Enhancing NO degradation in visible light through plasma-treated photocatalytic substrates featuring TiO2@g-C3N4 Z-scheme structure
Article information
Abstract
In this study, a novel approach is introduced involving grafting and dip coating on photocatalytic substrates to create a Z-Scheme structure of TiO2@g-C3N4 (TC). The synthesized photocatalyst undergoes comprehensive characterization through techniques such as XRD, SEM, XPS, FTIR, Contact Angle Goniometer, and BET analysis. Two configurations, TC coated on PVDF membrane (TC-PVDF) and TC coated on Film (TC-Film), are evaluated for the photodegradation of NOx under visible light. Compared to the benchmark material TiO2, TC-PVDF nanoparticles demonstrate outstanding performance, boasting a specific surface area of 60.93 m2/g and minimal toxic NO2 byproducts, with concentration as low as 3.54 ppb, constituting only 0.46% during NOx remediation. These nanoparticles exhibit remarkable stability, with less than a 10% decrease in efficiency after 5 cycles of visible light irradiation. Plasma treatment transforms TC-PVDF from superhydrophobic to hydrophilic, with a contact angle reduction from 124.7° to 0°. Notably, TC-PVDF shows substantial efficiency enhancements, reaching 84.04%. This study broadens insights into catalytic bag filters, providing practical implications for their use in NOx removal. It offers innovative solutions for air purification, addressing environmental challenges and advancing sustainable technologies.
Abstract
Graphical Abstract
1. Introduction
The escalating global energy demand, driven by rapid population growth and industrialization, is anticipated to double by 2050. However, the reliance on fossil fuels such as petroleum, coal, and natural gas not only contributes to their depletion but also exacerbates environmental pollution. Consequently, the need to address both the impending energy crisis and environmental challenges has compelled the exploration of sustainable technologies. Semiconductor-based photocatalysis has emerged as a promising solution, concurrently offering renewable energy production and effective pollutant degradation by harnessing solar energy and abundant water resources [1].
Over the years, a range of semiconductor photocatalysts has been studied extensively for pollutant degradation, with titanium dioxide (TiO2) garnering significant attention. TiO2 exhibits notable qualities such as high photocatalytic activity, chemical stability, low toxicity, and cost-effectiveness. When exposed to sunlight, TiO2 initiates photochemical reactions by generating electron-hole pairs, which participate in oxidation and reduction processes to form highly reactive radical species. However, the intrinsic limitations of TiO2, including its wide band gap and rapid recombination of charge carriers, hinder its full potential for efficient solar energy conversion [2].
The Z-scheme structure in the photocatalytic field refers to a specific type of heterostructure that consists of two semiconductors with staggered energy band alignments. This structure allows for efficient charge separation and transfer, leading to enhanced photocatalytic activity. The Z-scheme heterostructures can be constructed by anchoring different semiconductors onto each other, creating well-defined chemical bonding and intimate interfacial contact between the components. The interaction between the semiconductors in the Z-scheme structure facilitates interfacial charge transfer channels, reduces energy barriers, and promotes the overall stability and reusability of the photocatalyst. The Z-scheme mechanism enables the transfer of photogenerated electrons and holes between the semiconductors, preventing recombination and improving the efficiency of photocatalytic reactions. This design strategy has been successfully applied in various photocatalytic systems, such as CO2 reduction, water splitting, and pollutant degradation [3–7].
Addressing these limitations requires strategies to extend TiO2 light absorption into the visible spectrum and enhance charge carrier separation. Among the various methods, the creation of Z-scheme with other semiconductors stands out as a promising approach. In this context, graphitic carbon nitride (g-C3N4) emerges as a remarkable visible light-responsive and robust photocatalyst [8]. Notably, g-C3N4 possesses attributes such as non-toxicity, thermal stability, chemical inertness, and a moderate band gap suited for solar spectrum utilization [9–11].
To bridge the gap between efficient solar energy conversion and effective pollutant degradation, researchers have explored the coupling of TiO2 and g-C3N4 to form Z-scheme. This strategy leverages the complementary attributes of both materials to enhance photoactivity and charge separation [12]. The Z-scheme design combines the UV activity of TiO2 with the visible light response of g-C3N4, expanding the light absorption range and enabling the utilization of solar radiation for photocatalytic reactions. Furthermore, the Z-scheme architecture provides an increased surface area, enabling enhanced reaction kinetics [13, 14]. The principle involves the excitation of electrons in the valence band, driven by photons, to jump to the higher conduction band. Notably, the band gap of g-C3N4 is 2.7 eV, enabling excitation by visible light, while TiO2, with a 3.2 eV band gap, requires UV light for electron excitation [15, 16]. To enhance photocatalytic activity, recent attention has turned to Z-scheme systems, with g-C3N4 gaining prominence for its visible-light activity, stability, and ease
of manufacture. The limitations of TiO2, such as a wide bandgap and narrow UV light response, have spurred the development of heterojunction nanocomposites with g-C3N4 to improve sunlight utilization [17]. However, g-C3N4 faces challenges like fast charge carrier recombination, mitigated by combining it with TiO2 in devices that exhibit both visible light absorption and efficient carrier separation [18]. The Z-scheme structure in TiO2 and g-C3N4 refers to a type of heterojunction where the photoexcited electrons in the two materials are transferred in a sequential manner, resulting in enhanced charge separation and improved photocatalytic performance [19]. This structure has been observed in various composite systems involving TiO2 and g-C3N4. For example, Ding et al. constructed a ternary TiO2/MIL-88A(Fe)/g-C3N4 heterojunction, which exhibited a desirable Z-scheme band alignment and improved separation and transfer efficiency of photoinduced charge carriers [20, 21]. Similarly, Wang et al. synthesized TC composites and found that the Z-scheme interaction at the heterojunction played a crucial role in enhancing the photocatalytic hydrogen production rates [22]. Another study by Dai et al. reported the fabrication of a Z-scheme P-TC heterostructure, which showed excellent photocatalytic degradation performance against sulfonamide antibiotics [23, 24]. These findings highlight the significance of the Z-scheme structure in promoting efficient charge transfer and improving the photocatalytic properties of TiO2 and g-C3N4 composites. As demonstrated by various studies, the TC Z-scheme exhibits enhanced photocatalytic performance for applications such as hydrogen generation, environmental purification, and pollutant degradation [25]. The morphology-based TC heterostructures, including mesoporous, core-shell, and 2D architectures, have shown superior activity compared to individual TiO2 or g-C3N4 components. Mechanistic insights into the photodegradation of pollutants have been gained through radical trapping experiments and spectroscopic techniques [25]. The stability and reproducibility of the TC system have also been investigated, providing critical insights for practical applications.
In the pursuit of enhancing the efficiency of NOx removal, a novel approach involving the integration of TC Z-scheme photocatalytic systems has emerged. This innovative strategy harnesses the complementary attributes of TiO2 and g-C3N4, capitalizing on their combined potential for solar energy utilization and pollutant degradation. To further optimize this efficiency, a substrate transformation becomes pivotal. By introducing Polyvinylidene Fluoride (PVDF) and film as the substrate and functionalizing it with polyacrylic acid (PAA), a robust framework for NOx removal is envisioned. This multifaceted approach encompasses the creation of TC Z-scheme systems, the engineering of substrate materials, and the incorporation of functional polymers. These concerted efforts hold the promise of reshaping the landscape of NOx abatement technologies.
2. Method
2.1. Materials and Chemicals
All compounds were of analytical grade and used without further refinement. Any necessary dilutions were carried out using deionized water sourced from the laboratory water purification system (DI water, Ultrapure PURELAB Flex 3, ELGA Lab Water, UK). The TiO2 (with an assay of 99.4%) was provided by PANREAC QUIMICA S.L.U. located in Polígono Pla de la Bruguera, Castellar del Vallès (Barcelona). Urea (NH2CONH2, product number 635724, with an assay exceeding 99.5%) was procured from Sigma-Aldrich® Solutions in Germany, while Nitrogen (NO + N2) gas was obtained from the Ming Yang special gas company. Ethyl Alcohol (Choneye Pure Chemicals, purity ≥ 95%) is used for cleaning substrates and dip coating solution. Acrylic acid (98% stabilized) was purchased from Acros Organics.
2.2. Preparation of TC Suspension on The Substrates
2.2.1. TC synthesis
In order to create the TC composite, a meticulous procedure was adhered to, as illustrated in Fig. S1. Initially, a composite blend was meticulously crafted by combining a measure of 30 grams of Urea with a precisely calculated quantity of TiO2. This amalgamation was then subjected to a rigorous grinding process before being carefully placed within a crucible, ensuring even distribution. The crucible, now loaded with the synthesized mixture, underwent a controlled heating process. This involved subjecting it to elevated temperatures within an oven set to precisely 550°C, and this thermal exposure was maintained for a period of 120 minutes, allowing for the crucial interaction and transformation of the components as in previous research [26].
Following the heat treatment, the crucible, now housing the potential TC composite, was gently transferred to a furnace oven for a subsequent phase. Here, a patient approach was adopted, as the crucible was allowed to gradually cool down within the furnace over an extended span of 12 hours. This slow cooling process was vital to prevent abrupt temperature changes that could potentially compromise the integrity of the composite material.
The methodology employed for the fabrication of the TC composites can be characterized as a one-step pyrolysis technique, a process that involves a singular and continuous transformation of the precursor materials. This innovative method allowed for the manipulation of the final composition by utilizing different weight ratios of TiO2 and Urea. Through this versatile approach, a range of composite variations could be achieved, each possessing distinct properties and potential applications.
2.2.2. Cleaning substrate
The plastic substrates used were sheets of Polyester, with a thickness of 1 mm and cut into circular pieces of 150 mm in diameter (referred to as D150 mm). The initial treatment involved cleansing the substrates using a combination of water and detergent, followed by rinsing with deionized (DI) water. Subsequently, the substrates underwent ultrasonic cleaning with Ethyl Alcohol, followed by another DI water rinse. To complete the process, the substrates were dried at 60°C for 30 minutes, ensuring thorough drying without distortion or damage [27].
2.2.3. Substrate modification procedures
To establish the required conditions, a vacuum pump was employed to evacuate the chamber, while a mass flow controller was used to maintain the Argon (Ar) gas flow at a rate of 20 standard cubic centimeters per minute (sccm), referred to as MPC. Utilizing a plasma reactor depicted in Fig. S2 (Dressler PFG-600, USA), a configuration involving two parallel electrodes was set up to generate plasma. The impact of varying combinations of plasma power (set at 100 W) and exposure durations (fixed at 4 minutes) on the polymer engraftment process was systematically examined. In order to initiate the generation of free radicals, Oxygen (O2) gas was introduced at a flow rate of 50 sccm for a duration of 10 minutes after the plasma treatment [28].
2.2.4. Dip coating process
In Fig. S4, the plasma-pretreated substrates were submerged in an aqueous solution of Acrylic Acid (20%) purified with nitrogen, initiating graft polymerization. To facilitate this process, the temperature was consistently maintained using a water bath set at 70°C for a designated duration of 30 minutes. In preparing the dip-coating solution, 0.2 grams of TC was mixed with 25 mL of deionized (DI) water and an additional 25 mL of 98% Ethyl Alcohol. Subsequently, the plasma-treated substrates were immersed in this solution. Upon deposition of the material layer and the complete withdrawal of the substrate from the solution, a drainage and drying phase ensued, lasting approximately 10 minutes. This step aimed to remove any surplus solvent, allowing it to drain back into the beaker. Between subsequent depositions of layers, intermediate heat treatments were implemented. These involved placing the substrate onto a ceramic base supported by metal stands and positioning it within a preheated and stabilized oven at 60°C for a span of 60 minutes [29].
Following the prescribed duration of dry treatment, the film was carefully detached from the substrate holder. It was then permitted to cool down to the surrounding temperature before being repositioned within the substrate holder, marking the commencement of the next layer deposition.
2.3. Photocatalytic Activity Test
The appraisal of the photocatalytic efficacy involves the breakdown of gaseous NOx under the influence of visible light generated by an OSRAM Xenon lamp with a 300 W output. A specifically prepared substrate containing the previously mentioned material is cautiously positioned within a chamber reactor crafted from stainless steel. The stainless-steel reaction chamber undergoes a thorough examination, and subsequently, the upper part of the reactor is sealed using glass to facilitate the passage of light. This reactor, with a volumetric capacity of 4.5 liters and dimensions measuring 30 cm in length, 10 cm in width, and 15 cm in height, accommodates the suspended lamp, placed at a distance of 40 cm above the reaction chamber (Fig. S3).
Once the concentration of NOx reaches the critical threshold of 500 ppb, humidity levels are regulated to 30% using a model 42c apparatus from Thermo-science. The Model 42i-TL Low Source NO-NO2-NOx Analyzer is renowned for its robust performance within the 0.01 to 1 ppm range (with a 10 to 300-second averaging time). This adheres to the U.S.A. EPA Reference Method RFNA-1289-074 as NOx analysis standard procedure. Following this, the UV lamp is activated, its emissions passed through a visible light filter, and directed at the sample for a period of 30 minutes. Prior to this light exposure, the sample is left in darkness to allow for the absorption of NOx gas. The concentration of NOx is consistently monitored by a chemiluminescence NOx analyzer, specifically the Sabio 6042 model, all the while maintaining a constant flow of NOx at both the inlet and outlet. This meticulously controlled experimental setup enables the systematic evaluation of the photocatalytic performance in the degradation of gaseous NOx under visible light conditions [26].
The NO photocatalytic efficiency, NO2 conversion, and apparent quantum efficiency (AQE) were calculated by Eq. (1), Eq. (2), and Eq. (3), respectively [30, 31]:
where
CNO (ppb) is NO concentration.
CNO2 is NO2 concentration, i is the initial concentration, and f is the final concentration.
NA (mol−1) is the Avogadro constant 6.02214086×1023 mol−1.
Vt (L min−1) is NO flow rate.
M (g mol−1) is NO molar mass.
The photon flux is 2.72×1019 cm−2 min−1.
The surface area of diameter dish is 113.1 cm2.
2.4. Reusability of Photocatalysts
The photocatalytic examinations were repeated five times for NOx degradation. Prior to each successive photocatalytic test, the sample underwent a double deionized water wash and was subsequently dried in an oven at 80°C for 60 minutes after the initial nitric oxide recycling experiment. The degradation efficiency mirrored that of the prior test as the NOx concentration attained 500 ppb [26]. Following five successive cycles, the photocatalytic breakdown of NOx exhibited consistent results, underscoring the catalyst's stability and aptitude for NOx degradation. Each recovery test lasted for a duration of 30 minutes.
2.5. Characterizations
The crystal structures were confirmed using X-ray diffraction (XRD) patterns on a PANalytical X'Pert PRO X-ray diffractometer with Cu K radiation (= 1.5406, 40 kV, 40 mA) in the 10°–80° range. Fourier-transform infrared spectroscopy (FTIR, Jasco FT/IR-4700 spectrometer) was used to detect structural vibrations. The powder samples were thoroughly mixed with KBr before being molded into pill shapes (99 wt% of KBr). The FTIR profiles were taken with wavenumbers ranging from 4000 cm−1 to 400 cm−1. SEM images taken with a FE-SEM S-4800 N, Hitachi, were used to examine the morphology of the materials. Specific surface area (BET-Tristar) used for molecules adsorption of gas on a solid substrate. Contact Angle Goniometer (Sindatek, Model 100 SB) is used for measuring the water contact angle.
3. Results and Discussion
3.1. Analysis of Structure and Composition
The X-ray diffraction (XRD) analysis revealed that the primary peaks of g-C3N4 exhibited a decrease in peak intensity after being mixed with TiO2. In TC, the majority of these peaks corresponded to the (0 0 4), (1 0 1), (1 0 5), (2 0 0), and (2 1 1) planes of TiO2 anatase structure. The existence of a minor quantity of rutile phase in TiO2 was suggested by a weak diffraction peak observed at 63.4° (0 0 2). Notably, the combination of anatase and rutile phases in TC may result in enhanced photocatalytic performance compared to the individual phases [32].
As shown in Fig. 1a, there is a prominent peak at 2θ = 19.9° and three smaller peaks at 2θ = 18.3°, 25.9°, and 39.0°. The peak at 2θ = 19.9° corresponds to the (1 1 0) crystal structure peak of the γ-PVDF phase. Similarly, the peak at 2θ = 39.0° corresponds to the (2 1 1) crystal structure peak of the γ-PVDF phase. Additionally, the peaks at 2θ = 17.9° and 25.8° are associated with the (0 2 0) and (0 2 1) crystal structure peaks of the α-PVDF phase, respectively [33].
The FTIR spectra of TC, g-C3N4, and TiO2 mixtures are presented in Fig. 1b. In g-C3N4, there is an absorption band observed between 1220 and 1650 cm−1, which corresponds to the characteristic C-N heterocycle stretching vibration [34]. The presence of Ti-O bonds is responsible for the absorption band within the 400-800 cm−1 range, as stated by [35]. The stretching modes of g-C3N4 can be identified by peaks observed at 809 cm−1 and 1220–1650 cm−1 [36], with the characteristic vibrational mode of triazine observed around 809 cm−1 [37]. Wide-ranging vibrations of O-H bonds are detected within the spectrum of 2900-3500 cm−1. It is noteworthy that the FTIR spectrum of the TC mixture exhibits characteristic features of both TiO2 and g-C3N4, indicating the successful combination of the two materials.
3.2. Morphology Characterizations
The SEM technique was utilized to examine the morphological variances between the altered and unaffected membranes. The surface, cross-section, and internal porous characteristics of the film and membrane are depicted in Fig. 2 and Fig. 3, respectively by the previous report [38]. The top-view SEM images vividly display PVDF membranes with a three-dimensional microporous structure. Fig. 2 illustrates the typical SEM images of TC-modified substrates. A rough layer can be observed on the smooth surface of the PAA-grafted substrate. The coating on the prepared substrates is evenly distributed and maintains the porous structure. The coating comprises TC nanoparticles with a Z-scheme structure, as depicted in the magnified image of the membrane in Fig. 2a, b. The size of this material in Fig. 2a, c is considerably larger than the diameter of the PVDF membrane pores (ranging from 200 to 750 nm), thus ensuring the photocatalytic performance of the catalyst [39].
The TC material was present on both the film and membrane surfaces, while the membrane substrate exhibited micro-pores, as depicted in Fig. 2a–c, and d–f, respectively. Furthermore, the pore width of the modified PVDF membranes shown in Fig. 2b was narrower compared to that of the unmodified membranes illustrated in Fig. 2c. Both membranes exhibited asymmetric morphologies, characterized by finger-like pores interconnected by sponge-like walls. The distribution of TC nanowires within the micro-pores, as shown in Fig. 2a, d, was observed to be uniform, which proved beneficial in enhancing the mechanical strength of the membrane. It was observed that the modified membrane material nanoparticles tended to aggregate and partially dissolve [38]. Upon close examination of Fig. 2a, d, it is evident that the catalyst dispersion on the PVDF membrane manifests a more homogeneous distribution as compared to the Film. This observed uniformity in dispersion can be ascribed to the inherent morphological features of the PVDF membrane. Specifically, Fig. 2c reveals a pronounced surface roughness coupled with the presence of voids, which likely facilitate enhanced catalyst distribution. In stark contrast, Fig. 2f delineates a more planar surface devoid of discernible voids, as corroborated by SEM analysis.
Utilizing cross-sectional SEM analysis, we were able to determine the thickness of both the substrates and associated catalyst layers. Specifically, the thickness of the PVDF membrane, as depicted in Fig. 3a, was quantified to be 125 μm, while that of the Film, illustrated in Fig. 3c, was identified to be 102 μm. When examining the coated material on these substrates, a thickness of 9.78 μm was discerned for the material layered on the PVDF membrane (refer to Fig. 3b and a measurement of 11.38 μm for that on the Film (as illustrated in Fig. 3d). It should be further noted that employing the RF plasma and PAA fabrication techniques had implications on the substrate morphology, resulting in increased surface roughness. Intriguingly, within the TC-PVDF sample, there appeared to be a dispersion of TC incorporated into the membrane structure.
In order to ensure the secure bonding of TiO2 with the g-C3N4 layer, an element mapping of TC was conducted. Fig. S5–S7 presents an EDS view of TC, specifically towards the edge, revealing its layered structure. The elements C, N, Ti, and O are evenly dispersed across the composite's surface, even at the boundary, confirming the mixture of TiO2 and g-C3N4 within the sample as predicted. The signals of Ti and O are significantly stronger than those of C and N, which can be attributed to the lower content of g-C3N4 in the composite, as explained by [40]. The consistent distribution of C, N, Ti, and O elements throughout the region indicates the successful development of a composite and heterostructure between g-C3N4 and TiO2.
XPS spectra was utilized to investigate the chemical composition and states of the TC coatings on PVDF and Film substrates. The survey spectrum presented in Fig. 4a shows distinct peaks for C 1s, N 1s, Ti 2p, and O 1s, with binding energies around 285 eV, 397 eV, 459 eV, and 530 eV, respectively. The content of each element is summarized in Table S1.
Further examination of the high-resolution C 1s XPS spectrum (Fig. 4b) reveals peaks at binding energies of 283 eV, 285 eV, 286 eV, and 288 eV for both coatings. Notably, the peak at 288 eV is consistent with carbonyl (C=O) or carboxylic (O–C=O) functionalities. The 286 eV peak suggests a C–O bond, and the 285 eV peak corresponds to sp3 hybridized carbon in hydrocarbons [41, 42]. The N 1s spectrum (Fig. 4c) displays peaks at 397 eV and 399 eV for both coatings. The 397 eV peak is characteristic of graphitic or pyridinic nitrogen configurations in g-C3N4 structures, while the 399 eV peak may indicate amino (−NH2) or amide (−CONH–) groups [43, 44]. The O 1s spectrum for the TC-PVDF (Fig. 4d) exhibits peaks at 527 eV, 530 eV, and 532 eV. The 532 eV peak could relate to oxygen in organic compounds. Conversely, the Film substrate displays peaks at 528 eV and 531 eV, likely representing hydroxyl groups (O–H) or adsorbed water [45]. The Ti 2p spectrum for PVDF (Fig. 4e) shows peaks at 456 eV, 459 eV, and 464 eV. The first two peaks fall within the range of Ti4+ in TiO2, suggestive of the anatase or rutile phase [46]. For the Film substrate, peaks are observed at 456 eV and 462 eV, potentially indicating different interactions or phases of TiO2. Generally, the XPS analysis offers a detailed insight into the surface chemistry and elemental states of the TC coatings on both PVDF and Film substrates, highlighting the importance of their chemical interactions.
In comparison to one-dimensional photocatalysts, the two-dimensional g-C3N4, consisting of a few ultrathin layers, typically exhibits a large specific surface area (SBET), leading to a higher number of available active sites on its surface. Therefore, the Barrett-Joyner-Halenda (BJH) and Brunauer-Emmett-Teller (BET) techniques were employed to examine the physical surface properties of TC. Notably, the SBET parameters of g-C3N4 and its pore characteristics are of significance. Liu (2022) demonstrated the TC/RGO ternary composites exhibited a specific surface area of up to 31.4782 m2/g, which contributed to their good catalytic activity and stability [47]. The TC hybrid with an optimal mass ratio of 12:1 showed nearly 1.9 times higher catalytic performance compared to conventional TC hybrids [48]. The TC photocatalyst with a large specific surface area of 59193 m2/g demonstrated excellent visible-light photocatalytic activity and good stability [49]. The self-assembled g-C3N4 samples with tailored morphologies exhibited increased BET surface area and 7 to 13 times higher photocatalytic performance compared to conventional bulk g-C3N4. The TC composites with different mass ratios showed enhanced photocatalytic activity, with the highest degradation efficiency achieved at a mass ratio of 1:3, attributed to the formation of a Z-scheme heterojunction [50]. According to the findings in Table S2, the SBET of g-C3N4 is approximately 105.63 m2/g, which is considerably higher than that of TiO2 at 47.41 m2/g. When the TC composite was evaluated on different substrates, distinct variations in surface properties emerged. For instance, as depicted in Fig. S8a, the pure TiO2 exhibited an SBET of 47.41 m2/g, a pore volume (VPore) of 0.2808 cm3/g, and a predominant pore diameter (DPore) of 22.21 nm. When integrated with a film substrate, as depicted in Fig. S8b, these values shifted to an SBET of of 54.60 m2/g, a VPore of 0.3634 cm3/g, and a DPore of 22.91 nm. In the case of the PVDF substrate, the observed values were an SBET of 60.93 m2/g, a VPore of 0.3524 cm3/g, and a DPore of 20.10 nm. However, the composite supported on a glass substrate recorded an SBET of 73.17 m2/g, a VPore of 0.5896 cm3/g, and a DPore of 31.56 nm. These findings underscore the influence of the substrate on the surface properties of TC, with potential implications for its photocatalytic efficacy.
The observed morphological features of TiO2 and g-C3N4, such as pore size and distribution, can significantly impact their photocatalytic performance. For g-C3N4, a hydrothermal treatment can modify its morphology by building extended melem units with more oxygen functional groups at the defect edges, resulting in lower photoluminescence emission intensity and more efficient charge separation, leading to higher photocatalytic hydrogen evolution rates [51]. In the case of TC heterojunctions, the well-dispersed TiO2 nanoparticles on g-C3N4 nanosheets enhance the photocatalytic activity for the degradation of monochlorophenols. The nitrogen and Ti3+ defects and oxygen vacancies in the TC nanocomposites improve the light-harvesting ability and prevent rapid electron-hole recombination, enhancing the photocatalytic performance [52]. Additionally, the loose and porous TC heterojunction composite photocatalysts prepared using a template method exhibit a greatly increased specific surface area, improved light absorption performance, and suppressed recombination of photogenerated electrons and holes, resulting in excellent photocatalytic performance [53].
3.3. Wettability of The Substrates
Fig. 5a illustrates the results of the controlled experiment on the water contact angle of the Film. The average water contact angle was measured to be 82.46°, and it decreased after plasma treatment regardless of the type of oxygen gas used. Specifically, the water contact angle for O2 plasma decreased to 73.00°, 62.65°, 43.05°, 24.50°, and 10.05°, respectively. For samples treated with O2 plasma + O2 exposure, the water contact angle decreased to 44.30°, 31.03°, 13.05°, 4.50°, and 0°, respectively. The water contact angle is directly related to the surface characterization, and it is further influenced by the structural and chemical composition of the surface [54].
The water contact angle (WCA) exhibited a consistent decrease of 106°, 106.3°, 106.7°, 92.08°, and 97.4° when subjected to O2 plasma for a plasma exposure duration of 5 minutes. The magnitude of RF power applied had an impact on the extent of reduction in the water contact angle. With the application of O2 plasma + O2 exposure, the contact angles gradually decreased as the plasma exposure time increased from 0 to 2 minutes. Following 1 to 3 minutes of treatment with O2 plasma, the water contact angles of the samples reached a saturated value without any significant further fluctuations. Fig. 5a demonstrates that after 2 minutes of plasma treatment, the contact angles of PVDF substrates treated with varying plasma exposure times were measured to be 92.08° when using O2 plasma + O2 exposure [55].
As depicted in Fig. 5a, extending the reaction time from 1 to 5 minutes (with an O2 plasma flow rate of 20 sccm) results in a decrease in the contact angle from 124.5° to 89.7°, as reported by [56]. However, it should be noted that if the plasma treatment is unsuccessful, a phenomenon called "hydrophobic recovery" may occur, causing the grafting process to be reversed. Excessive plasma exposure can also lead to damage to the polymer of the membrane, as highlighted by [57].
Following RF plasma treatment, the WCA of the PVDF membrane surface undergoes a change, as depicted in Fig. 5b. The plasma treatment leads to a transformation of the film surface, transitioning it from superhydrophobic to hydrophilic. The initial contact angle on the film surface experiences a significant reduction, dropping from 124.7° to 0°. This decrease in contact angle is comparable to the findings observed in La-doped TiO2 subjected to a cold radio frequency (RF) discharge plasma at low pressure. In the case of the PVDF membrane surface, the contact angle decreased from 121.1° to 9.2°, as measured using waste fluorescent powder [58].
3.4. Photocatalytic Performance
The relationship between the concentration of NO and the duration of irradiation in the presence of various materials is illustrated in Fig. 6a. When exposed to visible light, TC on different substrates such as TC-Film, and TC-PVDF exhibited the ability to remove 87.77%, and 84.49% of NO, respectively, within a 30-minute period. In comparison, pure TiO2 and g-C3N4 only achieved NO removal rates of 71.56% and 76.34%, respectively. This indicates that the choice of substrate had minimal impact on the photocatalytic efficiency. The photoactivity of the samples was found to be influenced by the proportion of TiO2 and g-C3N4 on the substrate, as suggested by [59]. The outcomes indicate that TC demonstrates significant potential for repeated use. Table 1 offers a comprehensive analysis comparing the findings of this study with previous research conducted on TC.
The rate of the photocatalytic reaction for the tested photocatalysts was determined using the Langmuir-Hinshelwood function and observed within a five-minute period, as illustrated in Fig. 6b. The performance of photocatalysis is closely linked to the reaction rate of the materials. In the case of TiO2, g-C3N4, TC, TC-Film, and TC-PVDF, the reaction rate constants were found to be 2.60×102 min−1, 1.13×102 min−1, 1.10×102 min−1, 1.55×102 min−1, and 1.42×102 min−1, respectively. The corresponding reaction rate constants (k) are 0.01781, 0.01365, 0.01542, 0.01671, and 0.01472 min−1, respectively. These values are also illustrated in Fig. S9. Although the reaction rate constant of TiO2 is the highest, being twice that of g-C3N4, the photodegradation effectiveness of TiO2 does not show a significant difference compared to g-C3N4 within the first five minutes. This suggests that TC decomposes NO more rapidly than both TiO2 and g-C3N4 [60]. Additionally, when transitioning from a glass substrate to film and membrane substrates, the efficiency of the photocatalysis improves as well.
In terms of NO degradation, g-C3N4 exhibited the lowest efficiency, with only a 42.42% conversion rate to green products. On the other hand, TC-PVDF demonstrated a significant increase in NO to nitrogen species conversion, reaching 84.04% with lowest NO2 conversion (0.46%). Although this value surpassed that of TiO2, it still outperformed g-C3N4, which achieved a conversion rate of 24.86%. Fig. 6c illustrates that the increase in NO to NO2 conversion may be attributed to the accumulation of photocatalytic byproducts, including NO3, HNO3, NH4, and NO2, within the chamber reactor. The high NO to nitrogen species conversion and low NO to NO2 conversion observed with TC on the membrane substrate indicate a robust oxidation of NO to NO3 through photocatalysis [60].
In the photocatalyst process to degrade NOx pollutants, it is necessary to determine the optimum catalyst dosage. The optimal catalyst dosage must be determined to avoid excessive use of the catalyst to degrade NOx gas, effectively. As shown in Fig. 6d, as the amount of photocatalyst increases from 0.02 to 0.4 grams, the degradation rate of NOx is lower at a high dosage of TC. In addition, low doses produce almost complete degradation. This is because as catalyst dosage increases, which could be caused by the active sites on the surface of the TC catalyst increase. Furthermore, a further increase in the amount of catalyst exceeding 0.05 grams may reduce the activity due to the obstacle of the light path to the gas molecules, which may be due to the light scattering phenomenon dominating under higher catalyst loading.
In order to comprehend the reasons behind the higher photocatalytic efficiency of TC samples compared to g-C3N4, the AQE was determined, as depicted in Fig. S10. The AQE values of the photocatalytic materials closely correspond to the mechanism of photocatalytic NO elimination, as shown in Fig. 7. The material exhibiting the highest efficacy in the NO elimination test, which is TC on film substrate, displayed an AQE result of 14.637×104%. In contrast, g-C3N4 yielded a comparatively lower AQE result of 10.6×104% [61]. g-C3N4 exhibited a relatively lower AQE result in comparison to TiO2 for NOx degradation, primarily due to factors such as lower NO conversion efficiency and increased formation of toxic NO2. In contrast, TC hybrid photocatalysts demonstrated superior NOx conversion, with a hybrid structure allowing efficient NOx adsorption and storage of the non-toxic NO3 −, while also facilitating visible light absorption. The enhanced photocatalytic activity of TC composites was attributed to the formation of a Z-scheme heterojunction between g-C3N4 and TiO2 [62, 63]. Furthermore, the band gap energy of g-C3N4, being higher than that of TiO2, resulted in reduced visible light absorption. Additionally, higher charge transfer resistance and electron recombination rates in g-C3N4 contributed to slower electron transport and decreased photodegradation efficiency [64, 65].
The specific surface area of g-C3N4 varied with calcination temperatures, ranging from 116.3 m2/g to 142.1 m2/g. In contrast, TC heterostructured photocatalysts exhibited a significantly larger specific surface area of 59193 m2/g [63]. Studies on pollutant degradation demonstrated that TC nanocomposites exhibited enhanced photocatalytic activity compared to pure g-C3N4, showcasing the potential of these materials in pollutant removal. For instance, TC/Ag film displayed excellent photocatalytic performance with a 64.95% photodegradation rate for Rhodamine 6G (R6G) [49]. These findings provide valuable quantitative insights into the specific surface area of g-C3N4 and TiO2 materials and their efficacy in degrading pollutants like NO2. The stability of the TC catalyst on different substrates (glass, film, and membrane) is demonstrated in Fig. S11. Over the course of 5 rounds of testing, the degradation efficiency of NO by the TC Z-scheme structure has remained relatively stable. There was a slight decrease in efficiency by 4%, 3%, and 8% when using glass, film, and PVDF substrates, respectively. Furthermore, the FTIR analysis of the TC catalyst before and after photocatalysis showed that the characteristic vibration peaks of TC remained unchanged in terms of their positions. However, there was a slight decrease in the intensity of these peaks, indicating a minor alteration in the samples [66].
3.4. Photocatalytic Mechanism
The enhanced performance of TC composites compared to pure TiO2 and g-C3N4 can be attributed to several specific features and mechanisms. Firstly, the formation of a Z-scheme heterojunction between g-C3N4 and TiO2 in TC composites leads to improved photocatalytic activity [50, 67]. This heterojunction structure delays the recombination of holes and electrons, enhancing the photocatalytic efficiency under UV-light irradiation [67]. Furthermore, the high light absorption intensity, large specific surface area, and small particle size of TC composites contribute to their superior photocatalytic activity. Overall, the combination of these features and mechanisms in TC composites leads to their enhanced performance compared to pure TiO2 and g-C3N4.
TC outperforms pure TiO2 and g-C3N4 due to its unique hierarchical macro-mesoporous structure, high adsorption capacity, and low band gap [62]. This structure provides aligned macroporous channels and mesoporous walls, enabling efficient NOx adsorption and storage of NOx oxidation product while allowing visible light absorption [68]. TC showed four times higher NO conversion compared to TiO2 and two times higher compared to g-C3N4 [69]. Additionally, TC exhibited high NOx adsorption capacity and converted NO to nontoxic NO3 − with minimal formation of NO2, making it a more effective photocatalyst for NOx removal [70]. The enhanced photocatalytic activity of TC can be attributed to the formation of a double Z-scheme heterojunction, which amplifies the absorption edge and effectively separates electron-hole pairs through the influence of an internal electric field [53]. The synergistic effect of the heterojunction between g-C3N4 and TiO2 and the Schottky barrier presented among TiO2 and Ti3C2 in the Ti3C2@TC ternary photocatalyst also contributes to its superior performance. Considering the aforementioned results and discussions, the photocatalytic mechanism for the TC sample is proposed and illustrated schematically in Fig. 7. When using pure TiO2, only a small proportion of the photogenerated electrons and holes actively participate in the photocatalytic process, leading to low overall efficiency [60, 71].
Through the integration of g-C3N4 in the TC composite, specifically at a lower TC content, a portion of the TiO2 nanoparticles' surface is effectively coated with smaller-diameter g-C3N4 particles. This results in the formation of a TC photocatalytic reactor with a Z-scheme structure. During UV irradiation, the holes generated by the light source tend to remain in the valence band of TiO2, while the electrons jump from the conduction band of TiO2 to the valence band of g-C3N4. Subsequently, these electrons in the valence band of g-C3N4 are further promoted to its conduction band.
As a consequence of this phenomenon, the photo-induced charge carriers become spatially separated. When holes (h+) in the valence band of TiO2 interact with adsorbed H2O molecules (or surface hydroxyls) at the surface, hydroxyl radicals (•OH) are formed. The oxidation of NO to NO3 and the inhibition of NO2 are facilitated by the generated •O2 [72]. Moreover, the hole present in the valence band of TiO2 engages in a reaction with H2O, leading to the formation of •OH radicals. As a result, the •OH radicals play a role in converting NO gas into alternative nitrogen species.
4. Conclusions
To conclude, this study has successfully achieved the synthesis of TC using a simple pyrolysis process. A groundbreaking approach, involving grafting and dip coating on photocatalytic substrates, has been introduced, resulting in a unique Z-Scheme structure featuring the synthesized TC. Of utmost significance is the remarkable ability of the synthesized material to facilitate the photocatalytic removal of NOx under UV–vis illumination, displaying a notable preference for the production of the non-toxic byproduct NO2. These exceptional characteristics position the material as a promising and viable photocatalyst for a range of applications. Regardless of the substrate, whether Film or PVDF, both variants exhibit substantial efficacy, achieving the elimination of 87.77% and 84.49% of NO respectively within a 30 minute timeframe. Strikingly, the choice of substrate yields minimal impact on the photocatalytic efficiency, whereas the distinct Z-scheme structure significantly influences the performance of these materials. This structure adeptly mitigates the recombination of electron-hole pairs, thereby extending light absorption by the composite material under visible light. This reduction in recombination not only curbs photocatalytic inactivation but also enhances the overall photocatalytic performance. Impressively, the nanoparticles produced on both TC-PVDF and TC-Film demonstrate exceptional effectiveness, boasting specific surface areas of 54.60 m2/g and 60.93 m2/g respectively, alongside negligible generation of the toxic byproduct NO2 (a mere 0.46%) and consistent stability across multiple cycles with marginal efficiency decline of 3% and 8% respectively. Notably, the photocatalytic performance of the PVDF-based variant outperforms that of TiO2, achieving a remarkable NO conversion efficiency of 84.04% compared to TiO2 (67.28%). The conversion of NO to nitrogen species, coupled with the minimal formation of NO2, particularly observed with the material on the film substrate, signifies robust NO oxidation to NO3 through the photocatalytic process. Additionally, the application of plasma treatment to the film-based material induces a significant shift from superhydrophobic to hydrophilic behavior, culminating in an impressive efficiency of 87.77%. These findings underscore the substantial influence of the substrate on the surface characteristics of the synthesized material, consequently shaping its photocatalytic performance. This research harbors significant potential for elevating the efficacy of these materials in diverse applications, particularly in enclosed settings like industrial power plant exhausts, where the synthesized material exhibits substantial promise for NOx removal with minimal generation of toxic byproducts.
Supplementary Information
Acknowledgments
We would like to thank Chung Yuan Christian University (Taiwan): Department of Civil Engineering, Department of Environmental Engineering, and the Center for Environmental Risk Management; and Center for Environmental Toxin and Emerging-Contaminant Research (Taiwan).
Notes
Conflict-of-Interests Statement
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Author Contributions
H.Q.L. (PhD student) wrote, revised the manuscript and conducted all the experiments. M.T.P. (Postdoc Fellow) supervised, wrote and revised the manuscript. Y.F.W. (Professor) supervised, wrote the manuscript. S.J.Y. (Professor) concepted, wrote the manuscript.