Environment pollution and the energy crisis are two major challenges currently confronting human society. Recently, photocatalytic technologies have been shown to effectively decompose hazardous organic contaminants and address environmental and energy concerns to some extent [1]. Traditional semiconductor photocatalysts mostly comprise inorganic compounds, such as metal oxides, sulfides, nitrides, phosphides and their complexes. Most of them contain expensive metal elements [2]. Research into new types of inexpensive non-metal photocatalysts with high quantum efficiencies, high visible light usage, and high stability has garnered considerable interest worldwide in the field of photocatalysis [3].
Among the various available photocatalysts, graphitic carbon nitride (g-C3N4) is a promising non-metal photocatalyst that has received significant attention owing to its suitable band gap energy (2.7 eV), high thermal stability, excellent electronic properties [4], as well as the unique 2D structure favorable for hybridization [5]. Therefore, g-C3N4 is multifunctional with broad applications in water splitting activity [6, 7], energy conversion [8], and the degradation of environmental contaminants [9-11]. However, the photocatalytic efficiency of pure g-C3N4 is far from ideal owing to its low surface area and the fast recombination of photogenerated electron-hole pairs, which could be attributed to the classical thermal polymerization synthesis method [12]. Hence, it is necessary to develop strategies to improve the photocatalytic activity of g-C3N4 by balancing the relationship between electron transportation and light absorption. Therefore, extensive efforts have been devoted towards modifying g-C3N4 to enhance the photocatalytic activities, such as by doping with metal or non-metal elements [13-15], developing a heterojunction structure [16-19], introducing noble metals [20], and copolymerizing with other semiconductors [21]. Among them, copolymerization has been demonstrated to be efficient in modulating the electronic structure of g-C3N4 and the photocatalytic performance [22].
With regard to copolymerization, conducting polymers can serve as good candidates for the effective separation of photogenerated carriers from g-C3N4. Conducting polymers are important organic semiconductors with a unique delocalized π-π* conjugated electronic structure and excellent electrical, electrochemical, and optical properties; they have been applied in various electrocatalysts and photocatalysts, such as dye sensitized solar cells [23] and biosensors [24]. For example, He et al. [25] synthesized polyacrylonitrile (g-PAN)/g-C3N4 composites through the thermal condensation of PAN and melamine to enhance the visible light photocatalytic performance for H2 evolution. Hu et al. [26] developed a simple sonochemical approach to prepare g-C3N4 and polythiophene (Ptp) nanocomposites, which manifested better photocatalytic activity and stability than pure g-C3N4, owing to the strong interaction between polythiophene and g-C3N4. Ge et al. [27] synthesized polyaniline (PANI)/g-C3N4 compounds by in-situ deposition polymerization, which led to an improved photogenerated carrier separation. Among the aforementioned polymers, polypyrrole (PPy) possesses good environmental stability, high conductivity, and interesting redox properties, and it is easy to fabricate under various conditions [28]. Such characteristics make it a promising material that could be applied in photocatalyst modification to enhance pollutant degradation. For example, Torki et al. [29] investigated the photocatalytic performance of a NiS and NiS-immobilized magnetite PPy core/shell (Fe3O4@PPy) in cephalexin degradation. Wang et al. [30] found that PPy-modified Bi2O2CO3 showed enhanced photocatalytic performance compared with the pure Bi2O2CO3. Pruna et al. [31] introduced a simple two-step electrochemical approach to fabricate ZnO/PPy-GO hybrid shell arrays with better photoactivity. In addition, some research on PPy/g-C3N4 composites with Ag nanoparticles has also been conducted [32, 33]. The Ag nanoparticles anchored between g-C3N4 and PPy acted as electron transfer mediators that could facilitate the charge carrier separation. Nevertheless, some of the modification methods for PPy/g-C3N4 composites include complicated synthetic steps or introduce other semiconductors and noble metals, thus decreasing their practical utility. Therefore, systematic research on the modification of g-C3N4 with PPy is of great significance.
Herein, a series of novel PPy/g-C3N4 composites were prepared through a facile in-situ polymerization method; they exhibited improved photocatalytic activity in the degradation of methylene blue (MB) under visible light illumination compared with that of g-C3N4. The prepared composites were characterized by X-ray diffraction (XRD), Fourier-transform infrared spectroscopy (FT-IR), scanning electron microscopy (SEM), transmission electron microscopy (TEM), Brunauer-Emmett-Teller (BET) specific surface area spectrometry, ultraviolet-visible diffuse reflection spectroscopy (UV-vis DRS), photoluminescence (PL) spectroscopy, and photocurrent response. The effect of sodium dodecyl benzenesulfonate dose of PPy in the fabrication of the composites was also evaluated. Moreover, the main reactive species in the degradation reaction were detected by electron spin resonance (ESR) spectroscopy, and a possible mechanism for the enhanced visible light photodegradation of MB by PPy/g-C3N4 composites was proposed. The present work provides new insights into the mechanistic understanding of PPy in PPy/g-C3N4 composites for environmental applications.
Analytically pure chemicals, including melamine, ferric chloride hexahydrate, sodium dodecyl benzenesulfonate (SDBS), methylene blue (MB), and absolute ethanol, were all purchased from Chengdu Kelong Chemical Agents (China) and used as received without further purification. Pyrrole (chemically pure grade) was from Shanghai Aladdin Reagent Co., Ltd. and distilled twice to avoid coloration under lower pressures (vacuum). Deionized water was used in all of the experiments.
A certain amount of melamine was dissolved in deionized water, and ultrasonically dispersed for 1 h. The dispersion was dried at 60 ℃ for 12 h, and then calcined in an alumina crucible at 520 ℃ for 5 h (10 ℃ min-1) to give g-C3N4 as a powder. Subsequently, 4 g of the prepared g-C3N4 powder and 0.2 g SDBS were dissolved in 100 mL deionized water, and ultrasonically dispersed for 20 min. Then a specific amount of pyrrole was added to the mixture under constant stirring in 200 mL ice water (0 ℃). After an appropriate contact time, 10 mL of a specific concentration of aqueous ferric chloride was added drop-wise into the aforementioned solution. After 2 h, a significant change in the color of the solution from light yellow to light gray was observed. The mixture was still precipitated for 1 h, then filtered under vacuum and washed several times with 20 mL absolute ethanol and 100 mL deionized water. Finally, the composites were dried in an oven for 12 h at 60 ℃. The resulting composites were labeled as XPPy/g-C3N4, where X represents the mass ratios of pyrrole and g-C3N4 for 1, 0.75, 0.5, and 0.25 wt%. PPy/g-C3N4 composites with different amounts of added SDBS (e.g., 0.1, 0.2, and 0.3 g) were also prepared using the same method.
The crystal structures and phase composition were characterized by X-ray diffraction with Cu Kα radiation (XRD-6100, Shimadzu, Japan). The chemical bonds of the samples were analyzed by FT-IR (IR Prestige-21, Shimadzu, Japan). The morphology and microstructure of the prepared products were measured by SEM (JSM-7000F, JEOL, Japan) and TEM (JEM-2100F, JEOL, Japan). The specific surface area and pore diameter distribution of the as-prepared samples were calculated by N2 adsorption-desorption analysis (BET-BJH, ASAP 2020, Micromeritics, USA) at -196 ℃. The optical properties and band gaps were examined by UV-vis DRS (UV-2550, Shimadzu, Japan) using BaSO4 as the reference [16]. The PL (Hitachi F-7000, Japan) spectra of the samples were measured at room temperature under an excitation wavelength of 350 nm. The photocurrent response analysis was carried out by an electrochemical workstation (CHI 660E, China) with a three-electrode cell system equipped with an LED light as the visible light source. The working electrode with an area of 1 cm2, Pt wire and saturated calomel electrode were used as the counter and reference electrodes, respectively, and 0.1 mol L-1 Na2SO4 acted as the electrolyte. The ESR signals of the radicals spin-trapped with 5, 5-dimethyl-1-pyrroline-N-oxide (DMPO) were measured using an ESR JES-FA200 spectrometer (JEOL, Japan). A methanol dispersion was used for DMPO-•O2- and an aqueous dispersion was used for DMPO-•OH.
The photocatalytic degradation of MB was conducted using a 12-W LED lamp (light intensity was 28 mW cm-2 with a light distance of 15 cm). The desired catalyst (0.05 g) was added into 50 mL of the dye aqueous solution containing 10 mg L-1 MB. The mixture was magnetically stirred until a homogeneous suspension was obtained. MB molecules on the surface of the catalyst reached adsorption-desorption equilibrium after 1 h in a dark chamber where the concentration of the MB aqueous solution remained stable; the 2 h experiment under visible light was conducted immediately after equilibrium was reached. Meanwhile, 5 mL aliquots of the suspensions were collected at given time intervals for target dye concentration analysis immediately after centrifugation (5000 r min-1, 5 min). The MB concentration was determined by UV/vis spectroscopy at its characteristic wavelength of 662 nm [27]. The degradation rate (η) of MB was calculated as follows:
where C0 is the concentration of the dye solution adsorption-desorption equilibrium, and Ct is the concentration of the dye solution at time t after the photocatalytic reaction [30].
Fig. 1 shows the typical XRD patterns of the pure g-C3N4, 0.75PPy/g-C3N4, and PPy samples. Two distinct diffraction peaks of pure g-C3N4 appeared at 13.1° and 27.6° and were indexed as the (100) and (002) lattice planes, respectively [34]. The diffraction of the (002) lattice plane corresponded to the interlayer accumulation of stacked conjugated aromatic compounds, which commonly represents a graphitic structure. The (100) lattice plane corresponded to the tri-s-triazine unit structure, which was related to the experimental precursor of melamine [35]. The tri-s-triazine ring structure effectively reduced the carbon–nitrogen bond reaction barrier to promote the growth of the graphitic layer crystal structure [13]. The 0.75PPy/g-C3N4 sample exhibited similar peak positions to those of pristine g-C3N4, which demonstrated that modification with PPy did not change the lattice structure of g-C3N4, but affected the crystal growth and reduced the crystallinity, as noted by the decreasing diffraction intensity. In addition, no PPy diffraction peaks were observed, which should be ascribed to the small amount of the polymer with amorphous particles.
To further verify the interaction between PPy and g-C3N4, FT-IR spectra were obtained, as shown in Fig. 2, to investigate the chemical bonding. The same characteristic peaks appeared in pure g-C3N4 and the 0.75PPy/g-C3N4 composite. The peak at 807 cm-1 was assigned to the out-of-plane stretching of the tri-s-triazine unit. The series of peaks located at 1627, 1406, 1315 and 1246 cm-1 corresponded to the typical C(sp2)–N, C(sp2)=N stretching modes of g-C3N4 [36]. The peaks of PPy, located at 1529 and 1301 cm-1, were attributed to the symmetric and antisymmetric ring-stretching modes of the C=C and C–H vibrations in the pyrrole ring, respectively. The peak at 1035 cm-1 was attributed to the C-N heterocyclic stretching vibration and that at 916 cm-1 corresponded to the double polarization state of PPy [24]. The vibration peaks of the 0.75PPy/g-C3N4 composite were slightly stronger than those of pure g-C3N4, which was proposed to be because PPy was loaded on the g-C3N4 surface and the peaks of PPy and g-C3N4 overlapped with each other owing to their similar molecular structures. In short, FT-IR analysis supported that g-C3N4 and PPy co-existed in the as-prepared samples.
The surface morphologies of the samples were examined by SEM and TEM analysis. As shown in Fig. 3(a) and (b), the as-prepared pure g-C3N4 exhibited a smooth bulk layer structure, which was similar to the analogue graphite [37] with an irregular shape [38]. The PPy particle size was much smaller than that of g-C3N4. As for pure PPy and the 0.75PPy/g-C3N4 composite, small amorphous PPy particles were observed on the thin sheet-like g-C3N4 surface, which confirmed the composite structure illustrated in Fig. 3(e) and (f). No obvious agglomeration was observed in Fig. 3(c) and (d). The junction between g-C3N4 and PPy can increase the contact area [13], which may facilitate the charge transfer and inhibition of electron-hole recombination.
The specific surface area and pore structure of the as-selected samples were determined using nitrogen adsorption-desorption isotherms, as shown in Fig. 4. The 0.75PPy/g-C3N4 and g-C3N4 samples exhibited type-Ⅳ isotherms with H3 hysteresis loops at high relative pressures (i.e., between 0.5 and 1.0), which demonstrated the presence of mesopores (2–50 nm) and macropores (>50 nm) (Fig. 4) [26, 39]. Meanwhile, the relative pore-size distribution curves, which are presented in the inset of Fig. 4, were determined to calculate the broader pore diameter distribution at 1–140 nm. The mesopores of pristine g-C3N4 and 0.75PPy/g-C3N4 were mainly observed with sizes of 3.48 and 3.50 nm, while the wider mesopores were observed with sizes between 20 and 40 nm. Moreover, PPy inhibited the aggregation and controlled the pore distribution of the g-C3N4 nanoparticles [40]. The specific surface areas, total pore volume and peak pore size of pure g-C3N4 and 0.75PPy/g-C3N4 are presented in Table 1. The specific surface area was 18 m2 g-1 for 0.75PPy/g-C3N4 and 11 m2 g-1 for pure g-C3N4, which illustrated that PPy introduced a porous structure and increased the surface area of pure g-C3N4. This interesting result agreed well with the SEM observations of the deposited amorphous PPy particles shown in Fig. 3.
The optical properties of the as-prepared PPy/g-C3N4 composites were characterized by UV-vis DRS and PL spectroscopy, as shown in Fig. 5. UV-vis DRS was employed to analyze the intensity of light absorption. In Fig. 5(a), no absorption edge was observed for black PPy because of the absorption over the entire spectrum [32]. For pure g-C3N4, the basal absorption edge was at 450 nm, and the absorption intensity over the whole visible light region for the composites gradually increased with increasing the PPy content from 0 to 1 wt%. A slight red-shift in the absorbance was observed, which illustrated the enhanced visible light absorption ability of the composites with the interaction between g-C3N4 and PPy semiconductors in the heterojunctions [10]. With a 0.75 wt% PPy loading amount, the absorption edge was at 490 nm, which resulted in a broader absorption than that of pure g-C3N4. Based on the above results, PPy served as a photosensitizer with a strong absorption coefficient and caused the photosensitization of g-C3N4 [28]. In comparison with the pristine counterpart, the composites modified by PPy were able to enhance the absorption in the visible light region, which could lead to the superior usage of solar light.
Fig. 5(b) shows the PL spectra of the as-synthesized composites under an excitation wavelength of 350 nm, where a lower PL intensity resulted from the higher separation efficiencies of the photogenerated electron-hole pairs [35]. In general, the fluorescence intensity of the composites decreased when the PPy content was increased from 0 to 1 wt%, as compared with that of g-C3N4 alone. Owing to the high conductivity of PPy, 0.75PPy/g-C3N4 had the weakest intensity among all of the samples, which suggested that the electrons stimulated from the pure g-C3N4 were efficiently transferred by the conjugated structure, thus preventing the recombination of electrons and holes [9]. As a result of the efficient electron separation, a host of electrons could react with the dissolved oxygen on the sample surface to generate superoxide radicals for pollutant degradation [41].
To further investigate the separation efficiency of the electron-hole pairs in the composites, the photocurrents generated from the photocatalysts were measured under visible light with several on-off cycles of intermittent irradiation. The photocurrent-time curves of the pure g-C3N4 and 0.75PPy/g-C3N4 photocatalysts are shown in Fig. 6. The photocurrent was observed to be reproducible and stable. The photocurrent response of 0.75PPy/g-C3N4 was constantly more than four-fold higher than that of pure g-C3N4. In general, an enhanced photocurrent is widely regarded as the most efficient evidence to demonstrate the separation and diffusion of photogenerated electrons and holes, which are transferred from the inner structure of the photocatalyst to the free charge acceptors on its surface. Furthermore, a high separation efficiency of photoexcited carriers plays a key role in the photocatalytic activity of photocatalysts. Therefore, PPy can effectively suppress electron-hole recombination by intensifying the charge transfer of g-C3N4 under visible light irradiation. This result was confirmed by PL spectroscopy, as shown in Fig. 5(b).
In this study, the organic acid anion surfactant SDBS was used as the dopant to improve the conductivity of PPy [23]. Pure PPy possesses a π-π* double conjugate structure and poor conductivity in a long chain state, thus the introduction of dopants in the polymer structure typically improves the conductivity [24]. However, polymer chains with conjugated structures undergo charge transfer or redox reactions with dopants, so different types of dopants will have different effects on the conductivity mechanism [26]. The visible light photocatalytic activities of the as-prepared samples with different amounts of added SDBS addition were evaluated in the degradation of MB as the target compound. As shown in Fig. 7(a), the composites with 0.1, 0.2, and 0.3 g of SDBS showed higher degradation rates than pristine g-C3N4 after 2 h of visible light irradiation. Particularly, 0.2 g SDBS resulted in a 2.4 times higher degradation efficiency based on the pseudo-first-order model (Fig. 7(b)), which will be discussed later. The composite containing 0.3 g SDBS was burdened with a small amount of pyrrole to form the PPy structure [42], while that containing 0.1 g SDBS was not suitable for the PPy rings. In the reaction, the polymer chains provide electrons and exhibit electropositivity; thus they combine with the anionic surface active agent directly. The addition of SDBS was essential to the internal π-π* conjugated structure of PPy as well as the effective electron transfer. The changed energy state of PPy led to the decreased band intensity and increased conductivity. Therefore, with SDBS addition, the primary chain of PPy was changed and easily reacted with the oxygen on the g-C3N4 surface, which resulted in more active substances to facilitate the photodegradation process.
Next, the visible light photocatalytic activities of the composites with different mass ratios of PPy were evaluated. As shown in Fig. 8(a), the composites exhibited better photoactivities than pure g-C3N4 after 2 h of irradiation with visible light. Moreover, the degradation of MB treated with g-C3N4 with 0.75 wt% PPy reached approximately 99% after 2 h of visible light irradiation. The degradation of MB by 0.75PPy/g-C3N4 increased by 24% when compared with that with g-C3N4. Moreover, 0.75PPy/g-C3N4 showed strong adsorption in the dark experiment, mainly because of the increased surface, which was supported by the BET measurements (Fig. 4).
With respect to the reaction kinetics of the MB degradation, a pseudo-first-order model was adopted as shown by the linear fitting of ln(C0/Ct) versus t, where k is the apparent pseudo-first-order kinetic constant (min-1) [38]. The kinetic plots and corresponding k values are shown in Fig. 8(b). Apparently, the loaded quantities of PPy greatly influenced the photodegradation rate. The sample with 0.75 wt% PPy exhibited the highest photodegradation efficiency with a k value of 0.03773 min-1, which was 3 times higher than that of pure g-C3N4 (0.01284 min-1). The PPy nanoparticles dispersed on the semiconductor increased the number of single active junctions and enhanced the light absorption of g-C3N4 [23], thereby improving the photodegradation rate and photocatalytic performance.
According to the aforementioned degradation studies, PPy/g-C3N4 composites exhibited better photoactivity than g-C3N4. Determining the major oxidative species in the photocatalytic processes may shed light on the photocatalysis mechanism [43]. The ESR spin-trap technique was employed to monitor the reactive species generated upon irradiation of the PPy and g-C3N4 composite system with DMPO in methanol and in water for superoxide radical (•O2-) and hydroxyl radicals (•OH), respectively [44]. The ESR results of pure g-C3N4 and 0.75PPy/g-C3N4 are shown in Fig. 9. Signals corresponding to DMPO-•O2- were detected in the two samples, as presented in Fig. 9(a). The intensity of the signal of the 0.75PPy/g-C3N4 increased slightly after 5 min of irradiation, which confirmed that •O2- serves as an effective radical for pure g-C3N4 and PPy/g-C3N4 composites photocatalysis. Meanwhile, the DMPO-•OH signal of 0.75PPy/g-C3N4 was observed after 5 min of irradiation (Fig. 9(b)). Moreover, the intensity of the •OH signal increased considerably after 15 min of irradiation. However, the pure g-C3N4 sample showed almost no active signal, which indicated that the addition of PPy enhanced the production of both active radicals by electrons separation through the formation of the composite structure. Moreover, it is illustrated that •OH cannot be regarded as the active radical for pure g-C3N4. Therefore, a reaction mechanism involving both •OH and •O2- was expected in the photocatalysis process for the composites, as they are the main oxidative species in the PPy/g-C3N4 composite systems.
A reasonable degradation mechanism of MB by PPy/g-C3N4 catalysts under visible light is depicted in Fig. 10. As an organic semiconductive material, PPy induces the transition of excited electrons from the highest occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO) because of the highly delocalized π-π* conjugate structure under visible light [45]. Meanwhile, the conduction band (CB, -2.2 eV) for electrons (e-) is formed on the HOMO and the valence band (VB, +0.18 eV) for holes (h+) is formed on the LUMO in PPy [46]. Moreover, the CB potential of PPy is more negative than that of g-C3N4 (CB, -1.3 eV) and thus is regarded as an electron acceptor [47], while PPy acts as an electron transfer channel to rapidly separate electrons injected from the excited state into the CB of pristine g-C3N4 [48]. In addition, g-C3N4 excited by visible light irradiation generates electron-hole pairs. The photogenerated electrons transit to the CB of g-C3N4 to overlap with the electrons from the π*-orbital of PPy, whereas the photogenerated holes migrate from the g-C3N4 VB (+1.4 eV) into the π-orbital of PPy (+0.8 eV) driven by the potential offset [26], which is the main driving force for efficient charge separation and transfer [49]. In the meantime, a steady internal electric field for the redistribution of electrons on one side of the g-C3N4 and holes on the opposite side (PPy) has been established, which ensures the effective separation and prolonging of the lifetime of photocarriers. As the photogenerated electrons and holes are spatially separated into two components, the electrons in the CB of g-C3N4 could be trapped by O2 to produce •O2- owing to the sufficiently positive potential that is higher than the reduction potential of the O2/•O2- couple (-0.33 eV vs. NHE) [6, 10]. However, the EVB of g-C3N4 is lower than the standard redox potential of H2O/•OH (+2.38 eV vs. NHE) and OH-/•OH (+1.99 eV vs. NHE), which indicated that the photogenerated holes of g-C3N4 could not oxidize H2O and OH- to the active species •OH [7, 8, 50]. The •OH radicals should be generated by the further reduction of •O2- with successive electrons and H+. The •O2- and •OH radicals can react with MB, which results in the decoloration of dye solution and the formation of CO2, H2O, and intermediate products in the degradation process [51]. The major reaction steps under visible light irradiation could be written as follows:
Therefore, the high charge separation and transfer efficiency are important factors for the enhanced photoactivity, which emphasizes the role of PPy in the enhanced photocatalytic activity of the composites.
PPy/g-C3N4 composites were prepared by the in-situ polymerization method. Amorphous PPy particles coated on the g-C3N4 surface increased the specific surface area, and facilitated the separation of charge carriers, as well as extended the lifetime of electron-hole pairs. The strong absorption coefficient of PPy could strengthen the visible light adsorption of the composites. The •OH and •O2- radicals were confirmed to be the main reactive species for the PPy/g-C3N4 composites in the photocatalysis process. PPy/g-C3N4 exhibited enhanced visible light photocatalytic activity, when compared with g-C3N4 owing to the designed heterostructure with band structure matching. The present work provides a new approach for the design of a high-performance g-C3N4-based photocatalyst for environmental applications.
2018,
Vol. 39 Issue (4): 831-840 
PDF
2018,
Vol. 39 Issue (4): 831-840 PDF