Synthesis and Characterization of a Photocatalytic Material Based on Raspberry-like SiO₂@TiO₂ Nanoparticles Supported on Graphene Oxide

Abstract

A raspberry-like SiO₂@TiO₂ new material supported on functionalized graphene oxide was prepared to reduce titania’s band gap value. The material was characterized through different analytical methods such as Fourier transform infrared spectroscopy (FTIR), X-ray diffraction (XRD), scanning electron microscopy (SEM), transmission electron microscopy (TEM), and high-resolution transmission electron microscopy (HR-TEM). The band gap value was studied via UV-Vis absorption spectra and determined through the Kubelka–Munk equation. A theoretical study was also carried out to analyze the interaction between the species.

1. Introduction

TiO₂ is a widely studied semiconductor material that exhibits outstanding properties such as high activity, high thermal and chemical stability, nontoxicity, low cost, and high reactivity [1,2,3]. Due to its properties, TiO₂ has been investigated in several interesting works, one of which is the work of Honda and Fujishima [4], who studied the photocatalytic decomposition of water in 1972. After this, significant research in this field has focused on TiO₂ nanomaterials because outstanding properties are found at this scale, specifically those related to optical and electronic behavior. Titanium is an early transition metal discovered by Gregor and Klaproth in the XVIII century [5]. This metal can be easily oxidized to yield TiO₂. This compound is widely used due to its multiple applications. For instance, it is well-known that white paint has its pristine color from TiO₂ dyes. Furthermore, derivatives of this substance are applied in rechargeable batteries and other electronic components.

Due to its characteristics, TiO₂ nanomaterials have been successfully used in multiple environmental applications [6,7,8]; for example, hydrogen production through the decomposition of water, CO₂ reduction, nitrogen fixation, and one of the most popular is photocatalysis as an advanced water treatment [9,10,11]. In this sense, TiO₂ photocatalysis allows the elimination of substances that are hard to remove through conventional treatments [12,13], such as organic pollutants. Among those compounds are found, for instance, pharmaceuticals, personal care products (PPCPs), endocrine disruptors, etc [14,15]. The above belong to the so-called emerging pollutants [14,15,16,17], which exhibit the potential to cause damage in aqueous environments and even to human health [13,14,18].

TiO₂ photocatalysis starts when this semiconductor is subjected to energy with a wavelength proportional to its band gap value [9,10,11]. The electrons of the valence band are excited and migrate to the conduction band $\left({\overline{e}}_{CB}\right),$ resulting in the reduction of dissolved oxygen to generate peroxide radicals. At the same time, the movement of electrons induces the formation of positive holes in the valence band $\left({h^{+}}_{VB}\right)$ where they oxidize adsorbed water to yield OH● radicals. The produced radicals participate in reactions that oxidize the organic structure of pollutants [2,3,9].

Crystalline TiO₂ can be found in three different phases, anatase, rutile, and brookite, whose band gap energy values are 3.2 eV, 3.0 eV, and 3.4 eV, respectively [19,20,21]. Among them, anatase is the most active phase, as it exhibits higher surface adsorption to hydroxyl radicals and a longer lifetime of the electron–hole pair species (${\overline{e}}_{CB}$ and ${h^{+}}_{VB}$) [21].

Despite its advantageous properties, TiO₂ shows some drawbacks when it comes to environmental applications. For example, it exhibits low adsorption capacity of organic pollutants, specifically those with hydrophobic characteristics [22]. Another limitation is the rapid recombination of the electron–hole pair ${(\overline{e}}_{CB}/{h^{+}}_{VB})$ that hinders the process [22,23,24] as the excited electrons return to the valence band. Additionally, the band gap energy value of TiO₂ restricts the semiconductor to be activated only by light in the UV range of the electromagnetic spectra, corresponding to 400 nm < λ. This makes the use of specific and expensive lamps necessary to excite electrons [22,23,24]. To overcome TiO₂ limitations, some strategies have been proposed [24,25,26]. In this sense, it has been reported that doping TiO₂ with non-metal elements such as B, N, C, and F [25,27,28] can improve TiO₂ performance. For instance, some studies [29,30] have reported that nitrogen-doped TiO₂ exhibits a decrease in band gap value due to the substitution of O atoms by N atoms; this interaction modifies the electronic structure of TiO₂ and, in addition to lowering the band gap value, it also allows TiO₂ to absorb light at larger wavelength values [25,29,30].

Carbonaceous nanostructures have been proposed as good candidates to show exceptional features that can be used to enhance TiO₂ properties [31]; it has been suggested that the formation of the Ti-O-C bonds modifies the band structure of TiO₂ decreasing the band gap value and thus extending the light absorption of TiO₂ to longer wavelengths corresponding to the visible light region. Besides the abovementioned, carbonaceous materials have been reported to improve TiO₂ conductivity and surface area, resulting in better photocatalytic properties [25,27,28,31]. Recently, several carbon nanostructures have been used for improving TiO₂ properties [32], for instance, carbon nanotubes, carbon quantum dots, and fullerene. Graphene and graphene oxide [31,32] have been shown to be appropriate candidates for TiO₂ doping due to their properties, such as large surface area, high electrical and thermal conductivity, and flexible structure. Indeed, several works have stated that graphene can reduce the recombination of the photogenerated electron–hole pairs, an explanation for this is the capability of graphene to accept electrons and promote their mobility; thus, the excited electrons of TiO₂ can easily migrate to the surface of graphene that acts as an electron sink [33,34,35,36,37], and as a consequence, reduce the recombination rate of the electron–hole pair.

Another studied strategy for improving TiO₂ properties is the use of noble and non-noble metal ions, such as Ag, Au, Pd, Pt Cu, Fe, Co, and Ni [22,24,26]. It has been suggested that metallic species provide stability to TiO₂ and can also form heterojunctions that modify its band structure, resulting in a lower band gap [26]. Metal oxides such as Al₂O₃, ZnO, ZrO₂, MoS₂, Fe₂O₃ [38], and SiO₂ [39,40,41,42,43,44] have also been used to improve the properties of TiO₂. In this sense, SiO₂ is one of the most used materials because it exhibits a large surface area, low toxicity, high thermal stability, and good mechanical strength. It has been established that when joined to SiO₂, TiO₂ [45,46,47] exhibits enhanced thermal, chemical, and mechanical stability due to the formation of Si-O-Ti bonds [40]. Additionally, it has been demonstrated that agglomeration is avoided when using support for TiO₂ nanoparticles such as SiO₂ [22]. Additionally, a higher surface area is achieved, leading to better adsorption of organic molecules [41,42,43,44,48], resulting in improved photocatalytic activity.

In the present work, it is reported that the synthesis of TiO₂ anatase nanoparticles deposited on SiO₂ in a raspberry-like nanostructure; then, the SiO₂@TiO₂ system was supported on a functionalized graphene oxide matrix via covalent bonding. Therefore, the system SiO₂@TiO₂ is expected to be able to receive the inductive effect from the graphene oxide surface. Consequently, the final material SiO₂@TiO₂/GO will exhibit a significant reduction in the band gap value. The properties of the new system were studied through different techniques. A theoretical study was also carried out to analyze the interaction between SiO₂@TiO₂ and graphene oxide.

2. Results and Discussion

The main purpose of this section is to discuss the properties of the synthesized SiO₂@TiO₂ and SiO₂@TiO₂/GO materials based on different characterization techniques. The experimental results related to the modification of the band gap value of titania will be deeply discussed and compared to those theoretically obtained. This approach to studying the properties of catalysts and other battery materials has been widely used in other works [49,50,51,52].

2.1. XRD Characterization

The powder patterns of XRD were obtained and analyzed at 2θ, first for SiO₂@TiO₂ and then for SiO₂@TiO₂/GO. Figure 1a shows the presence of amorphous silica at 2θ = 23, whereas the presence of anatase was confirmed at 25.3 [1 0 1], 37.9 [0 0 4], 47.9 [2 0 0], 54 [1 0 5], 55.5 [2 1 1], 63 [2 0 4], and 75 [2 1 5] [53]. Figure 1b shows the XRD pattern of SiO₂@TiO₂/GO. The peak at 2θ = 11.97 [0 0 1] corresponds to graphene oxide, and the characteristic peaks of anatase (in blue) are also evident [54]. The average grain size of TiO₂ nanoparticles was calculated using the Scherrer equation, resulting in 8 nm.

2.2. FTIR Characterization

Figure 2 shows the pattern for SiO₂@TiO₂ and SiO₂@TiO₂/GO. For SiO₂@TiO_2, the bands 3397 and 1631 correspond to the stretching vibration of -OH, whereas 1090, 957, and 436 can be associated with Si-O-Si, Ti-O-Si, and Ti-O-Ti, respectively. For the SiO₂@TiO₂ nanoparticle supported on graphene oxide, the bands 3390, 1571, and 1369 correspond to -NH, -CO-NH, and C-N. These results indicate that the SiO₂@TiO₂ NPs joined to graphene oxide via covalent bonds.

2.3. Thermogravimetric Analysis

Figure 3 shows the thermogravimetric analysis for the system SiO₂@TiO₂/GO. The first loss of weight at 120 °C can be attributed to the water that was either chemically or physically adsorbed during the synthesis process and corresponds to approximately 5% of the material. The second loss at 200 °C can be related to the decomposition of the functional groups epoxy, hydroxy, and carboxy, which conform to the surface of graphene oxide and contribute to the bond between the surface and SiO₂@TiO₂. Beyond 200 °C, the weight loss is attributed only to the decomposition via pyrolysis of the carbon surface [55] to the remotion of stable oxygen groups such as phenol, carbonyl, and quinine [56].

2.4. SEM and HR-TEM Characterization

Figure 4 corresponds to SEM images of the synthesized samples. Figure 4a shows that SiO₂ nanoparticles are homogeneous spherical particles with an average size of 250 nm. These characteristics indicate that SiO₂ nanoparticles possess good nucleation centers for TiO₂ nanoparticles. Figure 4b shows the SiO₂@TiO₂ system; it is observed that TiO₂ nanoparticles are homogeneously dispersed on SiO₂, yielding a raspberry-like nanoparticle. The final SiO₂@TiO₂/GO material is shown in Figure 4c. The joint of the nanoparticles to graphene oxide sheets is evident. This is an important result because the direct interaction between graphene oxide and TiO₂ plays an important role in decreasing the band gap value of TiO₂.

2.5. TEM and HR-TEM Analysis

A more detailed visualization of the SiO₂@TiO₂ and SiO₂@TiO₂/GO heterostructures was obtained through TEM imaging. Figure 5a corresponds to a low-magnification TEM micrograph of isolated SiO₂@TiO₂ nanostructures, where the TiO₂ nanoparticles joined to the SiO₂ particles in a raspberry-like morphology are clearly observed (inset). The average size of the particles was 7.7 ± 1.3 nm, in accordance with the calculated crystallite size through XRD. A selected area electron diffraction (Figure 5b) performed on these particles showed a ring pattern consistent with polycrystalline TiO₂. The measured ring diameters of 0.350, 0.237, 0.188, 0.168, and 0.148 nm can be related to the crystal planes (101), (004), (200), (211), and (204) of the anatase phase, in agreement with the XRD results.

HR-TEM imaging showed that TiO₂ in the TiO₂@SiO₂ structures consisted of small monocrystalline particles, see Figure 5c. The measured lattice spacing of 0.350 nm corresponds to the (101) lattice plane of the anatase TiO₂. The morphological features of the TiO₂@SiO₂/GO heterostructures were also imaged in Figure 5d. The TiO₂@SiO₂ particles are observed, with large GO sheets involving them to form the heterostructure.

2.6. UV-Vis Absorption Spectra

In a semiconductor material, the band gap value indicates the energy needed to excite an electron from the valence band to the conduction band. Determining this parameter provides information on the electronic behavior of the material and its applications. The Tauc method relates the absorption coefficient $\alpha$ and the band gap energy ($Eg$) as follows [57]:

$$\alpha hv\propto {(hv-Eg)}^n$$

(1)

where $hv$ is the energy of the incident photon, $n$ can take the values ½, 3/2, 2, and 3, depending on the kind of electronic transition, $Eg$ is the band gap, and its value is directly obtained from the plot, as the linear part of the plot is extrapolated to the x-axis. The Tauc method works well with samples with no light scattering; however, when working with powder samples, light scattering should be considered, and the Tauc plot cannot be applied directly [58]. In this case, diffuse reflectance spectroscopy (DRS) is a better option. In this method, the experimentally obtained reflectance is transformed through the Kubelka–Munk function into values of absorption coefficient α. As a first step, the Kubelka–Munk or reemission function $f\left(R_{\mathrm{\infty }}\right)$ is calculated, Equation (2) [59].

$$f\left(R_{\infty }\right)=\frac{K}{S}=\frac{{(1-R_{\infty })}^2}{2R_{\infty }}$$

(2)

where $\left(R_{\mathrm{\infty }}\right)=\frac{R_{Sample}}{R_{Reference}}$ is the reflectance from an infinitely thick specimen, $K$ represents the absorption coefficient, and $S$ is the scattering coefficient.

Finally, the calculated $f\left(R_{\mathrm{\infty }}\right)$ can take the place of α in the Tauc equation as follows:

$$f\left(R_{\mathrm{\infty }}\right)hv\propto {(hv-Eg)}^n$$

(3)

This work measured diffusion reflectance for the samples in a UV-Vis Cary 5000 spectrophotometer at a wavelength ranging from 700 to 200 nm. Results were analyzed via the Kubelka–Munk method. The band gap was then obtained through the Tauc plot.

Figure 6a shows the Tauc plot for the SiO₂@TiO₂ system; the band gap value 3.2 eV corresponds to the TiO₂ anatase phase, while Figure 6b shows the Tauc plot for SiO₂@TiO₂/GO. In this case, the band gap value is 2.7 eV. This is a remarkable result; as previously mentioned, TiO₂ exhibits some drawbacks related to the position of its conduction and valence bands that define the band gap value and, at the same time, restrict TiO₂ from being activated with the energy of only a portion of the electromagnetic spectra, the UV region and, as a consequence, limits its application. The reported decrease in the band gap value indicates that the presence of graphene oxide modifies the band structure of TiO₂ since graphene oxide exhibits high conductivity, and its band gap is narrower than that of anatase. Thus, when associating graphene oxide and the raspberry-like nanoparticles, the highly populated HOMO from graphene oxide favors the attraction of virtual molecular orbitals of it and those coming from TiO₂. It is expected that the decrease of the band gap provokes a value that shifts the adsorption range of TiO₂ to a higher wavelength corresponding to the visible portion of the spectrum, which also enlarges the possible applications of the material.

Another expected improvement is related to the high electron mobility exhibited by graphene oxide, as this species is well-known due to the π–π interactions taking place in its structure. The functional groups such as amine, epoxy, and hydroxyl provoke an inductive effect that improves the electron mobility of graphene oxide [60]. Thus, the excited electrons of TiO₂ are easily transferred to graphene oxide, avoiding the fast recombination of those species and conferring titania a better photocatalytic activity [61]. The nature of those interactions will be discussed in the theoretical section.

2.7. Theoretical Studies

The band gap behavior was also studied through a theoretical DFT method. The molecular structure of all the involved species was optimized using the CAM-B3LYP method [62] included in the Gaussian 16 pack [63], and the calculations were carried out using the basis set 6-31g**. The stable minimum energy configuration of each molecule was approached by calculating their frequencies. The design of the fragment of modified GO is shown in Figure 7a.

There are two optimized structures to consider. The first one is shown in Figure 7a, where all the peripheral substituents are carboxylic groups and a species where some amine -NH₂ group replaces the -OH of some carboxylic groups. This last possibility was studied because there are reports using this functionalization to link this kind of nanoparticles to graphene derivatives [64], and even more importantly, this is the synthesis followed in this work. Furthermore, a variation of the graphene itself consists of placing epoxy groups on the inner surface, searching for interaction with terminal titanium atoms. The designed structure shown in Figure 7b was conceived following other interesting propositions [65] where the stability and chemical properties of mixed Ti and Si oxides are tested. Some Si atoms of a SiO₂ crystal cell were substituted by Ti atoms. Figure 7c shows the first simulation, which involves the SiO₂@TiO₂ raspberry-like nanoparticle interacting on the periphery of the functionalized graphene oxide derivative.

A fact to highlight is that the interaction among titanium atoms from the nanoparticle and the carboxylated graphene oxide peripheral groups can occur either on oxygen or nitrogen terminals. They are real coordinated covalent bonds with the lengths 2.04 Å and 2.08 Å, respectively. The found values match well with the reported lengths for this kind of bond [66,67]; additionally, they show Wiberg indexes of 0.552 and 0.511, respectively, which suggests a real bond [68].

The purpose of studying the interaction between TiO₂ and other compounds is to significantly reduce titania´s band gap value. With that in mind, the calculated band gap of the alone mixed-oxide species and the same for the GO-TiSi complex were calculated in the same conditions. The first one yields a value of 3.9 eV, but the same value for the complex is 2.3 eV. This is a demonstration of the validity of the proposition; the last value experiment decreases nearly half of the original one.

The observed decrease of the band gap value of TiO₂ is the result of the interaction of the molecular orbitals of both joined structures, graphene oxide and the SiO₂@TiO₂ raspberry-like nanoparticles. One feature that has been signaled about the nature of the association between functionalized graphene and nanoparticles of a catalyst is the new behavior of the species as a semiconductor; in fact, the new complex combines the characteristics of both parent molecules. Regarding the energy gap, a peculiar phenomenon is observed; the SiO₂@TiO₂ raspberry-like nanoparticle is expected to be a semiconductor itself with a discrete HOMO-LUMO set; however, these frontier molecular orbitals experiment a notorious change when the large reservoir of orbitals coming from the graphene oxide influence on them. Figure 8 shows the frontier molecular orbitals of the studied species, the HOMO in Figure 8a shows a strong influence of graphene oxide, whereas the LUMO in Figure 8b is focused on titanium atoms; this arrangement is one of the main reasons leading to the change in the energy gap value.

3. Materials and Methods

3.1. Chemicals

There are different precursors for tetraethyl orthosilicate TEOS, titanium isopropoxide TTIP, NH₄OH, HCl, EtOH, APTES, NaOH, and ClCH₂COONa (Sigma–Aldrich, St. Louis, MO, USA) were used as received. Graphene oxide (Sigma–Aldrich) was modified to create more active sites.

3.2. Synthesis

3.2.1. SiO₂ Nanoparticles

The sol–gel synthesis of SiO₂ nanoparticles starts with a hydrolysis reaction of alkoxy groups contained in the precursor. In this work, tetraethyl orthosilicate was chosen as a precursor because it contains enough alkoxy groups to hydrolysate [69]. According to the synthesis established by Stöber [70], the process started with the hydrolysis of TEOS in basic media and was followed by a series of condensation and hydrolysis reactions to finally obtain the desired SiO₂ nanoparticles. Briefly, 3 mL of TEOS was quickly added to a solution of NH₄, H₂O, and ethanol (4 mL:15 mL:100 mL) and left under magnetic stirring for 1 h. The mixture was neutralized with HCl. The result was centrifuged at 3500 rpm for 10 min. The precipitate was washed with water and dried at 70 °C for 20 h.

3.2.2. SiO₂@TiO₂ Raspberry-like Nanoparticles

TiO₂ nanoparticles were also obtained via a sol–gel synthesis; again, a precursor containing alkoxy groups was chosen. In this case, the process started with the hydrolyzation reaction of titanium isopropoxide (TTiP) and was followed by condensation and hydrolysis reactions. A total of 0.2 g of SiO₂ NPs was dispersed in 15 mL of isopropanol and left under magnetic stirring. On the other hand, TTiP was dissolved in EtOH and was slowly added (approximately 1 mL/min) to the dispersion. The mixture was left under magnetic stirring for 20 h. Then, a solution of water–isopropanol (3 mL:6 mL) was added to promote hydrolysis and condensation reactions. The result was left under magnetic stirring for 2 h. Then, it was transferred to a Teflon reactor for a thermic treatment 24 h, 10 °C. The product was recovered via centrifugation, washed with water, and dried for 24 h at 100 °C. The product was calcined at 500 °C for 3 h to obtain the anatase phase of TiO₂ [43,44].

3.2.3. SiO₂@TiO₂ Raspberry-like Nanoparticles Supported on Graphene Oxide

SiO₂@TiO₂ raspberry-like nanoparticles were supported on graphene oxide through chemical bonding. It is said that once obtained, the system SiO₂@TiO₂ and the graphene oxide layer were functionalized to create the appropriate sites for the bonding [71]. SiO₂@TiO₂ NPs were suspended in a mixture of 20 mL of EtOH and 4 mL of deionized water. Then, 0.5 mL of NH₄OH and 50 µL of APTES were added to create the amino (-NH₂) group over the SiO₂@TiO₂ surface. Furthermore, it was left under magnetic stirring for 24 h. The product SiO₂@TiO₂ -NH₂ was washed with EtOH and dried at 60 °C for 2 h. The graphene oxide surface was modified to obtain a more carboxylic acid group species GO-COOH. Graphene oxide was suspended in water using ultrasonication, NaOH and ClCH₂COONa were added, and the reaction was left under magnetic stirring for 3 h. The product was washed using a dialysis tube for 72 h. The precipitate was dried at 90 °C for 12 h. To obtain the species SiO₂@TiO₂/GO, a mixture of 20 mL of deionized water and 20 mg of EDC-HCl was prepared, and 10 mg of GO-COOH was dispersed using ultrasonication for 2 h. A total of 10 mg of SiO₂@ TiO₂-NH₂ was added, and the reaction was left under magnetic stirring for 24 h. The product was washed with deionized water and EtOH and dried at 60 °C for 12 h [48]. Figure 9 shows the steps followed to obtain the SiO₂@TiO₂ raspberry-like nanoparticles and the final species SiO₂@TiO₂/GO.

3.3. Characterization

The crystalline structure of the nanoparticles was studied using X-ray diffraction (XRD), and the spectra were obtained to analyze the crystalline phases of the samples using a Bruker D8 Advance diffractometer with Cu-Kα radiation (λ = 0.154 nm). The average size of the crystal lattices of TiO₂ was calculated using Scherrer’s formula [72]. Fourier transform infrared spectra were collected on a Nicolet 510P spectrometer.

To characterize the crystalline structure, morphology, and average diameter of the obtained compounds, a scanning electron microscope (SEM) and a high-resolution transmission electron microscope (HR-TEM) analysis were performed using a JEOL 7600F and a JEOL ARM200F operated at 20 keV, respectively. The electronic diffusion spectra were obtained using a UV-Vis Cary 5000 spectrophotometer, and the band gap was determined through Tauc’s plot using the Kubelka–Munk method. Thermogravimetric analysis, TGA, was performed using a TGA Q5000 V3.17 Build 265 equipment in a nitrogen atmosphere.

4. Conclusions

A new nanomaterial based on SiO₂@TiO₂ nanoparticles supported on graphene oxide was synthesized and characterized through XRD, FTIR, SEM, and HR-TEM. The band gap value was also determined using the Kubelka–Munk method.

The SiO₂@TiO₂ nanoparticle exhibits a raspberry-like morphology, suggesting that TiO₂ nanoparticles are exposed in such a way that they can be considered for catalytic processes. It was found that when joining SiO₂@TiO₂ to graphene oxide layers, the band gap of TiO₂ was lowered by almost 1 eV with respect to the band gap value of anatase. This result indicates that the conductor properties of graphene oxide endow TiO₂ with new properties.

Theoretical calculations of discrete molecules were carried out to evaluate the improvement of the electronic activity using graphene surfaces. The theoretical band gap value is in agreement with the experimental one, validating the research base.