*2.2. Photocatalytic Activity Measurements*

Figure 9a–d present the ability to photoreduce carbon dioxide with water vapor to carbon monoxide (CO), methane (CH4), and hydrogen (H2), respectively, with reaction times given in μmol/gphotocatalyst/dm3.

The addition of carbon spheres in an amount of 0.05 g per 1 g of TiO2 resulted in a slight increase in the photocatalyst activity (Figure 9b) compared to pure P25 (Figure 9a). The comparison of the photoactivity towards CO production is presented in Figure 10. The amount of carbon monoxide produced decreased slightly from 55.75 μmol/gphotocatalyst/dm<sup>3</sup> for P25 and 55.94 μmol/gphotocatalyst/dm<sup>3</sup> for P25 + C 1/0.05 to 43.62 μmol/gphotocatalyst/dm<sup>3</sup> for the sample P25 + C 1/0.1 and to 46.79 μmol/gphotocatalyst/dm3 in the case of the P25 + C 1/0.5 material. At the same time, the amount of methane decreased significantly and gradually from 9.02 μmol/gphotocatalyst/dm<sup>3</sup> (pure P25) through 4.76 and 4.02 μmol/gphotocatalyst/dm3 for materials P25 + C 1/0.05 and P25 + C 1/0.1, respectively; to practically 0.00 μmol/gphotocatalyst/dm3 in the case of the P25 + C 1/0.5 sample.

**Figure 7.** SEM images of silica cloth coated by TiO2 P25 combined with different amount of carbon spheres; (**A**) P25 + C 1/0.05; (**B**) P25 + C 1/0.1; (**C**) P25 + C 1/0.5.

From the results obtained, it can be seen that the amount of methane captured decreased with the increasing carbon content in the material (Figure 11). The same dependence was also observed for the hydrogen production (Figure 12).

Two factors may contribute to the decrease in the photoactivity for CO, CH4, and H2 with the increasing carbon sphere content. The first aspect is the well-known electron interaction between amphoteric electrons from the surface of the graphite spheres and the functional groups of the TiO2 surface. This prevents the recombination of the electron-hole pair. It is known that even a small interaction, specifically doping [14], is necessary to achieve a positive effect. The addition of large amounts of carbon from 0.1 g/1 g TiO2 to 0.5 g/1 g TiO2 causes the scattering of electrons in the structure with a large amount of carbon.

The second effect is caused by the increased CO2 adsorption by the carbon spheres, as shown in Table 1. TiO2 adsorbs only 0.72 mmol CO2/g at 30 ◦C (Table 3), while carbon spheres adsorb 2.43 mmol CO2/g. A large addition of carbon spheres increases the sorption and the presence of CO2 on the surface of the photocatalyst, which forms CO3 <sup>2</sup><sup>−</sup> ions in the presence of H2O vapors, which are strong electron scavengers. The further addition of carbon spheres did not positively affect the activity. In the case of sample P25 + C 1/0.5 (Figure 9d), the only product of the process was carbon monoxide. The higher content of carbon spheres blocked the UV-Vis radiation, which can generate the TiO2 electron-hole pair and, at the same time, scatter and disperse the excited electrons in the crystal lattice of the carbon spheres.

**Figure 8.** Example of EDX chemical element mappings of silica cloth coated by TiO2 P25 combined with carbon spheres for sample P25 + C 1/0.5.

**Table 4.** Surface chemical element compositions from EDX of silica cloth coated by TiO2 P25 combined with carbon spheres for sample P25 + C 1/0.5.


It can be observed that the main products are CO, CH4, and H2. Carbon monoxide is also the dominant component of the postreaction gases, regardless of the amount of carbon spheres added as a modifier.

The reactivity of the photocatalysts is probably limited by the reduction of water to hydrogen, which is the first stage of the whole complex process. In our case, with the exception of sample P25 + C 1/0.5, there is some excess of hydrogen, which is a positive phenomenon.

Hydrogen is formed in the reactions at the photocatalysts, as reported by Peng Wang et al. [14] and Minoo Tasbihi et al. [24]:

$$\mathrm{H\_2O} + 2\mathrm{h^+} \rightarrow \frac{1}{2}\mathrm{O\_2} + 2\mathrm{H^+} \tag{1}$$

$$2\text{H}^+ + 2\text{e}^- \rightarrow 2\text{ H}\_2\tag{2}$$

**Figure 9.** *Cont*.

**Figure 9.** Total content of the products of the photoreduction of CO2 in the process with: (**a**) P25, (**b**) P25 + C 1/0.05, (**c**) P25 + C 1/0.1, (**d**) P25 + C 1/0.5.

Carbon monoxide is the result of an easier two-electron reduction of carbon dioxide:

$$\rm{CO}\_2 + 2H^+ + 2e^- \rightarrow \rm{CO} + \rm{H}\_2\rm{O} \tag{3}$$

The lower efficiency of the reaction towards methane production is the result of the more difficult eight-electron CO2 reduction reaction:

$$\text{CH}\_2 + 8\text{H}^+ + 8\text{e}^- \rightarrow \text{CH}\_4 + 2\text{ H}\_2\text{O} \tag{4}$$

On the homogeneous surface of the photocatalyst, reactions No. 1 to No. 4 are the twoelectron production of hydrogen (reactions 1–2) and a two-electron to carbon monoxide (reaction No. 3), which could be considered as an intermediate step to the 8-electron reaction of No. 4 towards methane.

**Figure 10.** Total content of carbon monoxide of the photoreduction of CO2 for the studied samples: P25; P25 + C 1/0.05; P25 + C 1/0.1, and P25 + C 1/0.5.

**Figure 11.** Total content of methane in the photoreduction of CO2 for the following photocatalysts: P25; P25 + C 1/0.05; P25 + C 1/0.1, and P25 + C 1/0.5.

It is known that the hydrogen production reactions (reactions No. 1–2 water splitting) take place in the semiconductor valence band (on holes), and the CO2 reduction reactions take place in the conduction band (with the participation of electrons).

For this reason, the production of hydrogen controls all CO2 reduction reactions. In this case, only the CO formation reaction (No. 3) and the methane forming reaction (No. 4) are linearly competing with each other. Indeed, as shown in Figure 13, with the addition of carbon spheres, excess hydrogen decreases, and at the same time, methane decreases, which requires more hydrogen than the formation of carbon monoxide in reaction No. 3.

**Figure 12.** Total content of hydrogen in the photoreduction of CO2 for subsequent photocatalysts: P25, P25 + C 1/0.05, P25 + C 1/0.1, and P25 + C 1/0.5.

**Figure 13.** Selectivity and yield rates of the products after 4 h of the process.

The collected postreaction gas composition is different from that obtained on photocatalysts on mineralogical ground fibers, where only methane was obtained and no other components were identified [6]. In our case, by using silicon fibers, it is possible to obtain a mixture of CO, CH4, and small amounts of hydrogen, with high selectivity to CO. Figure 13 shows the calculated selectivity, and the obtained amounts of each component of the gas phase produced. In all the cases tested, high selectivity to carbon monoxide is observed—from about 83.40% to 100% for the sample P25 + C 1/0.5, where the hydrogen was entirely consumed by the two-electron reaction of reduction of CO2 to CO.

#### **3. Experimental**

#### *3.1. Preparation of the Samples*

In the experiments, photocatalysts consisting of TiO2 P25, manufactured by Degussa (Evonik Industries AG, Germany), and carbon material were tested. The carbon material was prepared with the use of resorcinol and formaldehyde. The description of the original method of producing carbon spheres used in this work is presented elsewhere [25]. Three powders with different mass ratios of titanium dioxide to carbon were prepared by grinding in a mortar. The mass ratios were: 1:0.05; 1:0.1, and 1:0.5. Aqueous suspensions of the prepared materials were applied to the fiberglass cloth. Glass fiber fabric with an area weight of 40 g/m2 was supplied by Fiberglass Fabrics (Opole, Poland). Then the fibers with the photocatalysts were dried at 110 ◦C for 1 h.

#### *3.2. Photoreduction Process*

Experiments were performed in a cylindrical quartz reactor with a working volume of 392 cm3 (Figure 14). Four Actinic BL TL-E Philips lamps were used with a total power of 88 W, emitting UV-A radiation in the wavelength range of 350–400 nm. The lamps were located outside the reactor and formed a ring. The reactor was sealed in a thermostatic chamber to exclude other light sources and ensure a stable process temperature. An amount of 1 cm3 of distilled water was poured into the reactor. A photocatalyst previously applied to glass fibers was added to the reactor. Then the whole system was purged with pure CO2 (Messer, Poland) for 30 min. After this time, the system was tightly sealed and the lamps were turned on. Both during purging and during the process, the gas was stirred using a pump with a flow rate of 1.6 dm3/h. The process was carried out for 4 h. Gas samples were collected every 2 h for analysis.

**Figure 14.** The scheme of the reactor for photocatalytic reduction of CO2 in the gas phase.

## *3.3. The Analysis of the Gas Phase*

The gas phase composition was analyzed using SRI 310C gas chromatograph (SRI Instruments, Torrance, CA, USA), equipped with a 5Ä molecular sieve column and an HID detector (Helium Ionization Detector). The carrier gas was helium. The analyses were performed under isothermal conditions at 60 ◦C. The gas flow through the column was 60 cm3/min, while the volume of the gas sample was 1 cm3. The content of each component in the gas phase was calculated in the subsequent measurements based on the calibration curve.

#### *3.4. XRD Analysis*

The crystalline structure of used titanium dioxide and carbon spheres was studied with X-ray powder diffraction (CuKα radiation, Malvern PANalytical B.V., The Netherlands). The mean crystallites sizes were calculated based on Scherrer's equation:

$$D = \frac{K\lambda}{\beta \cos \theta} \tag{5}$$

where:

*D*—mean crystallite size (nm),

*λ*—wavelength of Cu Kα radiation (nm),

*θ*—Bragg's angle (◦),

*β*—calibrated width of a diffraction peak at half maximum intensity (rad).

The percentage of anatase in the crystalline phase (%A) was calculated according to the equation:

$$\%A = \frac{I\_A}{I\_A + I\_R} \cdot 100\% \tag{6}$$

where *IA* and *IR* are the diffraction intensities of the anatase peak at 25.4◦ and rutile peak at 27.5◦, respectively.

#### *3.5. SEM/EDS Measurements*

The surface morphology of the samples was examined using a scanning electron microscope (SEM Hitachi SU 8020, Japan). The SEM and EDS analysis parameters were: acceleration voltage of 20 kV and current of 10 μA. The samples for investigation using SEM were firstly vapor-deposited with a 5 nm thin chromium layer to protect the samples from the electrical charge.
