The majority of global energy consumption is mainly supplied by fossil fuels combustion currently, which inevitably causes environmental problems and energy shortage, thus stimulating scientists for seeking renewable and green energy sources. Photocatalytic hydrogen evolution from water is a significant reaction for using renewable energy, which needs highly efficient and stable catalysts. During the past decades, various studies have been devoted to exploring highly efficient photocatalysts for H
2 evolution from water, including metal oxides [
1,
2,
3], metal chalcogenides [
4], and some other semiconductors [
5].
Although great achievements have been made in the development of H
2 evolution photocatalysts, there are still several challenges to its practical application. One of the main bottlenecks is the severe charge carrier recombination of the photogenerated electron-hole pairs during their transfer process [
6]. Designing a heterojunction structure is widely used to improve the electron-hole separation and suppress the recombination behavior of charge carriers [
7]. Recently, a new step-scheme (S-scheme) heterojunction with a staggered band structure has been proposed [
8], which is composed of two n-type semiconductor photocatalysts for reduction and oxidation reactions, respectively. In the S-scheme heterojunction, the electrons in the oxidation photocatalyst and the holes in the reduction photocatalyst will recombine in the interface, leaving the more powerful photogenerated electrons in the conduction band (CB) of the reduction photocatalyst, and the more powerful photogenerated holes in the valence band (VB) of oxidation photocatalyst, respectively. Therefore, the charge transfer route in this system makes the heterojunction has a strong redox ability. Due to these advantages, photocatalysts with the S-scheme structure have been utilized in many fields, such as water splitting [
7,
8,
9,
10,
11,
12], pollutant degradation [
13,
14,
15,
16,
17,
18], CO
2 photoreduction [
19,
20,
21,
22,
23,
24,
25,
26,
27,
28], and inactivation of bacteria [
29,
30,
31,
32]. These investigations indicate that the construction of S-scheme heterojunction can be an ideal option to design high-performance photocatalysts.
Among various photocatalysts, metal sulfides have attracted widespread attention in recent years [
33,
34]. Especially, molybdenum disulfide (MoS
2), as a typical transition metal sulfide, has been considered as a promising hydrogen evolution reaction (HER) catalyst because of its low energy barriers for hydrogen adsorption and desorption. MoS
2 has a layer-dependent band gap; for bulk the value is 1.2 eV and it increases to 1.9 eV for monolayer [
34], which is suitable for visible light-driven photocatalysis. However, bare MoS
2 usually has poor photocatalytic HER activity from pure water splitting because of its relatively low conduction band potential −0.16 eV [
35,
36], which is not sufficient for H
+ reduction to produce H
2 (−0.41 V vs. NHE at pH 7). SnS
2, an inexpensive and nontoxic semiconductor with a band gap of 2.0–2.25 eV [
4,
37], has recently been proven to be a relatively stable and efficient visible-light-driven photocatalyst which has a higher conduction band edge position than that of MoS
2 [
38]. Nevertheless, the practical applications of single SnS
2 have also been limited by the high electron-hole recombination rate and the inferior charge transfer rate [
39]. Up to the present, considerable research efforts have been devoted to exploring S-scheme heterojunction with SnS
2 and MoS
2 for their photocatalytic applications, such as MoS
2/SnS
2/r-GO for CO
2 reduction [
40], 2D/2D SnS
2/MoS
2 nanosheets for methylene blue decomposition [
41] and CNT@MoS
2/SnS
2 nanostructure for efficient Cr(VI) reduction [
42]. These studies proved the existence of S-scheme heterojunction between SnS
2 and MoS
2 semiconductors because of their increased electron-hole separation and improved photocatalytic activities in heterojunction structure. However, besides the photocatalytic activity, selectivity is another important factor that influences the performance of a photocatalyst, which has rarely been studied in S-scheme heterojunction photocatalysts. Previous studies have shown that accumulated electrons on the catalyst surface facilitate multi-electron reduction reactions, meanwhile dispersed electrons are inclined to participate in single or two electron reduction processes [
43,
44,
45,
46]. In an S-scheme MoS
2/SnS
2 heterojunction structure, photo-excited electrons in the CB of MoS
2 will combine with holes in the VB of SnS
2 owing to internal electric field (IEF), leaving the CB electrons of SnS
2 and VB holes of MoS
2 highly dispersed on the heterojunction surface. From this viewpoint, we can use this type of MoS
2/SnS
2 hetero-structure to improve the selectivity of single or two electron reduction reactions, such as H
2 evolution from CO
2 saturated water. Furthermore, previous studies have demonstrated that the quantum-sized photocatalyst is more advantageous in photocatalysis when compared with the larger nanoscale photocatalyst, because of its increased redox potentials of photogenerated electrons and holes by quantum confinement effect [
47].
Inspired by the afore-mentioned unique features, we adopted a one-step *in situ* synthesis method, to design a 0D/2D S-scheme heterojunction involving MoS
2 nanosheets decorated with SnS
2 quantum dots (QDs) for H
2 evolution in CO
2 saturated water solution. A series of analytical techniques including XPS, TEM, XRD, and UV-vis diffuse reflectance measurements have been utilized to study morphology, structural and optical properties of photocatalysts. The photocatalytic efficiency is measured by H
2 evolution in CO
2 saturated water upon Xenon light irradiation. It was found the SnS
2/MoS
2 S-scheme heterojunction can significantly improve H
2 evolution performance. The H
2 evolution rate of SnS
2/MoS
2, SnS
2 and MoS
2 are 42.49, 22.09, and 11.19 µmol·g
−1·h
−1, respectively, in CO
2 saturated pure water under the conditions we used. Moreover, *in-situ* diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) was applied to explore the reaction pathways. A mechanism is proposed based on these characterizations.
*2.1. Preparations*
All reagents were of analytical grade and were used as received without further purification.
For preparation of pure MoS2 sample, 30 mmol thiourea was dissolved in 70 mL deionized water under vigorous stirring. After that, 5 mmol sodium molybdate dehydrate (Na2MoO4·2H2O) was added and then stirred for 0.5 h to form a homogeneous solution. The solution was then transferred into a 100 mL Teflon-lined stainless steel autoclave up to 70% of the total volume. The autoclave was heated at 220 °C for 18 h at autogenous pressure, and then cooled to room temperature naturally. The resulting sample was separated by centrifugation from the reaction solution and washed with deionized water and absolute alcohol several times. The obtained solid MoS2 slurry were dispersed in water and sonicated for 1 h to form a MoS2 suspension, which was then freeze-dried for further characterization.
For the SnS2 nanosheet sample, 15 mmol thiourea was dissolved in 70 mL deionized water under magnetic stirring for 0.5 h to form a transparent solution. 2.5 mmol SnCl4·5H2O was added to the solution followed by stirring 0.5 h to form a transparent precursor solution. The precursor suspension was then transferred to 100 mL Teflon-lined stainless-steel autoclave, maintained at 220 °C for 18 h. After that, the reactor was cooled down to room temperature naturally for about 12 h. Then, the solid product was collected by centrifugation and washed by ethanol and distilled water several times, and then freeze-dried for further characterization.
For preparation of the SnS2/MoS2 nanocomposite, 30 mmol thiourea was dissolved in 70 mL deionized water under vigorous stirring. After that, 2.5 mmol Na2MoO4·2H2O and 2.5 mmol SnCl4·5H2O were added and stirred for 0.5 h to form a homogeneous solution. This solution was transferred into a 100 mL Teflon-lined stainless steel autoclave up to 70% of the total volume. The autoclave was heated to 220 °C and maintained at 220 °C for 18 h. After that, the reactor was cooled down to room temperature naturally. The obtained SnS2/MoS2 solid product was separated by centrifugation from the reaction solution, washed with absolute ethanol and deionized water several times, and then freeze-dried for further characterization.
*2.2. Characterizations*
The crystal structures were characterized through powder XRD patterns, which were obtained using a Japan Rigaku X-ray diffractmeter/RINT2500HLR+ with Cu Kα radiation operating at 40 kV and 80 mA (Rigaku, Tokyo, Japan). The transmission electron microscope (TEM) analyses were carried out using a JEOL JEM-2100F field emission electron microscope (JEOL, Tokyo, Japan). X-ray photoelectron spectroscopy (XPS) spectra were collected using a PHI5000VersaProbe spectrometer (ULVAC PHI, Chigasaki, Japan). UV-vis absorption spectra of the samples were measured by a Shimadzu UV-3600 Spectrophotometer (Shimadzu, Kyoto, Japan). The *in-situ* DRIFT measurement was performed on a Nicolet 8700 DRIFT spectrometer using KBr window (Thermo Fisher Scientific, Waltham, MA, USA).
*2.3. Photocatalytic Test*
Photocatalytic H2 evolution in CO2 saturated water was performed under a 300 W Xenon light (full spectrum) at atmospheric pressure. Typically, a total of 0.025 g catalyst was added to a homemade quartz reactor (600 cm3), which was located approximately 10 cm from the lamp. 50 mL of deionized water was injected into the reactor and sonicated for 0.5 h for the dispersion of the catalyst. After that, pure CO2 (99.99%) was slowly bubbled through the reactor for 20 min to remove the air in the reactor. Then the reactor was sealed and irradiated under Xenon light for various times. During the experiment, the temperature of the reactor was maintained at 15 °C by providing a flow of cooling water. All products were quantified by gas chromatography (GC) equipped with a thermal conductivity detector and flame ionization detector.
*2.4. Electrochemical Measurements*
Electrochemical measurements were performed on a CHI 700E electrochemical workstation using a standard three-electrode cell. The working electrodes were prepared from the photocatalysts. A platinum mesh and a standard saturated calomel electrode (SCE) were used as a counter electrode and a reference electrode, respectively. To prepare the working electrodes, 5 mg of the photocatalyst was added into 0.1 mL of 1% Nafion solution in ethanol and sonicated for 1 min. After that, the obtained slurry was dispersed onto a 2 cm × 1.5 cm FTO glass by dip coating and dried at 25 °C.
*3.1. Catalyst Characterizations*
The crystal structures of MoS
2, SnS
2, and SnS
2/MoS
2 composites were analyzed by XRD, as shown in . The XRD pattern of the as-synthesized MoS
2 sample can be indexed to the hexagonal structure of MoS
2 (JCPDS No. 75-1539). For the as-prepared MoS
2, the detected peaks at 2θ = 14.1°, 32.9°, 39.5°, 49.4°, and 58.7° are assigned to the (002), (100), (103), (105), and (110) planes of the hexagonal phase. The diffraction peaks of SnS
2 located at 2θ = 15.0°, 28.3°, 32.2°, 41.9°, 50.1°, and 52.6° can be indexed to (001), (100), (011), (012), (110), and (111) planes of hexagonal SnS
2 (JCPDS No.83-1705). Both the diffraction peaks of SnS
2 and MoS
2 can be observed in the case of SnS
2/MoS
2 composite materials. Besides that, an additional peak at 26.2° is observed in the XRD pattern of SnS
2/MoS
2, which can be indexed to a tiny amount of SnO in the interface of SnS
2/MoS
2. The formation of SnO is inevitable during the preparation of SnS
2/MoS
2 heterojunction. When SnS
2 closely contacted with MoS
2, the electrons in SnS
2 would flow to MoS
2 spontaneously to maintain Fermi energy at the same level. After that, the surface adsorbed O
2 molecules would donate electrons to SnS
2 and lead to the formation of tiny SnO species.
. XRD patterns of MoS2, SnS2 and SnS2/MoS2.
The detailed morphology of the as-prepared catalysts was further determined by transmission electron microscopy (TEM) measurement. a shows the low-magnified TEM image of the as-prepared MoS
2. b is the low-magnified TEM image of the hydrothermally synthesized SnS
2. Hexagonal nanosheets with an average size of 200~500 nm can be observed in b. The selected area electron diffraction (SAED) from one single SnS
2 hexagonal nanosheet is shown in c, which indicates the single crystallinity of the as-prepared SnS
2. This is further confirmed from its high-resolution TEM image in d. As shown in d, two lattice distances of 0.32 nm are observed from (010) and (100) planes of hexagonal SnS
2, indicating that the surface plane of hexagonal SnS
2 is (001) plane. e shows the low-magnified TEM image of the SnS
2/MoS
2 nanocomposite, which indicates a quantum dot decorated sheet like structure. The particle size of SnS
2 quantum dot in the nanocomposite is about 4~7 nm. The average thickness of the MoS
2 nanosheet in the nanocomposite is 3~10 nm. From the TEM image, it is obvious that the composite structure restricted the crystal growth of SnS
2. HRTEM image of the SnS
2/MoS
2 nanocomposite in f further confirmed the quantum dots decorated sheet-like nanostructure.
. (a) TEM image of MoS2, (b) TEM image of SnS2, (c) SAED pattern of SnS2 on one single crystal, (d) HRTEM image of SnS2, (e) TEM image of SnS2/MoS2 nanocomposite, (f) HRTEM image of SnS2/MoS2 nanocomposite.
The chemical state of the as-prepared materials was further investigated by XPS analysis. a–c shows the XPS spectra of Sn 3d, Mo 3d and S 2p in different samples. The peak positions of Sn 3d
3/2 and Sn 3d
5/2 in pure SnS
2 are located at 495.2 and 486.7 eV, respectively, while these peaks shift to 495.6 and 487.1 eV, respectively in the SnS
2/MoS
2 nanocomposite. The increased binding energy of Sn 3d in the composite material may be ascribed to the electron transfer from Sn to Mo element in the interface of the SnS
2/MoS
2 nanocomposite. This electron transfer is further confirmed by the change of Mo 3d XPS spectrum. As shown in b, the binding energies of Mo 3d
3/2 and Mo 3d
5/2 in pure MoS
2 are 232.17 and 229.08 eV, while these values are shifted to 232.12 and 229.0 eV in the SnS
2/MoS
2 composite. The decreased binding energy of Mo 3d in the SnS
2/MoS
2 composite may be attributed to the electron transfer from Sn to the adjacent Mo element. The shift of the binding energies of Mo 3d and Sn 3d proved the strong chemical interaction between SnS
2 quantum dots and MoS
2 nanosheets in the composite material. There is no obvious differences in the S 2p XPS spectrum (c) between MoS
2 and SnS
2. However, both S 2p
1/2 and S 2p
3/2 shifted to higher binding energies in the SnS
2/MoS
2 composite, which further confirmed the strong chemical interactions between S, Sn, and Mo in the SnS
2/MoS
2 composite.
. (a) Sn 3d XPS spectra of the SnS2 and SnS2/MoS2 samples, (b) Mo 3d XPS spectra of the MoS2 and SnS2/MoS2 samples, (c) S 2p XPS spectra of SnS2, MoS2 and SnS2/MoS2 samples.
UV-vis spectroscopy was used to study the optical absorption of the samples. As shown in a, all of the samples exhibited visible light absorption property. The SnS
2 sample absorbs light from UV light to visible light with wavelength shorter than 550 nm, while the MoS
2 nanosheets show an intense light absorption from UV light to almost the entire visible light region with an absorption band edge around 700 nm, which is similar to our previous studies [
43]. Compared with bare SnS
2 and MoS
2 samples, the SnS
2/MoS
2 nanocomposite showed much increasing light absorption in the visible light region beyond 700 nm, indicating that the composition of SnS
2 and MoS
2 can extend the light absorption capability of photocatalysts. For a crystalline semiconductor, the optical absorption near the band edge follows the equation [
48]: ahυ = A(*hυ* − *E
g*)
*n*/2, where *a*, *υ*, A and *E
g* are the absorption coefficient, light frequency, proportionality constant and band gap, respectively. *n* depends on whether the transition is direct (*n* = 1) or indirect (*n* = 4). The values of n and Eg were determined by the following steps: at first, plot ln(*ahυ*) vs ln(*hυ* − *E
g*), using an approximate value of *E
g*, and then determine the value of *n* with the slope of the straightest line near the band edge; second, plot (*ahυ*)
1/*n* vs *hυ* and then evaluate the band gap *E
g* by extrapolating the straightest line to the *hυ* axis intercept [
49]. According to the equation, the value of n is estimated to be 1 or 4 for SnS
2 and MoS
2, respectively. As shown in b,c, the band gaps of SnS
2 and MoS
2 are estimated to be about 1.90 and 1.46 eV, respectively.
. (a) Ultraviolet-visible absorption spectra of the different samples, (b) *ahv*-*hv* curve of SnS2, (c) *ahv*-*hv* curve of MoS2.
*3.2. Photocatalytic Activity*
The as-prepared SnS
2/MoS
2 nanocomposite exhibited photocatalytic H
2 evolution performance in CO
2 saturated aqueous solution under 300 W Xenon light irradiation. Blank experiments revealed that there was no H
2 and hydrogenated carbon products before and after light irradiation in the reaction cell without catalyst or light irradiation. a–c showed the yields of H
2, CH
4, and CO product on SnS
2/MoS
2, bare SnS
2 and bare MoS
2 catalyst, respectively under light irradiation. It can be seen that the amount of these products increased almost linearly with increasing irradiation time. Although the catalyst was dispersed in CO
2 saturated water, H
2 is the majority product from SnS
2/MoS
2 and pure SnS
2 photocatalyst. Only a small amount of CH
4 and CO products are observed. Compared with SnS
2/MoS
2 and pure SnS
2, the amount of CH
4 and CO are obviously increasing on pure MoS
2 catalyst. The increasing yield of CH
4 in the case of MoS
2 can be attributed to the accumulated electrons from charged excitons on catalyst surface [
43], which facilitates muti-electron reduction reactions, such as CO
2→CH
4 (8 electron transfer). Correspondingly, the production of H
2 on MoS
2 by the single electron reduction process (H
+ + e→H·) is much lower than that on SnS
2 as reported [
44,
45,
46]. According to a, the total amounts of H
2, CH
4, and CO generated on SnS
2/MoS
2 catalyst were 915.2, 81.52, and 33.0 μmol/g within 22 h (a). Therefore, the selectivity of the photogenerated electrons for H
2, CH
4, and CO generation are 81.1%, 11.7%, and 7.2%, respectively. Under the same condition, the total amount of H
2, CH
4, and CO generated on the SnS
2 catalyst were 463.1, 47.5, and 18.7 µmol/g within 22 h (b), corresponding to a selectivity of 68.9%, 28.3%, and 2.8% of the photogenerated electrons for generating H
2, CH
4, and CO, respectively. c shows the H
2, CH
4, and CO obtained on the pure MoS
2 during 22-h irradiation in CO
2 saturated water. The total amounts of H
2, CH
4, and CO obtained are 238.2, 170.4, and 63.8 μmol/g, responding to a selectivity of 24.2%, 69.3%, and 6.5% of the photogenerated electrons for H
2, CH
4, and CO production, respectively. Photocatalytic performances of pure MoS
2, pure SnS
2, and SnS
2/MoS
2 nanocomposite indicate that the SnS
2/MoS
2 nanocomposite can increase both the productivity and selectivity of H
2 from CO
2 saturated water under light irradiation. Especially, the amount of H
2 obtained on SnS
2/MoS
2 is almost double of that on SnS
2 within 22 h. The much improved H
2 productivity and selectivity of the SnS
2/MoS
2 composite catalyst may be ascribed to its superior electron-hole separation and faster charge transfer at the interface of an S-scheme heterojunction structure, which will be further discussed in the following sections.
. Photocatalytic performances of (a) SnS2/MoS2, (b) pure SnS2 and (c) pure MoS2 in CO2 saturated aqueous solution. (d) Comparison of photocatalytic H2 evolution performance of SnS2/MoS2 in different reaction mediums.
To investigate the role of CO
2 in H
2 evolution from CO
2 saturated water, comparative experiments were conducted in N
2 saturated pure water and N
2 saturated hydrochloric acid solution (pH 4). As shown in d, no H
2 was observed when CO
2 was replaced by N
2 to fill in the reaction cell. Only trace amount of H
2 was obtained in acidic N
2 saturated water. These results revealed that H
2 evolution may come from CO
2 reduction intermediates but not from direct water splitting.
*3.3. Photocatalytic Mechanisms*
To further understand the much higher performance and the mechanism of hydrogen evolution on SnS
2/MoS
2, an electrochemical analysis was conducted. a displays the electrochemical impedance spectroscopy (EIS) of SnS
2, MoS
2 and SnS
2/MoS
2. Compared with other samples, SnS
2/MoS
2 exhibited the minimum arc radius of the impedance spectrum, suggesting the more efficient charge transfer in the heterojunction structure, which could better promote the separation and migration of photoinduced charge carriers [
50,
51]. Linear sweep voltammetry (LSV) scans were performed to directly determine the over potentials for water reduction by different samples on FTO conducting substrate in 0.1 M Na
2SO
4 solution at pH 7. As shown in b, cathodic scans revealed the over potentials for water reduction were located at about −0.37 and −0.3 V vs. NHE for the MoS
2 and SnS
2 samples, respectively. Under the same conditions, the SnS
2/MoS
2 nanocomposite exhibited much lower reduction over potential of about −0.006 V, indicating the much easier water reduction kinetic on SnS
2/MoS
2 nanocomposite. The information on water reduction kinetics of these photocatalysts can be further proved by Tafel plots stemming from the corresponding LSV curves in b. As shown in c, the Tafel slopes for SnS
2, MoS
2 and SnS
2/MoS
2 samples are 0.3776, 0.4355, and 0.3696 Vdec
−1, respectively. Lower Tafel slope theoretically indicates faster water reduction kinetics, which facilitates photocatalytic water reduction and CO
2 hydrogenation over the SnS
2/MoS
2 nanocomposite.
. The photo-electrochemical performances of MoS2, SnS2 and SnS2/MoS2 in 0.1 M Na2SO4. (a) Impedance spectroscopy plots of MoS2, SnS2 and SnS2/MoS2 under light irradiation, (b) Cathodic current-potential scans of different samples in the dark, (c) Tafel plots of MoS2, SnS2 and SnS2/MoS2 obtained from the cathodic current-potential scan, (d) Mott-Schottky plots of the synthetic MoS2 and SnS2.
The flatband potentials (*E*
fb) of the SnS
2 and MoS
2 electrodes are measured by Mott-Schottky analysis, which was generated from the capacitance values measured at 1000 Hz in dark. As shown in d, the positive slopes of the plots indicate the as-prepared SnS
2 and MoS
2 are n-type semiconductor and then electrons are the majority charge carriers. The *E*
fb value was calculated from the intercept of the axis with potential values, which is −0.49 and −0.31 vs. NHE at pH 7 for SnS
2 and MoS
2, respectively. *E*
fb is strongly related to the bottom of the conduction band (*E*
cb) and is usually considered to be about 0.1 V below the *E*
cb for many n-type semiconductors [
52]. Therefore, the *E*
cb values for SnS
2 and MoS
2 are estimated as −0.59 and −0.41 vs. NHE at pH 7, respectively. These *E*
cb values enabled the thermodynamic feasibility of the conduction band electrons in SnS
2 and MoS
2 for H
+ reduction to H
2 because the E(H
+/H
2) = −0.41 V at pH 7.
. Schematic band energy alignment of SnS2 and MoS2 before contacting (a) and band bending at the interface (b).
According to the band positions calculated from the electrochemical analysis and the work functions of fewer layer MoS
2 and hexagonal SnS
2 [
53,
54], S-sheme or Type-I heterojunction between MoS
2 and SnS
2 may be formed [
8]. However, if Type-I was formed, the photogenerated electrons and holes in MoS
2 and SnS
2 will recombine in the interface and then largely suppress the photocatalytic activity. Therefore, S-scheme heterojunction should be constructed between MoS
2 and SnS
2 based on the photocatalytic performance, as shown in . According to the band structures of MoS
2 and SnS
2 (a), MoS
2 can be considered as an oxidative photocatalyst with a lower *E*
f, and SnS
2 is a reductive photocatalyst with a higher *E*
f. In this case, when the MoS
2 photocatalyst closely contacted with SnS
2, the electrons in SnS
2 spontaneously flowed to the MoS
2 until their Fermi level are the same. After that, an inner electric field (IEF) at the interface of SnS
2/MoS
2 heterojunction was produced and then impeded the continuous flow of electrons from SnS
2 to MoS
2. Under the impact of such an electric field, the band edge of SnS
2 would bend upward because of losing electrons, and the band edge of MoS
2 would bend downward because of accumulating electrons. Under light irradiation, the photogenerated electrons in the CB of MoS
2 will combine with the photogenerated holes in the VB of SnS
2 owing to inner electric field, leaving the powerful CB electrons of SnS
2 and VB holes of MoS
2 in the composite photocatalyst (b). As mentioned above, the *E*
cb of SnS
2 QDs and MoS
2 nanosheet were estimated to be about −0.59 and −0.41 vs. NHE at pH 7, respectively. A higher *E*
cb facilitates a faster hydrogen evolution process. Therefore, a S-scheme heterojunction formed in the interface of SnS
2/MoS
2 not only effectively separated the photogenerated electron-hole pairs, but also provided maximized redox ability of the photogenerated charge carriers, responsible for the more efficient hydrogen evolution performance of SnS
2/MoS
2.
*3.4. Photocatalytic Reaction Pathways*
To further investigate the mechanism of hydrogen evolution in CO
2 saturated solution under light irradiation, DRIFTS was employed to monitor the surface reaction intermediates. Before introducing CO
2, we analyzed the DRIFT spectrum in vacuum in the dark (a). Surfaces of the as-prepared MoS
2, SnS
2, and SnS
2/MoS
2 samples were dominated by strongly adsorbed H
2O and CO
2 even in the vacuum state. The absorption band around 1617 and 3300 cm
−1 were ascribed to bending vibration of H
2O and stretching vibration of –OH [
55], respectively, as shown in a. Additionally, the distinct peak of 666 and 2359 cm
−1 were ascribed to bending vibration and asymmetrical stretching vibration of CO
2, respectively [
56]. Additionally, the absorption band appeared around 1456 and 1650 cm
−1 were attributed to symmetrical and asymmetrical OCO stretching vibration in HOCOO− (HCO
3−) species [
56]. The absorption bands from bidentate or monodentate carbonate (–CO
32−) species appeared around 1561 cm
−1 [
43]. Unsaturated C–H bond absorption at 2847 and 2922 cm
−1 may be come from the S–C–H on the surface of catalysts, which come from surface organic contaminant during the sample preparation [
43]. A weak absorbance around 2514 cm
−1 was assigned to stretching vibration of S–H [
43].
*In situ* DRIFT spectra were recorded after introducing CO
2 and H
2O to further study CO
2 reduction mechanism under light irradiation. b showed the change of surface functional groups on pure MoS
2 within different irradiation time. Along with the increased irradiation time, six positive absorption bands emerged around 1065, 1285, 1358, 1528, 1783, and 2497 cm
−1, which were ascribed to the stretching vibration of C–O bond in –HC–OH [
43], the stretching vibration of C–O bond in –COOH group [
43], the stretching vibration of C–O bond in HCOO– [
57], the stretching vibration of C=O bond in HCOO– [
57], the stretching vibration of C=O bond in –COOH [
43] and the surface S–H group, respectively. Simultaneously, four negative absorption peaks appeared at 1439, 1608, 1671 and 3400 cm
−1, indicating the decreasing concentrations of the surface HCO
3− and H
2O species, respectively. The change of the absorption bands in DRIFT spectra revealed the reaction of HCO
3− and H
2O to generate –HC–OH, –COOH, HCOO–, and –H species under light irradiation on the surface of pure MoS
2. H
2 evolution during the CO
2 hydrogenation process may be mainly ascribed to the decomposition of these hydrogenated intermediates and the release of H radicals from –H under light irradiation.
The *in situ* DRIFT spectra on SnS
2 under the same condition are shown in c. Similar to the pure MoS
2, the absorption bands from HCO
3− (1420 and 1656 cm
−1) and H
2O (1602 and 3239 cm
−1) are gradually decreased along with the light irradiation increasing time. By contrast, six positive absorption bands emerged around 1112, 1288, 1347, 1523, 1794, and 2500 cm
−1, which are ascribed to the stretching vibration of C–O bond in –HC–OH, the stretching vibration of C–O bond in –COOH, the stretching vibration of C–O in HCOO–, the stretching vibration of C=O bond in HCOO–, the stretching vibration of C=O bond in –COOH and the surface S–H group, respectively. This suggested that HCO
3− and H
2O react with each other to generate –HC–OH, –COOH, HCOO– and –H intermediates on the SnS
2 surface, similar to that on MoS
2. However, there are some differences on the *in situ* DRIFT spectrum between MoS
2 and SnS
2 for CO
2 reduction. First of all, the absorption intensity from S–H group in SnS
2 is obviously larger than that in MoS
2 under the same condition. The better H
2 evolution performance of SnS
2 may partially originate from the higher dissociation of H from S–H.
DRIFT spectra in d revealed the CO
2 hydrogenation process on SnS
2/MoS
2. Under light irradiation, the absorption bands from HCO
3− (1435 and 1635 cm
−1) and surface –OH (3353 cm
−1) are dramatically decreased along with the obviously increasing absorption bands from –COOH and HCOO– species, indicating the reaction of surface HCO
3− and –OH to generate –HC–OH, –COOH and HCOO– intermediates. Different from MoS
2 and SnS
2, it was found the absorption band from the surface S–H group did not emerge after light irradiation on SnS
2/MoS
2. The reasons for the disappearance of the S–H absorption band on SnS
2/MoS
2 may be ascribed to the faster reaction kinetics in S-scheme structure for H radials release from S–H species once they are formed on the surface. Furthermore, it was found the absorption bands from –COOH species on SnS
2/MoS
2 are broader and stronger compared with that on bare MoS
2 and SnS
2, indicating the diversity of the reactive sites for –COOH generation. Considering the appearance of –HC–OH, –COOH and HCOO– species on the three samples and their increasing concentrations under light irradiation, H
2 evolution may partly arise from the decomposition of these carbon-hydrogen species. Both the enhanced concentrations of carbon-hydrogen intermediate species and the faster dissociation of H from S–H on the SnS
2/MoS
2 surface may be responsible for the higher H
2 evolution performance.
. CO2 adsorption and activation on different samples revealed by *in-situ* DRIFT observation. (a) DRIFT spectra of different samples under vacuum in the dark. b, c, d. *In situ* DRIFT spectra recorded after ambient temperature adsorption of CO2/H2O on pure MoS2 (b), pure SnS2 (c) and SnS2/MoS2 (d) under 300 W Xe arc lamp irradiation.
. Schematic illustration of H2 generation process in CO2 saturated solution on SnS2/MoS2 catalyst.
The phenomena observed from in situ DRIFT spectra on MoS
2, SnS
2, and SnS
2/MoS
2 confirmed the reasons for the better photocatalytic activity of the SnS
2/MoS
2 nanocomposite. The whole H
2 generation process on SnS
2/MoS
2 is illustrated in . Under light irradiation, CO
2 is reduced in water by photogenerated electrons to generate carbon-hydrogen intermediates such as –HC–OH, –COOH and HCOO– species, as well as –H species on the surface. Most of these intermediates are not stable, and their decomposition generates H
2 under light irradiation. Furthermore, S-scheme structure decreased the electron-hole recombination and increased the redox ability of the photogenerated charge carriers for both CO
2 reduction and H
2O oxidation, respectively. All of the above reasons endowed the higher efficiency of H
2 evolution by use of SnS
2/MoS
2 in CO
2 saturated solution.
Conceptualization, S.S.; Methodology, X.C., X.L. and L.L.; Validation, X.L.; Formal Analysis, X.C., X.L. and L.L.; Investigation, X.C. and X.L.; Resources, S.S. and X.C.; Writing-Original Draft Preparation, X.C. and X.L.; Writing-Review & Editing, S.S., X.C., J.P. and X.L.; Supervision, S.S.; Project Administration, S.S.; Funding Acquisition, S.S.
This work was financially supported by Natural Science Foundation of Shanghai (No. 21ZR1401600).
Not applicable.
Not applicable.
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.
