Porous, Tremella-like NiFe₂O₄ with Ultrathin Nanosheets for ppb-Level Toluene Detection

Abstract

As a typical spinel ferrite, NiFe₂O₄ is suitable for use in gas sensors. Herein, we report the fabrication of porous, tremella-like NiFe₂O₄ assembled using porous, ultrathin nanosheets via the coordination of Ni²⁺ and Fe²⁺ with 1,4-phenylenediboronic acid. The optical band gap of the NiFe₂O₄ is estimated to be about 1.7 eV. Furthermore, the NiFe₂O₄ sensor annealed at 400 °C exhibits a low detection limit of 50 ppb, a fast response/recovery time (11.6 s/41.9 s to 10 ppm toluene), good reproducibility, and long-term stability at 220 °C. The suitable sensing performances can be attributed to the good catalytic activity of NiFe₂O₄ to toluene oxidation. Moreover, the ultrathin nanosheets with porous structures provide a large number of active sites to significantly favor the diffusion and adsorption/desorption of toluene molecules. This current work provides an insight into fabricating NiFe₂O₄ using 1,4-phenylenediboronic acid, which is promising for ppb-level toluene detection.

1. Introduction

Toluene (C₇H₈) is a typical volatile organic compound (VOC) which is extensively used in medicine, leather, pesticides, laboratory solvents, pharmaceuticals, and so on [1,2,3]. However, toluene is harmful for both human health and the environment. Exposure to toluene could pose a risk to human health by damaging the central nervous system, liver, kidneys, heart, etc. [4,5,6]. It is reported that short-term exposure to 70 ppb of toluene may negatively affect the cognitive function of the brain [7]. Furthermore, toluene is reported to be a vital biomarker in the diagnosis of lung cancer. The concentrations of toluene in patients with lung cancer were reported to range from 80 to 100 ppb [8,9]. Hence, rapid toluene detection with excellent selectivity at an early stage is very promising for multiple applications in health care, environmental protection, and medical diagnoses. Among various techniques, resistive gas sensors fabricated with semiconductor metal oxides have attracted wide attention due to their advantages of easy fabrication and low cost [10,11]. However, it is still challenging to improve their sensing performance, especially for selective and sensitive ppb-level toluene detection.

Recently, spinel ferrite materials (AB₂O₄) have shown significant advantages in the detection of various gases due to their unique chemical compositions and crystal structures in which two cation sites are occupied by either transition or post-transition cations [12,13]. The combination of various transition and post-transition metal cations in spinel ferrites could be favorable for a highly sensitive and selective gas sensor. Thus far, sensitive materials such as NiFe₂O₄, ZnFe₂O₄, ZnCo₂O₄, NiCo₂O₄, and NiMn₂O₄ have been developed for gas sensors [14,15,16]. NiFe₂O₄ is one of the particularly important spinel ferrite materials which has recently been used in a gas sensor for the detection of toluene. Various synthesis methods have been developed to tune the morphology of NiFe₂O₄ to improve its sensing properties. For example, Lai et al. used KIT-6 as a template to successfully obtain mesoporous NiFe₂O₄, which exhibited an excellent toluene-sensing property [17]. Kim’s group synthesized 3D hierarchical NiFe₂O₄ porous microspheres with the assistance of urea and glucose. The NiFe₂O₄ sensor achieved a good toluene gas response at a low working temperature [18].Yang and his co-workers prepared NiFe₂O₄ nanomaterials using a co-precipitation method; these materials exhibited good sensing properties in their sensitivity and selectivity to toluene detection [19]. Paula et al. prepared NiFe₂O₄ nanoparticles via levitation-jet synthesis which were shown to be remarkably sensitive to toluene gas. In our previous work, Ni/Fe-MOF-derived 1D porous NiFe₂O₄ nanorods showed outstanding selectivity and sensitivity to toluene [20]. Even so, NiFe₂O₄-based gas sensors usually exhibit unsatisfied sensing properties in their low gas responses and high detection limits. Therefore, a novel synthesis method is highly desirable for fabricating NiFe₂O₄ with a unique microstructure to enhance its toluene detection to the ppb level. Recently, using an organic acid as both a coordinating reagent and a soft template has been an effective way of fabricating metal oxides [21,22]. Studies have shown that boric acid is capable of coupling with the metal ions of Pb²⁺, Cd²⁺, and Fe²⁺ to form nanosheet complexes [23,24,25]. Carboxylphenyl boronic acid was reported to complex with metal ions (such as Zn²⁺, Cd²⁺, Ni²⁺, abd Ln³⁺) to form 2D coordination polymer nanosheets [26,27]. These pioneered works provide us a clue for synthesizing NiFe₂O₄ nanosheets by taking 1,4-phenylenediboronic acid to coordinate with both Ni²⁺ and Fe²⁺ ions.

Herein, we successfully prepared porous, tremella-like NiFe₂O₄ with ultrathin nanosheets by using 1,4-phenylenediboronic acid to coordinate Fe²⁺ and Ni²⁺. Our prepared NiFe₂O₄ possesses good crystallinity, catalytic activity, a porous, ultra-thin nanosheet structure, and a large specific surface area (77.8 m²/g). The sensor based on NiFe₂O₄ annealed at 400 °C exhibits good toluene-gas-sensing properties, especially with a low detection limit of 50 ppb. The findings of this study demonstrate the possibility of using 1,4-phenylenediboronic acid to fabricate NiFe₂O₄ nanosheets which are then suitable for application in trace toluene detection.

2. Experimental Section

2.1. Fabrication of NiFe₂O₄

All the starting reagents were of analytical grade. Deionized water (18 MΩ cm at room temperature) was used for this work. The tremella-like NiFe₂O₄ was synthesized via a hydrothermal process followed a thermal treatment. Specifically, 0.5 mmol (0.1188 g) NiCl₂·6H₂O (AR 99.9%) and 1 mmol (0.1988 g) FeCl₂·4H₂O (AR 98%) were completely dissolved into the mixture of 50 mL N,N-dimethylformamide (DMF) (AR 99.5%) with 5 mL anhydrous ethanol and 5 mL deionized water. Then, 3 mmol (0.4973 g) of 1,4-phenylenediboronic acid (C₆H₈B₂O₄) (AR 95%) was further added under magnetic stirring for 30 min to form a homogenous solution. After a hydrothermal reaction at 130 °C for 12 h, the obtained product was separated via centrifugation and rinsed with deionized water and ethanol. Next, the precursor products were collected after being dried at 60 °C for 6 h. Finally, the precursor was annealed in air at 350 °C, 400 °C, and 450 °C for 2 h with a heating rate of 3 °C/min, and the corresponding NiFe₂O₄ samples were labeled S-350, S-400, and S-450, respectively.

2.2. Characterization

The crystal phase of the NiFe₂O₄ was analyzed on a Rigaku TTRIII X-ray diffraction (XRD) equipped with Cu K_α radiation (1.54056 Å) from a scanning angle of 10° to a scanning angle of 70°. A field emission scanning electron microscope (FESEM, FEI QUANTA 200, Hillsboro, OR, USA) was utilized to analyze the morphological features of all samples. Transmission electron microscopy (TEM) images were taken on JEOL JEM-2100 microscope at an accelerating voltage of 100 kV. The corresponding elemental mapping image was acquired by the EDS attachment on the FESEM. A thermogravimetric analysis (TGA-DSC) was performed to investigate the decomposition of the precursor at a heating rate of 10 °C/min (TA SDT-2960, New Castle, DE, USA). X-ray photoelectron spectra (XPS) were characterized to measure elemental valence states, and the electron binding energy was calibrated by referring to the C1s peak at 284.6 eV. The UV–visible absorption spectra were tested on a UV–VIS spectrophotometer (SolidSpec-3700). The Brunauer–Emmett–Teller (BET) surface area was derived from nitrogen isotherms measured on a surface adsorption instrument (ASAP 2020 Plus).

2.3. Sensor Fabrication and Testing Properties

The gas sensor was fabricated and tested using our general procedure [28,29]. The as-synthesized NiFe₂O₄ powders were dispersed into a certain amount of deionized water and sonicated into a floating solution. The above suspension solution was then coated onto the outer surface of the alumina tube and dried at 120 °C for 2 h. A pair of Au electrodes and corresponding Pt wires at both ends of the alumina tube were set. To ensure good contact between the gold electrode and the sensing material, the sensors were annealed in air at 300 °C for 2 h. The gas-sensing performances were tested using a WS-30 A system (Weisheng Instruments Co., Zhengzhou, China) with a volume of the closed chamber of 18 L. The desired gas concentration was obtained based on the static liquified gas distribution method [28]. The response (β) of the sensor is defined as:

β = R_g/R_a

(1)

where R_g and R_a stand for the resistance of the sensor in the target gas and air, respectively. In addition, the response and recovery time were described as the time taken for 90% of the entire resistance change to be achieved. During the test of the gas-sensing properties, the relative humidity (RH) was maintained around 30%.

3. Results and Discussion

The crystalline structures were examined via X-ray diffraction (XRD). The XRD curves of the NiFe₂O₄ samples are shown in Figure 1. In comparison with the standard card (JCPDS: 54-0964), a cubic NiFe₂O₄ structure (space-group: $Fd\overline{3}m$) can be confirmed. Specifically, the distinct peaks are detected at 2θ of 30.3°, 35.7°, 37.3°, 43.3°, 57.3°, and 63°, corresponding to the Bragg planes of (220), (311), (222), (400), (511), and (440), respectively. The crystallite size of the NiFe₂O₄ is estimated by using Scherrer’s equation (Equation (2)):

$$\mathrm{D}=\mathrm{k}\frac{\mathsf{\lambda }}{\mathsf{\beta }\cos \mathsf{\theta }}$$

(2)

where D, k, λ, β, and θ are the crystallite size, the Scherrer constant (0.89), the X-ray wavelength (1.54056 Å), the peak width at half, and Bragg’s diffraction angle, respectively. It can be calculated that the grain sizes of the S-400 NiFe₂O₄ and S-450 NiFe₂O₄ are 13.1 and 16.2 nm. Due to the poor crystallinity of the S-350, its grain size cannot be calculated. The X-ray diffraction peak of the materials has no peaks for other impurities, indicating that the NiFe₂O₄ is of a high quality. In Figure 1, the XRD peak becomes sharp as the annealing temperature increases, indicating strong crystallinity. The band gap is an important parameter for a metal oxide semiconductor. Therefore, the UV-vis absorption spectra of NiFe₂O₄ materials have been studied (Figure 2), and they are used to calculate the band gap using the following equation [29,30]:

$${\left(\alpha hv\right)}^2=A(hv-{\mathrm{E}}_{\mathrm{g}})$$

(3)

where α, h, ν, E_g, and A represent the absorption coefficient, Planck’s constant, light frequency, band gap energy, and a constant, respectively. The Tauc-plot of (αhv)² versus the photo energy (hv) of the NiFe₂O₄ (S-350, S-400, and S-450) is as shown in Figure 2b–d. In general, the intercept between the linear portion of the (αhv)² to hv plots on the hv axis represents the value of the band gap energy (E_g). From Figure 2b–d, it can be seen that the band gap is estimated to be 1.73 eV, 1.70 eV, and 1.72 eV for the S-350, S-400, and S-450 samples. Although the band gap depends strongly on the defect level and microstructure derived from the different synthetic methods, our values are comparable with some reported values of the ban gaps for pure NiFe₂O₄, as summarized in

Table 1. Comparision of band gaps of NiFe₂O₄ synthesized with different methods.

Materials	Synthesis Method	Morphology	Annealing (°C)	Band Gap (eV)	Ref.
NiFe2O4	Co-precipitation	Nanoparticles	700	1.59	[31]
			800	1.67
			900	1.71
NiFe1.95Y0.05O4			800	1.62
NiFe1.9Y0.1O4				1.59
NiFe1. 85Y0.15O4				1.59
NiFe1.8Y0.2O4				1.57
NiFe2O4	Co-sputtering (Ar:O2 = 40:60)	Nanoparticles		1.785	[32]
	Co-sputtering (Ar:O2 = 30:70)			1.78
	Co-sputtering (Ar:O2 = 20:80)			1.779
NiFe2O4	Magnetron sputtering	Thin films		1.78	[33]
			400	1.88
			500	2.46
			700	2.72
NiFe2O4	Co-precipitation	Nanoparticles	700	1.59	[34]
NiFe2O4	CVD	Thin films		2.34 /2.78	[35]
NiFe2O4	PLD	Thin Films	690	2.36	[36]
NiFe2O4	Hydrothermal synthesis	Nano-sheets	350	1.73	This work
			400	1.70
			450	1.72

.

The morphology and microstructure of the NiFe₂O₄ were observed via SEM. As can be seen from Figure 3a,b, the NiFe₂O₄ precursor shows a nanosheet structure. After being annealed in air, the nanosheets became wrinkled into a tremella-like structure. Similar morphologies are observed after annealing at different temperature of 350 °C (Figure 3c,d), 400 °C (Figure 3e,f), and 450 °C (Figure 3g,h). The NiFe₂O₄ exhibits a tremella-like shape with a diameter of 3–4 μm (Figure 3e). In the highly magnified SEM image in Figure 3f, the NiFe₂O₄ consists of closely interlaced ultrathin nanosheets, resulting in a porous nanostructure.

The TEM characterization provides more details about the microstructures. It can be seen that the tremella-like S-400 NiFe₂O₄ is composed of subunits of ultrathin nanosheets (Figure 4a,b). The ultrathin nanosheet structure is the result of the complex formation of boric acid ligands with Fe²⁺ and Ni²⁺ [37]. The homogeneous distributions of the Ni, Fe, and O elements on the tremella-like NiFe₂O₄ are illustrated in Figure 4c. Interestingly, a large number of holes can be clearly observed from Figure 4d. These pores range in diameter from 5 nm to 40 nm. Some typical NiFe₂O₄ nanoparticles are from 9.7 nm to 14.2 nm in diameter, as shown in Figure 4e. The HRTEM image (Figure 4f reveals a lattice spacing of 0.25 nm, which is ascribed to the (311) plane of NiFe₂O₄, and the results are well matched with XRD analysis.

Taking the NiFe₂O₄ annealed at 400 °C as an example (S-400), the textural properties of the tremella-like NiFe₂O₄ nanosheets were further studied via a nitrogen adsorption–desorption test, as shown in Figure 5. Obviously, the NiFe₂O₄ adsorption isotherm belongs to type IV, according to the IUPAC classification [38]. Moreover, the material has an H3-type hysteresis loop, which may be due to the accumulation of several nanosheets forming mesoporous properties. As a result, the unique NiFe₂O₄ possesses a high specific surface area of 77.80 m²/g through the N₂ adsorption–desorption measurement. In addition, the pore distribution can be estimated by the Barrett–Joyner–Halenda (BJH) method, which shows that the average pore size of the NiFe₂O₄ is centered at about 18.5 nm (inset in Figure 5). The TEM images in Figure 4d,e reveal that the NiFe₂O₄ comprises numerous nanoparticles, giving rise to the formation of pores. The specific area of the NiFe₂O₄ is strongly dependent on the synthesis methods and morphologies, as shown in

Table 2. Comparision of specific surface areas and crystal sizes of NiFe₂O₄ synthesized with different synthesis methods and morphologies.

Materials	Synthesis Method	Morphology	Surface Area(m2/g)	Crystal Size (nm)	Ref.
NiFe2O4	MOFs templates	Nanocubes	33.72		[40]
NiFe2O4	MOFs templates	Hexagonal biyramids	64.78		[41]
NiFe2O4	MOFs templates	Fusiformis	61.50		[42]
NiFe2O4	Carbon sphere templates	Nanospheres	86.9	16	[43]
NiFe2O4	Ultrasound assistance	Nanoparticles	17.178	17	[44]
NiFe2O4	Co-precipitation	Nanoparticles	5.32		[45]
NiFe2O4	Electrospinning	Nanofibers	54	20	[46]
NiFe2O4	Hydrothermal	Microspheres	46.44	46.4	[18]
NiFe2O4	Hydrothermal	Nanosheet	77.80	13.1	This work

. Notably, our synthesized tremella-like NiFe₂O₄ with nanosheets exhibits a larger specific surface area than common 0D (nanoparticles), 1D (nanofibers and fusiformis), and 3D (nanocubes, hexagonal bipyramids, nanospheres and microspheres) NiFe₂O₄ materials. Thus, we provide a facile hydrothermal method using 1,4-phenylenediboronic acid to fabricate porous NiFe₂O₄ with high specific surface area which will facilitate the molecules adsorption and desorption of gas molecules as well as oxidation on the surface, leading to an good gas-sensing performance [39]. Therefore, NiFe₂O₄ is expected to show good gas-sensing properties towards toluene detection.

1,4-phenylenediboronic acid is rich in carboxyl groups which will easily coordinate with the metal ions of Fe²⁺ and Ni²⁺ to form complexes [47]. In addition, 1,4-phenylenediboronic acid may serve as a capping agent or template for the formation of a nanosheet structure during the synthesis process. In an atmosphere in which DMF is used as a complex solvent, the borate anions in 1,4 p-phenylboronic acid are drawn together with Fe²⁺ and Ni²⁺ via electrostatic interaction and form complexes of specific two-dimensional structures [48]. With subsequent heat treatment, the pyrolysis process occurs in the complex. The NiFe₂O₄ was then synthesized after thermal treatment, which was studied via a thermogravimetric analysis (TGA), as shown in Figure 6. It can be seen that the weight loss of the sample is aroud 3.5% below 150 °C, which may originate from the loss of absorbed water in the precursor. As the temperature rises further to 350 °C, the decomposition of the precursor results in a weight loss of approximately 18.4%. The weight loss at this stage is ascribed to the thermal decomposition of the complexes formed from Ni²⁺, Fe²⁺ and 1,4-phenylenediboronic acid. There is an endothermic peak at a temperature of 257.08 °C, which may be the peak for the decomposition of organic matter to absorb energy. The decomposition of organic matter will release a large amount of carbon dioxide and nitrogen oxide gases [49]. The kinetics of this process are very intense, resulting in a large number of holes in most of the product. After that, a slight weight loss could orignate from the removal of the residual carbon and the further crystallization of the NiFe₂O₄.

To identify the elemental compositions and chemical bonding states, the NiFe₂O₄ nanosheets were subjected to an XPS analysis. As shown in Figure 7a, the coexistence of Ni, Fe, and O elements on the S-400 NiFe₂O₄ was confirmed. The Ni 2p spectrum of the S-400 NiFe₂O₄ (Figure 7b) exhibits two spin-orbit doublets characteristic of Ni 2p_3/2 and Ni 2p_1/2 located at 854.11 eV and 873.36 eV, respectively, as well as satellite peaks centered at 861.15 eV and 878.90 eV, corresponding to the characteristic of Ni²⁺ [50,51]. Meanwhile, the Ni 2p_3/2 and Ni 2p_1/2 peaks can be deconvoluted into two peaks at 856.06 eV and 872.69 eV, which are attributed to Ni³⁺ species. Similarly, in the Fe 2p spectrum of the S-400 NiFe₂O₄ (Figure 7c), peaks at 713.85 eV and 726.10 eV can be assigned to Fe³⁺. The binding energies of 710.53 eV and 723.40 eV correspond to Fe²⁺. The coexistence of Ni²⁺/Ni³⁺ and Fe²⁺/Fe³⁺ cations is essential for catalyzing toluene to enhance the sensor’s gas-sensing performance [52]. In particular, the formation of high-oxidation states of Ni³⁺ and Fe³⁺ could favor the transport of electrons and the formation of abundant oxygen ions to catalyze the reaction of the target gas molecules for a high degree of sensitivity [53]. The O 1s peak in XPS (Figure 7d–f) is asymmetric in shape and can be divided into two major peaks. Taking the S-400 sample as an example (Figure 7e), the low binding energy of 530.06 eV at the peak is characteristic of lattice oxygen (O_lattice), which does not react with the target gas. Conversely, at relatively high binding energies (532.00 eV), a well-resolved peak is assigned to surface-adsorbed oxygen, which is thought to react with the target gas molecules to determine the gas-sensing characteristics [54]. Although a similar behavior of O 1s was observed, the ratios of adsorbed oxygen were estimated to be 13.12%, 23.79%, and 20.16% for S-350, S-400, and S-450 NiFe₂O₄ (Figure 7d–f), respectively. These results indicate that the S-400 NiFe₂O₄ could show a better sensing property because the adsorbed oxygen will evolve into active oxygen species to enhance the reaction with the target gas molecules [8].

The gas-sensing performance of NiFe₂O₄ is evaluated by the change in the conductance of the sensor in different concentrations of toluene. Figure 8a displays the responses of the sensor based on NiFe₂O₄ annealed at 400 °C at a temperature from 180 °C to 260 °C to toluene concentrations of 10 ppm and 30 ppm. Regardless of the toluene concentrations, the sensor shows an increase in its response to the maxmium value at 220 °C and then drops with the further increase in temperature. Due to the insufficient catalytic reaction of the toluene molecules on the NiFe₂O₄ surface, the response values are relatively lower at both lower and higher temperatures [55]. Therefore, the following toluene gas-sensing tests were performed at an optimal temperature of 220 °C. Figure 8b shows the dynamic responses of the NiFe₂O₄ annealed at 350 °C, 400 °C, and 450 °C to toluene at a concentration of 30 ppm at 220 °C. The S-400 based sensor exhibits a better response of 5.8 compared to the S-350 sensor (2.1) and the S-450 sensor (1.9). Therefore, the gas-sensing performance of the following study is mainly focused on the S-400. As can be seen from Figure 8c, the S-400 sensor shows a good dynamic response to a ppb level of toluene. The NiFe₂O₄ sensor (S-400) still shows a response of 1.1 to 50 ppb of toluene, suggesting a low limit of detection (LOD) at the ppb level. Additionally, the dynamic response of the NiFe₂O₄ sensor to various toluene concentrations (1–50 ppm) was examined, as shown in Figure 8d. In addition, the relationship between the responses and the toluene concentrations is shown in Figure 8e. We can see that the sensor shows a good dynamic response to toluene at low concentrations from 50 ppb to 5 ppm. A linear relationship between the response (β) and toluene concentration (C) is fitted by the following equation (Equation (4)), with a relative correlation coefficient of R² = 0.9943:

β = 0.5287C + 1.1563

(4)

When the tolenene concentration exceeds 5 ppm, the S-400 NiFe₂O₄ sensor gradually becomes saturated. The NiFe₂O₄ sensor exhibits a response of 4.7 with a fast response (11.6 s) and recovery (41.9 s) time (Figure 8f) to 10 ppm of toluene. The fast response recovery time is attributed to the loose, porous structure of the ultra-thin nanosheets.

Figure 9a shows the response and recovery time of the NiFe₂O₄ gas sensor corresponding to the transient curve in the 0.05–50 ppm concentrations of toluene. The response time of the gas sensor decreases with the increase in toluene concentration. In addition, it shows a longer response time at lower concentrations. Kim et al. proposed a nonlinear reaction–diffusion model to explain the relationship between the response/recovery time and gas concentration [56], as shown in following equation:

τ = kx₀²C₀^r−1/D

(5)

where k, x₀, C₀, and D represent the reaction rate constant, film thickness, gas concentration, and diffusion coefficient, respectively. According to Equation (5), the gas sensor exhibits a fast response in a high concentration of toluene due to the rapid diffusion via large kinetics. Due to the slower kinetic process of adsorption–desorption and the ionization of surface oxygen during recovery [57], a longer recovery time is needed for a higher gas concentration. In brief, our NiFe₂O₄ sensor shows a comparable sensing performance for toluene detection compared with typical sensing materials in the literature, as summarized in

Table 3. Comparision of sensing performances to toluene based on typical sensing materials.

Materials	Microstructures	T (°C)	Response	LOD	τres/τrecov (s)	Ref.
NiO	Nanopaticle	250	1.5@300 ppm	100 ppm	-	[59]
NiO	Flower-like	250	1.73@0.5 ppm	0.5 ppm	13.6/24.5@2 ppm	[60]
In2O3	Nanotube	340	10.9@100 ppm	1 ppm	17/40@40 ppm	[61]
WO3	Nanopaticle	225	132 @100 ppm	2 ppm	2/6@100 ppm	[62]
TiO2	Nanopaticle	327	5.15@1 ppm	5 ppb	-	[63]
Zn2SnO4	Nanosheet	280	4.7@5 ppm	5 ppm	1/3.55@10 ppm	[64]
SnO2	Sphere	400	20.2@50 ppm	-	52/74@50 ppm	[65]
NiFe2O4	Nano-octahedron	260	6.41@100 ppm	1 ppm	25/40@100 ppm	[66]
NiFe2O4	Tremella-like	220	4.67@10 ppm	50 ppb	11.6/41.9@10 ppm	This work

T: temperature; LOD: limit of detection; Ref.: references.

. On the other hand, the responses toward other tested gases with concentrations of 10 ppm, such as xylene, acetone, ethanol, methanol, isopropanol, n-butyl alcohol, formaldehyde, and ammonia, have been examined, as shown in Figure 9b. Obviously, the response to toluene is about 2–3 times than that of the interfering gases due to the catalytic feature of p-type NiFe₂O₄ to toluene at the optimal operating temperature of 220 °C [58].To investigate the short-term repeatability, five consecutive tests of 10 ppm and 30 ppm of toluene were verified, as shown in Figure 9c, which shows a good repeatability. Finally, the long-term stability of one month was validated in Figure 9d. For instance, the responses to 30 ppm of toluene fluctuate around 6.1, suggesting a good stability over a long period of time. The sensor based on porous, tremella-like NiFe₂O₄ with nanosheets has the characteristics of a high response, a low detection limit, and fast response and recovery times, which are conductive to practical applications in toluene detection.

As a widely accepted sensing mechanism, the adsorbeed oxygen species (O⁻, O²⁻ and O₂⁻) play an crucial role in determining the sensing performance of a sensor based on semiconductor oxides. The NiFe₂O₄ is a p-type semiconductor whose conductivity carriers come from holes. A porous, tremella-like NiFe₂O₄ nanosheet reacts with adsorbed oxygen and its toluene target gas as follows [1,67]:

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

(6)

$${\mathrm{O}}_2\left(\mathrm{adsorbed}\right){ + \mathrm{e}}^{-}\to {\mathrm{O}}_2^{-}$$

(7)

$${\mathrm{O}}_2^{-}{ + \mathrm{e}}^{-}\to 2{\mathrm{O}}^{-}$$

(8)

$${\mathrm{O}}^{-}{ + \mathrm{e}}^{-}\to {\mathrm{O}}^{2-}$$

(9)

$${\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{CH}}_3+{\mathrm{O}}^{-}\to {\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{CHO}}^{-}{+ \mathrm{H}}_2\mathrm{O}\hspace{1em}$$

(10)

$${\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{CH}}_3+{\mathrm{O}}^{2-}\to {\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{CHO}}^{-}{+ \mathrm{H}}_2{\mathrm{O}+\mathrm{e}}^{-}$$

(11)

$${\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{CHO}}^{-}\to {\mathrm{C}}_6{\mathrm{H}}_5{\mathrm{CHO}+\mathrm{e}}^{-}$$

(12)

When exposed to air, the formation of oxygen species, as confirmed by the above XPS analysis, will capture free electrons from its conductance band. Upon exposure to toluene, the toluene molecules will react with these oxygen species, resulting in the release of the electrons. Accordingly, the change in conductance of the sensor occurs, achieving the detection of toluene. The true resistance change in the NiFe₂O₄ sensor at 1 ppm of toluene is shown in Figure 10a. The resistance of around 7.7 kΩ and 13.7 kΩ are tested in the air and in 1 ppm of toluene for the NiFe₂O₄ sensor, producing a response of 1.8. Figure 10b shows the reaction of the target gas toluene molecules in the porous nanosheets. First, oxygen is adsorbed onto the surface of the porous nanosheet, which then reacts with electrons in the conduction band, giving rise to active oxygen species (O⁻, O²⁻ and O₂⁻). After exposure to the target gas, the toluene molecule will react with active oxygen species to produce benzaldehyde and H₂O molecules, as illustrated by the above equations (Equations (6)–(12)) [60,68].

The gas-sensing performance of this sensor is mainly ascribed to the following aspects. Fistly, the gas-sensing performance is derived from the unique spinel structure of the NiFe₂O₄, which possesses good catalytic activity. NiFe₂O₄ with a spinel structure shows a reversible redox reaction (Ni²⁺ + h⁺ ↔ Ni³⁺, Fe²⁺ + h⁺ ↔ Fe³⁺), resulting in the catalytic ability for the toluene to react on the surface of the NiFe₂O₄ at moderate temperatures [69]. Secondly, the unique, tremella-like structure assembled by the ultrathin nanosheets of NiFe₂O₄ favors a toluene-gas-sensitive performance. The unique morphology not only possesses a large specific surface area to provide abundant active sites but also provides abundant mesoporous structures as effective channels for gas diffusion. Finally, the porous, ultrathin NiFe₂O₄ nanosheet structure plays an important role in the highly sensitive detection of toluene, especially at low concentrations. The porous nanosheet structure not only provides a large number of active sites for gas reactions but also gas molecule transport channels, offering a preference for the detection of low concentrations of toluene.

4. Conclusions

In summary, the porous, tremella-like NiFe₂O₄ assembled with ultrathin nanosheets was successfully synthesized with the assistance of 1,4-phenylenediboronic acid. The band gap of the as-synthesized NiFe₂O₄ is around 1.7 eV. The NiFe₂O₄ annealed at 400 °C shows better performance at sensing toluene. It exhibits a good response of 4.7, fast response and recovery times of 11.6 s and 41.9 s to 10 ppm of toluene at 220 ℃, and a low detection limit of 50 ppb. Our good sensing performance originates from the original catalytic activity of NiFe₂O₄, the high surface area (77.80 m²/g), and the unique porous nanosheet structure. The as-prepared porous, tremella-like NiFe₂O₄ with ultrathin nanosheets using 1,4-benzenediboronic acid as a ligand provides a feasible strategy for the fabrication of NiFe₂O₄ to detect ppb-level toluene. Finally, it is also worth noting that the response time and selectivity of the NiFe₂O₄ sensor toward toluene could be further enhanced by doping, coupling with other sensing materials, or functionalizing noble metal nanoparticles for practical applications.