Advanced Hydrogenation Technologies for Photoelectrochemical Hydrogen Production

Abstract

Since it was first reported at Science in 2011, the hydrogenation technique to prepare black TiO₂ has received great attention worldwide. However, most of the reported hydrogenation techniques require harsh conditions and/or high costs that seriously limit their practical applications. In response to overcome the above technical bottleneck, two advanced hydrogenation technologies, low temperature hot wire H hydrogenation and room temperature H⁺ hydrogenation, are developed. The chapter focuses on the two technologies to hydrogenate TiO₂ nanorods achieving the highest photocurrent density of black TiO₂ reported to date under photoelectrochemical (PEC) hydrogen production conditions, while simultaneously addressing issues like high temperature, high pressure and prolonged treatment as well as chemical residues with traditional hydrogenation approaches. Comparisons of the proposed technologies with conventional methods are conducted: Our advanced hydrogenations demonstrate more efficient and low-cost technologies beyond those of state-of-art hydrogenations, enabling them to move from basic research to large-scale practical application.

1. Introduction

The “green hydrogen” prepared by photoelectrochemical (PEC) water splitting absorbs sunlight through semiconductor photoanode, generates photogenerated electrons and holes to split water into hydrogen and oxygen, and efficiently converts solar energy into hydrogen energy for storage, which has great scientific significance and broad application prospects [1, 2].

Metal oxide semiconductors have been widely studied due to their excellent PEC stability, low cost, favorable band-edge position and bandgap, among which the most common oxide semiconductors such as Fe₂O₃ and BiVO₄ have suitable bandgaps, but due to their short hole diffusion length, the solar energy conversion efficiency is low, which seriously limits their practical applications. Among many semiconductors photocatalytic materials, TiO₂ has become one of the most promising semiconductor photocatalysts because of its stable chemical properties, strong oxidation-reduction and low cost, but it is a wide bandgap semiconductor material that can only absorb ultraviolet light in the solar spectrum, so improving the light response of TiO₂ to visible light is a research hotspot in this field. In recent years, in order to improve the response of TiO₂ to visible light, the surface doping of various elements has been reported, but the photocurrent is less than 1 mA cm⁻² [1]. Therefore, new doping technology to improve the response of TiO₂ to visible light for enhanced solar-to-hydrogen conversion efficiency is highly demanded [1, 2, 3, 4, 5, 6, 7, 8, 9, 10].

2. Experimentation with different hydrogenation technologies

The acquisition of black TiO₂ by high-pressure hydrogenated TiO₂ has attracted great interest [11]. In 2011, American scientists Xiaobo Chen and Samuel Mao published a study in the journal Science detailing their novel method for synthesizing black TiO₂ through high-pressure hydrogen treatment. Figure 1 illustrates the distinctive core-shell structure of black TiO₂ and the corresponding energy diagram depicting a reduced band gap. The researchers confirmed that black TiO₂ possesses a broader band for light absorption, enabling it to absorb not only visible light but also ultraviolet and infrared light. Furthermore, black TiO₂ exhibits exceptional photocatalytic properties. Since its discovery, black TiO₂, with its combination of high sunlight absorption and ultra-high photocatalytic capabilities, has garnered significant attention in the scientific community.

Figure 1.

(A) Core-shell crystal structure and narrowing band gap model; (B) photo: Hydrogenation of white pristine TiO₂ to black TiO₂; (C) and (D) high-resolution transmission electron microscope (HR-TEM) images of pristine TiO₂ and black TiO₂, respectively, confirming the core-shell structure of black TiO₂ by comparison.

Compared with the traditional elemental doped TiO₂, the hydrogenation method of hydrogen doping has a unique advantage: Because it causes the original crystal structure to be distorted without destroying the original structure, resulting in a new microstructure and electronic structure, which can change its band gap, conductivity and electrochemical properties, and has a wide range of application prospects [12]. Then, most of the hydrogenation methods reported in the existing literature, such as high-temperature hydrogen annealing and hydrogen plasma annealing, require high temperature or high pressure, long treatment time, high dose of H₂ or H⁺ and other harsh conditions. In order to easily and economically prepare black TiO₂ with disordered shells, the researchers also developed a chemical reduction method. However, due to surface defects and chemical residues, the PEC activity of black TiO₂ is relatively low, and the typical photocurrent density is less than 2 mA cm⁻² [13]. In the face of this bottleneck, we have invented two advanced hydrogenation methods, low temperature hot wire H hydrogenation and room temperature H⁺ hydrogenation [14, 15], which successfully overcome not only the bottlenecks of the traditional hydrogenation process such as high temperature, high pressure and high energy consumption but also promoted the development of the world’s hydrogenation field, so that the hydrogenation technology can truly move from basic research to large-scale practical application.

3. Attempts with advanced H and H⁺ hydrogenation approaches

3.1 H hydrogenation

3.1.1 Work principle, experimental set-up and H treatment

Molecular hydrogen (H₂) has been experimentally shown to undergo catalytic dissociation into atomic hydrogen (H) when in contact with tungsten wires heated to a temperature exceeding 1000°C. Langmuir conducted research on the efficiency of tungsten wires in facilitating the dissociation of H₂ into H atoms in 1912 [16]. The ability of hot wires to dissociate hydrogen is effectively utilized in hot wire activated chemical vapor deposition (HWCVD) processes. The process of HWCVD is fundamentally distinct from plasma enhanced chemical vapor deposition (PECVD) due to the absence of hydrogen ions or energetic particles in the generation of atomic hydrogen H in HWCVD [14].

Inspired by the above previous work, we developed a low-temperature H hydrogenation method for the first time and studied the effect of hydrogenation parameters on enhanced PEC performance of hydrogenated TiO₂ photoanode materials [14].

Figure 2 illustrates the experimental set-up of low-temperature H hydrogenation method. A parallel array consisting of 10 tungsten wires, each 600 mm in length and 0.53 mm in diameter, was installed within a vacuum chamber. The spacing between the wires was set at 50 mm, resulting in a total activated area of 600 × 450 mm². The samples underwent treatment at varying wire temperatures (1600, 1700 and 1800°C), with a separation of 75 mm between the wires and the sample. Throughout the 20 mins H treatments, the sample temperature rises to 245, 265 and 290°C at T_wire = 1600, 1700 and 1800°C, respectively.

Figure 2.

Experimental set-up of low-temperature H hydrogenation.

The flux rate of H hydrogen can be accurately controlled and adjusted by the wire temperature, as it is dependent on both the wire temperature and pressure, which the pressure is usually preset at 1 Pa controlled by a gate valve. The hydrogen treatments, encompassing all processing parameters such as wire temperature, pressure, hydrogen flux and time, were meticulously regulated by a programmable recipe, enabling reproducible hydrogenations on TiO₂ nanorods.

In the subsequent sections, we have conducted a systematic investigation on the impact of wire temperatures of 1600, 1700 and 1800°C on the resulting structural, optical and PEC properties of H-treated TiO₂ nanorods.

3.1.2 Influence of wire temperature on the structural, optical, electrical and PEC properties

The investigation of the microstructure of H-TiO₂ nanorods was conducted using scanning electron microscopy (SEM) and high-resolution transmission electron microscopy (HR-TEM). The morphology of H-TiO₂ treated at a wire temperature of 1700°C closely resembles that of pristine TiO₂. The TEM and HR-TEM images indicate that the pristine TiO₂ exhibits a single-crystalline structure, whereas H-TiO₂ treated at a wire temperature of 1700°C composed of a core-shell structure which the thickness of the disordered shell is around 2 nm (Figure 3a–f). To investigate compositional changes of resulting disordered shell, the [O]/[Ti] ratio was determined through EELS analysis of Ti-L_2,3 and O-K spectra. It indicates that the [O]/[Ti] ratio of pristine TiO₂ remains constant at approximately 2 throughout the core and surface of the TiO₂ nanorod, whereas the [O]/[Ti] ratio of H-TiO₂ at a wire temperature of 1700°C is ca. 2 in the core region and decreases gradually in the disordered region (Figure 3g–i).

Figure 3.

TEM, HR-TEM images and fast Fourier transform (FFT) patterns of (a–c) pristine TiO₂ and (d–f) H-TiO₂ nanorods treated at T_wire = 1700°C. (g–i) the corresponding [O]/[Ti] ratios obtained from the quantification of the acquired electron energy loss spectroscopy (EELS) spectra.

The optical absorption spectra of pristine TiO₂ and H-TiO₂ were analyzed in order to investigate the improved optical absorption properties of H-TiO₂. The results presented in Figure 4a indicate a red-shift in the band edge of H-TiO₂ as wire temperatures are elevated. Additionally, within the visible spectrum ranging from 420 to 850 nm, the absorption of H-TiO₂ demonstrates a gradual increase with rising wire temperatures. Further, Tauc plots depicted in Figure 4b illustrate that the band edge of H-TiO₂ is decreased to 2.9 eV relative to pristine TiO₂, which has a band edge of 3.0 eV, confirming the band gap narrowing of H-TiO₂.

Figure 4.

(a) Optical absorption and (b) Tauc plots of optical absorption curves for pristine TiO₂ and H-TiO₂ nanorods treated at T_wire = 1600, 1700 and 1800°C, respectively.

In order to investigate the improved PEC properties of H-TiO₂ nanorods, photocurrent density versus potential (J-V) measurements were conducted using a three-electrode electrochemical set-up (Ag/AgCl in 3 M KCl as reference electrode and platinum wire as counter electrode). A Newport solar simulator 150 W Xe lamp with AM 1.5G filter acted as a light source for the PEC test. The lamp power was adjusted using a reference silicon solar cell to obtain 100 mW cm⁻² (1 sun). All J-V curves of pristine TiO₂ and H-TiO₂ with an active area of 1 cm² were recorded in 1 M KOH using a Princeton 2273 potentiostat with a scan rate of 10 mV s⁻¹. A comparison of the J-V curves of pristine TiO₂ and H-TiO₂ samples was performed to examine the impact of wire temperature on the photocurrent density (Figure 5a). The photocurrent density demonstrates a noticeable increase following treatment at T_wire = 1600°C, reaching its peak value at T_wire = 1700°C, and subsequently decreasing at T_wire = 1800°C (Figure 5a). Specifically, the photocurrent density of H-TiO₂ at T_wire = 1700°C attains approximately 2.5 mA cm⁻² (at 1.23 V vs. RHE), surpassing that of pristine TiO₂ by more than threefold.

Figure 5.

(a) J-V curves of pristine TiO₂ and H-TiO₂ nanorods in 1 M KOH solution in the dark and under solar illumination. (b) Mott-Schottky plots of pristine TiO₂ and H-TiO₂ nanorods treated at T_wire = 1600, 1700 and 1800°C, respectively.

The impact of wire temperature on the conductivity of H-TiO₂ was investigated through Mott-Schottky (M-S) measurements (dark, 1 M KOH) performed on the electrochemical workstation (Gamry, USA) (Figure 5b). The donor density of pristine TiO₂ and H-TiO₂ can be obtained by the M-S equation (Eq. 1):

$$N_d=\left(\frac{2}{{\text{e}}_0{\epsilon }_0{\epsilon }_{\text{r}}}\right){\left[\frac{d\left(1/C^2\right)}{dV}\right]}^{-1}$$E1

Here N_d is the donor density, e₀ is the electron charge, ε₀ is the permittivity of vacuum, ε_r is the dielectric constant of TiO₂ nanorods, C is capacitance and V is the applied bias voltage. It is anticipated that the slope of H-TiO₂ will decrease as the wire temperature increases, suggesting an increase in donor density and conductivity in accordance with the M-S equation. The donor densities of H-TiO₂ were calculated to be 2.5 ± 0.1 x 10¹⁷ cm⁻³ (1600°C), 4.6 ± 0.1 × 10¹⁷ cm⁻³ (1700°C) and 2.1 ± 0.2 × 10¹⁸ cm⁻³ (1800°C), surpassing those of pristine TiO₂ (1.5 ± 0.1 × 10¹⁷ cm⁻³), indicating a trend of increased conductivity in H-TiO₂ samples with higher wire temperatures.

The optical absorption spectra indicate that the H hydrogenation significantly influences the reduction of the band gap from 3.0 to 2.9 eV. The J-V curves demonstrate that samples treated at a wire temperature of 1700°C exhibit the highest PEC activity, reaching 2.5 mA cm⁻² at 1.23 V vs. RHE. The IPCE curves (not shown) further support H-TiO₂ at a wire temperature of 1700°C exhibit the most pronounced red-shift of the band edge, resulting in band gap narrowing.

In order to understand why H-TiO₂ treated at T_wire = 1700°C exhibits the highest photocurrent density, XPS analysis was conducted on pristine and H-TiO₂ samples at various wire temperatures. The XPS O1s spectra of the hydrogen-treated TiO₂ samples revealed two distinct peaks at binding energies of 530 and 531.5 eV, corresponding to TiO₂ and Ti-OH species, respectively (Figure 6a). The intensity of the Ti-OH peak was observed to increase as the wire temperature was raised from 1600 to 1700°C, but decreased as the temperature was further increased to 1800°C. The diminished Ti-OH peak observed in H-TiO₂ at a temperature of 1800°C suggests a reduced presence of Ti³⁺, aligning with the findings of the EELS analysis. The XPS Ti2p spectra of both pristine and H-TiO₂ samples exhibit similar characteristics (Figure 6b). Specifically, the Ti2p XPS spectra display binding energies of 458.5 eV for Ti 2p_3/2 and 464.3 eV for Ti 2p_1/2, which align with the standard values for TiO₂. Notably, the presence of a Ti-H peak at 456.5 eV, rather than metallic Ti at 453.8 eV, is observed in the sample treated at Twire = 1800°C. The creation of surface Ti–H bonds in H-TiO₂ treated at a temperature of 1800°C is accompanied by a further reduction in surface Ti-OH bonds, leading to a decrease in the concentration of Ti³⁺ in the H-TiO₂ sample (Figure 6b). It is confirmed that H-TiO₂ treated at T_wire = 1700°C with the highest photocurrent density is attributed to the highest concentration of Ti³⁺.

Figure 6.

(a) O1s and (b) Ti2p XPS spectra of pristine and H-TiO₂ samples treated at T_wire = 1600, 1700 and 1800°C, respectively.

The findings indicate that low-temperature H hydrogenation results in the efficient and straightforward development of a disordered shell on the H-TiO₂ nanorods, leading to a reduction in band gap, enhancement in donor density, conductivity and concentration of Ti³⁺ and consequently, a notable enhancement in PEC activity.

3.2 H⁺ hydrogenation

3.2.1 Work principle, experimental set-up and H⁺ treatment

H⁺ hydrogenation utilizing kinetic hydrogen ion species has garnered recent interest due to its independence from thermal activation. However, traditional H⁺ hydrogenation methods are hindered by the requirement of high temperatures or the need for high power output from plasma equipment to generate high-energy hydrogen ions, resulting in inefficient and uncontrollable hydrogenation processes [17, 18].

The magnetic field generated by an electromagnetic coil in hydrogen plasma exerts distinct effects on hydrogen ions and free electrons. While the magnetic field can alter the trajectory of electrons, it is unable to significantly impact the motion of hydrogen ions due to their significantly greater mass. As free electrons orbit the magnetic field lines in a circular manner, their path is elongated, resulting in a heightened likelihood of hydrogen ionization and consequently an augmented density of hydrogen ions.

In contrast to traditional plasma hydrogenation methods, based on the above working principle, we incorporated a circular electromagnetic coil into the substrate electrode of a radio frequency (RF) plasma physical vapor deposition (PVD) system to generate a magnetic field (Figure 7) [15]. This magnetic field serves to enhance the concentration of hydrogen ions, and reduce RF power and self-bias voltage, thereby facilitating the production of controllable kinetic low-energy hydrogen ions. In our experimental set-up, a low self-bias voltage (V_bias) of −250V was achieved by applying 20 W of RF power to the substrate during the H⁺ hydrogenation process.

Figure 7.

(a) Experimental set-up of room temperature H⁺ hydrogenation. (b) the role of a circular electromagnetic coil in the H⁺ hydrogenation process.

A programmable recipe was utilized to meticulously regulate the processing parameters of RF power, magnetic field strength, hydrogen flow rate, processing pressure and treatment duration. In the context of the H⁺ hydrogenation process, an RF power of 20 W was administered at the substrate electrode to attain a low self-bias voltage of −250V, with a magnetic field strength of 30 mT, hydrogen gas flow rate of 50 sccm and processing pressure of 1.5 Pa. To study the influence of treatment time on the structural, optical, electrical and PEC properties of H-TiO₂ nanorods, H-TiO₂ samples were synthesized at time intervals of 2.5, 5, 10, 20 and 40 mins, respectively.

3.2.2 Influence of treatment time on the structural, optical, electrical and PEC properties

SEM was used to analyze morphological changes in TiO₂ before and after H⁺ treatment. No significant morphological changes were observed in SEM images of pristine TiO₂ and H-TiO₂ at 5 and 40 mins. X-ray diffraction (XRD) patterns showed only rutile phase with (101) and (002) peaks, indicating no phase change in either pristine TiO₂ or H-TiO₂ samples.

TEM images show that in contrast to the unaltered TiO₂, H-TiO₂ samples exhibit a composite structure consisting of a crystalline core and a disordered shell (Figure 8). Analysis of the FFT image reveals that the core structure of H-TiO₂ remains single-crystalline, while the shell thickness measures approximately 2 nm for H-TiO₂ treated at 5 mins and increases to 5 nm for H-TiO₂ treated at 40 mins.

Figure 8.

TEM, HR-TEM and FFT images of H-TiO₂ at (a–c) 5 and (d–f) 40 mins treatment time, respectively.

Figure 9 displays line-scan EELS data showing changes in [O]/[Ti] ratio in pristine TiO₂ and H-TiO₂ samples. The [O]/[Ti] ratio in the shell of H-TiO₂ decreases over time, indicating the formation of oxygen vacancies. In the core of H-TiO₂ treated at 5 mins, no oxygen vacancies are observed.

Figure 9.

[O]/[Ti] composition ratios of the (a) pristine TiO₂ and H-TiO₂ at (b) 5 and (c) 40 mins treatment time, respectively.

The composition of the disordered shell was analyzed using XPS. Figure 10 displays the O1s spectra of pristine TiO₂ and H-TiO₂ treated at 5 and 40 mins, respectively. The O1s peak at 530 eV corresponds to Ti-O, while the peak at 531.5 eV is indicative of Ti-OH, suggesting the presence of Ti-OH in the shell of H-TiO₂. Additionally, Figure 10 illustrates the Ti2p spectra of pristine TiO₂ and H-TiO₂ at 5 and 40 mins, respectively. It is observed that Ti³⁺ is present in H-TiO₂ treated at 5 mins, and significant reduction (Ti⁴⁺ → Ti³⁺ → Ti²⁺) is evident in H-TiO₂ treated at 40 mins.

Figure 10.

(a) O1s and (b) Ti2p XPS spectra of pristine TiO₂ and H-TiO₂ at 5 and 40 mins treatment time, respectively.

We measured optical absorption spectra to compare the optical properties of pristine TiO₂ and H-TiO₂. The absorption band edges of H-TiO₂ are red-shifted and the absorption in the visible to infrared range increases with longer treatment time (Figure 11a). The band gap narrows from 3.0 eV in pristine TiO₂ to 2.91 eV in H-TiO₂ after 5 mins of treatment (Figure 11b).

Figure 11.

(a) Optical absorption spectra and (b) Tauc plot of pristine TiO₂ and H-TiO₂ at different treatment times.

We calculated the photocurrent density of H-TiO₂ treated at 5 mins by integrating the measured IPCE spectrum (Figure 12a) with measured standard AM 1.5G solar spectrum [19]. Figure 12b shows that the calculated photocurrent density of H-TiO₂ at 5 mins is ca. 2.5 mA cm⁻², which is consistent with the measured photocurrent density, indicating the reliability of our PEC data. We note that a tiny contribution to the photocurrent density originates from absorbed photons with wavelengths larger than 427.5 nm. The formation of occupied Ti³⁺ midgap in the band gap promotes the visible-light absorption to enhance the visible-light-driven PEC activity (Inset, Figure 12a) [1, 14, 20].

Figure 12.

(a) The IPCE spectrum of H-TiO₂ treated at 5 mins. (b) Calculated photocurrent density of H-TiO₂ treated at 5 mins by integrating the measured IPCE spectrum with measured standard AM 1.5G solar spectrum.

Figure 13a demonstrates an increase in photocurrent density of H-TiO₂ with increasing treatment time up to 5 mins, followed by a decrease in treatment time exceeding 5 mins. Specifically, at the 5-mins mark, a photocurrent density of 2.55 mA cm⁻² at 1.23 V_RHE is attained, representing the highest value reported in studies of H-TiO₂.

Figure 13.

(a) J-V curves of pristine TiO₂ and H-TiO₂ at different treatment times in 1 M KOH solution in the dark and under solar illumination. (b) M-S plots of pristine TiO₂ and H-TiO₂ at different treatment times.

Figure 13b shows that an increase in treatment time results in a rise in donor density but a decline in depletion region width. In comparison to pristine TiO₂, the negative shift in flat band potential of H-TiO₂ at 2.5 and 5 mins is attributed to a significant increase in donor density, subsequently causing a shift in the Fermi level toward the conduction band. We determined the band bending by flat band potential from M-S plots, the largest band bending of 1.1 V is achieved for H-TiO₂ at 5 mins.

The experiments on the hydrogenation-properties relation indicate that the increased photocurrent density of H-TiO₂ at 5 mins can be attributed to the elevated levels of Ti³⁺ within the disordered shell, which subsequently narrows the band gap and facilitates the generation of electron-hole pairs. EELS observes that oxygen vacancies are exclusively generated in the disordered shell, suggesting that the suppression of bulk defects enhances charge transport and facilitates the larger band bending to further promote charge transport and injection. Our conclusion is that the increased PEC performance can be primarily attributed to the narrower band gap, greater band bending and bulk defect suppression.

3.3 Comparison of H and H⁺ methods with traditional hydrogenation methods

Compared with the conventional physical and chemical hydrogenation methods [11, 17, 18, 19, 21], we utilized low-temperature H and room-temperature H⁺ approaches to achieve the highest photocurrent density of H-TiO₂ (Figure 12). Although H₂ hydrogenation can achieve high photocurrent densities of H-TiO₂, but high temperature and prolonged treatment time are needed (Table 1). Chemical reduction can prepare H-TiO₂ at room temperature, but the photocurrent density of H-TiO₂ is limited due to chemical residues and resulting charge carrier recombination centers. Although the conventional plasma hydrogenation can efficiently fabricate H-TiO₂, but high temperature or high power is required (Table 1). It is evident that we made a breakthrough in the development of efficient hydrogenation technology (Figure 14, Table 1).

Hydrogenation method	Temperature	RF power	Treatment time, pressure	Photocurrent density of H-TiO2 at 1.23 VRHE	Features and references
High pressure H2	200°C	N/A	5 days, 20 bar	N/A	xHigh pressure; prolonged treatment [11]
High temperature H2	400°C	N/A	1 h, 1 bar	2.4 mA cm^−2	xHigh temperature; prolonged treatment [19]
Low temperature H	265 °C	N/A	20 mins, 1 Pa	2.5 mA cm^−2	√ Low temperature; short treatment [14]
Atomic H via spillover	250 °C	N/A	10 mins, 1 bar	1.5 mA cm^−2	xHigh cost: Noble metals; chemical residues [21]
High temperature H^+	500°C	15 W	3 h	0.2 mA cm^−2	xHigh temperature; prolonged treatment [17]
High temperature H^+	425°C	200 W	1 h	0.9 mA cm^−2	xHigh temperature and power; prolonged treatment [18]
Room temperature H^+	25°C	20 W	5 mins, 1.5 Pa	2.55 mA cm^−2	√ Room temperature; short treatment [15]

Table 1.

A comparison of the low temperature hot wire H hydrogenation and room temperature H⁺ hydrogenation with state-of-the-art hydrogenations.

Figure 14.

Comparison of photocurrent density of black TiO₂ by our advanced hydrogenation approaches with state-of-the-art hydrogenations.

4. Conclusions

From the above discussion of hydrogenation efficiency, we experimentally confirmed that the low-temperature H and room-temperature H⁺ approaches fundamentally break through the intrinsic limitations of poor conductivity and large band gap of TiO₂, and greatly improve its PEC properties. Compared to conventional hydrogenation methods, the two advanced hydrogenation technologies we have invented have the advantages of low cost, uniform hydrogenation, low temperature and low energy consumption, which can realize the optimal control of the physical and chemical properties of oxide photoanode materials, and have broad application prospects. Our findings indicated that two advanced hydrogenation technologies are future directions of development of practical hydrogenation technology and the influence of hydrogenation processing parameters on the structural, optical and PEC properties provides experimental guidance for the rapid synthesis of black TiO₂ with desirable PEC activity, which will advance the significant development of the hydrogenation technology to promote black TiO₂ for PEC hydrogen production from research to practical application.

Acknowledgments

X. Wang acknowledges the support of the research fund by the Institute of Carbon Neutrality, ShanghaiTech University. H. Shen and B. Wang thank the support of the research fund by Yiwu Research Institute, Fudan University.

Conflict of interest

The authors declare no conflict of interest.