Effect of temperature on the synthesis of g-C₃N₄/montmorillonite and its visible-light photocatalytic properties

Abstract

Graphite phase carbon nitride (g-C₃N₄) is a non-metal semiconductor material with a suitable band gap (2.7 eV) for visible photocatalysis. However, the high cost of relevant synthesis methods and poor adsorption performance have limited its practical applications. The objective of the present study was to mitigate these problems by synthesizing the g-C₃N₄ in the presence of exfoliated montmorillonite (Mnt). Compared with bulk montmorillonite, the specific surface area of exfoliated two-dimensional Mnt layers was significantly increased. As a result, the light transmittance of the lamella improved noticeably due to the fact that a freshly exposed surface had a large number of active reaction sites, making Mnt an excellent carrier for the photocatalyst g-C₃N₄. In order to improve the photocatalytic performance of g-C₃N₄, a series of g-C₃N₄/Mnt composites was prepared by a wet chemical method using Mnt nanolayers as the matrix. X-ray diffraction, infrared spectroscopy, Brunauer-Emmett-Teller nitrogen adsorption/desorption, transmission electron microscopy, and ultraviolet-visible diffuse reflectance spectroscopy were used to analyze the phase structure, the chemical bonds, the specific surface area and pore sizes, the morphology, and the light absorption characteristics of the composites, respectively. Rhodamine B (RhB) served as the target dye to test the photocatalytic degradation performance of the composites under visible light. According to the findings, the surface of the Mnt nanolayers was densely and uniformly covered by g-C₃N₄, forming a multi-layered stack structure. An increase of the calcination temperature improved the crystallinity of g-C₃N₄, leading first to densification and then to relaxation of the layered composite structure. Conversely, the band gap of the composite gradually decreased from 2.56 to 2.4 eV. Furthermore, temperature exposure changed the photocatalytic performance of the composite drastically. While the largest photocatalytic activity was observed at 610°C, it started to decrease with further heating of the composite. The complete degradation of RhB solution occurred after 2 h of visible light irradiation. The findings of the current study provide a scientific basis for the synthesis of a new generation of photocatalysts.

Introduction

Photocatalysis is of great importance in the fields of energy and environmental applications, and is regarded as a revolutionary technology in the 21st Century. Most photocatalysts (such as TiO₂ and GaN) exhibit wide band gaps within the range 2.0–3.4 eV (Fujishima et al., 2008). This enables their photocatalytic activity under ultraviolet light (<385 nm) to be enhanced, which, meanwhile, accounts for only 3% of the solar energy (Herron et al., 2015). In that regard, searching for a highly active photocatalyst with visible light response and its practical application has attracted the attention of many scholars(Dong et al., 2014; Fatimah et al., 2011; Thomas et al., 2008).

Graphite phase carbon nitride (g-C₃N₄) is a non-metal semiconductor photocatalytic composite with a band gap (2.7 eV) suitable for the initiation of catalytic reactions under visible light. Moreover, g-C₃N₄ is an ecofriendly, non-toxic, non-polluting material that possesses stable physical and chemical properties, thus offering promise for a wide range of photocatalytic applications (Dong et al., 2014; Kumar et al., 2018; Wang et al., 2009).

Methods of synthesis of g-C₃N₄ include physical vapor deposition (PVD), high-temperature chemical vapor deposition (CVD), and hot melt techniques (Thomas et al., 2008; Li, Y.P., et al., 2014). However, all of them are costly and require stringent reaction conditions. In addition, unmodified g-C₃N₄ has no adsorption properties, and is easy to agglomerate, resulting in small specific surface area, high electron-hole recombination rate, and poor light utilization efficiency. These shortcomings reduce the photocatalytic efficiency of g-C₃N₄, thereby limiting its practical application (Xue et al., 2015).

Recently, research has been focused on supporting the multi-layered silicate mineral matrix to solve the problem of easy agglomeration and small specific surface area of photocatalysts and to extend their potential application range (Cheng et al., 2011; Fatimah et al., 2011; Peng et al., 2016). Increasing attention has been paid to montmorillonite as a promising carrier for photocatalysts due to its large specific surface area and good adsorption performance (Li, C. et al., 2016, 2018; Li et al., 2017; Sonawane et al., 2010; Sun et al., 2002). Also, attempts have been made to combine Mnt with graphite phase carbon nitride and to obtain materials with increased photocatalytic properties. For example, Li, S. et al. (2016) prepared a g-C₃N₄/Mnt composite with a g-C₃N₄ mass fraction of 31.8%, whose photocatalytic degradation rate in relation to RhB was four times greater than that of pure g-C₃N₄. Furthermore, combining g-C₃N₄ with Mnt as a carrier allowed those authors to solve the problem of g-C₃N₄ agglomeration effectively. According to Li et al. (2017), the g-C₃N₄/Mnt composite possessed a large specific surface area, good temperature resistance, and easy recycling, which help to improve its photo absorption characteristics. Moreover, the photocatalytic activity of the g-C₃N₄/Mnt composite prepared exceeded that of pure g-C₃N₄.

When water molecules get into the interlayer space of the Mnt 2:1 layers through an aqueous solution, the distance between the Mnt lamellae enlarges because of increasing interlayer penetration depth and peeling away of hydration layers. In this respect, compared with bulk montmorillonite, the specific surface area of exfoliated two-dimensional Mnt layers is increased significantly. This improves noticeably the light transmittance of the lamellae due to the fact that a freshly exposed surface has a large number of active reaction sites.

Given the above information, Mnt has the prospect of being a good photocatalytic support. However, some aspects still need to be clarified. Among these are the need to confirm an increase in specific surface area of exfoliated Mnt as the support for the g-C₃N₄ photocatalyst, and to study the photocatalytic effect of g-C₃N₄ in the g-C₃N₄/Mnt composite and its related mechanism. The objective of the present study was, therefore, to find answers to these questions by measuring the effect of temperature on the photocatalytic performance of a series of g-C₃N₄/Mnt composites, prepared by a wet chemical method using exfoliated Mnt as the matrix.

Experimental

Raw Materials and Reagents

The Mnt was prepared and exfoliated as follows. Bentonite (Santai City, Sichuan Province, China) was ground as a dry powder and passed through a 200-mesh sieve (74 μm), 10 g of which was combined with 0.6 g of anhydrous sodium carbonate in a 500 mL beaker, mixed by stirring, and then 160 mL of deionized water (with a resistivity of 18.2 MΩ • cm) was added, giving a a solid-liquid ratio of 1:10. The mixture was then stirred for 30 min in a water bath at 80°C to effect dispersion and sodium saturation. After that, the solid:liquid ratio was adjusted to 1:20 using deionized water, and the suspension was stirred in a high-speed homogenizer at a rotational speed of 10,000 rpm for 8 h. The resulting suspension was centrifuged at 600 rpm (30 g) and the resulting sedimented gel was frozen (–80°C) and freeze dried. The resulting dry powder was labeled “M”. A portion of the powder was also resuspended in water to give a 20% suspension.

Raw materials were chemically pure melamine, analytically pure anhydrous sodium carbonate, and RhB, all purchased from Chengdu Kelong Chemical Reagent Factory (Chengdu, China).

Catalyst Preparation

Melamine (9.0 g) was added to 1000 mL of deionized water and stirred with a magnetic stirrer at 80°C to prepare a hot melamine solution, which was then dropped into 1000 mL of the Mnt suspension using a peristaltic pump at a speed of 1.2 mL/min. The combined suspension was stirred for 12 h at 80°C in a water bath, then dried in a vacuum freeze dryer to obtain a precursor composite. To obtain the g-C₃N₄/Mnt composites, the precursor was placed in a 50 mL corundum crucible, sealed with tin foil, and calcined for 4 h at 490, 520, 550, 580, 610, or 650°C. The heating rate was 2.3°C/min. After natural cooling, the composites were labeled as CNM-T, where CNM stands for g-C₃N₄/Mnt and T is the corresponding calcination temperature.

The same steps were taken to prepare a pure g-C₃N₄ specimen at 550°C, referred to as CN-550°C.

For the comparative study, 0.1 g of M and 1.68 g of CN-550°C were mixed and ground to obtain a physically mixed composite CN+M.

Photocatalysis Experiment

A 300 W xenon lamp, purchased from Beijing China Education Au-Light Co., Ltd, was used as the visible light source within a wavelength range of 400–800 nm. The photocatalyst (50 mg) was dispersed in RhB solution (20 mg/L, 100 mL) and then stirred for 1 h in the dark on a magnetic stirrer to reach the adsorption-desorption equilibrium. After that, the light source was turned on. Aliquots (6 mL) of the RhB solution were taken every 10 min. An Evolution 300 UV-Visible spectrophotometer (Shimadzu, Kiya-cho, Nijiyanan, Japan) was used to test the absorbance of the supernatant at the maximum absorption wavelength of RhB (554 nm), and the degradation rate of RhB was calculated according to the Beer-Lambert law (Swinehart, 1962) as follows:

(1) $A=log\left(1/T\right)=\mathrm{Kbc}$

(2) $D=\left(A_0-A_t\right)/A_0=\left(C_0-C_t\right)/C_0$

where A is the absorbance; T is the transmittance; K is the molar absorption coefficient; b is the thickness of the absorbing layer; c is the concentration of the light-absorbing substance (mol/L); A ₀ is the absorbance of the solution in the dark at adsorption equilibrium; A _t is the absorbance of the solution at time t of light radiation; C ₀ is the solution concentration in the dark at adsorption equilibrium (mg/L); and C _t is the solution concentration at time t of the photocatalytic reaction (mg/L).

Characterization

Thermal performance analysis of samples was conducted in ambient atmospheric air using a STA449F5 synchronous thermal analyzer (TG-DSC) (NETZSCH, Berlin, Germany). Specimens were heated within a temperature range of 35–800°C at a heating rate of 20°C/min, and α-Al₂O₃ powder was taken as the reference material.

X-ray diffraction (XRD) patterns of samples were collected within a scanning angular range of 3 to 80°2θ by an Ultima IV X-ray diffractometer (PANalytical B.V., Eindhoven, The Netherlands) with a CuKα radiation source operating at a voltage of 40 kV and a tube current of 40 mA. The XRD patterns were analyzed using Jade 6.0 software (Huang & Li, 2012).

The micromorphology of specimens was probed by means of an Ultra 55 field emission scanning electron microscope (Carl ZeissNTS GmbH, Oberkochen, Germany) with a MnKα radiation source, operating at a voltage of 15 kV and an energy resolution >127 eV. The optical images were recorded at a count rate of 20,000 cps.

The specific surface areas of samples were determined using an Autosorb-1MP aperture analyzer (American Kangta Instruments Co., Ltd, Florida, USA). The experiments were made at a degassing temperature of 200°C for 6 h.

The optical properties were analyzed using a Shimadzu UV-3700 solid ultraviolet-visible-near-infrared spectrophotometer (Shimadzu, Kiya-cho, Nijiyanan, Japan) within the scanning range 200–800 nm at a resolution of 0.1 nm. BaSO₄ was employed as a reference sample.

The spectroscopic analysis was performed on powder specimens using a Frontier Fourier-transform Mid/Far Infrared spectrometer (PerkinElmer, Waltham, Massachusetts, USA). The data were collected over the range 4000–400 cm^–1.

The absorbance of RhB solution was analyzed by an Evolution 300 UV spectrophotometer (Shimadzu, Kiya-cho, Nijiyanan, Japan) using a CEL-HXF300F3 xenon lamp as the visible light source (300 W, 400–800 nm), purchased from Beijing China Education Au-Light Co., Ltd.

Results and Discussion

Thermal Performance

The differential thermal analysis curves of the M, CN-550°C, and CNM-550°C samples (Fig. 1) revealed an exothermic peak at 118°C associated with the evaporation exotherm of adsorbed water and exchangeable cation coordination water molecules on the surface and between the layers of Mnt (Ralph & Richards, 1942) The mass loss rate of M from 25 to 180°C was 13.6%. A similar result was obtained from the exothermic peak at 112°C in the DTA curve, caused by the removal of physisorbed water of Mnt in Ref. (Shmuel et al., 1989).

Fig. 1 Thermal curves of M, g-C₃N₄, and CNM-550°C

The mass loss of CN-550°C from 25 to 600°C was 5.3%, which was due to the evaporation of water adsorbed on the surface and the release of volatile substances (Chang et al., 2013). This could indicate that the structure of g-C₃N₄ underwent no noticeable changes up to 600°C. However, the emergence of the exothermic peak at 735°C in the DTA curve was a testimony to the decomposition of g-C₃N₄ into NH₃ and CO₂ gases (Yan et al., 2009), thus meaning the deterioration of the g-C₃N₄ quality at temperatures above 600°C. Moreover, almost no loss occurred between 750 and 800°C, which could indicate the complete decomposition of g-C₃N₄ into gases had already occurred.

The mass loss of CNM-550°C from 25 to 600°C was 7%, greater than that of pure g-C₃N₄. The appropriate peak was associated with the evaporation of adsorbed water and the dehydroxylation reaction of Mnt in the composite. Meanwhile, unlike the CN-550°C sample, a further increase in the temperature to 750°C did not impact the quality of CNM-550°C because of the significantly improved CN thermal stability after compounding with M.

Phase Analysis

The XRD patterns of a series of samples at various calcination temperatures are shown in Fig. 2. Almost all the specimens, except that labeled as M, revealed the presence of (100) and (002) reflections at 12.97 and 27.53°2θ, respectively, which were associated with a g-C₃N₄ phase (Shi et al. 2013). No diffraction peaks of melamine were observed, meaning that the latter was decomposed and converted into g-C₃N₄. At 490°C, the full-width at half maximum (FWHM) of the (002) peak was larger and the peak intensity was weaker, indicating a decrease in the crystallinity of g-C₃N₄. With a further increase in the calcination temperature, the FWHM of (100) and (002) XRD peaks of g-C₃N₄ gradually decreased at rising peak intensities, suggesting the improvement in crystallinity of g-C₃N₄ with increasing calcination temperature. On the other hand, a lack of peaks from Mnt indicated that Mnt was stripped and dispersed in g-C₃N₄ without forming a periodic structure (Zhao & Tan, 2018). This meant that the g-C₃N₄/Mnt nanolayer composite had been prepared successfully.

Fig. 2 XRD patterns of g-C₃N₄/Mnt composites at various calcination temperatures

Infrared Spectroscopy Data

The FTIR spectroscopic data in Fig. 3 reflected the chemical structure of the CNM-T samples. The peaks at 1640, 1569, 1326, 1412, and 1240 cm^–1 corresponded to the tensile vibration modes of C=N and C-N heterocycles (Li, Y., et al., 2014). Their intensities increased gradually with rising temperature, indicating the increase in the degree of crystallinity of g-C₃N₄. The broad band in the range 3000–3400 cm^–1 was due to the N–H stretching vibration and the adsorption of H₂O molecules (Li, C. et al., 2016). A strong absorption peak at 805 cm^–1 represented the out-of-plane bending vibration of the triazine ring (Liu et al., 2015).

Fig. 3 FTIR spectra of g-C₃N₄/Mnt composites at various calcination temperatures

Surface Topography

Scanning electron microscopy images of Mnt nanolayers, g-C₃N₄, and g-C₃N₄/Mnt composites at different calcination temperatures (Fig. 4) revealed that the Mnt was completely exfoliated into flakes with a thickness of 5–10 nm and their diameters were 2–5 um. (Fig. 4a). The thinner the flakes, the larger the specific surface area and the surface energy, causing the flakes to wrinkle and bulge. In turn, the adsorption performance of pollutants was also enhanced. However, a block structure of the pure g-C₃N₄, caused by the large surface energy, led to the formation of pure g-C₃N₄ agglomerates. Moreover, this was not conducive to the absorption of light, resulting in the poor photocatalytic performance of g-C₃N₄.

Fig. 4 SEM images of Mnt nano-particles, g-C₃N₄, and g-C₃N₄/Mnt composites at various calcination temperatures: a M, b g-C₃N₄, c CNM-490°C, d CNM-520°C, e CNM-550°C, f CNM-580°C, g CNM-610°C, and h CNM-650°C

The Mnt nanolayers were covered evenly and densely by g-C₃N₄. (Fig. 4c-h). The CN-490 sample exhibited a multi-layered stacked structure with relatively loose interlayers. According to the XRD data, this was due to the low crystallinity of g-C₃N₄ at 490°C. Furthermore, the sample surface possessed a fine granular structure owing to the impurities that remained unconverted into g-C₃N₄.

As the calcination temperature increased from 520 to 550°C, the layered structure of the composite became clear, dense, and smooth, indicating that the crystallinity of g-C₃N₄ increased. However, the compact stacking of the lamellae led to a block structure of g-C₃N₄, thereby reducing the specific surface area and hindering the light absorption and the photocatalytic effect. At 580°C, the laminate structure of g-C₃N₄ became loose due to the fact that the heat generated by the high-temperature calcination destroyed the interlayer van der Waals forces between the g-C₃N₄ particles. As a result, the layers were separated from each other and the spacing between them increased, thereby leading to the enlargement of the specific surface area. An increase in the calcination temperature to 610°C caused a further composite peel off and the emergence of large voids on the sample surface, further increasing the specific surface area. At 650°C, the composite material exhibited a fluffy layered structure, in which flakes were peeled off into the 1–5-nm-thick layers. Moreover, the excessively high temperature caused distortion of the triazine rings in g-C₃N₄, which would affect the photocatalytic performance of CNM-650.

Specific Surface Area

The specific surface areas and the pore volumes of CNM-T, M, and CN-550°C specimens (Table 1) showed that these values increased gradually with increasing temperature. For example, the specific surface area and pore volume of CN-550°C were 5.61 m²/g and 0.008 cm³/g and those of CNM-550°C were 13.24 m²/g and 0.113 cm³/g, respectively. The maximum specific surface area and pore volume of CNM-610°C were 29.10 m²/g and 0.251 cm³/g, respectively. Therefore, more adsorption sites could be provided for pollutants, thereby increasing the reaction rates (Wang et al., 2018).

Table 1 Specific surface areas of g-C₃N₄/Mnt composites at various calcination temperatures and of pure Mnt

Optical Absorption Performance

The UV-Vis diffuse reflectance spectra (DRS) and the Kubelka-Munk function curves of CNM-t composites at various temperatures (Fig. 5) exhibited pronounced absorption within the visible range of 460–530 nm, which was interpreted as the electronic transition in the g-C₃N₄ semiconductor (Shi et al., 2016). Furthermore, the light absorption edge of samples gradually redshifted (to higher wavelength) with rising calcination temperature from 484 nm for CNM-490°C to 523 nm for CNM-610°C, which was due to increased pore size of specimens. This could be explained in terms of multiple reflections induced by the coarse-pore structure under light irradiation, thus allowing the catalyst to absorb more light energy and, respectively, to generate more electron-hole pairs (Liang, 2015).

Fig. 5 a UV-Vis absorption spectra and b Kubelka-Munk function curves of CNM composites at various calcination temperatures

In order to study further the optical properties of the composite materials, their band gap values (E _g ) were calculated according to the Tauc approach (Tauc, 1974) and the Davis-Mott model (Davis & Mott, 1970) as follows:

(3) ${\rm A}hv=k{\left(hv–E_g\right)}^{n/2}$

Here, A is the adsorption coefficient; h is Planck's constant; ν is the photon frequency; and E _g represents the band gap energy; k is a constant; and n is related to the type of semiconductor.

In this respect, the band gaps of CNM-490°C, CNM-550°C, CNM-610°C, and CNM-650°C were as large as 2.56, 2.5, 2.37, and 2.4 eV, respectively, proving that the E _g of CNM-t could be reduced effectively by increasing the calcination temperature.

Photocatalytic Performance

The photocatalytic degradation rates evaluated for CN-550°C, CN+MC, and CNM-550°C (Fig. 6) showed that the self-degradation rate of RhB solution, assessed through a blank experiment under visible light irradiation, was only 6.7%, indicating the light irradiation stability of RhB. After the dark reaction for 1 h, the absorption rate of the pure g-C₃N₄ was only 1%, but increased as soon as g-C₃N₄ was compounded with Mnt nanolayers. This was because of the electrostatic attraction that made the negatively charged Mnt adsorb a large number of positively charged RhB molecules (Al-Ghouti et al., 2009). In turn, the degradation rate of CNM-550°C (97.8%) was higher than that of CN-550°C (81.5%) and CN+M (76.2%). This meant that the photocatalytic performance of composites was improved due to the loading of Mnt, causing the interaction between the g-C₃N₄ interface and the Mnt (Li, J. et al., 2016).

Fig. 6 Photocatalytic degradation rate of CN-550°C, CN+M, and CNM-550°C samples as a function of irradiation time

The photocatalytic degradation rate of CNM-490°C after a 2-h exposure to visible light irradiation was only 65% (Fig. 7). Such a low value was due to the poor crystallinity of g-C₃N₄. The best photocatalytic activity was found for CNM-610°C, the degradation rate of which also reached >95% after 60 min of visible light irradiation. The improvement of photocatalytic efficiency within the temperature range of 490 to 610°C seemed to be on account of a gradual increase in the crystallinity of the composite along with a small band gap of g-C₃N₄. The changes in both the crystallinity degree and the band gap width arose from a porous sheet structure of g-C₃N₄ with a loose surface after thermal peeling. The degradation rate of the composites under higher calcination temperatures, however, underwent a sharp drop, achieving a value of 68.2% only for the CNM-650°C sample. This could be explained by the onset of decomposition of g-C₃N₄ above a threshold temperature, where a partial collapse of g-C₃N₄ nanolayers resulted in a rapid decrease in the specific surface area and pore volume, thus reducing the degradation rate and photocatalytic performance. The excellent photocatalytic effect of the composite could be attributed to two phenomena. First, the introduction of Mnt layers improved the adsorption capacity of the composite material in relation to target pollutants, promoting the interplay of g-C₃N₄ with pollutants and thereby improving the degradation efficiency of the whole composite (Li et al., 2017). Second, close contact between g-C₃N₄ and Mnt could be achieved, inducing a positive synergistic effect. As a result, the electrons generated by light excitation of g-C₃N₄ formed an electrostatic attraction between the negatively charged Mnt and the positively charged RhB molecules, which further promoted the electron exchange reaction.

Fig. 7 Photocatalytic degradation of RhB by g-C₃N₄/Mnt composites at various calcination temperatures

Summary and Conclusions

A series of g-C₃N₄/Mnt nanolayers was prepared at various calcination temperatures, using melamine as the precursor and Mnt nanolayers as the matrix. The Mnt nanolayers in the composites were as thin as 5–10 nm and their diameters were 2–5 um. Their surface was uniformly and densely coated by g-C₃N₄ in a block-based manner, forming a multi-layered stacked structure.

An increase in temperature led to a gentle increase of the crystallinity of g-C₃N₄ along with a decrease in the band gap of the composite. Besides this, the light absorption edge gradually redshifted, and the multi-layered stack structure became dense. At 580°C, the interlayer van der Waals forces of g-C₃N₄ were destroyed by the thermal energy and the structure of the composite material began to loosen until it was peeled off.

The conclusion was, therefore, that the combination of Mnt nanolayers and g-C₃N₄ was beneficial to the transfer of photogenerated electrons, increasing the photocatalytic effect and the degradation rate of RhB of the composite material at temperatures up to 610°C. A further increase in temperature led to a structural collapse and decomposition of g-C₃N₄; as a result, the specific surface area and the pore volume of CNM-650°C decreased rapidly along with a reduction in the photocatalytic performance.