Sintered TiO₂/recycled glass composites designed for the potential degradation of waterborne pollutants

1 Introduction

In the last half-century, agricultural soil and ground water have been contaminated by a range of pollutants affecting human health and beneficial soil organisms. In recent years, more research efforts have focused on expanding and improving advanced oxidation technologies (AOTs) [1], which are based on hydroxyl radical degradation of organic pollutants in waters and contaminated soil remediation [1, 2]. These processes have promising applications because they can oxidize all organic pollutants through high reaction rates without the need for other chemicals assisting the reaction [1]. In particular, TiO₂ is being extensively studied for its photocatalytic properties and its effectiveness to produce hydroxyl radicals when it is irradiated with ultraviolet (UV) light [3]. It is generally known that the anatase polymorph of TiO₂ is more efficient than rutile as a photocatalyst [4]. TiO₂ band gap is within the 3.0 to 3.2 eV range and its wavelength is about 400 nm [5]. Therefore, when TiO₂ is irradiated with UV light with a wavelength shorter than 400 nm, its surface can achieve localized high temperatures [5]. Such temperature oxidizes all materials, while adjacent organic pollutants are decomposed completely into water and carbon dioxide [6]. In addition, TiO₂ has many advantages due to its low cost, low toxicity, and good chemical and thermal stability [7].

Based on these studies, the present work proposes a remediation design for contaminated water using TiO₂ composite technology. The method consists of TiO₂ nanoparticles sintered within a porous glass bed. To find alternative applications to the large amounts of waste glass, which takes millions of years to degrade, powdered recycled glass is selected as the structural grid to formulate the composite filters. In recent years, recycled glass has been assessed as a lightweight fill material for pavements and other civil engineering applications. According to these studies, recycled glass can be classified as a nonhazardous material for the environment [8, 9].

The production of TiO₂/recycled glass composites can be used as part of irrigation-filtration systems but need to be produced at low cost to handle large-scale agricultural uses. The porous media created by glass beads can prevent irrigation sprinklers plugging by preventing the accumulation of solids at the water distribution system. Moreover, the presence of UV-activated TiO₂ on the glass surface can enhance the degradation of organic waterborne pollutants, such as fertilizers, present in irrigation water.

The replica method is one of the several techniques used to produce macroporous ceramics. According to Studart et al., it consists of the infusion of a ceramic suspension with a cellular structure to obtain a macroporous ceramics [10]. The method can produce macroporous ceramics with porosity range between 70% and 90% [11]. Nonetheless and as aforementioned, for large-scale applications, an economical fabrication method becomes mandatory. For those reasons, we selected a pressureless sintering technique for the fabrication of the TiO₂/recycled glass filters. The selected technique also allowed for a more controlled sintering process to achieve a desired interconnected porosity.

To study the feasibility of this design, this research was focused on the filter structure, its mechanical characterization, and the subsequent immobilization of TiO₂ particles onto the porous glass structure. Porosity analysis and water percolation experiments would help characterize the filter structure. Additionally, compressive tests would permit to assess the viability of using this TiO₂/recycled glass filter under given water heads.

2 Materials and methods

2.2 Porosity

The material was produced by glass sintering at atmospheric pressure (i.e. pressureless sintering process). For surface porosity measurements, 1 g of each recycled glass was sintered at 700°C, 725°C, 750°C, 775°C, and 800°C for 10–30 min; each sample was removed from the furnace at 5 min intervals for testing. We should note that, upon preliminary trials, the recycled glass beads did not sinter properly at temperatures below 700°C. The starting mixtures consisted of 1 g of recycled glass and 5 ml of C₂H₆O, with 0.1–1 wt% of TiO₂. C₂H₆O was used as carrier of the TiO₂ particles onto the recycled glass beads surface.

In the case of the composite fabricated with glass powder and TiO₂, all samples were sintered at 750°C for 30 min. This sintering temperature was selected because at lower temperatures the composites did not sinter properly, something that had been corroborated with the sintered glass specimens. Fifteen samples of recycled glass and recycled glass/TiO₂ composites were sintered at specific times and temperatures for porosity measurements.

A Nikon (Melville, NY, USA) Epiphot 2 optical microscope allowed obtaining micrographs of the sintered specimens. Image J (public domain image analysis package) allowed analyzing micrographs obtained at a 5× magnification to assess the porosity percent. Figure 1A exemplifies the analysis of a sintered specimen. After image processing, the white regions in Figure 1B represent the sintered glass beads and the black region represents the voids or pores. The image analysis software measured the black area and calculated the resulting porosity as area percent. A similar analysis for porosity measurements was performed to E-glass/polypropylene composites by Santulli et al. [13].

Figure 1:

Representative sintered glass micrographs analyzed using Image J: (A) original optical micrograph and (B) processed image used for porosity measurement.

2.3 Percolation

To study the percolation characteristics of the composite filter, 30 g of MG30 and MG80 recycled glass were sintered using the same sintering schedule described previously (i.e. from 700°C to 800°C and for 10–30 min at 5 min intervals). Moreover, the TiO₂/recycled glass composites were obtained by mixing 30 g of recycled glass, 10 ml of C₂H₆O, and 20 ml of H₂O containing from 0.1 to 1 wt% of TiO₂. These composites were sintered at 800°C for 20 min for the percolation measurements. Figure 2 shows a schematic of the percolation apparatus in which a one-directional water flux flows through sintered recycled glass samples. Equation (1) computes the percolation (P_flux) with the elapsed time Δt (s), necessary to obtain a water volume of 600 ml in the water container [i.e. ΔV_w (ml). Fifteen samples of recycled glass and recycled glass/TiO₂ composites were sintered for percolation measurements.

Figure 2:

Schematic of the percolation apparatus.

(1)Pflux=ΔVwΔt

3 Results

3.1 Effects of recycled glass particle size

3.1.1 Porosity

As both sintering temperature and time increased, we observed a reduction in MG30 and MG80 surface porosity (Figures 3 and 4). MG30 recycled glass had higher porosity values compared to that of MG80. Larger voids formed among MG30 sintered glass than the ones created in the MG80 glass.

Figure 3:

Sintered glass micrographs of MG30 recycled glass sintered for 15 min at (A) 725°C, (B) 775°C, and (C) 800°C.

Figure 4:

Micrographs of MG80 recycled glass sintered for 15 min at (A) 725°C, (B) 775°C, and (C) 800°C.

Figure 5 presents a surface plot of the MG30 porosity as a function of sintering temperature and time, where an apparent maximum porosity of 62% and a minimum of 3% can be observed for the range of sintering parameters tested. Moreover, the surface graph of MG80 porosity shown in Figure 6 points at a maximum porosity of 68% and a minimum one of 4%, which demonstrates the pronounced effect of the sintering parameters used and glass particle size on the resulting porosity. This figure is a performance map that provides a selection mechanism for any desired porosity value required by a given filter design.

Figure 5:

MG30 porosity as a function of sintering time and temperature.

Figure 6:

MG80 porosity as a function of sintering time and temperature.

3.1.2 Percolation

The difference in recycled glass particle size as well as the interconnected porosity size between MG30 and MG80 recycled glass types are shown in Figures 7 and 8. MG30 samples presented higher percolation flux values compared to MG80 ones due to the presence of larger voids (pores) between MG30 particles (before sintering). Figures 9 and 10 show the surface plots of MG30 and MG80 percolation values as a function of sintering time and temperature. The highest and lowest percolation values for MG30 samples were 55 and 3 ml/s, respectively, whereas the highest and lowest percolation rates for the MG80 filters were 18 and 0.63 ml/s, respectively.

Figure 7:

SEM images of MG30 recycled glass sintered at 725°C for 25 min: (A) low and (B) high magnification.

Figure 8:

SEM images of MG80 recycled glass sintered at 750°C for 25 min: (A) low and (B) high magnification.

Figure 9:

MG30 filter percolation as a function of sintering time and temperature.

Figure 10:

MG80 filter percolation as a function of sintering time and temperature.

3.1.3 Compressive strength

Figure 11 displays the compressive strength of MG30 filters as a function of sintering temperature and time. At lower temperatures (700–750°C) for 10–25 min, we obtained brittle samples tolerating no compressive deformation. Furthermore, for sintering temperatures from 775°C to 800°C for 10–30 min, one can clearly observe an increase in compressive strength. The maximum compressive strength obtained was 80 MPa for samples sintered at 750°C, 775°C, and 800°C for 30 min.

Figure 11:

MG30 compressive strength as a function of sintering temperature and time.

Figure 12 shows the resulting compressive strength of the sintered MG80 glass. Brittle samples resulted at temperatures from 700°C to 725°C for 10–25 min. Moreover, an increase in the compressive strength was visible for specimens sintered from 750°C to 800°C for 10–30 min. In this case, the maximum compressive strength obtained for samples sintered from 775°C to 800°C for 20–30 min was 80 MPa. Clearly, MG80 samples showed higher compressive strength than MG30 samples for the same sintering parameters (i.e. temperature and time).

Figure 12:

MG80 compressive strength as a function of sintering temperature and time.

3.2 Effects of TiO₂ in the composites

3.2.1 Surface porosity

The experimental results in Figure 13 demonstrate that the addition of TiO₂ particles increased the MG30-based composite porosity. The TiO₂ particles are distributed randomly over the surface of the MG30 beads (Figure 18C) and we believe that the adverse effect on the porosity is caused by TiO₂ that lowered the viscous flow and hence the sintering rate of the MG30 beads. As a consequence, surface diffusion causes the bead surfaces to have a more spherical shape compared to the beads without TiO₂ particles. According to Greskovich and Lay, surface diffusion is the sintering mechanism responsible for the rounding of particles and the growth of necks between particles [14].

Figure 13:

Optical micrographs of MG30 recycled glass with (A) 0 and (B) 0.7 wt% TiO₂.All samples were sintered at 750°C for 30 min.

A significant reduction in the MG80-based composite porosity resulted from the addition of TiO₂ particles (Figure 14). We attribute this behavior to the smaller size difference between MG80 beads and TiO₂ particles. In this case, the TiO₂ particles are retained within the small voids created by MG80 particles, thus physically cutting down the porosity in the presintered and sintered materials. This appears to be sufficient to overcome the TiO₂ hindering the surface diffusion that facilitates neck growth and continuing sintering. Figure 15 summarizes the measured porosity with respect to the added percent of TiO₂ to MG30 and MG80 beads. For the sake of subsequent design purposes, a regression analysis was conducted on the data in Figure 15 and resulted in the fitted models presented in Equations (3) and (4). Both models adequately represent the experimental data, as proven by the high R² values of 91% for Equation (3) and 98% for Equation (4). In both equations, the TiO₂ variable represents the amount of dioxide (wt%).

Figure 14:

Optical micrographs of MG80 recycled glass containing (A) 0 and (B) 0.7 wt% TiO₂.All samples were sintered at 750°C for 30 min.

Figure 15:

Porosity changes as a function of TiO₂ percent in recycled glass-based composites.

(3)Porosity(%)MG30+TiO2=26.42∗TiO23−38.90∗TiO22 +20.27∗TiO2+35.45

(4)Porosity(%)MG80+TiO2=−18.22∗TiO2 3+53.74∗TiO22 −41.66∗TiO2+13.47 

3.2.2 Percolation

Figure 16 shows SEM images at two magnifications of MG30-based composites sintered with 1% TiO₂ at 800°C for 20 min. The dispersion of TiO₂ particles on the MG30 particles becomes apparent in Figure 16A–C. Figure 16 provides evidence to why the presence of TiO₂ decreased the percolation properties of MG30-based composite filters. SEM images revealed how TiO₂ particles increased the roughness of recycled glass beads, which could affect the laminar water flow through the filter. Figure 17 shows SEM images of MG80-based composites with 1% TiO₂ sintered at 800°C for 20 min.

Figure 16:

SEM images of MG30-based composites containing 1% TiO₂ sintered at 800°C for 20 min obtained at different magnifications, (A) low, (B) medium, and (C) high, on the surface of a sintered glass bead.

Figure 17:

SEM images of MG80 recycled glass beads mixed with 1% TiO₂ sintered at 800°C for 20 min at different magnifications, (A) low, (B) medium, and (C) high, on the surface of the sintered beads.

Figure 18 displays the percolation values of recycled glass-based composites as a function of the TiO₂ content. MG30-based composites percolation values decreased by 56% from 3.7 to 1.61 ml/s for 0% to 1% TiO₂. Moreover, MG80-based composites percolation values decreased by 78% from 3.26 to 0.72 ml/s, with the addition of TiO₂ from 0% to 1% TiO₂. The MG80-based composite percolation rates decreased more abruptly than that of MG30 due to the presence of the TiO₂ particles. The dioxide particles appear to locate themselves between MG80 particles, thus decreasing the filter percolation properties. Equation (5) (R²=97%) and Equation (6) (R²=99%) are models fitted via linear regression analysis and represented in Figure 18 of MG30- and MG80-based composites, respectively. In both equations, the TiO₂ variable represents the amount of dioxide (wt%).

Figure 18:

Measured percolation of recycled glass-based composites sintered at 800°C for 20 min as a function of the TiO₂ amount (wt%).

(5)P(mls)MG30+TiO2= 3.756∗exp(−1.008∗TiO2)

(6)P(mls)MG80+TiO2= 3.249∗exp(−3.044∗TiO2)

3.2.3 Compressive strength

Figure 19 shows the compressive strength values of MG30- and MG80-based composites bearing different TiO₂ levels sintered at 800°C for 20 min. The compressive strength of TiO₂/MG30 composites was adversely affected by the addition of TiO₂ particles, as shown in Figure 19. TiO₂ particles decreased the sintering ability of MG30 particles by effecting a slower surface diffusion rate and thus reducing the material compressive strength. For concentrations of 0.6 to 1 wt% TiO₂, one can observe a large compressive strength loss, nearly 90%. In the MG80-based composite, TiO₂ particles (average particle size of 12 nm) raised the composite compressive strength due to the lower porosity of the samples. As discussed previously, TiO₂ particles lodged within the voids of the MG80 glass particles, which bettered the load transfer among MG80 particles upon compression.

Figure 19:

TiO₂/MG30 and TiO₂/MG80 composites compressive strength as a function of TiO₂ percentage.

4 Discussion

Recycled glass filters function similar to sand filters, as they can retain large particulates and diminish water turbidity. However, in addition to reducing water turbidity, our approach included the degradation of pollutants by adding TiO₂ to the recycled glass. Porosity, percolation, and compressive strength tests performed on recycled glass filters demonstrated how the addition of TiO₂ affects these properties.

Because filters are selected according to the minimum allowable fluid rate and pollutant concentration, the careful monitoring of percolation is essential to avoid stagnation of the fluid to effectively degrade the pollutant. In our case, we addressed the possibility of using these composite filters for waterborne contaminants, which require different percolation rates to be effective. Thus, it is necessary to select a filter with a similar percolation value as the polluted water or soil to avoid groundwater stagnation. As a consequence, percolation performance maps presented in this work can serve as a selection tool of the optimum sintering parameters (i.e. temperature and time) for the fabrication of filters intended for a specific polluted system. Moreover, additional performance maps of porosity and compressive strength provide supplementary, yet important, information for design purposes. Also, the presented graphs and statistical models of porosity, percolation, and compressive strength of the composites provide ancillary parameter information to design systems based on these TiO₂/recycled glass composites.

Furthermore, photocatalytic degradation processes of organic pollutants depend on the type of pollutant and the photocatalyst. In our case, photocatalytic anatase is a typical degrader of organic pollutants in water. The reported TiO₂ dosages can be correlated with the required time for the degradation of pollutants using recycled glass/TiO₂ composites. Thence, in ensuing works, the percolation values of samples could be directly related to the required time for pollutant degradation. Our research renders models to describe the percolation changes of TiO₂/recycled glass composites as a function of the TiO₂ amount. In closing, these descriptive models bring out the required sintering parameters for the fabrication of filters for the remediation of a specific system.

5 Conclusions

1. As expected, in general, porosity and percolation properties of glass filters decrease for larger sintering time and temperature.

2. Concurrently, the compressive strength of the composites increased with longer times and higher temperatures of sintering.

3. MG30 composite sintering behavior was adversely affected by the addition of TiO₂ particles to the glass surface. Because of this, one can observe an increase in TiO₂/MG30 composite porosity accompanied with a loss in compressive strength.

4. It is believed that the small voids presented by MG80 recycled glass served for the agglomeration of TiO₂ particles. Accordingly, the TiO₂/MG80 composites presented lower porosity values and a higher compressive strength for larger amounts of TiO₂.

5. Finally, the recommended ranges of sintering parameters for TiO₂/recycled glass composites were from 750°C to 800°C for 20 to 30 min for the experimental conditions discussed in this work.