Enhancing the photoelectric performance of metal oxide semiconductors by introduction of dislocations†

Haoyu Zhang ^a, Shuang Gao ^a, Hongyang Wang ^a, Fangping Zhuo ^b, Qaisar K. Muhammad ^b, Xufei Fang ^bc, Jürgen Rödel ^b, Till Frömling *^b and Qi Li *^a
aKey Laboratory of Advanced Technologies of Materials (Ministry of Education), School of Materials Science and Engineering, Southwest Jiaotong University, Chengdu 610031, P. R. China. E-mail: qiliuiuc@swjtu.edu.cn
^bDepartment of Materials and Earth Sciences, Technical University of Darmstadt, 64287 Darmstadt, Germany. E-mail: till.froemling@mr.tu-darmstadt.de
^cInstitute for Applied Materials, Karlsruhe Institute of Technology, 76131 Karlsruhe, Germany

First published on 8th August 2024

---

---

1 Introduction

Photocatalysis has been a promising technology for energy and environmental applications since its discovery due to its environmental friendliness, high economic benefit, and long-term sustainability.^1,2 To optimize the photocatalytic performance for further applications, massive research efforts have been made to broaden the photo-response for higher solar energy utilization efficiency.^3–5 Particular efforts were directed to accelerate charge carrier transfer and suppress recombination for preferable separation⁶ and to strengthen the surface adsorption of reactants or regulate the surface pH value to promote reactions.^7–9 Among these approaches, enhancing charge carrier separation and transfer is believed to be the most efficient and direct way to boost photocatalytic activity.¹⁰

To achieve the desired high charge carrier separation efficiency, various material design strategies have been implemented to improve photocatalytic performance, such as noble metal loading, heterojunction construction, defect control, etc.^{4,5,7,8,11–16} Among them, the defect control strategy has been rapidly developed in recent years. Current research in defect engineering focuses on 0D defects (vacancies and element doping, etc.).^14–16 These introduced 0D defects typically generate defect levels in the band gap of a semiconducting photocatalyst, which could serve as trapping centers to capture photogenerated charge carriers.¹⁷ This synergistic effect could enhance the electron–hole (e⁻–h⁺) pair separation efficiency to improve the photocatalytic performance.¹⁸ However, it is difficult to control the distribution of these 0D defects in a photocatalyst due to their highly random nature. An excess of 0D defects, however, tends to cause the reverse effect. The defects become recombination centers, deteriorating the charge carrier separation efficiency.¹⁹

Recently, introducing dislocations (1D defects) in functional ceramics has raised great attention.^20–24 The impact of dislocations on the electronic and ionic conductivity of metal oxides is of particular significance for photocatalysis and may surpass the effect of classical 0D chemical doping.^25–29 Therefore, using the term 1D doping²⁰ to introduce dislocations in these materials is warranted. As this approach does not require extra, potentially scarce elements, 1D doping was later also denoted as self-doping or sustainable doping.²⁷ It could already be illustrated that dislocations have an impact on the photoelectric properties of semiconducting ceramics. Previous experiments on the deformation of ZnS led to an apparent reduction in the optical band gap, which was attributed to the creation of trap states inside the dislocation core.^30–32 Localized photoelectric experiments with conductive atomic force microscopy on strontium titanate demonstrated a significantly higher laser-induced photocurrent in dislocation-rich regions.³³ These observations indicate that introducing dislocations can alter the electronic structure of materials locally, directly influencing the charge carrier separation and transfer in a semiconducting ceramic and subsequently affecting its photocatalytic performance. Ren et al. proposed this in a work where TiO₂ nanowires with different dislocation densities also exhibited varying photo-catalytic efficiency when considering dye degradation in a water solution.³⁴ However, the dislocation introduction was rather unintentional.

In this work, 1D defects of dislocations were introduced into bulk BaTiO₃ and TiO₂ single crystals through a high-temperature deformation process to explore the effect of dislocations on both their photoelectric and photocatalytic performances. The two materials have already been discussed for photocatalysis of various reactions. Single crystal samples were selected as templates for this study as ideal crystalline samples for which the optimal slip plane and load direction can be chosen for extensive plastic deformation.^24,35 The catalyzed reaction discussed in this work is the generation of H₂O₂ from oxygen-enriched water without sacrificial agents. Notably, for rutile TiO₂, it could be illustrated that hydrogen peroxide production dominated, particularly in oxygen-rich environments, instead of complete water splitting.^36,37 With an overall band gap of 3.0 eV, its valence band is at −0.5 V against the reversible hydrogen electrode.³⁸ This enables the oxygen reduction reaction (Fig. 1):

O2 + 2H+ + 2e− → H2O2	(1)

The corresponding oxidation reaction is the production of oxygen from water:

2H2O + 4h+ → O2 + 4H+	(2)

The same reactions could also be induced for BaTiO₃ catalysts.⁴⁰ However, it has not been proven in literature so far. Compared to rutile, the band gap of BaTiO₃ is 3.2 eV and there is a 0.2 eV shift of the valence band to higher energy.³⁸ Therefore, it will be part of the investigation of whether this impacts H₂O₂ production and how the dislocation affects this behavior. Hydrogen peroxide is a heavily used chemical currently synthesized by the industrial anthraquinone process.^39,41 However, this process is not environmentally friendly. Therefore, the H₂O₂ synthesis is an excellent benchmark for identifying changes in catalytic efficiency due to the presence of dislocations while also addressing the industrially relevant topic of sustainable synthesis.⁴²

2 Experimental

2.1. Dislocation imprint and microstructure characterization

BaTiO₃ and TiO₂ single crystals were synthesized by top-seeded solution-growth (TSSG). TiO₂ and BaTiO₃ single crystals were [010]-oriented and [110]-oriented with coordinate systems: X: [001]; Y: [100]; Z: [010] (TiO₂) and X: [10]; Y: [001]; Z: [110] (BaTiO₃) in the exact geometry of 4 × 4 × 8 mm³ (Electro-Optics Technology GmbH, Idar Oberstein, Germany). For BaTiO₃ single crystals, uniaxial compression in the direction of [110] was applied at a temperature of 1150 °C to activate the {100} 〈100〉 slip system with a Schmid factor of 0.5, which induced the generation of dislocations along [001] (see Fig. 2a).⁴³ To this end, samples were heated at a rate of 1 °C min⁻¹ to 1150 °C. After a thermal equilibrium for 30 min, a preload of 1.25 MPa was applied. Compression was conducted at 1150 °C with a loading rate of 0.2 N s⁻¹ (0.0125 MPa s⁻¹) using a load frame (Z010, Zwick/Roell, Ulm, Germany) equipped with a linear variable differential transformer (LVDT) for precise displacement measurement. Plastic deformation in single crystal BaTiO₃ can be achieved to several%.³⁶ A plastic deformation value of 2% was chosen to safely reach the regime of steady-state creep with equilibrium dislocation density but limited sample barrelling simultaneously.⁴⁴ When the deformation reached this value, the sample was carefully unloaded with an unloading rate of 0.5 N s⁻¹ (0.031 MPa s⁻¹) to avoid cracking. Afterwards, the sample was cooled to room temperature at 1 °C min⁻¹ under uniaxial compressive stress of 1.25 MPa. The orientation of the as-prepared samples was confirmed using Laue back-reflection (1001 Model, Huber, Rimsting, Germany). The surfaces of the (001) and (110) cut samples were then finely polished to a 0.5–1.0 mm thickness. The TiO₂ single crystals were deformed in a similar process. However, in this case, the crystals were deformed at 1050 °C along the [010] directions, which activates the {110} 〈110〉 slip system. Edge dislocations mostly occurred along [100] (see Fig. 2b).²⁷ Prior work had demonstrated that high-temperature compression is able to increase the dislocation density by ∼2 to 3 orders of magnitudes (from ca. 1011 m⁻² to 1013 m⁻²) for both materials.^27,36

Bright-field transmission electron microscopy (TEM) and scanning transmission electron microscopy (STEM) observations were conducted on a JEM-2100F TEM (JEOL, Tokyo, Japan). (001) and (110) cut samples were extracted with a 300 μm thickness and then polished using a Multi-Prep polishing system (Allied High Tech Products Inc., Compton, CA, USA) down to 20 μm. The as-polished thin slices were annealed at 200 °C for 30 min with a heating/cooling rate of 1 °C min⁻¹ to release the stress from the polishing procedure. The annealed TEM slices were mounted on supporting molybdenum grids of 100 mesh (Plano, Wetzlar, Germany) and thinned by Ar ions using a dual ion milling system (Gatan, Pleasanton, CA, USA) into electron transparency.

2.2. Light absorbance, electrochemistry, photoluminescence, and surface photovoltage measurements

A reflection method was adopted to quantify the light absorbance of the single crystal samples by an ultraviolet-visible spectrophotometer (UV-3600 Plus, Shimadzu Co., Ltd., Kyoto, Japan). The single crystal was embedded in the center of the BaSO₄ powder bed in the sample holder, and another pure BaSO₄ powder bed was used to determine the baseline. Electrochemical measurements were conducted in a neutral electrolyte (0.2 M Na₂SO₄ solution) at room temperature with a three-electrode configuration, in which a platinum wire served as the counter electrode and an Ag/AgCl electrode served as the reference electrode, and an electrochemical workstation (CHI760E, Shanghai Chenhua Instrument Ltd., P. R. China) was used. Single crystal samples with and without dislocations were mounted on electrode holders to serve as the working electrode. For the photocurrent measurement, an Xe arc lamp of 300 W (Beijing Perfect Light Technology Co., Ltd., China) was utilized to provide the simulated solar light with a range of wavelengths from ∼320 nm to ∼780 nm for several on–off illumination cycles with an interval of 20 seconds. Mott–Schottky (M–S) plots were measured at 1 kHz and 3 kHz, respectively, to ensure the data accuracy. Photoluminescence (PL) spectra were quantified by a fluorescence spectrometer (FLS1000, Edinburgh Instruments Co., Ltd., UK). The surface photovoltage (SPV) spectra were obtained on a self-built system from Taiyuan University of Technology. The system consisted of a 500 W xenon lamp (CHF-XM500W, Beijing Perfect Light Technology Co., Ltd., China), a monochromator (Omni-λ300, Zolix, China), a lock-in amplifier (SR830, Stanford Research Systems, USA), optical chopper (SR540, Stanford Research Systems, USA), a photovoltaic cell, and a data processing computer.

2.3. Photocatalytic H₂O₂ production experiments

Photocatalytic H₂O₂ production was quantified in a sealed glass vessel. Throughout the whole process, a single crystal was placed at the bottom of the glass reactor. Then, 25 mL of pure deionized (DI) water was added to the reactor without the addition of any sacrificial agents. Subsequently, either dry air or pure oxygen was bubbled into DI water for 30 min to expel dissolved impurity gas. Finally, the Xe arc lamp was activated for solar light simulation. Continuous mechanical stirring was applied to the DI water, and a circulating cooling water system kept the reaction at room temperature. As the reaction progressed, 3 mL reaction solution with the H₂O₂ product was sampled every 10 min for subsequent chromogenic experiment.

The horseradish peroxidase (POD) chromogenic method was used to evaluate the photocatalytic H₂O₂ production performance. 10 mg POD was dissolved in 10 mL DI water. The obtained POD solution and 0.1 g N,N-diethyl-p-phenylenediamine (DPD, Sinopharm Chemical Reagent Co., Ltd., Shanghai, P. R. China) were added into two 10 mL sulfuric acid solutions (0.05 M) with continuous mechanical stirring, respectively. A phosphate buffer solution was prepared by mixing 0.5 M di-potassium hydrogen phosphate with 0.5 M potassium dihydrogen phosphate. Subsequently, 3 mL sampled reaction solution was mixed with 0.3 mL phosphate buffer solution. 30 μL sulphated DPD solution and 30 μL sulphated POD solution were then added to the mixed phosphoric acid buffer solution. The color was gradually observed in the mixed solution for a duration of 20 seconds. The light absorbance of the chromogenic solution was quantified by an ultraviolet-visible spectrophotometer (UV-3600 Plus, Shimadzu Co., Ltd., Kyoto, Japan) with a testing wavelength of 450–650 nm and an absorption characteristic peak at 551 nm.

3 Results and discussion

3.1. Introduction and distribution of dislocations in BaTiO₃ and TiO₂ single crystals

In contrast to metals, ceramics have been commonly considered as brittle without any dislocation activity. However, especially uniaxial deformation of ceramic single crystals at high temperatures allows for the activation of defined slip systems.^{26,27,36,43,45–47}Fig. 2a and b feature the related schematics of tetragonal BaTiO₃ and rutile TiO₂ samples used in this work before and after deformation, respectively. Uniaxial deformation of BaTiO₃ was conducted at 1150 °C while a slightly lower temperature of 1050 °C was sufficient to plastically deform TiO₂.²⁶ The high-temperature compression activated the {100} 〈001〉 and {110} 〈110〉 slip systems, which generated dislocations primarily along [100] and [001] in BaTiO₃ and TiO₂ single crystals, respectively.^26,36Fig. 2c–e feature the bright-field STEM images of the imprinted dislocations in BaTiO₃ crystals viewed on (010), (110), and (001) planes, which facilitates the understanding of the spatial distribution of the dislocations. Clearly, those dislocations extending along the [001] direction in BaTiO₃ were parallel to (010) and (110) planes and perpendicular to the (001) plane. Fig. 2f highlights the imprinted dislocations in TiO₂ crystals lying parallel on the (011) plane, which are the reported slip planes for rutile TiO₂.²⁷

3.2. Dislocation effects on optical properties and energy band structures of BaTiO₃ and TiO₂ single crystals

The optical properties of samples were quantified by diffuse reflectance with wavelengths ranging from 300 nm to 800 nm. Their light absorbance spectra could be approximated from their measured diffuse reflectance by the Kubelka–Munk (K–M) function as given by eqn (3):

F(R) = (1 − R)2/2R	(3)

where F(R) refers to the K–M function, and R is the diffuse reflectance.⁴⁸ Tauc plots ((F(R)hν)^1/2vs. hν) were converted from their light absorbance data, and linear fitting at the absorption edge of the plots revealed the sample's band gap values. Fig. 3a and d depict the light absorbance spectra of BaTiO₃ and TiO₂ crystals with and without dislocations, respectively. The results reveal that the plastically deformed, dislocation-rich BaTiO₃ and TiO₂ samples did not exhibit a drastic impact on their light absorption ranges, and their light absorbance edges were at ∼498 nm and ∼438 nm for BaTiO₃ and TiO₂ crystals, respectively, consistent with previous reports.^49–52 It is well known that single crystal samples and powder samples may have different light absorbance behavior, and single crystal samples with different orientations may also have different light absorbance behavior. Fig. 3c and d display Tauc plots of BaTiO₃ and TiO₂ crystals with and without engineered dislocations constructed from their light absorbance data. The linear fits of the plots revealed that band gap values of BaTiO₃ single crystal without and with uniaxial deformation were determined at ∼2.39 eV and ∼2.43 eV, respectively. In comparison, that of TiO₂ single crystal without and with dislocation engineering were determined at ∼2.83 eV and ∼2.87 eV, respectively. The very similar band gap values demonstrate that dislocations induced by uniaxial deformation do not significantly impact the macroscopic apparent band gap. This makes the materials TiO₂ and BaTiO₃ behave differently as compared to ZnS.^30–32 Even though a change in the electronic properties at dislocations could be established for the two materials in this work, a modification of the electronic structure cannot be measured globally by light absorption spectroscopy. A reduction of the band gap could also be beneficial for using energy from light in the visible range for photocatalysis.³⁹ However, there is always a trade-off between a catalyst's reduction potential and light absorption range. Considering light absorbance, the samples with dislocations generally had higher light absorbance intensities compared to their counterparts without dislocations, which might be induced by the enhancement of light scattering and subsequent absorption by their imprinted high-density dislocations.⁵³

To elucidate the dislocation impact on energy band structures of BaTiO₃ and TiO₂ single crystals, M–S measurements were conducted at 1 kHz and 3 kHz, respectively, as shown in Fig. 3b, c, e and f. The flat-band potentials of BaTiO₃ and TiO₂ single crystals could be determined from the intercepts of the linearly fitted M–S curves with x-axis. Notably, the M–S analysis revealed that the introduction of dislocations did not alter their flat-band potentials. Consequently, the dislocation-containing samples and their pristine counterparts exhibited identical conduction band potentials. For n-type semiconductors, the conduction band potential (E_CB) is typically 0.1–0.3 V more negative than the flat-band potential (E_FB).^54,55 Thus, their E_CB values could be calculated combined with the Nernst equation by eqn (4):⁵⁶

ECB(NHE) = EFB(vs. Ag/AgCl) + 0.1975 − 0.2	(4)

The E_CB(NHE) values of BaTiO₃ and TiO₂ single crystals were determined at ∼−0.77 V and ∼−0.65 V, respectively, no matter dislocations were induced or not. By combining their E_CB(NHE) values with their band gap values, their E_VB(NHE) values could be obtained and their schematic energy band structures were displayed in Fig. 3g. For BaTiO₃ single crystals, the E_VB(NHE) values were calculated at ∼1.62 V and ∼1.66 V for the dislocation and reference samples, respectively. For TiO₂ single crystals, the E_VB(NHE) values were calculated at ∼2.28 V and ∼2.31 V for the dislocation and reference samples, respectively.

Thus, these results indicated that the introduction of dislocations only induced a slight narrowing of their band gaps and subsequently a slight reduction of their oxidation potentials. This effect can be attributed to the compressive stress exerted on the valence band by the dislocations.^57,58 Notably, this band gap modification occurs without significant alteration of the materials' light absorption.

3.3. Photogenerated charge carrier separation and transfer in deformed BaTiO₃ and TiO₂

Photocurrent measurements with light on–off cycles were conducted on a three-electrode configuration, as depicted schematically in Fig. 4a. This setup only assesses a current in case a photoelectrocatalytic reaction is induced. Therefore, it is possible to extract the impact of dislocations on photocatalytic properties. Under illumination, photo-excited charge carriers migrate directionally to the photocatalyst surface to generate a photocurrent, which decays rapidly when the illumination is switched off. At this point, the focus is not on the respective induced reaction but on the fact that a reaction causes a photocurrent in general. Fig. 4b and c contrast obtained photocurrents of BaTiO₃ and TiO₂ single crystals with and without engineered dislocations, respectively. It could be illustrated that the photocurrent values of both samples were higher by a factor of 3 to 5 for samples with high dislocation density. Thus, dislocations induced by uniaxial deformation significantly affected the materials' photogenerated charge carrier separation. Photogenerated charge carrier separation and transfer properties of BaTiO₃ and TiO₂ were further investigated through photoluminescence spectroscopy (PL) and surface photovoltage spectroscopy (SPV) analysis. Fig. 5a and b contrast PL curves of BaTiO₃ and TiO₂ single crystals with and without engineered dislocations, respectively. For both BaTiO₃ and TiO₂ single crystals, their measured PL intensities decreased when dislocations were introduced, revealing a suppression of photogenerated charge carrier recombination by dislocations.⁵⁹Fig. 5c and d reveal SPV curves of BaTiO₃ and TiO₂ single crystals with and without dislocation engineering, respectively. For both BaTiO₃ and TiO₂ single crystals, the onset of the photo response was generally similar, independent of dislocation density. This observation is consistent with the fact that dislocations did not have a noticeable effect on the absorption edge in Fig. 3. However, their SPV signal intensities were quite different when dislocations were present. In their pristine forms, BaTiO₃ and TiO₂ single crystals used in this work are n-type conductors. Single crystals can obtain impurity doping by foreign elements, giving them a certain preferred type of conductivity. Especially for TiO₂, this has been evaluated in depth.^26,27 SPV is a measurement that is sensitive to the minority charge carriers. In n-type semiconductors, photogenerated holes thus transfer to surfaces of BaTiO₃ and TiO₂ single crystals, causing the measured positive SPV signals.⁶⁰ The SPV signal intensity of samples with high dislocation density decreases compared to the ones with low density, which suggests that more photogenerated electrons move to their surfaces to reduce positive SPV signal.

Thus, SPV results provide strong evidence that dislocations not only enhanced the photogenerated charge carrier separation and transfer efficiencies but also helped to generate a directional transfer of photogenerated electrons to surfaces of BaTiO₃ or TiO₂ single crystals. As discovered in previous reports by us and other research groups, positively charged cores of dislocations in n-type semiconductors are compensated by a high electron concentration in the space charge region.^26,27,61,62 The field in that region certainly contributes to increased exciton separation and lifetime. For bundled dislocations, overlapping of negative space charge zones could form percolating paths, which were found to be responsible for enhanced electronic conductivity along dislocations.^27,63 Thus, the dislocations in n-type BaTiO₃ or TiO₂ single-crystal photocatalysts could serve as fast electron pathways for photogenerated electrons. Separating e⁻–h⁺ pairs and increasing their respective transport is generally seen as one of the most essential strategies to improve the photocatalytic performance of semiconducting ceramics.^39,41

3.4. Dislocation-enhanced photocatalytic H₂O₂ synthesis

Fig. 6 demonstrates the impact of dislocations on the photocatalytic H₂O₂ generation of BaTiO₃ and TiO₂ single-crystal samples. Different reaction conditions without any sacrificial agents were considered for both low and high dislocation density. Fig. 6a and b contrast the photocatalytic H₂O₂ production of BaTiO₃ single crystal samples with either air or O₂ saturation in water under simulated solar illumination. It could be found that the photocatalytic H₂O₂ production of the BaTiO₃ sample increased significantly due to dislocations, regardless of whether air or O₂ bubbling was utilized. For example, the H₂O₂ concentration in the air-bubbled pure de-ionized (DI) water using the dislocation-rich BaTiO₃ reached ∼6.6 μmol L⁻¹ after just 30 min treatment. In comparison, the photocatalytic reaction with the pristine single crystal sample without mechanically induced dislocations resulted in just ∼3.5 μmol L⁻¹. When O₂ was introduced instead of air, the H₂O₂ production efficiency was further improved because the O₂ reduction reaction with two photogenerated electrons depends on oxygen concentration. For example, the H₂O₂ generation in the O₂-enriched pure DI-water catalyzed by the BaTiO₃ single crystal sample with dislocations reached ∼22.2 μmol L⁻¹, more than a factor of two higher as compared to that with the pristine sample (∼9.7 μmol L⁻¹). To test the reproducibility of the processes, Fig. 6c summarizes the photocatalytic H₂O₂ production performances of the dislocation-rich BaTiO₃ for three consecutive runs. It demonstrates that the photocatalytic H₂O₂ production performances are stable for multiple uses. Fig. 6d illustrates the H₂O₂ production catalyzed with TiO₂ samples in O₂-enriched pure water. It also demonstrates that dislocations in TiO₂ single crystals largely enhance its photocatalytic H₂O₂ production. Fig. 6e underscores that the behavior is stable for at least three consecutive runs. Fig. S1 and S2† exhibit the comparative XRD and UV-vis analysis results of dislocation-rich BaTiO₃ and TiO₂ single crystal samples before and after three consecutive runs, respectively. These results provide compelling evidence for their structural and optical stabilities throughout the experimental process, which subsequently result in their photocatalytic H₂O₂ production performance.

From Fig. 6, it is generally possible to derive that the photocatalytic performance of TiO₂ is higher than that of BaTiO₃ and that the changes in the photoelectric properties due to dislocations evaluated in Section 2.3 resulted in a further increase in performance. This indicates a less optimal band gap position of BaTiO₃ with respect to H₂O₂ synthesis, as hypothesized in the introduction. The amount of produced H₂O₂ with BaTiO₃ is, nevertheless, remarkable. The potential of dislocation-enhanced catalysis, even if only demonstrated on single crystals, can be contrasted to the H₂O₂ yield with the current literature. For the dislocation-rich TiO₂ sample, the yield amounts to 80.625 mmol h⁻¹ m⁻² in pure water saturated with oxygen, while the one for BaTiO₃ is lower with 69.375 mmol h⁻¹ m⁻² in pure water saturated with oxygen. These results are very promising as they are much better than other optimized catalysts (some also with defect engineering treatment) in pure water and even with sacrificial agents, as summarized in Table 1.^64–74

---

Apart from the higher photocurrent for the oxygen photoreduction due to the higher exciton lifetime, the change in surface structure surrounding the dislocations could also have impacted the reaction. It is known that relatively stable Ti–OOH species can develop and inhibit further reaction to hydrogen peroxide.⁷⁵ A disruption of surface structure, such as oxygen vacancies, was already demonstrated to support the hydrogen peroxide formation.⁷⁶

It will be of high interest to elucidate the impact of doping and initial defect chemistry on the photocatalytic properties. These factors affect space charges and, thus, will modify charge separation properties. Additionally, there will be a dependence on the type and mesoscopic structure of dislocations and on the structure of the material itself. Further current projects focus on imprinting surfaces of polycrystalline ceramics, as applicable for mass production, to transfer this new mechanism into industrial processes. The use of metal oxide semiconductors like TiO₂ and BaTiO₃ for the production of hydrogen peroxide may be even more feasible when combining the introduction of dislocations with other strategies to enhance the photocatalytic activity (e.g., formation of heterojunctions).^41,42

4 Conclusions

Dislocations primarily along 〈001〉 and 〈100〉 orientations were successfully introduced into BaTiO₃ and TiO₂ single crystals by uniaxial compression at enhanced temperature. It was found that their light absorbance properties were not significantly affected by these dislocations. At the same time, their charge carrier separation/transfer behaviors were enhanced to reduce the recombination of photogenerated electron–hole pairs effectively. Primarily, the dislocations served as fast pathways for photogenerated electrons from the photocatalyst bulk to the surface. Thus, dislocations essentially elevated the amount of photogenerated electrons on the photocatalyst surface, which is beneficial for electron-dominated photocatalytic reactions. Hence, these dislocation-rich BaTiO₃ and TiO₂ single crystal samples demonstrated enhanced photocatalytic H₂O₂ synthesis over their counterparts without dislocations. The amount of produced H₂O₂ even exceed multiple approaches from literature despite no facilitation of sacrificial agents. This work opens a novel approach to utilize 1D defects in bulk photocatalysts to accelerate the photocatalytic reactions, which could be readily applied to various photocatalytic applications.

Data availability

The manuscript provides most of the relevant data, which mostly refer to the respective procedures or are TEM images. The Tauc analysis and the surface photovoltage measurements can be made available upon request.

Conflicts of interest

S. Gao, J. Rödel, T. Frömling, and Q. Li have filed a patent application under PCT/CN2023/142482.

Acknowledgements

We are grateful to the National Natural Science Foundation of China (Grant No. 52272125) and the Deutsche Forschungsgemeinschaft (DFG, Project No. 414179371) for funding. We would also like to thank the Analysis and Testing Center of the Southwest Jiaotong University for its assistance in material characterization.

References

1. X. Yang and D. Wang, ACS Appl. Energy Mater., 2018, 1, 6657–6693 CrossRef CAS .
2. D. Ravelli, D. Dondi, M. Fagnoni and A. Albini, Chem. Soc. Rev., 2009, 38, 1999–2011 RSC .
3. M. Cargnello and B. T. Diroll, Nanoscale, 2014, 6, 97–105 RSC .
4. Q. Xu, L. Zhang, B. Cheng, J. Fan and J. Yu, Chem, 2020, 6, 1543–1559 CAS .
5. H.-L. Wang, L.-S. Zhang, Z.-G. Chen, J.-Q. Hu, S.-J. Li, Z.-H. Wang, J.-S. Liu and X.-C. Wang, Chem. Soc. Rev., 2014, 43, 5234–5244 RSC .
6. F. Chen, T.-Y. Ma, T.-R. Zhang, Y.-H. Zhang and H.-W. Huang, Adv. Mater., 2021, 33, 2005256 CrossRef CAS .
7. H. Park, H. Kim, G. Moon and W. Choi, Energy Environ. Sci., 2016, 9, 411–433 RSC .
8. C. Boerigter, R. Campana, M. Morabito and S. Linic, Nat. Commun., 2016, 7, 10545 CrossRef CAS .
9. J. Fu, J. Yu, C. Jiang and B. Cheng, Adv. Energy Mater., 2018, 8, 1701503 CrossRef .
10. M. R. Al-Mamun, S. Kader, M. S. Islam and M. Z. H. Khan, J. Environ. Chem. Eng., 2019, 7, 103248 CrossRef CAS .
11. K. Li, S. Zhang, Y. Li, J. Fan and K. Lv, Chin. J. Catal., 2021, 42, 3–14 CrossRef CAS .
12. Y. Chen, Y.-N. Wang, W.-Z. Li, Q. Yang, Q.-D. Hou, L.-H. Wei, L. Liu, F. Huang and M.-T. Ju, Appl. Catal., B, 2017, 210, 352–367 CrossRef CAS .
13. S. Linic, P. Christopher and D. B. Ingram, Nat. Mater., 2011, 10, 911–921 CrossRef CAS .
14. J. Di, J.-X. Xia, M. F. Chisholm, J. Zhong, C. Chen, X.-Z. Cao, F. Dong, Z. Chi, H.-L. Chen, Y.-X. Weng, J. Xiong, S.-Z. Yang, H.-M. Li, Z. Liu and S. Dai, Adv. Mater., 2019, 31, 1807576 CrossRef .
15. A. Kumar and V. Krishnan, Adv. Funct. Mater., 2021, 31, 2009807 CrossRef CAS .
16. Y.-Q. He, Q. Lei, C.-G. Li, Y. Han, Z. Shi and S.-H. Feng, Mater. Today, 2021, 50, 358–384 CrossRef CAS .
17. J. Yin, H.-Z. Yang, K.-P. Song, A. M. El-Zohry, Y. Han, O. M. Bakr, J.-L. Brédas and O. F. Mohammed, J. Phys. Chem. Lett., 2018, 9, 5490–5495 CrossRef CAS PubMed .
18. H.-N. Shi, S.-R. Hu, S. Long, J.-G. Hou, W.-J. Ni, C.-S. Song, K. G. G. Gurzadyan and X.-W. Guo, Appl. Catal., B, 2019, 245, 760–769 CrossRef CAS .
19. B. Dong, T. Liu, C. Li and F. Zhang, Chin. Chem. Lett., 2018, 29, 671–680 CrossRef CAS .
20. K. K. Adepalli, M. Kelsch, R. Merkle and J. Maier, Phys. Chem. Chem. Phys., 2014, 16, 4942–4951 RSC .
21. C. Rodenbücher, S. Menzel, D. Wrana, T. Gensch, C. Korte, F. Krok and K. Szot, Sci. Rep., 2019, 9, 2502 CrossRef .
22. K. Szot, C. Rodenbücher, G. Bihlmayer, W. Speier, R. Ishikawa, N. Shibata and Y. Ikuhara, Crystals, 2018, 8, 241 CrossRef .
23. A. Nakamura, X.-F. Fang, A. Matsubara, E. Tochigi, Y. Oshima, T. Saito, T. Yokoi, Y. Ikuhara and K. Matsunaga, Nano Lett., 2021, 21, 1962–1967 CrossRef CAS .
24. L. Porz, T. Frömling, A. Nakamura, N. Li, R. Maruyama, K. Matsunaga, P. Gao, H. Simons, C. Dietz, M. Rohnke, J. Janek and J. Rödel, ACS Nano, 2021, 15, 9355–9367 CrossRef CAS PubMed .
25. K. Szot, W. Speier, G. Bihlmayer and R. Waser, Nat. Mater., 2006, 5, 312–320 CrossRef CAS .
26. Q. K. Muhammad, H. Bishara, L. Porz, C. Dietz, M. Ghidelli, G. Dehm and T. Frömling, Mater. Today Nano, 2021, 100171 Search PubMed .
27. Q. K. Muhammad, L. Porz, A. Nakamura, K. Matsunaga, M. Rohnke, J. Janek, J. Rödel and T. Frömling, Nano Energy, 2021, 85, 105944 CrossRef CAS .
28. Q. K. Muhammad, M. Scherer, A. K. Opitz, S. Taibl, C. Boehme, M. Rohnke, J. Janek, S. Gao, J. Fleig and T. Frömling, ACS Nano, 2022, 16, 16655–16667 CrossRef CAS .
29. J. M. Börgers, J. Kler, K. Ran, E. Larenz, T. E. Weirich, R. Dittmann and R. A. D. Souz, Adv. Funct. Mater., 2021, 31, 2105647 CrossRef .
30. Y. Oshima, A. Nakamura and K. Matsunaga, Science, 2018, 360, 772–774 CrossRef CAS .
31. Y. Ogura, A. Nakamura, T. Kameyama, Y. Kurokawa, E. Tochigi, N. Shibata, T. Torimoto, S. Hoshino, T. Yokoi and K. Matsunaga, J. Am. Ceram. Soc., 2024, 107, 2040–2047 CrossRef CAS .
32. K. Matsunaga, S. Hoshino, M. Ukita, Y. Oshima, T. Yokoi and A. Nakamura, Acta Mater., 2020, 195, 645–653 CrossRef CAS .
33. M. Kissel, L. Porz, T. Frömling, A. Nakamura, J. Rödel and M. Alexe, Adv. Mater., 2022, 34, 2203032 CrossRef CAS PubMed .
34. P. Ren, M. Song, J. Lee, J. Zheng, Z.-X. Lu, M. Engelhard, X.-C. Yang, X.-L. Li, D. Kisailus and D.-S. Li, Adv. Mater. Interfaces, 2019, 6(17), 1901121 CrossRef .
35. M. Höfling, L. Porz, M. Scherer, S. Gao, F.-P. Zhuo, D. Isaia and J. Rödel, J. Mater. Res., 2022, 37, 737–746 CrossRef .
36. Y. Guo, X. Tong and N. Yang, Nano-Micro Lett., 2023, 15, 77 CrossRef CAS .
37. X. Zeng, Y. Liu, X. Hu and X. Zhang, Green Chem., 2021, 23, 1466–1494 RSC .
38. S. Li, F. Chen, R. Schafranek, T. J. M. Bayer, K. Rachut, A. Fuchs, S. Siol, M. Weidner, M. Hohmann, V. Pfeifer, J. Morasch, C. Ghinea, E. Arveux, R. Günzler, J. Gassmann, C. Körber, Y. Gassenbauer, F. Säuberlich, G. V. Rao, S. Payan, M. Maglione, C. Chirila, L. Pintilie, L.-C. Jia, K. Ellmer, M. Naderer, K. Reichmann, U. Böttger, S. Schmelzer, R. C. Frunza, H. Uršič, B. Malič, W.-B. Wu, P. Erhart and A. Klein, Phys. Status Solidi RRL, 2014, 8, 571–576 CrossRef CAS .
39. S. Wu and X. Quan, ACS ES&T Eng., 2022, 2, 1068–1079 Search PubMed .
40. K. Wang, M.-Q. Zhang, D.-G. Li, L.-H. Liu, Z.-P. Shao, X.-Y. Li, H. Arandiyan and S.-M. Liu, Nano Energy, 2022, 8, 107251 CrossRef .
41. W. Yu, C. Hu, L.-Q. Bai, N. Tian, Y.-H. Zhang and H.-W. Huang, Nano Energy, 2022, 104, 107906 CrossRef CAS .
42. X. Zeng, Y. Liu, X. Hu and X. Zhang, Green Chem., 2021, 23, 1466–1494 RSC .
43. M. Höfling, X.-D. Zhou, L. M. Riemer, E. Bruder, B.-Z. Liu, L. Zhou, P. B. Groszewicz, F.-P. Zhuo, B.-X. Xu, K. Durst, X.-L. Tan, D. Damjanovic, J. Koruza and J. Rödel, Science, 2021, 372, 961–964 CrossRef .
44. W. Blum, P. Eisenlohr and F. Breutinger, Metall. Mater. Trans. A, 2002, 33, 291–303 CrossRef .
45. W. Sigle, C. Sarbu, D. Brunner and M. Rühle, Philos. Mag., 2006, 86, 4809–4821 CrossRef CAS .
46. D. Brunner, Acta Mater., 2006, 54, 4999–5011 CrossRef CAS .
47. S. Stich, K. Ding, Q. K. Muhammad, L. Porz, C. Minnert, W. Rheinheimer, K. Durst, J. Rödel, T. Frömling and X.-F. Fang, J. Am. Ceram. Soc., 2022, 105, 1318–1329 CrossRef CAS .
48. J. Tauc, R. Grigorovici and A. Vancu, Phys. Status Solidi B, 1966, 15, 627–637 CrossRef CAS .
49. S. Wang, S.-N. Tang, H.-J. Gao, X.-Y. Chen, H. Liu, C. Yu, Z.-J. Yin, X.-X. Zhao, X.-D. Pan and H. Yang, Opt. Mater., 2021, 118, 111273 CrossRef CAS .
50. A. U. L. S. Setyadi, Y. Iriani and F. Nurosyid, Mater. Sci. Eng., 2018, 333, 012035 Search PubMed .
51. Y. Ishibai, J. Sato, T. Nishikawa and S. Miyagishi, Appl. Catal., B, 2008, 79, 117–121 CrossRef CAS .
52. X.-W. Cheng, X.-J. Yu and Z.-P. Xing, Appl. Surf. Sci., 2012, 258, 3244–3248 CrossRef CAS .
53. L.-J. Liu and C.-T. Chen, J. Cryst. Growth, 2006, 292, 472–475 CrossRef CAS .
54. H.-Y. Zhang, S. Gao, H.-T. Guan, W.-Y. Yang and Q. Li, J. Adv. Ceram., 2023, 12(8), 1641–1654 CrossRef CAS .
55. L.-Z. Lu, H.-Y. Zhang, J.-H. Wang, H.-L. Wang, Q. Chen, J.-B. Xue, Q.-Q. Shen, C.-L. Zhou, J.-K. Shang and Q. Li, Appl. Catal., B, 2024, 356, 124240 CrossRef CAS .
56. D. S. Ginley and M. A. Butler, J. Electrochem. Soc., 1978, 125, 1968 CrossRef CAS .
57. P. Ren, M. Song, J. Lee, J. Zheng, Z.-X. Lu, M. Engelhard, X.-C. Yang, X.-L. Li, D. Kisailus and D.-S. Li, Adv. Mater. Interfaces, 2019, 6, 1901121 CrossRef .
58. S. Bai, N. Zhang, C. Gao and Y.-J. Xiong, Nano Energy, 2018, 53, 296–336 CrossRef CAS .
59. F. Deschler, E. D. Como, T. Limmer, R. Tautz, T. Godde, M. Bayer, E. V. Hauff, S. Yilmaz, S. Allard, U. Scherf and J. Feldmann, Phys. Rev. Lett., 2011, 107, 127402 CrossRef PubMed .
60. L. Kronik and Y. Shapira, Surf. Sci. Rep., 1999, 37, 1 CrossRef CAS .
61. D.-C. Sun, W.-Y. Yang, L. Zhou, W.-Z. Sun, Q. Li and J.-K. Shang, Appl. Catal., B, 2016, 182, 85–93 CrossRef CAS .
62. V. Duzhko, V. Y. Timoshenko, F. Koch and T. Dittrich, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 64, 075204 CrossRef .
63. K. K. Adepalli, Adv. Funct. Mater., 2013, 23, 1798–1806 CrossRef CAS .
64. M. C. Blount and J. L. Falconer, Appl. Catal., B, 2002, 39, 39–50 CrossRef CAS .
65. X. Ma and H.-F. Cheng, Chem. Eng. J., 2022, 429, 132373 CrossRef CAS .
66. N. Lu, N. Liu, Y. Hui, K.-F. Shang, N. Jiang, J. Li and Y. Wu, Chemosphere, 2020, 241, 124927 CrossRef CAS PubMed .
67. Q. C. Chen, C.-J. Lu, B.-Y. Ping, G.-Y. Li, J.-Y. Chen, Z.-M. Sun, Y.-J. Zhang, Q.-S. Ruan and L. Tao, Appl. Catal., B, 2023, 324, 122216 CrossRef CAS .
68. R. Boro, N. Kim, J.-S. Kim, R. Paul, Y. Nailwal, Y. Choi, D.-H. Seo, J. Mondal and J. Ryu, J. Colloid Interface Sci., 2023, 652, 1784–1792 CrossRef PubMed .
69. X.-Y. Li, P.-H. Li, Y.-J. Li, H.-T. Liu, Z.-Z. Yang, Y.-Y. Chen and X.-Y. Liao, J. Environ. Chem. Eng., 2023, 11, 110594 CrossRef CAS .
70. Y. Hong, Y.-J. Cho, E.-M. Go, P. Sharma, H. Cho, B. Lee, S. Lee, M. S. O. Park, M. Ko, S.-K. Kwak, C. Yang and J.-W. Jang, Chem. Eng. J., 2021, 418, 129346 CrossRef CAS .
71. W. Gan, X.-C. Fu, J.-C. Jin, J. Guo, M. Zhang, R.-X. Chen, C.-S. Ding, Y. Q. Lu, J. R. Li and Z.-Q. Sun, J. Colloid Interface Sci., 2024, 653, 1028–1039 CrossRef CAS PubMed .
72. C. Zhu, M.-M. Zhu, Y. Sun, Y.-J. Zhou, J. Gao, H. Huang, Y. Liu and Z.-H. Kang, ACS Appl. Energy Mater., 2019, 2, 8737–8746 CrossRef CAS .
73. J. Luo, Y.-N. Liu, C.-Z. Fan, L. Tang, S.-J. Yang, M.-L. Liu, M.-E. Wang, C.-Y. Feng, X.-L. Ouyang, L.-L. Wang, L. Xu, J.-J. Wang and M. Yan, ACS Catal., 2021, 11, 11440–11450 CrossRef CAS .
74. S.-N. Li, G.-H. Dong, R. Hailili, L.-P. Yang, Y.-X. Li, F. Wang, Y.-B. Zeng and C.-Y. Wang, Appl. Catal., B, 2016, 190, 26–35 CrossRef CAS .
75. M. Teranishi, S. Naya and H. Tada, J. Am. Chem. Soc., 2010, 132, 7850–7851 CrossRef CAS .
76. K. Dong, J. Liang, Y.-Y. Wang, Q. Li, D.-W. Ma and X.-P. Sun, Chem Catal., 2021, 1, 1437–1448 CrossRef CAS .

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ta03786c

This journal is © The Royal Society of Chemistry 2024