabilizing at a crystallite size of around 7–8 nm. Moreover, it could be observed that among the samples, Ce-BDC-1 and its derived CeO₂-C-1 exhibited the smallest change in crystallite size (from 8.67 to 7.96 nm), which might be beneficial for CeO₂ to maintain superior physical properties.

Therefore, the N₂ adsorption–desorption characteristics of the Ce-BDC derived CeO₂ samples were investigated. As shown in Figure 3c, all CeO₂-C-X samples displayed a certain microporosity but were mainly dominated by mesopores. Interestingly, hysteresis loops began to appear at lower relative pressures for CeO₂-C-0.5 and CeO₂-C-1, while CeO₂-C-1.5 began to form them at p/p₀ = 0.64, and CeO₂-C-2 only began to form them at a higher p/p₀ of 0.87. This indicated the presence of mesopores distributed across different pressure ranges in the various CeO₂-C-X samples. Moreover, the differences in mesopore distribution could be observed from the pore size distribution calculated by the BJH method, which was depicted in Table S1 and highlighted section of Figure 3d. The pore diameters of CeO₂-C-0.5 and CeO₂-C-1 were significantly smaller than those of CeO₂-C-1.5 and CeO₂-C-2, which may be attributed to the pronounced structural collapse induced by the larger particle sizes of CeO₂-C-1.5 and CeO₂-C-2. In addition, the specific surface area of CeO₂-C-X follows the order: CeO₂-C-1 > CeO₂-C-1.5 > CeO₂-C-2 > CeO₂-C-0.5, which is favorable for efficient gas adsorption during subsequent NH₃ sensing.

To further investigate the structural characteristics of the derived CeO₂, Raman spectroscopy was employed, and the results are displayed in Figure 3e,f. The strong vibrational peak at 463 cm⁻¹ was attributed to the symmetric stretching vibration (F_2g mode) of Ce-O coordination, which dominated in all CeO₂-C-X samples [31]. The intensity of the F_2g peak clearly decreased with the increase in X in CeO₂-C-X, especially for CeO₂-C-1.5 and CeO₂-C-2, which had larger precursor grain sizes. It could be inferred that the significant structural collapse in CeO₂-C-1.5 and CeO₂-C-2 led to reduced symmetry of the Ce-O lattice, resulting in more disordered lattice vibrations. In addition, weak and broad bands observed at 595 cm⁻¹ and 1174 cm⁻¹ could be assigned to the Frenkel defect-induced mode (D band) and the second-order longitudinal optical mode (2LO band), respectively [32]. These bands were considered to be related to the concentration of oxygen vacancies in CeO₂, among which CeO₂-C-1 likely contained the highest amount of oxygen vacancies. By calculating the ratio of the D + 2LO bands to the F_2g peak, the effect of oxygen vacancies in CeO₂ can be estimated. Among the samples, CeO₂-C-1 exhibited the highest ratio of 0.0515, which was significantly greater than other Ce-BDC-derived CeO₂-C. Due to differences in precursor particle size, Ce-BDC derived CeO₂ exhibited significant structural variations. Among them, CeO₂-C-1 demonstrated the best combination of properties, including the highest specific surface area, the smallest mesopore distribution, and the greatest effect of oxygen vacancies, while maintaining the CeO₂ crystal structure. This could be attributed to the minimal particle size change before and after derivation, which helped preserve the morphology of its physical structure.

3.2. XPS Analysis

XPS was used to analyze the surface elements and chemical states of CeO₂-C-X gas sensing materials. As shown in Figure S4a, the survey spectra of CeO₂-C-X (X = 0.5, 1, 1.5, 2) displayed characteristic peaks of Ce 3d and O 1s. The Ce/O atomic ratios calculated from peak intensities are shown in Figure S4b. It was observed that CeO₂-C-0.5 and CeO₂-C-1 exhibited Ce/O ratios close to 2, while those of CeO₂-C-1.5 and CeO₂-C-2 were below 1, suggesting the formation of unstable CeO₂ structures with more lattice defects. Further deconvolution of the Ce 3d spectra, as shown in Figure S4c, revealed that three pairs of peaks (V and U, V″ and U″, V‴ and U‴) corresponded to tetravalent Ce⁴⁺ (CeO₂), while two pairs (V₀ and U₀, V′ and U′) were attributed to trivalent Ce³⁺ (Ce₂O₃) [33]. The calculated Ce⁴⁺ contents were: CeO₂-C-1 (76.49%) > CeO₂-C-0.5 (76.04%) > CeO₂-C-1.5 (74.80%) > CeO₂-C-2 (73.29%). The significantly lower Ce⁴⁺ content in CeO₂-C-1.5 and CeO₂-C-2 compared to CeO₂-C-0.5 and CeO₂-C-1 indicated their inferior redox performance [34]. Further deconvolution of the O 1s spectra was performed to identify the different oxygen species, as shown in Figure 4. The peak around 532 eV was attributed to surface hydroxyl groups (O_H), the peak near 531 eV corresponded to surface adsorbed oxygen (O_C), and the peak around 529 eV was assigned to lattice oxygen (O_L) [35]. The O_L proportion in CeO₂-C-1.5 (Figure 4c) and CeO₂-C-2 (Figure 4d) was significantly lower than that in CeO₂-C-0.5 (Figure 4a) and CeO₂-C-1 (Figure 4b), indicating that the excessive structural defects caused by the calcination of large-particle Ce-BDC led to a decline in potential gas sensing performance. Notably, XPS analysis revealed that CeO₂-C-1.5 and CeO₂-C-2 possessed excessive surface defect concentrations, which were actually unfavorable for enhancing NH₃ sensing performance. In contrast, although CeO₂-C-1 at Figure 4b exhibited the lowest adsorbed oxygen content (14.05%) and defect concentrations, Raman analysis still indicated a relatively pronounced influence of oxygen vacancies, and its proper surface Ce/O ratio with a moderate level of oxygen defects was more conducive to NH₃ adsorption and conversion.

3.3. Gas Sensing Properties

A series of CeO₂-C derived from Ce-BDC precursors with different particle sizes and crystallinity was synthesized and assembled into sensors for NH₃ detection, and the relevant procedures were described in Supplementary Materials. Under common environmental conditions, the NH₃ sensing performance of CeO₂-C was tested at RT = 25 ± 3 °C, using air as the recovery gas and maintaining the relative humidity (RH) at 40% ± 2% [36]. CeO₂ exhibited relatively stable resistance values in the current-voltage (I–V) measurements (Figure S5). As shown in Figure 5a, at around 100 s, 100 ppm NH₃ gradually reached the different CeO₂-C-X sensors, leading to a change in resistance and thus a variation in response value. The four CeO₂-C-X sensors exhibited distinct response behaviors, among which CeO₂-C-1 (82%) showed a higher response than CeO₂-C-1.5 (68%), CeO₂-C-2 (59%), and CeO₂-C-0.5 (49%). The response time (τ_res) and recovery time (τ_rec) were measured to be 22 s and 66 s, respectively. The LOD of CeO₂-C-1 was calculated to be 0.31 ppm using the method of IPUAC and Burgués et al. in Supplementary Materials [37]. Further six-cycle repeatability tests were conducted to observe stability during repeated response-recovery in Figure 5b, the changes in resistance clearly reflected the NH₃ adsorption and desorption processes, indicating the good cycling stability of the CeO₂-C-X sensor series. The responses to various inorganic gases, excluding VOCs, were evaluated as shown in Figure 5c, with H₂S and NO₂ considered major interfering gases. CeO₂-C-1 demonstrated the best anti-interference performance of low response to H₂S (27%) and NO₂ (11%) at RT. The NH₃ sensing performance of CeO₂-C-X at concentrations from 10 to 500 ppm is shown in Figure 5d. As NH₃ concentration increased, the response values rose accordingly, with CeO₂-C-1 consistently showing the highest response throughout the range. Moreover, CeO₂-C-1 could achieve a response of 19.7% even at 10 ppm NH₃. Significantly, the CeO₂-C-X sensor series showed a near-exponential distribution of response values, suggesting higher sensitivity in the low-concentration range of NH₃ (Figure 5e). The response of CeO₂-C-1 to varying NH₃ concentrations could be accurately fitted by y = 143.3 − 129.5·0.9924^x, with an excellent correlation coefficient (R² = 0.9956). To evaluate the long-term stability of the CeO₂-C-1 sensor, cyclic stability tests were conducted under the same conditions after 30 days. As shown in Figure S6, CeO₂-C-1 still exhibited a response of about 80%, demonstrating its excellent long-term stability. When compared with recently reported room-temperature NH₃ sensors in

Table 1. NH₃ sensing performance at RT compared with recently reported work.

Materials	RH (%)	Conc. (ppm)	RS (%)	τres/τrec (s)	LOD	Ref
CeO2 film	0	100	35	19/66	NA	[38]
SnS2/graphene	NA	100	62.36	63/586	0.1 ppm	[39]
MoSe2/VTiO2	50	6	2.78	94/50	0.3 ppm	[40]
Au/PANI/WS2	51	100	286.1	24/26	13.8 ppb	[41]
NiO@CeO2	0	400	97	79/21	NA	[42]
CeO2/Ti3C2Tx	0	10	28.36	145/875	NA	[43]
PAni-CeO2	0	100	80	9.31/531.60	NA	[44]
CeO2-C-1	40	100	82	22/66	0.31 ppm	This work

, the CeO₂-C-1 developed in this study showed competitive NH₃ sensing performance among comparable materials.

The O₂ content and H₂O are typically key factors affecting gas sensing performance [36]. Initially, the NH₃ sensing behavior of the material under oxygen-free conditions was examined. As shown in Figure 6a, the gas sensing performance of the CeO₂-C-1 sensor was tested under an Ar atmosphere with RH maintained at 40%. Although both the response and recovery times were delayed and the response value decreased, the material still exhibited τ_res and τ_rec of 26 s and 74 s, respectively, with a substantial response intensity of 64%. In six consecutive cycles of response testing in Figure 6b, the response showed slight fluctuations of 14.4%, indicating instability and a decline in sensing performance under oxygen-free conditions. As illustrated in Figure 6c, the NH₃ response under varying concentrations still followed an exponential relationship. Humidity also influenced the resistance of CeO₂-C-1, as shown in Figure 6d–f. It was observed that under low humidity, the CeO₂-C-1 sensing material exhibited higher response values, though the response and recovery times were also prolonged. When the humidity reached 50%, the response value decreased but still remained at 68%. Overall, CeO₂-C-1 maintained good gas-sensing performance under both oxygen-free conditions and varying humidity levels, which further broadened its potential applications.

3.4. Gas-Sensing Mechanism

CeO₂ is a typical n-type wide-bandgap semiconductor, and its NH₃ gas-sensing mechanism primarily relies on the gas adsorption-reaction process on the surface and the resulting changes in the carrier concentration of the material [45]. The CeO₂-C obtained from Ce-BDC contains residual carbon, which is beneficial for improving the electrical conductivity of CeO₂ and mitigating the influence of humidity [46,47]. The possible NH₃ sensing mechanism on CeO₂-C is illustrated in Figure 7. In air, the surface of CeO₂-C can adsorb oxygen molecules to form chemically adsorbed oxygen species (such as O₂⁻, O⁻, and O²⁻) [48]. At RT, O₂ molecules adsorb on the surface of CeO₂-C and form O₂⁻ species as illustrated in Equations (1) and (2). These O₂⁻ species capture electrons from the conduction band of CeO₂-C, resulting in the formation of a surface depletion layer (L_d) and an increased resistance [49,50].

$${\mathrm{O}}_2 \left(\mathrm{gas}\right)\to {\mathrm{O}}_2 \left(\mathrm{ads}\right)$$

(1)

$${\mathrm{O}}_2{ \left(\mathrm{ads}\right)+\mathrm{e}}^{-}\to {\mathrm{O}}_2^{-} \left(\mathrm{ads}\right), \mathrm{T}<150 °\mathrm{C}$$

(2)

As shown in Equations (3) and (4), when NH₃ molecules interact with the CeO₂-C surface, it reacts with the surface-adsorbed O₂⁻ species, releasing electrons back into the conduction band. This reduces the width of the depletion layer, leading to a decrease in resistance [51,52]. When re-exposed to air, O₂⁻ species are re-formed, and the material returns to its high-resistance state.

$${\mathrm{NH}}_3 \left(\mathrm{gas}\right)\to {\mathrm{NH}}_3 \left(\mathrm{ads}\right)$$

(3)

$${\mathrm{NH}}_3 \left(\mathrm{ads}\right)+3{\mathrm{O}}_2^{-}\to {2\mathrm{N}}_2+{6\mathrm{H}}_2\mathrm{O}+3{\mathrm{e}}^{-}$$

(4)

In addition, in the presence of ambient humidity, H₂O molecules may adsorb on the CeO₂-C surface. Meanwhile, NH₃ tends to generate surface electrolyte ions, which reduce the resistance of the nano-sized CeO₂-C (Equation (5)). When this sensing mechanism predominates, the sensor typically exhibits an integrative-type response, as the generated electrolyte ions are difficult to release [53]. Consequently, external stimuli such as thermal activation or UV irradiation are generally required to recover baseline resistance. The semiconductor-based sensing mechanism, primarily governed by depletion layer modulation, still predominates. It is evident that H₂O and NH₃ continue to compete for adsorption sites on the surface of CeO₂-C, with increasing humidity leading to a slight decrease in the NH₃ response.

$${\mathrm{NH}}_3 \left(\mathrm{ads}\right)+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{NH}}_4^{+}+{\mathrm{OH}}^{-}$$

(5)

The superior gas-sensing performance of the CeO₂-C sensor for NH₃ detection at RT can be attributed to several factors. Among the series, the precursor Ce-BDC-1 and its derivative CeO₂-C-1 exhibit the smallest particle size variation, preserving a stable nanostructure with the highest specific surface area. CeO₂-C-1 also possesses the smallest average mesopore size, and this combination of limited particle growth and refined porosity ensures abundant active sites and efficient gas diffusion, thereby enhancing sensing performance and selectivity. Although the derived CeO₂-C shows a high Ce³⁺ concentration, its sensing performance is inferior to that of certain other CeO₂-C materials with lower Ce³⁺ content. This counterintuitive behavior may arise from excessive defect clustering, structural instability, or less favorable surface morphology.

4. Conclusions

In summary, a series of nano Ce-BDC materials and their derived CeO₂-C-X material were successfully synthesized and systematically studied for room-temperature NH₃ sensing. Among them, Ce-BDC-1-derived CeO₂-C-1 showed minimal particle size variation and the highest specific surface area. It possessed 82% response to 100 ppm NH₃, with a response/recovery time of 22 s/66 s at 25 °C and RH = 40%. It also exhibited a low detection limit (LOD = 0.31 ppm), along with excellent long-term stability and selectivity, thereby confirming the potential applicability of MOF-derived materials in gas sensing. Its superior sensing performance can be attributed to its well-developed mesoporous structure, which likely arises from the minimal particle size variation, allowing it to retain favorable physical properties. The NH₃ sensing of CeO₂-C at RT is primarily governed by depletion-layer modulation, where surface-adsorbed oxygen species capture and release electrons upon NH₃ exposure. These findings highlight that careful synthesis control of the precursor Ce-MOF facilitates the design of CeO₂-based gas sensors with optimized physical properties and stable gas-sensing performance.