Revisiting the Key Optical and Electrical Characteristics in Reporting Photocatalysis of Semiconductors

Revisiting the Key Optical and Electrical Characteristics in Reporting Photocatalysis of Semiconductors Dai-Phat Bui, Minh-Thuan Pham, Hong-Huy Tran, Thanh-Dat Nguyen, Thi Minh Cao, Viet Van Pham* Photocatalysis Research Group (PRG), Faculty of Materials Science and Technology, University of Science, VNU–HCM, 227 Nguyen Van Cu Street, District 5, Ho Chi Minh City, 700000, Viet Nam Corresponding authors: daiphatbui301196@gmail.com (Dai-Phat Bui) and pvviet@hcmus.edu.vn (Viet Van Pham)


Introduction
Global warming has been becoming increasingly serious that triggers extreme concerns across the globe. It comes from environmental pollution, including air, soil, and water pollution 1 . In the scientific domains, environmental pollution alters the environment's chemical, physical and biological nature; it is the reason for an estimated 12.6 million deaths each year 2 . Thus, finding new technologies and methods to address environmental pollution is critical. Various methods are performed to address pollutants, including ion exchange, flotation, media filtration, centrifugation, and adsorption approaches. However, there are limitations due to high operating costs, low efficiency, poor recyclability, and toxic by-products 3,4 .
Photocatalysis has been studied and considered a green and effective approach in addressing environmental pollution 4 . Photocatalysis has attracted for its versatility in the environment and energy sector. In the environmental field, photocatalysts can be used to remove pollutants in the air and water. The photocatalytic process can eliminate the total toxicity with products of CO2, water, and other fewer substances 5 . In the energy field, photocatalysts can generate alternative fuels such as H2, O2, and bioethanol via splitting H2O or reducing CO2 6 . In the photocatalytic process, photocatalysts are activated by a suitable light source to induce photocatalytic reactions. The use of light gives photocatalytic technology a great advantage compared to other waste treatment methods such as adsorption or filtration. However, the catalysts' optical and electrical properties and photocatalytic mechanisms are not paid needed attention to compared to the sole efficiency. 7 This causes difficulties to understand the mechanism of decomposition of pollutants and future photocatalytic application.
This work highlights key techniques to study optical and electrical properties in reporting photocatalysis. In addition, the current mistakes and inconsistencies in interpreting these properties are presented and revised. Here, TiO2, g-C3N4, and SnO2 are used as representatives because TiO2 is the most studied photocatalyst (accounting for 52.5% of studies in this field, Web of Science, March 2021), g-C3N4 is an emerging nonmetallic photocatalyst (7.8%), and SnO2 is an emerging photocatalyst (1.6%). The bandgap of TiO2, SnO2, g-C3N4 is determined by diffuse reflectance spectroscopy (DRS), choosing the compatible light source becomes more straightforward. The conduction band (CB) positions and the valence band (VB), which determine the redox abilities, are obtained by the Mott-Schottky technique. The supporting states in the bandgap of the materials are investigated by Photoluminescence (PL). The band structure of their composites is studied by X-ray Photoelectron Spectroscopy (XPS) and Ultraviolet Photoelectron Spectroscopy (UPS). The understanding of optical and electrical features of photocatalysts not only supports research expansion into new materials/composites and but also enables researchers to design wastewater or air pollution treatment systems with the suitable operation time, area, light source, and conditions, which helps to optimize the photocatalytic activity of photocatalysts and systems.

Diffuse Reflectance Spectroscopy Analysis
It is essential to know the bandgap energy to determine the right light source for activating photocatalysis. The photon energy should be higher than the bandgap energy of the photocatalyst to generate electron-hole pairs. The DRS is often used to determine the bandgap energy of photocatalysts. DRS results are plotted with reflectance or absorbance as the vertical axis vs. wavelength as the horizontal axis. To determine bandgap, the eqn.
(1) is used to transform wavelength into energy as a new horizontal axis. The Tauc equation (eqn. (2)) and Kubelka Munk equation (eqn. (3)) transform the absorbance and the reflectance into a new vertical axis, respectively [8][9][10] . The parameters of these equations are calculated as eqn. (4)eqn. (8). The bandgap energy is determined by the tangent of the plot with Ox. It is useful for single-component photocatalysts and heterojunctions [11][12][13] . However, the absorbance should not be converted into reflectance or vice versa by eqn. (9) or eqn. (10) due to the significant change of bandgap energy. Also, photocatalysts with similar bandgap energies could behave differently. This shows that the photocatalytic behaviors depend not only on the bandgap value but also on the positions of the conduction band and valence band of materials, which DRS cannot determine. The next sections will explain these phenomena and discuss the band structure of photocatalysts.
where E is Photon energy (eV), h is Planck's constant (4.132  10 -15 eV.s), υ is the frequency (s -1 ), c is photon velocity (nm), λ is the wavelength (nm), α is absorption coefficients from absorption mode, B is constant, and Eg is bandgap (eV), A is absorbance value, l is sample width (cm), R is reflectance value, K is absorption coefficient from reflectance mode, and S is scattering coefficient from reflectance mode. For direct and indirect bandgaps, r is equal to 2 and ½, respectively.      conversion of wavelength to energy should preserve the emission spectra area as eqn. (15) because the intensity vs. energy system (IE) is higher at lower energy and vice versa. In other words, the intensity vs. wavelength system (Iλ) should transform into IE as eqn. (16) - (17). Thirdly, IE could be calculated as eqn. (18) Where IE is signal per energy unit, Iλ is the signal per wavelength unit. Figure 3 indicates the PL spectra of TiO2, SnO2, and g-C3N4; the spectra undergo Gaussian peak fitting to find component peaks. Figure 3  Obviously, with the DRS, Mott-Schottky plots, and PL, the basic properties of the pure TiO2, SnO2, and g-C3N4 are being clarified. The next sections describe the changes of band alignment for the composites between them.

X-ray Photoelectron Spectroscopy
For determining the composition of the photocatalysts, X-ray photoelectron spectroscopy (XPS) is well known as a powerful tool, but another application of XPS that should be considered is to determine the band offset of heterojunctions. Comparing to the mentioned approach of Mott-Schottky, evaluating band shifting with XPS is more accurate because XPS records the elements' core levels, which are more stable than the surface value. In other words, the VB offset could be accurately determined when the deviation of the core levels before and after contact is measured. The VB offset is calculated as eqn. (20) or (21), so the CB is calculated as eqn. (22) 28 . The exact position of band structure vs. NHE is useful in studying and applying for suitable applications, as mentioned above.   Valence band edge spectra of TiO2 (c) and SnO2 (d). Band alignment of materials before contact (e) and after contact (f) 27 .

Ultraviolet Photoelectron Spectroscopy
Fermi energy level (EF) is an important part of the band structure; it helps determine the The UPS results of TiO2, g-C3N4, and TiO2/g-C3N4 in Figure 5 (a) reveal the relative position of HOMO and VB to EF of them. Combining with DRS and Mott-Schottky results, the work function of these materials can be calculated by the difference between the Vacuum level and EF. The electrons will transfer from the material having the lower work function (g-C3N4) to the material having the higher work function (TiO2) to balance the EF. This is consistent with the work function of TiO2/g-C3N4 in Figure 5 (a). The band alignment of materials before and after contact is shown in Figure 5 (b). Due to the transfer of electrons from g-C3N4 to TiO2, the band is bending, and an electric field is formed at the interface of TiO2/g-C3N4. This is a direct method for confirming the latest type of photocatalysts, known as Z-Scheme and S-Scheme.

Conclusion and perspectives
In this report, we have discussed the key optical and electrical characteristics in studying photocatalyst. The highlights and perspectives for future work are shown in the graphical abstract and described as follows: (1) Bandgap energies determine the light source for activating photocatalysts. The DRS should be used to determine the bandgap energy of photocatalysts.
(2) Positions of conduction band minimum (CBM) and valence band maximum (VBM) determine the redox ability. The Mott-Schottky technique is used to determine the CB position. Combining with the bandgap from DRS, the VB position is also determined. (4) Comparing to the mentioned approach of Mott-Schottky, evaluating band shifting with XPS is more accurate because XPS records the elements' core levels, which are more stable than the surface value. XPS that should be considered is to determine the band offset of heterojunctions.

(5)
Fermi energy level (EF) may determine the semiconductor type of photocatalysts and predict the transport of electrons in heterojunctions. UPS is the technique of recording the VB and EF of photocatalysts.
Graphical Abstract: Summary of approaches for study on photocatalysis