Abstract

A novel micro-mesostructured SiO2/reduced graphene oxide (RGO) nanocomposite was successfully synthesized by means of simple one-step hydrothermal method under acidic conditions using tetraethoxysilane (TEOS) and graphene oxide (GO) as the raw material. The nanocomposites were characterized by TEM, XRD, FT-IR, TG-DSC, and N2 adsorption-desorption. The results showed that GO was partially reduced to RGO without adding any reducing agent and SiO2 nanoparticles (ca. 10 nm) were uniformly anchored on the surface of RGO. The optimized composite contained 75 wt.% SiO2 and possessed hierarchical micro-mesoporous structure with surface area of 676 m2/g. The adsorption performance of synthesized SiO2/RGO samples was investigated by removal efficiency of Cr(VI) ions in wastewater. The Cr(VI) adsorption reached equilibrium in 30 min and 98.8% Cr(VI) adsorption efficiency was achieved at pH = 2 at 35°C. Stability tests showed that SiO2 nanoparticles effectively prevented RGO from the restacking. The mechanisms of composite formation and for Cr(VI) adsorption were suggested.

1. Introduction

Graphene has attracted significant research attention since its first report in 2004 [1], owing to the unique two-dimensionally hexagonal structure [1] and extraordinary properties [24]. Graphene has been known as promising adsorbent with great potential in wastewater treatment [57] because of its large theoretical surface area (2630 m2/g) [8], remarkable electronic properties [9] and mechanical properties [10], and high ability of modification [11, 12]. However, it is difficult to disperse graphene in water due to its hydrophobicity, agglomeration, and restacking [1315], limiting its use in aqueous systems for decontamination.

Graphene oxide (GO), a highly oxidative form of graphene containing of a variety of oxygen functional groups, can solve the problem because it has similar properties to graphene but is more hydrophilic, which is a prerequisite as adsorbent used in aqueous system [1618]. However, GO has very weak binding affinity for anionic compounds as there is a strong electrostatic repulsion between them [7]. Furthermore, the potential separation problem still limits the application of graphene and GO. Therefore, one possible route to improve these properties for applications is to combine graphene materials with other particles to form nanocomposites, which has been proven to be an effective and feasible method to avoid the aggregation of individual graphene sheets and nanoparticles, thus overcoming the separation problem [1921].

Due to their advantages such as environmental friendliness, hydrophilicity, chemical resistance, high strength, and low cost [2224], SiO2 nanoparticles are undoubtedly a good candidate to decorate graphene materials to overcome the shortage of individual sheets. Thus, SiO2/RGO nanocomposites were selected as the promising adsorbents in this study.

In recent years, SiO2/RGO nanocomposites have been fabricated via different methods. For instance, Hao et al. [25] reported the preparation of SiO2/graphene composites through a two-step reaction, including the synthesis of SiO2/GO and the reduction of GO. Zhou and Shi [26] studied a more simplified hydrothermal synthesis of SiO2-graphene hybrid but without investigation of application. Liu et al. [27] developed a simple method to construct PEI@SiO2@rGO composites to remove Hg2+ and dyes from contaminated water. Through typical sol-gel technique, Zeng et al. [28] synthesized SiO2-coated GO sheets in a water-alcohol-ammonia mixture to prepare SiO2-GO-MIPs sensor. Li et al. [29] prepared SiO2/RGO nanosheet composites through a modified sol-gel process with NH2NH2·H2O as reductant. Overall, most synthesis processes used in the abovementioned previous studies are relatively complex, including at least two main steps, or reduced by adding reductants. And some synthesis methods with low adsorption efficiencies are rarely used in the actual treatment process.

Herein, we report an efficient one-step hydrothermal method to fabricate SiO2/RGO nanocomposites through the reaction of silica sol and GO in a sealed stainless steel reactor with PTFE lining. This facile fabrication procedure combines the reduction of GO and the preparation of SiO2/RGO nanocomposites in one step. The adsorption performance of SiO2/RGO nanocomposites for hexavalent chromium (Cr(VI)) was investigated in aqueous solution. It is well known that Cr(VI) is mainly caused by dragout, chromate passivation solutions, and rinsing of plated articles, which has high toxicity and high mobility in water and causes great harm to the environment and human health. Currently, adsorption method has been widely used for Cr(VI) removal [3032]. However, the low efficiency and complicated producing process of these adsorbents limits its application [33, 34]. Therefore, it is necessary to explore more suitable adsorbents with high efficiency, facile synthesis process, and low cost for Cr(VI) removal. In this study, a new method has been developed to synthesize SiO2/RGO nanocomposite by one-step process. The obtained nanocomposites with large specific surface area (676 m2/g) and unique hierarchical micro-mesoporous structure show superior performance in adsorption of Cr(VI). The hierarchical micro-mesoporous SiO2/RGO nanocomposites present great potential in the fields of environmental remediation.

2. Experimental

2.1. Materials

Tetraethyl orthosilicate (TEOS, 28%), nitric acid (65%), diphenylcarbohydrazide, hydrochloric acid (36~38%), potassium permanganate, hydrogen peroxide, and phosphoric acid (>85%) (Sinophare Chemical Reagent Co., Ltd.) were all of analytical grade and potassium dichromate (Shanghai Chemical Reagent Co., Ltd.) and natural flake graphite (Sinophare Chemical Reagent Co., Ltd.) with particle size below 30 μm were of chemical grade. All reagents were obtained from commercial sources and used without further purification.

2.2. Synthesis of SiO2/RGO Nanocomposites

Through a one-step method, SiO2/RGO nanocomposite samples were synthesized by hydrolyzing tetraethyl orthosilicate (TEOS) directly in a GO aqueous dispersion, which was prepared according to the Hummers method through the oxidation of natural flake graphite [35]. The synthesized GO suspension (9.0 g, 1 wt.%) was diluted with 100 mL deionized water and then was ultrasonicated for 15 min. The pH of the solution was adjusted to 1.0 with 0.1 M HNO3. TEOS solution with different volume (1 mL, 2 mL, and 3 mL) was then slowly added to the above suspension with magnetic stirring at 40°C. Such obtained mixture with different mass ratio of GO and SiO2 (1 : 3, 1 : 6, and 1 : 9) underwent hydrothermal reaction in a 200 mL cylindrical sealed stainless steel reactor with polytetrafluoroethylene (PTFE) lining at 120°C for 12 h. During the process of hydrothermal reaction, GO was reduced while a portion of functional oxygen groups was removed to recover the densely packed benzene-ring structure partially. Meanwhile, along with the process of hydrolysis of TEOS, different forms of silica were loaded on the surface of GO sheets through covalent bond linkage [36]. As a result, the amorphous SiO2 particles were incorporated onto the surface of the RGO films. Finally, the synthesized composite samples were collected through centrifuge for 5 min at 4000 rpm and washed with deionized water until neutral. The samples were then dried overnight at 60°C in air. The samples were as assigned as SiO2/RGO-1, SiO2/RGO-2, and SiO2/RGO-3 corresponding to GO and SiO2 mass ratio of 1 : 3, 1 : 6, and 1 : 9, respectively. For comparison, the pure SiO2 was obtained by the hydrolysis of TEOS under the same procedure of SiO2/RGO preparation but without adding GO.

2.3. Cr(VI) Adsorption and Method of Analysis

The adsorption experiments were carried out in a 200 mL glass jacketed batch reactor operated at 180 rpm rotating speed. A certain amount of SiO2/RGO nanocomposites were dispersed into the solution, and the aqueous solutions of Cr(VI) were adjusted to different concentrations using K2Cr2O7. The pH was controlled using 0.5 M HCl and 0.5 M NaOH and the temperature was regulated by super-thermostatic water bath. The SiO2/RGO nanocomposites were separated by centrifugation and the concentration of Cr(VI) at different time intervals in the batch reactor was determined using 1,5-diphenylcarbohydrazide (DPC) spectrophotometry method and the measuring wavelength is 540 nm. The adsorption efficiency (E, %) was calculated by the following equation:where and (mg/L) are the concentrations of Cr(VI) at initial time and the time of sampling, respectively. All experiments were repeated twice to ensure the reproducibility.

2.4. Characterization of the SiO2/RGO Nanocomposites

The morphology of GO and SiO2/RGO nanocomposite was investigated by a transmission electron microscopy (TEM, JEM-2100, JEOL). The functional groups of the samples were analyzed by Fourier transform infrared (FT-IR) spectroscopy using Nicolet Nexus 470 FT-IR Spectrometer in the range of 400~4000 cm−1. The X-ray diffraction (XRD) patterns of the synthesized samples were obtained using a D8 Advance diffractometer (Bruker, Germany) with Cu K radiation (0.15418 nm) and scanning rate 5°/min. The samples were also studied by a TG-DSC (Thermogravimetric and Differential Scanning Calorimetric Method) in oxygen at heating rate of 10°C/min. N2 adsorption-desorption isotherms were measured using a Quantachrome Autosorb-iQ-MP (USA) analyzer at −196°C with a relative pressure ranging from 0.0112 to 0.9948. The specific surface area was calculated by a Brunauer Emmett Teller (BET) model.

3. Results and Discussion

3.1. Material Characterization

The morphologies of GO nanosheets and SiO2/RGO-2 nanocomposites samples were fully analyzed by TEM in Figure 1. Figure 1(a) shows the TEM images of the GO membrane with a lateral size of several micrometers. The wrinkled morphology was compatible with the mechanisms of the stability of 2D membranes that had intrinsic ripples owing to thermal fluctuations [37]. As shown in Figure 1(b), SiO2 nanoparticles were uniformly distributed on the surface of the corrugated platelet of RGO, which was obtained from the reduction of GO through hydrothermal reaction, consistent with the FT-IR results. Compared with the image of GO in Figure 1(a), restacking was much less due to the loading of SiO2 nanoparticles. It is obvious that nanoparticles of SiO2 have been loaded uniformly onto the RGO membrane.

FT-IR spectra of GO, RGO, SiO2, and SiO2/RGO-2 are shown in Figure 2. The following functional groups are identified in the GO sample in Figure 2(a): a broad and strong peak appearing at 3386 cm−1 assigned to O-H stretching vibration; the C=O stretching vibration of -COOH groups at 1732 cm−1; the O-H bending vibration and sp2 C=C stretching of skeletal ring at around 1624 cm−1; the adsorption at 1215 cm−1 and 1172 cm−1 attributed to C-O vibration and C-O-C vibration, respectively. The peak at 864 cm−1 and 590 cm−1 is possibly attributed to C-H bonds out-plane bending vibration. It is obvious that the well-oxidized GO has been synthesized and -OH, -COOH, -CHO, and -C-O-C-functional groups existed in GO sample [38, 39]. After hydrothermal treatment, as shown in Figure 2(b), the peak at 1732 cm−1 (C=O stretching vibration of -COOH groups) disappears completely and other characteristic peaks at 1630 cm−1 (C=C in phenol ring) and 1150 cm−1 (C-O-C vibration) become weak, confirming the reduction of GO. Meanwhile, a new band displays at 1439 cm−1 which is contributed from the C-H bending vibrations of the reduced graphene lattice. In Figure 2(c), the characteristic peaks of SiO2 are as follows: the broad adsorption at 3454 cm−1 is attributed to the stretching vibration of -OH in structural water, and the peak at 1636 cm−1 is related to the H-O-H bending vibration of residual water molecules [40, 41]. Similar to the spectra of RGO, several peaks assigned to oxygen functional groups at 3386, 1732, 1630, and 1150 cm−1 become weak or disappear, which implies that GO was partially reduced. In Figure 2(d), the sharp adsorption at 1101 cm−1 is due to Si-O-C or Si-O-Si stretching vibration and Si-OH bending vibration lies at 968 cm−1 [42]. The peaks at 796 cm−1 and 471 cm−1 are assigned to the stretching and bending vibrations of Si-O bonds, respectively. It is therefore quite certain that SiO2 nanoparticles existed in the nanocomposites synthesized in this study. Furthermore, the fact that the spectra of RGO become weak or unclear in Figure 2(d) is not only due to the partial reduction of GO, but also due to the reaction between oxygen functional groups and Si-OH, which can be indirectly proved by the formation of a covalent bonding Si-O-C between RGO sheets and SiO2 nanoparticles [43]. Based on the above analysis, it can be concluded that SiO2 nanoparticles are well-dispersed on the surface of RGO sheets.

Figure 3 shows the XRD patterns of graphite, GO, RGO, SiO2, and SiO2/RGO-2 nanocomposites samples. It can be seen that graphite and GO samples exhibit two sharp and intense peaks centered at 26.5° (Figure 3(a)) and 9.4° (Figure 3(b)), respectively. Both of them are the typical (002) diffraction peaks consistent with previous studies [8, 22]. According to Bragg equation (), the corresponding lattice parameter of GO is determined to be 0.94 nm, which is larger than that of graphite (0.334 nm), due to the existence of oxygen functional groups. After hydrothermal reduction, GO nanosheets were partially reduced to RGO nanosheets. This is evidenced by the appearance of the (002) diffraction peak at 24.9° and the (100) diffraction peak at 43.0° (Figure 3(c)). The broad diffraction peak in the range of 15~35° and centered at 21.9° (Figure 3(d)) is assigned as amorphous SiO2 obtained by the hydrothermal method. In Figure 3(e), the diffraction pattern of SiO2/RGO-2 is similar to that of amorphous SiO2, but the center of the newly obtuse peak has shifted to 23.3° because of the influence of RGO with lattice parameter of 0.380 nm, which is slightly bigger than that of graphite, due to remaining oxygen functional groups. Moreover, the (100) diffraction peak at 43.0° disappears due to the coverage of SiO2 nanoparticles.

Figure 4 shows TG-DSC profiles of the as-prepared SiO2/RGO-2 nanocomposite sample. It can be seen that the weight loss of the SiO2/RGO-2 nanocomposite is divided into three stages and the curve of DSC exhibits two major distinct peaks. The first low-temperature weight loss under 100°C is primarily due to the evaporation of adsorbed free water and water held on the surface by hydrogen bonding, which is an endothermic process as illustrated by the endothermic peak of DSC curve. Weight loss in the range of 100–450°C is attributed to the decomposition and removal of oxygen functionalities including hydroxyl, carboxylic, and epoxide groups [40]. And the weight loss from 450 to 550°C is likely attributed to the decomposition of the carbon skeleton of RGO, corresponding to the obvious exothermic peak of DSC curve. Further weight loss above 550°C is minimal and the remaining mass is SiO2 particles [42, 44]. The content of SiO2 loaded on RGO in the nanocomposite is 74.68%.

The N2 adsorption-desorption isotherms and pore size distributions of SiO2/RGO nanocomposites are presented in Figure 5 and Table 1. For SiO2/RGO-1, it has hierarchical micro-mesoporous structure including 63% micropores and 37% mesopores as listed in Table 1, but a typical type II isotherm with a small hysteresis phenomenon corresponds to a relatively low porosity material in agreement with the lower specific surface area (514 m2/g). For SiO2/RGO-3, a typical type IV isotherm with an obvious hysteresis phenomenon indicates the existence of abundant mesopores (658 m2/g) and the ratio of mesopores () reaches up to 98.09% (Table 1). In comparison, the N2 adsorption-desorption isotherms of SiO2/RGO-2 exhibit combined characteristics of type II and type IV, which shows a noticeable nitrogen adsorption increment in the low pressure region from 0.0 to 0.5 (Figure 5(a)) followed by a gradual increase with increasing the relative pressure and displays a long hysteresis at almost the whole range of relative pressure [25]. The results suggest that SiO2/RGO-2 possesses rich microporous and mesoporous structure, which is also proved by the data of Table 1 (especially ) and shown by the pore distribution curve of Figure 5(b). Meanwhile, the long hysteresis loop belongs to typical H4 model, suggesting that the pores in the composites mainly consisted of slit apertures formed by the stack of layered materials, different from the only packing of particles. It can be concluded that the adsorption mainly takes place on the surface and the interlayer of plate-like nanocomposites [26, 45].

Furthermore, the increase of the proportion of SiO2 will reduce the cost of the production of SiO2/RGO nanocomposites, but the excessive SiO2 led to unreasonable pore structure, as illustrated in Table 1. Although the BET surface area of SiO2/RGO-3 increased to 691 m2/g, the surface area of micropores decreased dramatically to 33 m2/g, which undermines the adsorption of Cr(VI) ions. In comparison, the BET surface area and pore volume calculated by Barret-Joyner-Halenda (BJH) method of SiO2/RGO-2 were 676 m2/g and 0.4265 cm3/g, respectively. It has not only higher surface area than SiO2/RGO-1 but also more reasonable hierarchical micro-mesoporous structure than SiO2/RGO-3, including mesoporous surface area 371 m2/g and microporous surface area 305 m2/g. Mesopores not only provide active sites but also facilitate ions diffusion and enhance micropores supply higher adsorption sites. This hierarchical micro-mesoporous structure is beneficial to the adsorption of Cr(VI) ions in wastewater. From the above analysis, Due to the hierarchical micro-mesoporous structure and the high surface area, SiO2/RGO-2 nanocomposite has greater potential for Cr(VI) adsorption.

3.2. Cr(VI) Adsorption Performance

Using Cr(VI) ions solutions to simulate wastewater, the adsorption performance of the as-prepared SiO2/RGO samples was evaluated through the removal efficiency of Cr(VI) ions. The adsorption activity was examined under different conditions. 100 mL initial solution containing Cr(VI) with the concentration of 10 mg/L was charged into the jacketed batch reactor operated at 20°C. Different samples (100 mg) were dispersed in the solution by stirring at 200 rpm and the pH = 2. The adsorption curves were obtained through UV-vis spectrophotometer by measuring the absorbance of Cr(VI) in the samples gathered at different time intervals. The changes of adsorption efficiencies are given in Figure 6.

The adsorption performance of SiO2/RGO nanocomposites was assessed. As shown in Figure 6(a), the results suggest that the adsorption efficiency of all SiO2/RGO nanocomposites (>90%) was higher than that of the pure SiO2 nanoparticles (26.5%) and pure RGO (56.7%) with same dosage. Obviously, the composite was not just a physical mixture of two substances but a new nanocomposite material. The prevention of the restacking of sheets and the aggregation of SiO2 nanoparticles contributed to the improved Cr(VI) removal efficiency.

Meanwhile, the curves in Figure 6(a) showed that the adsorbent of SiO2/RGO-2 exhibited better adsorption efficiency than that of SiO2/RGO-1 and SiO2/RGO-3 samples. The uptake of Cr(VI) by SiO2/RGO-2 adsorbents increased dramatically in the first 10 min and reached the equilibrium in 60 min with a adsorption efficiency of 98.8%, which demonstrated superior adsorption rate and efficiency [46, 47]. In addition, the Cr(VI) concentration after treatment with SiO2/RGO-2 adsorbent drops below 12 μg/L, which completely meets the standards for drinking set by the World Health Organization [48]. This phenomenon was attributed to higher specific surface area than SiO2/RGO-1 and more reasonable hierarchical micro-mesoporous structure than SiO2/RGO-3, which was consistent with the analysis of N2 adsorption-desorption. Furthermore, as shown in Figure 6(b), the stability of SiO2/RGO-2 nanocomposites was studied by comparing the adsorption performance between the fresh sample (SiO2/RGO-2 (new)) and the aged sample that was kept for six months after preparation (SiO2/RGO-2 (aged)). The SiO2/RGO-2 (aged) adsorbent had the adsorption efficiency with 98.6%, nearly the same as that of the freshly prepared adsorbent samples. So it is believed that the SiO2/RGO nanocomposite has more superior stability than the pure SiO2 nanoparticles and pure RGO sheets.

3.3. Mechanism of SiO2/RGO Formation and Cr(VI) Adsorption

The composites of SiO2/RGO were successfully synthesized through one-step hydrothermal method. The recombined process was illustrated by the schematic of Figure 9 step A. At the beginning, using TEOS as silica source SiO2 nanoparticles formed through hydrolysis and dehydrate process [49]. Simultaneously, superheated H2O promotes acid-catalyzed reaction of organic compounds because of a sufficiently high H+ concentration compared to normal liquid phase [50, 51]. The H+-catalyzed dehydration could make GO be reduced partially [52, 53] and make part of oxygen functional groups break away from the skeleton [54]. At last, the SiO2 nanoparticles were anchored onto the surface of RGO owing to electrostatic interactions and covalent reaction. Similar to graphene sheets, the surface charge containing delocalized electrons was regarded as negative [55]. In contrast, the zeta potential of SiO2 nanoparticles was positive when pH value was less than 3 (Figure 7). Therefore, the RGO sheets possessed the ability to attach SiO2 nanoparticles through electrostatic interaction. Additionally, silicon hydroxyls easily interact with oxygen functional groups (e.g., -COOH, -OH) through covalent reaction (Figure 9 step A). During this process, RGO sheets were further unfolded and improved the stability because of the electrostatic repulsion among SiO2 nanoparticles.

The adsorption mechanism was schematically illustrated in step B and step C of Figure 9. The Cr(VI) adsorption capacity of SiO2/RGO was closely related to pH value due to its influence on the surface properties of the adsorbent and different ionic forms of the chromium solutions [56]. Chromium existed in different oxidation states and the stability of these forms is dependent upon the pH of the system [57]. The equilibrium between different ionic chromium species can be described as follows:

With increasing pH value, the chromium ions were converted from into and . The effect of pH on adsorption of Cr(VI) was illustrated in Figure 8. It is obvious that the adsorption efficiency of Cr(VI) decreased sharply as the pH increased from 1.0 to 8.0, showing the strong dependence of Cr(VI) removal performance on the system’s pH. The highest efficiency with ~92% Cr(VI) removal was achieved in the pH of 1~2, in which adsorption capacity increased to the maximum value without apparent change of chromium form, mainly . It is therefore obvious that the low pH leads to an increase in the concentration of H+ ions [58]. In highly acidic media, the adsorbent surfaces might be highly protonated to positive charges [59], as shown in Figure 9 step B. The positive charges represent the protonation of -OH [60]. Therefore, as illustrated in Figure 9 step C, strong electrostatic attraction exists between the positively charged nanocomposites surface and , which was the dominant form of Cr(VI) at pH less than 2.0. Unlike , only need one adsorption site to facilitate them, leading to high adsorption efficiency at pH less than 2.0.

4. Conclusions

A low cost SiO2/RGO nanocomposite adsorbent with hierarchical micro-mesoporous structure was successfully synthesized by means of a feasible one-step hydrothermal method. The adsorbent can remove Cr(VI) ions from aqueous solution effectively. SiO2 nanoparticles (ca.10 nm) were uniformly distributed on the membrane of RGO, which not only improved the stability of composites owing to preventing the restacking of RGO sheets, but also made the separation easier due to the relatively high intensity of SiO2. The adsorption efficiency of Cr(VI) ions reached 98.8%, and the adsorption reached equilibrium in 30 min. The results suggest that SiO2/RGO nanocomposites adsorbent is suitable for effective and rapid removal of Cr(VI) ions from wastewater.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.

Acknowledgments

This work was supported by the National Natural Science Foundation of China (51404097, 51404098, 51472075, and U1361119), International Science and Technology Cooperation Project of Henan province (152102410047), and National Science Research Program of Henan Provincial Education Department (no. 2011B440006).