Boron doped 1T phase MoS2 as a cocatalyst for promoting photocatalytic H2 evolution of g-C3N4 nanosheets

Pengyun Qiu, Yn An, Xinyu Wng, Shnn An, Xioli Zhng, Jin Tin,*,Wen Zhu

a School of Materials Science and Engineering, Shandong University of Science and Technology, Qingdao 266590, China

b State Key Laboratory of Materials Processing and Die & Mould Technology, School of Materials Science and Engineering, Huazhong University of Science and Technology, Wuhan 430074, China

c School of Materials Science and Engineering, Zhengzhou University, Zhengzhou 450001, China

Keywords:1T phase MoS2 g-C3N4 nanosheets Photocatalytic H2 evolution Active sites Boron doping

ABSTRACT As one of the 2D transition metal sulfides, 1T phase MoS2 nanosheets (NSs) have been studied because of their distinguished conductivity and suitable electronic structure.Nevertheless, the active sites are limited to a small number of edge sites only, while the basal plane is catalytically inert.Herein, we report that boron (B) doped 1T phase MoS2 NSs can replace precious metals as a co-catalyst to assist in photocatalytic H2 production of 2D layered g-C3N4 nanosheets (g-C3N4 NSs).The H2 evolution rate of prepared B-MoS2@g-C3N4 composites with 15 wt% B-MoS2 (B-MoS2@g-C3N4–15, 1612.75 μmol h-1 g-1) is 52.33 times of pure g-C3N4 NSs (30.82 μmol h-1 g-1).Furthermore, the apparent quantum efficiency (AQE) of B-MoS2@g-C3N4–15 composites under the light at λ=370 nm is calculated and reaches 5.54%.The excellent photocatalytic performance of B-MoS2@g-C3N4–15 composites is attributed to the B ions doping inducing the distortion of 1T phase MoS2 crystal, which can activate more base planes to offer more active sites for H2 evolution reaction (HER).This work of B-MoS2@g-C3N4 composites offers experience in the progress of effective and low-price photocatalysts for HER.

Recently, g-C3N4nanosheets (NSs) have attracted much attention to hydrogen production due to excellent chemical stability,suitable band structure, simple syntheses, low cost and special two-dimensional (2D) layered structure [1–3].Besides, the special nitrogen-rich polymeric structure of 2D layered g-C3N4NSs could provide numerous active sites for HER [4,5].Nevertheless, the photocatalytic hydrogen evolution performance of pure 2D layered g-C3N4NSs is unsatisfactory because of fast electron-holes pairs recombination and insufficient absorption of light [6–8].Hence, it is essential to exploit novel methods to enhance the H2generation performance of 2D layered g-C3N4NSs.The construction of an internal electric field (IEF) is an effective strategy to enhance photocatalytic hydrogen evolution performance because of its key role in photo-induced carrier separation [9–11].Precious metals act as cocatalysts is a common strategy to construct IEF [12,13].Yet, the widespread application of precious metals is limited by high cost[14].Hence, the development of an inexpensive and efficient cocatalyst is crucial for enhancing the photocatalytic activity of 2D layered g-C3N4NSs.

Among various cocatalysts, transition-metal chalcogenides, such as MoS2, receive widespread attention ascribed to the superior 2D layered structure [15–19].MoS2has both semiconductor 2H phase and metallic 1T phase.Among them, metallic 1T phase MoS2NSs with octahedral coordination can improve the transfer and capture of photogenerated carriers to boost the photocatalytic H2evolution activity of 2D layered g-C3N4NSs, attributing to excellent conductivity [20–22].In addition, the Gibbs free energy of 1T phase MoS2NSs for H+absorption is near-zero, which is suitable for HER[6,23].However, the basal plane of MoS2NSs is inert, which limits the photocatalytic hydrogen evolution reaction [6].Recent researches show that MoS2′s basal plane can be activated through doping atoms to design the active sites of MoS2, attributed to the local electronic structure modulation [6,24–26].Thus, boron (B)ions are incorporated into the lattice of 1T phase MoS2could activate the basal plane, which can improve photocatalytic H2production performance.Hence, 2D-layered g-C3N4NSs modified by doping B into 1T phase MoS2NSs (B-MoS2NSs) could exhibit better photocatalytic performance.

Fig.1.(a) X-ray diffraction patterns of pure g-C3N4 NSs and B-MoS2@g-C3N4 composites.(b) C 1s, (c) N 1s, (d) Mo 3d, (e) S 2p and (f) B 1s XPS spectra of B-MoS2@g-C3N4–15 composites.

In this work, we propose B-MoS2@g-C3N4composites for photocatalytic H2evolution, in which triethanolamine acts as the sacrificial agent.The B into 1T phase MoS2NSs are powerfully connected with 2D layered g-C3N4NSs through an easy hydrothermal method.The synthesized B-MoS2@g-C3N4composites with 15 wt% B-MoS2(B-MoS2@g-C3N4–15) display an efficient rate of hydrogen evolution (1612.75 μmol h-1g-1), which is 52.33 times as much as pure g-C3N4NSs (30.82 μmol h-1g-1).In addition,the apparent quantum efficiency (AQE) of pure g-C3N4NSs and BMoS2@g-C3N4–15 composites are 0.41 and 5.54% under the light atλ=370 nm.The loading of B-MoS2NSs improves the light absorption of g-C3N4to stimulate more photogenerated carriers.In addition, the incorporation of B ions into the lattice of 1T phase MoS2NSs can provide more active sites and speed up the photocatalytic hydrogen evolution reaction.Thus, B-MoS2@g-C3N4composites display excellent photocatalytic H2production activity.

The fabrication of B-MoS2@g-C3N4composites is displayed in Scheme S1 (Supporting information).Firstly, pure g-C3N4NSs are synthesizedviaa direct thermal polymerization way of urea.During heating, urea first reacts to form bulk g-C3N4when the muffle furnace temperature is kept at 550 °C, and then g-C3N4NSs with 2D layered structure are formed at the muffle furnace temperature of 500 °C.Afterward, g-C3N4NSs, ammonium tetrathiomolybdate and boric acid are added to the 70 mLN,N-dimethylformamide solution and evenly dispersed.Then, B-MoS2NSs are grown on the surface of g-C3N4NSs by hydrothermal method to obtain BMoS2@g-C3N4composites.

Fig.2.SEM images of (a) pure g-C3N4 NSs and (b) B-MoS2 NSs.(c) TEM and (d)HR-TEM images of B-MoS2@g-C3N4–15 composites.

For pure g-C3N4NSs and B-MoS2@g-C3N4composites, two characteristic peaks are located at 13.1° and 27.4° (green, pink,blue, yellow and purple curves in Fig.1a), attributing to (100) and(002) planes of g-C3N4(JCPDS No.87–1526) [27–29].Simultaneously, the peaks at 13.1° (100) and 27.4° (002) of g-C3N4in BMoS2@g-C3N4composites become weaker after B-MoS2NSs loading on g-C3N4NSs, attributing to the fact that the order degree of g-C3N4NSs is decreased by B-MoS2NSs incorporation [6].As shown in black curve in Fig.1a, there are two peaks at 10.2° and 32.5° indexed to (002) and (100) plane of MoS2.Compared with pure MoS2(black curve in Fig.1a), the peak at 10.2° of MoS2shifts to a lower degree of 8.8° in contrast to the 10.2° (002) peak of MoS2due to the B ions incorporating into the lattice of MoS2cause the distortion of MoS2crystal (red curve in Fig.1a) [6].In addition, the peak of B-MoS2@g-C3N4composites at 8.8° is similar to B-MoS2, indicating that B-MoS2and g-C3N4coexist.

To further analyze the chemical bonding state, the XPS spectra of B-MoS2@g-C3N4composites are tested.As shown in Fig.S1(Supporting information), the survey XPS spectrum indicates that B-MoS2@g-C3N4composites consist of C, N, Mo, S and B.As shown in Fig.1b, the C 1s XPS spectrum is deconvolved into two peaks at 284.8 and 288.4 eV, corresponding to C=C and C-(N)3bonds, respectively [30,31].Fig.1c indicates that the N 1s spectrum is deconvolved into three peaks at 398.2, 399.1 and 400.6 eV, attributing to C=N–C, N-(C)3and C–NHx, respectively [32,33].As shown in Fig.1d, there are two green peaks at 227.8 and 230.9 eV, assigning to Mo 3d5/2and Mo 3d3/2of 1T phase [34].The two pink peaks at 228.7 and 232.2 eV are attributed to Mo 3d5/2and Mo 3d3/2of 2H phase [34].In addition, the peaks at 225.2 and 234.9 eV are assigned to S 2s and oxidation of Mo [6].As for the S 2p XPS spectrum (Fig.1e), two green peaks at 160.7 and 162.1 eV are attributed to S 2p3/2and S 2p1/2of 1T phase [35].Simultaneously, there are two pink peaks at 161.5 and 163.4 eV, assigning to S 2p3/2and S 2p1/2of 2H phase [35].Besides, according to Mo 3d and S 2p spectra, the proportion of 1T phase MoS2is about 71.2%, indicating that 1T phase MoS2in B-MoS2@g-C3N4composites is the main phase.Fig.1f exhibits a visible peak of B element at 184.6 eV, illustrating the successful doping of B ions in MoS2[6].Besides, the Raman spectrum of B-MoS2(Fig.S2 in Supporting information) is measured to determine the MoS2phase.There are three peaks at 147,237, and 335 cm-1, respectively, which correspond to the J1, J2, and J3modes of 1T phase MoS2[16].

As shown in Fig.2a, g-C3N4NSs display a special 2D layered structure, and the lamellae of g-C3N4NSs present irregular wrinkled sheet morphologies.Pure B-MoS2displays flower-like assemblies composed of numerous small nanosheets (Fig.2b).In Fig.2c, B-MoS2@g-C3N4composites still keep typical 2D sheet-shaped morphology.As shown in Fig.2d, B-MoS2NSs are assembled on the g-C3N4NSs, and the lattice space distance (0.98 nm) is attributed to the (002) plane of MoS2[6].The close connection between g-C3N4NSs and B-MoS2NSs facilitates the fast transfer of photogenerated electrons from g-C3N4NSs to B-MoS2NSs, which can effectively inhibit electron-hole pairs recombination.Fig.S3(Supporting information) shows the EDX mapping images of BMoS2@g-C3N4–15 composites, which display even distribution of C,N, Mo, S and B, indicating the coexistence of g-C3N4and B-MoS2.

As shown in Fig.3a, an obvious absorption of g-C3N4NSs and B-MoS2@g-C3N4composites is observed, and the absorption edge isca.430 nm.The band gap energy (Eg) of pure g-C3N4NSs is obtained through the formula (αhν)1/2∝hν-Eg, and theEgof pure g-C3N4NSs is calculated and the value is 2.61 eV (Fig.S4 in Supporting information).Significantly, the optical absorption of B-MoS2@g-C3N4composites is stronger than that of pure g-C3N4NSs, which indicates that B-MoS2NSs loading onto g-C3N4NSs can effectively enhance the light absorption ability of the catalyst.Among these, the light absorption of B-MoS2@g-C3N4–15 composites is the strongest, which can boost the production of photogenerated carriers.To further research the role of B ions doping,the UV–vis DRS absorption spectra of MoS2@g-C3N4–15 and BMoS2@g-C3N4–15 composites are shown in Fig.3b.After B ions are doped in 1T-MoS2NSs, the light absorption of B-MoS2@g-C3N4–15 composites is improved (yellow curve in Fig.3b), indicating that doping B into MoS2can boost the utilization of light.

To study the photogenerated charge separation and transfer properties of pure g-C3N4NSs, MoS2@g-C3N4–15 and B-MoS2@g-C3N4–15 composites, the photoelectrochemical (PEC) analysis is performed (Figs.3c and d).Fig.3c displays that pure g-C3N4NSs,MoS2@g-C3N4–15 and B-MoS2@g-C3N4–15 composites present the photocurrent responses on each illumination [6].In addition, the photocurrent values of MoS2@g-C3N4–15 and B-MoS2@g-C3N4–15 composites are higher than that of pure g-C3N4NSs, which indicates that 1T-MoS2assembled on the g-C3N4NSs can effectively improve the generation and separation of photogenerated carriers.Notably, the photocurrent value of B-MoS2@g-C3N4–15 composites is superior to MoS2@g-C3N4–15 composites, indicating that B ions doping could inhibit the recombination of electron-hole pairs.The charge transfer activity of pure g-C3N4NSs, MoS2@g-C3N4–15 and B-MoS2@g-C3N4–15 composites is further explored through EIS measurement (Fig.3d).MoS2@g-C3N4–15 and B-MoS2@g-C3N4–15 composites present a smaller arc radius than that of pure g-C3N4NSs, indicating 1T-MoS2can accelerate carrier separation (Fig.3d).Besides, B-MoS2@g-C3N4–15 composites display the smallest arc radius, indicating the separation of photoexcited carriers of B-MoS2@g-C3N4–15 composites is most effective.Hence, B ions doped into 1T-MoS2and MoS2NSs assembled on the g-C3N4NSs can synergistically promote photocatalytic activity.

To explore the possibility of hydrogen production, Mott-Schottky plots are tested to estimate the conduction band (CB) potential of pure g-C3N4NSs.As shown in Fig.S5 (Supporting information), g-C3N4is identified as an n-type semiconductor because of the positive slope of the curves in the Mott-Schottky plots.Through extrapolation to the x-intercept in Mott-Schottky plots,the flat band potential (EFB) of g-C3N4is obtained (-0.42 eVvs.Ag/AgCl).The obtainedEFBis converted to a potentialvs.standard hydrogen electrode (NHE), and then the value is subtracted by 0.2 eV to obtain anECBvs.NHE of the sample.Therefore, theECBof g-C3N4is -0.4 eVvs.NHE.By the valence band potential(EVB)=Eg+ECBandEgresults (Fig.S4 in Supporting information),theEVBof g-C3N4is 2.21 eVvs.NHE (Fig.S6 in Supporting information).Based on the above research, theECBof as-prepared g-C3N4NSs is lower than 0 eV, which indicates that prepared g-C3N4NSs can conduct photocatalytic hydrogen production.

Fig.3e indicates that the photocatalytic hydrogen production of all photocatalysts is linear with time, indicating that the catalyst has stable photocatalytic H2evolution performance.As shown in Fig.3f, bare g-C3N4shows an unacceptable photocatalytic hydrogen evolution performance (30.82 μmol h-1g-1),which is attributed to the fast recombination of carriers and low light utilization.Yet, B-MoS2@g-C3N4–15 composites present excellent photocatalytic hydrogen production and H2evolution rate, indicating that adding B-MoS2NSs as cocatalysts can promote the photocatalytic activity of catalysts.The H2evolution rate of B-MoS2@g-C3N4–15 composites (1612.75 μmol h-1g-1) is 52.33, 1.3, 1.15 and 1.31 times of bare g-C3N4NSs (30.82 μmol h-1g-1), B-MoS2@g-C3N4–5 (1236.04 μmol h-1g-1), B-MoS2@g-C3N4–10 (1405.12 μmol h-1g-1), B-MoS2@g-C3N4–20 composites (1238.47 μmol h-1g-1), respectively.The improved photocatalytic HER performance of B-MoS2@g-C3N4composites indicates that B-MoS2cocatalyst loading onto g-C3N4improves the utilization of light to stimulate more photogenerated electrons, accelerates carrier separation, and inhibits electron-hole pairs recombination.However, increasing B-MoS2NSs content from 15% to 20%, a decrease in photocatalytic performance is detected, owing to the overmuch B-MoS2NSs loading on g-C3N4NSs impediment the photo-absorption of g-C3N4NSs.As shown in Fig.S7 (Supporting information), the photocurrent values of B-MoS2@g-C3N4–15 composites is higher than that of B-MoS2@g-C3N4–20 composites,which indicates that the overmuch B-MoS2NSs loading on g-C3N4NSs in B-MoS2@g-C3N4–20 is adverse for the photocatalytic performance of photocatalyst.To study the effect of B ions doping, we measure the hydrogen production of MoS2@g-C3N4–15 composites.As shown in Fig.S8 (Supporting information), the photocatalytic hydrogen production amount and rate of B-MoS2@g-C3N4–15 composites (1612.75 μmol h-1g-1) is higher than that of MoS2@g-C3N4–15 composites (1370.68 μmol h-1g-1), attributing that B ions doping into MoS2can activate the base planes of MoS2and offer more active sites.To further evaluate the H2production cycle property of B-MoS2@g-C3N4–15 composites, the photocatalytic H2evolution performance is tested for 15 h (Fig.S9 in Supporting information).Almost 92.4% of the incipient property is kept,and the micromorphology of B-MoS2@g-C3N4–15 composites after the cycle test (Fig.S10 in Supporting information) does not change significantly, indicating the good stability of B-MoS2@g-C3N4–15 composites.We further determine the apparent quantum efficiency(AQE) of pure g-C3N4NSs and B-MoS2@g-C3N4–15 composites under light atλ=370 nm irradiation.As shown in Fig.S11 (Supporting information), the AQE value of B-MoS2@g-C3N4–15 composites(5.54%) is higher than that of bare g-C3N4NSs (0.41%), which illustrates the optical utilization of B-MoS2@g-C3N4–15 composites is higher than that of pure g-C3N4NSs.

Scheme 1.Schematic illustration of the photocatalytic H2 evolution mechanism of B-MoS2@g-C3N4 composites.

According to the above studies, we propose a possible photocatalytic hydrogen evolution mechanism to explicate the reason for the improved photocatalytic hydrogen evolution performance of BMoS2@g-C3N4composites (Scheme 1).The loading of B-MoS2NSs improves the light absorption of g-C3N4to stimulate more photogenerated carriers and offer more active sites for HER.In addition,B-MoS2as a cocatalyst can capture photoinduced electrons, accelerate electron transfer and inhibit recombination of electron-hole pairs.B ions are doped into the lattice of 1T-MoS2, which can activate more base planes of 1T-MoS2and is suitable for the HER to activate H+.The photoexcited electrons produced under the sunlight through g-C3N4moved to B-MoS2, and then reduced water to hydrogen.Concurrently, TEOA as a sacrificial agent consumed the holes.Hence, the close combination of B-MoS2and g-C3N4boosts the photocatalytic hydrogen evolution activity of photocatalysts.

In conclusion, we have successfully synthesized an efficient B-MoS2@g-C3N4composite for photocatalytic H2evolutionviacocatalyst and doping strategy.The as-prepared B-MoS2@g-C3N4composites with 15 wt% B-MoS2(B-MoS2@g-C3N4–15) present an extremely improved photocatalytic H2evolution rate of 1612.75 μmol h-1g-1, which is 52.33 times of bare g-C3N4NSs(30.82 μmol h-1g-1).The above experimental results confirm that the enhanced photocatalytic activity of B-MoS2@g-C3N4–15 composites may be assigned to the following factors: (1) As a cocatalyst, B doped 1T phase MoS2NSs greatly improves the light utilization of photocatalyst and stimulates more photogenerated carriers;(2) B-MoS2NSs with excellent conductivity are closely connected onto g-C3N4NSs, which can accelerate electron transfer and inhibit carrier recombination; (3) The base planes are activated through doping B ions into the lattice of MoS2, which can induce the distortion of MoS2crystal and provide more active sites for HER.This easy assembly strategy offers guidance for rationally constructed photocatalysts based on B-doped 1T phase MoS2as a cocatalyst for H2production.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

The authors are thankful for fundings from the National Natural Science Foundation of China (No.51872173), Taishan Scholars Program of Shandong Province (No.tsqn201812068), Natural Science Foundation of Shandong Province (No.ZR2022JQ21), and Higher School Youth Innovation Team of Shandong Province (No.2019KJA013).The authors would like to thank Shiyanjia Lab (www.Shiyanjia.Com) for the XPS analysis.

Supplementary materials

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2023.108246.