3D/2D Heterojunction of CeO2/Ultrathin MXene Nanosheets for Photocatalytic Hydrogen Production.

Two-dimensional (2D) nanomaterials benefit from the high specific surface area, unique surface properties, and quantum size effects, which have attracted widespread scientific attention. MXenes add many members to the 2D material family, mainly metal conductors, most of which are dielectrics, semiconductors, or semimetals. With excellent electron mobility, beneficial to electron-hole separation, and large functional groups that can be tightly coupled with other materials, MXenes have broad application prospects in photocatalysis. Meanwhile, the application of CeO2-based materials in organic catalysis, photocatalytic water splitting, and photodegradation of organic pollutants has been extensively explored, and studies have found that CeO2-based materials show good photocatalytic performance. In view of this, we synthesized regular octahedral CeO2 with a homojunction in one step by a hydrothermal method and compounded it with ultrathin 2D material MXene, which exhibited fast carrier migration efficiency and a good interfacial effect, making the material show excellent photocatalytic activity. The results showed that the photocatalytic H2 evolution performance of the MXene/CeO2 heterojunction was significantly improved. In this study, a low-cost catalyst with high photocatalytic activity was prepared, presenting a new research idea for achieving a combined 3D/2D photocatalytic system.

Introduction

The massive global energy consumption has led to severe environmental problems. People urgently need a clean energy source to solve the increasingly severe ecological problems. Photocatalytic hydrogen production is one of the ideal ways to harness solar energy to solve the growing energy crisis and environmental problems.[1,2] To date, using semiconductor photocatalysts to decompose water to produce hydrogen is one of the most attractive methods of solar energy conversion. The critical points of the practical application of photocatalytic H2 evolution technology are high efficiency, high stability, and low cost.[3−5] However, one of the modification methods to improve the hydrogen production performance of the catalyst system is to synthesize various noble metals, such as Pt, Pd, Au, Ag, etc., on the surface. However, the cost consumed in the actual production of a photocatalyst is exceptionally high, and at the same time, heavy metals pollute the environment to a certain extent. Therefore, the exploration of non-precious metal catalysts is significant for achieving improved photocatalyst performance to split water to produce hydrogen.[6,7]

As the most abundant rare earth oxide, n-type cerium oxide (CeO2) is widely used in high-efficiency photocatalysts because of its ease of converting Ce4+ to Ce3+. Cerium dioxide (CeO2) is a widely utilized catalyst material because of its excellent photocatalytic performance, simple synthesis method, and environmental friendliness.[8,9] Typically, non-semiconductor catalysts exhibit superior photocatalytic activity and reaction stability after forming heterojunctions with CeO2.[10,11] With the recent research on rare earth metal oxides, there has been many reports in the literature on the modification of CeO2 materials to improve their photocatalytic activity. Li et al. studied the activity and stability under different reaction conditions by synthesizing the exposed surface of CeO2 nanoparticles with a controllable shape.[12] Jiang et al. investigated the non-homogeneous reactivity of CeO2 NCs with different crystalline surface shapes in other photocatalytic processes. They found that surface defect structures (e.g., Ce3+ ions and O vacancies) significantly affected the surface properties of CeO2 nanomaterials.[13] Zheng et al. successfully synthesized CeO2 nanocrystals with specific shapes with clear crystal faces and a hierarchical porous structure.[14]

Two-dimensional materials are excellent complexes for the construction of CeO2-based heterojunctions because of their excellent optical and electrochemical performances.[15] MXene is a new type of two-dimensional transition metal carbide/carbonitride/nitride, which has attracted more and more attention since the discovery of Ti3C2 in 2011.[16,17] MXenes are usually obtained by selective etching of the MAX phase (M stands for transition metals, A stands for IIIA/IVA elements, and X stands for C/N elements).[18] On the one hand, MXenes have excellent carrier mobility, a large specific surface area, and many active sites, leading to a large number of reactions occurring on the surface, reducing the carrier migration distance. On the other hand, MXenes have good elasticity and a unique layered structure, making them easy to compound on other photocatalyst surfaces.[19] MXenes also show excellent potential in photocatalysis and can be used as a synergistic catalyst because the above advantages can effectively separate/transfer light-induced carriers and many active surface sites for catalytic reactions.[20,21] Owing to the ideal photoresponsiveness and electronic coupling properties, the migration efficiency of photo-generated electron–hole pairs at the heterojunction interface is improved, which can greatly enhance the photocatalytic activity of the material.[22] Recently, it has been reported that ideal nanostructures and good interfacial effects between materials are more likely to accelerate the migration rate of photo-generated carriers and thus improve the catalytic activity of PHE.[23]

In this study, a 2D MXene/3D CeO2 photocatalyst was successfully prepared by synthesizing hexahedral prism-anchored octahedral CeO2 and compounding a thin layer of two-dimensional material MXenes on the surface. This photocatalyst has the advantages of being non-toxic and harmless and a simple synthesis method. The highest yield of photocatalytic H2 evolution is 454.32 μmol·g–1·h–1. Finally, the possible reaction mechanism of photocatalytic H2 evolution is explained. The increase in photocatalytic activity may be due to the interface effect of 2D/3D materials, resulting in many active sites on the surface of the material. This work provides new ideas and directions for researching other 2D/3D heterojunction photocatalyst materials.

Experimental Section

Materials

The whole chemicals used in this research were obtained from commercial suppliers and used without further distillation. Ti3AlC2 (Macklin, 98%), NaH2PO4·2H2O, Ce (NO3)3·6H2O, DMSO, and hydrofluoric acid (HF, content ≥40.0%) were supplied by Aladdin Reagent Co., Ltd.

Synthesis of Ti3C2 Nanosheets

Ti3C2 nanosheets were prepared according to the method reported elsewhere with some modifications.[6] 1.0 g of Ti3AlC2 powders were immersed in 20 mL 40% HF aqueous solution and stirred at room temperature (RT) for 72 h to etch out Al in Ti3AlC2. The suspension was filtrated, rinsed with deionized water until the pH of the supernatant became near 7, and then dried at 80 °C for 12 h. To achieve the exfoliation, the obtained multilayered Ti3C2 powder was added to 20 mL DMSO and stirred for 24 h. The mixture obtained was centrifuged and washed with deionized water more than three times to obtain a solid powder. The collected solid powder was then dispersed into deionized water again and layered by ultrasonication. After ultrasonication for 1 h, the suspension was removed by centrifugation at 5000 rpm for 1 h to remove the unexfoliated Ti3C2. Finally, the supernatant of Ti3C2 nanosheets was obtained.

Synthesis of Hexahedral Prism-Decorated CeO2 Octahedron

Referring to the method synthesized by Zou et al., hexahedral prism-anchored octahedral CeO2 was synthesized.[24]

Preparation of CeO2–Ti3C2T MXene

First, multilayered Ti3AlC2 two-dimensional material was synthesized, and Ti3AlC2 was etched into few-layer MXene nanosheets by adding hydrofluoric acid and ultrasonic treatment.[25] Second, according to the reported method to synthesize hexahedral prism-anchored octahedral CeO2, a certain amount of CeO2 and a specific concentration of MXene nanosheet suspension were mixed in a beaker, stirred for 30 min, washed with distilled water, centrifuged several times, and dried in vacuum for 12 h. Finally, the CeO2/MXene solid powder catalyst can be obtained.

Results and Discussion

The preparation method of the CeO2/MXene catalyst is illustrated in Scheme , and the specific steps are shown in Section . Zeta potential measurements determined the surface charge properties of CeO2 and MXene. As indicated in Figure S1, the zeta potential value of MXene is −31.05 mV, which suggests that the MXene surface is negatively charged, while the zeta potential value of pure CeO2 is 46.03 mV. Therefore, the 3D/2D CeO2/MXene heterostructure can be formed by electrostatic interaction self-assembly between CeO2 and MXene.

Scheme 1

Illustration of the Controllable Synthesis of CeO2 and CeO2/MXene

The SEM images show that the synthesized CeO2 presents an octahedral shape with an occasional texture on the surface, forming nanorod-like arms with a length of about 150 nm (Figure a–d). The synthesized nanocrystals have a uniform morphology and a particle size of about 500–600 nm. The facet-based homojunction formed on the surface of CeO2 nanocrystals provides more active sites for photocatalytic reactions.[24] Supplementary SEM images are shown in Figure S2.

Figure 1

Low-magnification (a,b) and high-magnification (c,d) SEM images of hexahedral prism-anchored octahedral CeO2.

The morphology and thickness of the prepared Ti3C2 were observed using a field emission transmission electron microscope and an atomic force microscope. As shown in AFM images and the corresponding height profiles in Figure a–c, Ti3C2 has a typical nanosheet-like morphology with a thickness of about 4.35 nm). Meanwhile, Figure d,e shows that the Ti3C2 MXene has a typical sheet shape with slight wrinkles on the edges. The unetched MXene can be seen in Figure S2, which is many times thicker than the etched MXene layer and is not conducive to the transport of photogenerated charges. The exfoliated nanosheets can also have the optical path length significantly reduced, the photoexcitation range increased, and the bandgap value reduced. The TEM images of MXene without HF etching are shown in Figure S3.

Figure 2

AFM (a–c) and FETEM (d,e) images of MXene.

The synthesis conditions of CeO2/Ti3C2 nanocomposites are simple, mild, and easily reproducible. SEM images in Figures a–d and S4a–f prove that the MXene material covers the surface of CeO2, and the Mxene material is thin and has a smooth surface. The contact area between MXene and CeO2 is relatively wide and conducive to the transfer of photo-generated carriers. The TEM image in Figure e further confirms the formation of the 3D/2D heterojunction. Moreover, there is no obvious MXene material on the surface of CeO2.[26] This is due to the ultra-thin layer spacing of MXene which causes the morphology to be not clearly shown in TEM images. The corresponding enlarged images (Figures h and S5c) show the different interfaces between CeO2 and MXene, and the staggering lattice fringes prove that the heterojunction is formed between CeO2 and MXene. Therefore, the SEM and TEM results show that CeO2 and MXene are easy to composite. Consistent with this, the HAADF images of the sample (Figures i and S6) indicate that the material contains all elements such as Ce, O, C, and Ti.

Figure 3

SEM images of CeO2/MXene (a–d), TEM images of CeO2/MXene (e–h), and the element maps of Ce, O, and Ti of the selected area (i).

As shown in Figure a, the crystal structure of the synthesized material was analyzed by XRD. The typical diffraction peaks at 2θ = 28.55, 33.08, 47.48, and 56.33 correspond to CeO2 (111), (200), (220), and (311) planes, respectively, by comparing with the standard card (JCPDS 34-0394).[27,28] The XRD pattern of CeO2/MXene showed that the XRD peaks did not change much after MXene was compounded, which proved that the crystal structure of CeO2 did not change significantly. It may that the ultrathin characteristics of MXene led to the little XRD diffraction peaks.[29] The removed O-terminus on the surface of MXene may lead to the formation of O vacancies, and after recombination with CeO2, the concentration of O vacancies may increase, and the peak value of the EPR signal may become higher. The existence of oxygen defects in CeO2 is demonstrated by electron paramagnetic resonance spectroscopy, and pure CeO2 exhibits an insignificant symmetry signal at g = 2.001, which is due to the capture of unpaired electrons on the oxygen defects.[30,31] Such oxygen vacancies can more effectively separate photo-generated electron–hole pairs, resulting in good photocatalytic activity of the material.[32] After complexation with MXene, the signal peak becomes more intense, implying an increase in the concentration of oxygen vacancies, which may be because MXene itself contains specific oxygen vacancies.[33]

Figure 4

XRD patterns of (a) CeO2 and CeO2/MXene and (b) EPR signals of CeO2 and CeO2/MXene.

The XPS technology was used to study the surface chemical composition and the valence state of the CeO2/MXene composites. The Ce 3d spectrum (Figure b) can be fitted with eight peaks at 884.3, 884.2, 887.5, 893.6, 898.7, 902.4, 907.3, and 916.4 eV, which can be divided into two groups of spin–orbit multiplets representing Ce 3d3/2 and Ce 3d2/5. u and υ bands are from Ce3+. In contrast, the other bands are attributed to Ce4+.[28,34] The presence of the Ce3+/Ce4+ oxidation state mixture on the CeO2/MXene surface indicates that the oxidation state may not be fully formed on the sample surface, which may be caused by the oxygen vacancies present in the CeO2 crystal structure. The O 1s spectra of CeO2 and CeO2/MXene (Figure c) are observed at peaks of 529.3 and 531.6 eV, respectively. The lower binding energy (BES) peaks are considered to be due to the Ce–O bonds, while the higher peaks are considered to be due to the surface-adsorbed oxygen and H2O.[27,35] As shown in Figure d, the two peaks in the Ti 2p XPS spectra at 458.6 and 462.09 eV are attributed to the Ti–C bond, while two peaks at 467.64 eV are attributed to the Ti–O bond, which come from the surface attached oxygen-containing functional groups on MXene.[6] The C 1s spectra of MXene can be fitted into four peaks located at 281.3, 284.8, 286.7, and 288.6 eV which are assigned to C–Ti, C–C, C–O, and C–F bonds, respectively, and the spectra are shown in Figure S7.[36]

Figure 5

Full spectrum (a) and high-resolution XPS spectra of Ce 3d (b), O 1S (c), and Ti 2p (d).

Figure a shows the transient photocurrent responses of CeO2 and CeO2/MXene samples. By comparing the photocurrent densities of different samples, it is proved that CeO2/MXene has almost 1.25 times higher photocurrent density than CeO2, which indicates that the photo-generated carrier transfer rate can be significantly enhanced after compositing MXene.[23] The excellent electrical conductivity of MXene makes the photoluminescence intensity of CeO2/MXene the weakest in the photoluminescence spectrum (Figure b), indicating that CeO2/MXene has the best electron–hole separation performance.[37] The light absorption properties of the synthesized materials were tested by the UV–vis DRS technique, as shown in Figure . According to Figure c, pure CeO2 has a clear absorption spectrum in the visible range (390–780 nm). At the same time, CeO2/MXene samples’ absorption spectrum begins to red shift, indicating that MXene helps increase the capacity of light absorption. The samples’ electron transfer path and valence band position can be better studied by calculating the indirect band energy from the ultraviolet spectrum. The corresponding band gap energies were calculated from the Kubelka–Munk function αhν = A (hν – Eg)1/2, based on the tangent lines of (αhν)2 to hν plots, where α is the absorption coefficient and hν is the photon energy. Figure d presents Tauc’s plot for estimating the corresponding band gap energy (Eg). The Eg values of pure CeO2 and CeO2/MXene were calculated to be 3.81 and 4.06 eV, respectively, by linear extrapolation.[38] Compared with CeO2, the band gap of CeO2/MXene is obviously narrowed, which means that light can be more easily excited to promote the transfer of photo-generated electrons on the surface of the CeO2/MXene sample, thus making CeO2/MXene show better photocatalytic performance.[39,40] The valence band positions of the catalysts were analyzed using the VB-XPS spectra (Figure S8) and were 2.16 and 2.08 eV for CeO2 and CeO2/Mxene, respectively. The conduction band positions were −1.9 and −1.73 eV, respectively, calculated using the ECB = EVB – Eg equation.

Figure 6

(a) Transient photocurrent, (b) PL spectra, (c) UV–vis DRS spectra, and (d) Tauc’s plots.

Figure a,b shows that CeO2/MXene does not show DMPO-•O2– or DMPO-•OH signals under dark conditions, which indicates that photocatalytic reactions will not occur under conditions of no light. After 5 min of visible light irradiation, an unmistakable response signal of DMPO-•OH was detected. It proves two points: first, •OH and •O2– will be produced during the photocatalytic reaction. Second, as the illumination time increases, the signal becomes more intense, proving that the concentration of free radicals increases.[41,42]

Figure 7

(a) DMPO spin-trapping ESR spectra of CeO2/MXene in methanol dispersion for DMPO-•O2– and (b) aqueous dispersion for DMPO-•OH.

The samples were also used in a photocatalytic H2 evolution experiment to verify further the effect of a photo-generated electron in practical applications. The photocatalytic hydrogen evolution (PHE) performance with different catalysts is presented in Figure a,b. The results show that the PHE rate of CeO2/MXene is 454.32 μmol·g–1·h–1, while that of CeO2 is 223.15 μmol·g–1·h–1. Compared with CeO2, the PHE rate of CeO2/MXene is more than doubled. Because MXene’s efficient electron transfer capability, it may be that protons react more quickly with light-generated electrons to produce more H2. Meanwhile, the hydrogen evolution results show that MXene can effectively enhance the activity of photocatalysts. Figure S9 shows continuous H2 production with no noticeable decrease in the subsequent reactions, demonstrating that the synthesized material is reusable.

Figure 8

(a) Time-dependent H2 evolution. (b) H2 evolution rate of CeO2 and CeO2/MXene.

We propose a possible PHE mechanism based on these experimental and theoretical data (Figure ). First, MXene can form a heterojunction structure after being combined on the CeO2 surface. When the electrons in CeO2 are excited, the photo-generated electrons are first excited and jump from the valence band to the conduction band. Due to the smooth surface structure and excellent interlayer structure of MXene, photoexcited electrons can migrate rapidly from semiconductor CeO2 to metallic MXene through the tightly coupled interface. Meanwhile, the photo-generated holes remaining in the valence band of CeO2 are consumed by the sacrificial agent. Subsequently, photo-generated electrons are further reacted with the absorbed protons under light to produce H2.[43−46]

Figure 9

Schematic diagram of possible electron transfer paths and photocatalytic H2 evolution mechanism in CeO2/MXene.

Conclusions

In summary, we successfully prepared 3D CeO2/2D Ti3C2 MXene photocatalysts by compositing ultrathin exfoliated Ti3C2 MXene nanosheets on the CeO2 surface. The constructed 3D/2D CeO2/MXene structures have excellent interfacial effects, desirable photoresponsive properties, and high photo-generated carrier mobility. The process of photocatalytic hydrogen precipitation under light illumination can be realized by adding a small amount of CeO2/MXene photocatalyst. The experimental results show that the special two-dimensional structure, fast carrier migration rate, and ideal interfacial effects of Ti3C2 MXene have unique synergistic effects on photocatalytic H2 evolution. The Schottky contact between the n-type semiconductor CeO2 and the metal-like Ti3C2 can significantly promote the separation of photo-generated carriers and the transfer of the photo-generated electrons to MXene nanosheets, endowing the Schottky heterojunction with strong photocatalytic activity, which can be used for the photocatalytic H2 evolution process. The construction of heterojunction photocatalytic systems is an effective way to enhance photocatalytic performance because 2D semiconductors can act as electron acceptors to improve charge separation and prolong the carrier lifetime. In addition, the semiconductor material CeO2 transferred the photo-generated electrons to Mxene, which accelerated the reduction rate of protons to produce more hydrogen gas and improved the photocatalytic performance. It provides a new idea for fabricating rare earth metal oxide and 2D MXene composites and can be applied to other similar 3D/2D heterostructures.