Simultaneous Conduction and Valence Band Regulation of Indium-Based Quantum Dots for Efficient H2 Photogeneration

Indium-based chalcogenide semiconductors have been served as the promising candidates for solar H2 evolution reaction, however, the related studies are still in its infancy and the enhancement of efficiency remains a grand challenge. Here, we report that the photocatalytic H2 evolution activity of quantized indium chalcogenide semiconductors could be dramatically aroused by the co-decoration of transition metal Zn and Cu. Different from the traditional metal ion doping strategies which only focus on narrowing bandgap for robust visible light harvesting, the conduction and valence band are coordinately regulated to realize the bandgap narrowing and the raising of thermodynamic driving force for proton reduction, simultaneously. Therefore, the as-prepared noble metal-free Cu0.4-ZnIn2S4 quantum dots (QDs) exhibits extraordinary activity for photocatalytic H2 evolution. Under optimal conditions, the Cu0.4-ZnIn2S4 QDs could produce H2 with the rate of 144.4 μmol h−1 mg−1, 480-fold and 6-fold higher than that of pristine In2S3 QDs and Cu-doped In2S3 QDs counterparts respectively, which is even comparable with the state-of-the-art cadmium chalcogenides QDs.


Introduction
Artificial photosynthesis is regarded as a promising approach to convert solar energy into usable energy forms, such as molecular hydrogen (H 2 ), to resolve the energy crisis and environment pollution [1][2][3][4][5]. Hence, synthesis of efficient and low-cost photocatalysts is with vital importance for solar-to-fuel conversion. The semiconductor quantum dots (QDs) which the size is less than their exciton Bohr radius in three dimensions have been attracted tremendous attention in this field owing their unique properties, such as large surface area, abundant surface active sites, multiple exciton generation, short charge transfer distance and so on [6][7][8][9][10][11]. Among the popular chalcogenide semiconductor-based QDs in artificial photocatalysis, the indium-based chalcogenides seem to be more promising photocatalysts due to their less toxic, noble metal-free as well as the potential visible light harvesting [12][13][14][15]. However, the photocatalytic H 2 evolution activity of indiumbased photocatalysts is relatively faint so far, especially compared with their cadmium chalcogenide analogs [16][17][18][19], so more works remained to be done in optimizing the properties of indium-based QDs. In general, though the bandgap of bulk In 2 S 3 (~2.3 eV) is suitable for absorbing visible light, the bandgap of quantized indium chalcogenide nanoparticles is usually larger than 3.0 eV as a result of quantum confinement effect [19][20][21][22][23]. On the other hand, the more negative conduction band edge corresponds to the stronger thermodynamic driving force for proton reduction [24,25], which would result in the higher photocatalytic performance. To balance the above two contradictory effects of narrowing the bandgap and upshifting the conduction band edge, the conduction and valence band should be well designed.

Instruments and Characterizations
UV-vis spectra were measured with a Shimadzu UV-2600PC spectrophotometer (Shimadzu Corp., Kyoto, Japan). Fluorescence measurements were carried out with a Hitachi (model F-4600) spectrophotometer (Hitachi High-Tech Corp., Tokyo, Japan) at room temperature. TEM images were obtained on a JEM 2100 (JEOL Co. Ltd., Tokyo, Japan) operating at 200 kV. X-ray diffraction (XRD) pattern was obtained by using Bruker D8 Focus (Bruker Corp., Billerica, MA, USA) under Cu-K α radiation. X-ray photoelectron spectroscopy (XPS) measurements were taken on an ESCALAB 250 spectrophotometer (ThermoFisher Scientific Corp., Waltham, MA, USA) with Al-K α radiation. The binding energy scale was calibrated using the C 1s peak at 284.60 eV. All pH measurements were made with a Model pHS-3C meter (Mettler Toledo FE20, Mettler Toledo (Shanghai) Co. Ltd., Shanghai, China). The generated amount of H 2 was characterized by GC analysis (GC-2014 Shimadzu, Shimadzu Corp., Kyoto, Japan) using N 2 as the carrier gas with a molecular sieve column (5 Å; 30 m × 0.53 mm) and a thermal conductivity detector.

Synthesis of the Cu-Doped ZnIn 2 S 4 and In 2 S 3 QDs
The indium-based chalcogenides QDs were synthesized through a simple hydrothermal method referring to the previous reported method with some revisions [22]. Taking the Cu 0.4 -ZnIn 2 S 4 QDs (0.4 represents the precursor molar ratio of Cu with Zn) for example, 0.85 mmol of zinc acetate, 1.70 mmol of indium acetate, 0.34 mmol of copper acetate monohydrate and 3.25 mmol of L-cysteine were dissolved in 30 mL of ultrapure water and adjusted the pH to 9.0. Following this, 3.25 mmol of thioacetamide was added and the mixture was heated to 110 • C for 4 h after vigorous stirring. After reaction, the solution was swiftly cooled to room temperature, and then the product was precipitated and purified with isopropanol and water. The obtained Cu-doped ZnIn 2 S 4 QDs were dispersed in 40 mL of ultrapure water for further use.
The QDs with different contents of Cu and/or Zn were synthesized by regulating the amount of copper and/or zinc precursors (copper/zinc acetate), while the amount of thioacetamide was also changed with stoichiometric ratio. For the pristine In 2 S 3 QDs, the reaction was carried out with the same expect for the absence of copper and zinc precursors.

General Procedure for Photocatalytic H 2 Evolution
The photocatalytic reactions were performed under 460 nm LEDs irradiation with the Ni 2+ as cocatalysts and H 2 A as the electron donor. Generally, 1.6 mg of QDs, 10 µg of nickel acetate tetrahydrate and 400 mg of ascorbic acid were added to a 20 mL of Pyrex tube. The total reaction volume of the mixture was 6.0 mL by adding ultrapure water and the pH of the solution was adjusted to 5.2 by HCl or NaOH. Prior to irradiation, the sample was sealed and de-aerated by bubbling N 2 for 10 min to remove the dissolved oxygen, and then 1.0 mL of CH 4 was injected the system to serve as internal standard. The amount of evolved H 2 gas was quantified by comparing the area ratio of CH 4 to H 2 and the response factor of CH 4 /H 2 in the gas chromatography (GC).

The Calculated of AQY for Cu 0.4 -ZnIn 2 S 4 QDs
For the optimal photocatalytic H 2 evolution, the irradiation area was 3.0 cm 2 by 460 nm LEDs (light intensity 100 mW·cm −2 ). The calculation of apparent quantum efficiency (Φ) is according to the following equation: wherein n H 2 is the amount of photo-generated H 2 and N is the corresponding amount of incident photon. The amount of incident photon was calculated from the irradiation time, the irradiation area and the illumination power. It showed that about 5.5 mL of H 2 was generated after 1.0 h irradiation. From the combined measurements of the amount of photo-generated H 2 and the corresponding amount of incident photon, the apparent quantum yield was calculated to be 11.8%.

Microstructure Characteristics and Composition Analysis
The photocatalysts with different contents of Zn and/or Cu decorated In 2 S 3 QDs were synthesized by regulating the species and amount of precursors. According to the photocatalytic H 2 evolution performance, the as-prepared pristine In 2 S 3 , ZnIn 2 S 4 , Cu 0.4 -In 2 S 3 and Cu 0.4 -ZnIn 2 S 4 QDs were selected for the structure and morphology characterizations. As shown in Figure 1, the diffraction peaks of the pristine In 2 S 3 can be indexed as the tetragonal phase In 2 S 3 (JCPDS no. 25-0390) [27]. However, though the XRD patterns change clearly after the co-decoration of Zn and Cu ions (Cu 0.4 -ZnIn 2 S 4 ), the synthesized counterparts exhibit three characteristic peaks at about 28.0 • , 47.5 • and 55.3 • , which can be assigned to (112), (024) and (132) lattice plane of tetragonal phase of Cu 0.412 In 0.412 Zn 0.175 S (JCPDS no. 47-1371) [20]. What's more, the diffraction peaks of Zn or Cu decorated In 2 S 3 QDs (ZnIn 2 S 4 or Cu 0.4 -In 2 S 3 ) are well corresponding to that of Cu 0.4 -ZnIn 2 S 4 , demonstrating the identical crystal structure of the above QDs. Additionally, the varied contents of Zn or Cu hardly shifted the diffraction peaks of the as-prepared indium-based chalcogenide QDs ( Figure S1), which probably owes to the similar radius of Zn 2+ (0.74 Å), Cu + (0.74 Å) and In 3+ (0.76 Å) [28,29], demonstrating the preservation of crystal structure during the transition metal ions decoration. The size and morphology of the synthesized various QDs were also investigated. According to Debye-Scherrer formula, the broad diffraction peaks implied the ultra-small size of the nanocrystals in three dimensions [30,31]. While the similar full width at half maximum (FWHM) of different contents of Zn and/or Cu decorated In2S3 QDs indicated the approximate size ( Figure 1). It should be mentioned that FWHM of pristine In2S3 QDs was smaller than the other QDs, which implied the slightly larger size of In2S3 QDs. Indeed, the diameters in terms of XRD patterns for In2S3, Cu0.4-In2S3, ZnIn2S4 and Cu0.4-ZnIn2S4 QDs along the (024) plane is 3.8 nm, 2.6 nm, 2.6 nm and 3.1 nm, respectively (the difference of size could be ignored in the photocatalytic activity comparison owing to the seriously faint photocatalytic H2 evolution of In2S3 QDs). The diameters of various QDs were further characterized by the transmission electron microscope (TEM) images. As shown in Figure 2a, the In2S3 QDs displayed the near-spherical nanocrystals with the diameter of about 3.9 nm, while the other three QDs (Figure 2b-d) were also near-spherical nanocrystals with the diameter ranging from 2.7~3.0 nm, slightly smaller than the pristine In2S3 QDs. What's more, the high-resolution TEM (HR-TEM) images revealed that the lattice spacing of all the QDs was 3.1 Å [20,32,33], which was corresponding to the (112) plane of tetragonal phase of In2S3, ZnIn2S4 or their doped counterparts, further demonstrating that the crystal structure of tetragonal phase was well preserved after the introduction of Zn and/or Cu ions into the In2S3 QDs. The size and morphology of the synthesized various QDs were also investigated. According to Debye-Scherrer formula, the broad diffraction peaks implied the ultra-small size of the nanocrystals in three dimensions [30,31]. While the similar full width at half maximum (FWHM) of different contents of Zn and/or Cu decorated In 2 S 3 QDs indicated the approximate size ( Figure 1). It should be mentioned that FWHM of pristine In 2 S 3 QDs was smaller than the other QDs, which implied the slightly larger size of In 2 S 3 QDs. Indeed, the diameters in terms of XRD patterns for In 2 S 3 , Cu 0.4 -In 2 S 3 , ZnIn 2 S 4 and Cu 0.4 -ZnIn 2 S 4 QDs along the (024) plane is 3.8 nm, 2.6 nm, 2.6 nm and 3.1 nm, respectively (the difference of size could be ignored in the photocatalytic activity comparison owing to the seriously faint photocatalytic H 2 evolution of In 2 S 3 QDs). The diameters of various QDs were further characterized by the transmission electron microscope (TEM) images. As shown in Figure 2a, the In 2 S 3 QDs displayed the near-spherical nanocrystals with the diameter of about 3.9 nm, while the other three QDs (Figure 2b-d) were also nearspherical nanocrystals with the diameter ranging from 2.7~3.0 nm, slightly smaller than the pristine In 2 S 3 QDs. What's more, the high-resolution TEM (HR-TEM) images revealed that the lattice spacing of all the QDs was 3.1 Å [20,32,33], which was corresponding to the (112) plane of tetragonal phase of In 2 S 3 , ZnIn 2 S 4 or their doped counterparts, further demonstrating that the crystal structure of tetragonal phase was well preserved after the introduction of Zn and/or Cu ions into the In 2 S 3 QDs. To gain more insight into the chemical composition and valence state of metal elements in the prepared various QDs, the X-ray photoelectron spectroscopy (XPS) was explored. The peaks of Cu, Zn, In and S elements were clearly observed in the XPS survey spectrum of Cu0.4-ZnIn2S4 QDs [19,34,35], while the In2S3, ZnIn2S4 and Cu0.4-In2S3 also clearly manifested their respective peaks without the presence of other metal elements (Figure 3a), suggesting the rational chemical composition in the designed QDs. Furthermore, the binding energy of Cu 2p3/2 and Cu 2p1/2 in both Cu0.4-In2S3 and Cu0.4-ZnIn2S4 QDs was at 932.3 eV and 952.1 eV (Figure 3b), suggesting the monovalent Cu in the QDs, which could be further demonstrated by the absence of satellite peak [36]. The decrease of valence state of Cu probably owes to the reduction by L-cysteine. What's more, the binding energies of Zn 2p (1022.1 eV for 2p3/2 and 1045.2 eV for 2p1/2) and In 3d (444.6 eV for 3d5/2 and 452.2 eV for 3d3/2) in Cu0.4-ZnIn2S4 QDs (Figure 3c,d), suggested that the chemical states of Zn and In are +2 and +3, respectively, which were in good agreement with the previous reports on Zn-In-S ternary semiconductors [37,38]. Additionally, the features and positions of In 3d, Zn 2p and S 2p peaks (Figure 3c,d and Figure S2) remained the To gain more insight into the chemical composition and valence state of metal elements in the prepared various QDs, the X-ray photoelectron spectroscopy (XPS) was explored. The peaks of Cu, Zn, In and S elements were clearly observed in the XPS survey spectrum of Cu 0.4 -ZnIn 2 S 4 QDs [19,34,35], while the In 2 S 3 , ZnIn 2 S 4 and Cu 0.4 -In 2 S 3 also clearly manifested their respective peaks without the presence of other metal elements (Figure 3a), suggesting the rational chemical composition in the designed QDs. Furthermore, the binding energy of Cu 2p 3/2 and Cu 2p 1/2 in both Cu 0.4 -In 2 S 3 and Cu 0.4 -ZnIn 2 S 4 QDs was at 932.3 eV and 952.1 eV (Figure 3b), suggesting the monovalent Cu in the QDs, which could be further demonstrated by the absence of satellite peak [36]. The decrease of valence state of Cu probably owes to the reduction by L-cysteine. What's more, the binding energies of Zn 2p (1022.1 eV for 2p 3/2 and 1045.2 eV for 2p 1/2 ) and In 3d (444.6 eV for 3d 5/2 and 452.2 eV for 3d 3/2 ) in Cu 0.4 -ZnIn 2 S 4 QDs (Figure 3c,d), suggested that the chemical states of Zn and In are +2 and +3, respectively, which were in good agreement with the previous reports on Zn-In-S ternary semiconductors [37,38]. Additionally, the features and positions of In 3d, Zn 2p and S 2p peaks (Figure 3c,d and Figure S2) remained the same in the four QDs, indicating the similar lactic framework and coordination environment of the indium-based chalcogenide QDs. Combining all the above results, we believe that by regulating the species and amount of precursors, the Zn and/or Cu decorated In 2 S 3 QDs with the identical crystal structure and similar size have been successfully prepared. same in the four QDs, indicating the similar lactic framework and coordination environment of the indium-based chalcogenide QDs. Combining all the above results, we believe that by regulating the species and amount of precursors, the Zn and/or Cu decorated In2S3 QDs with the identical crystal structure and similar size have been successfully prepared.

Photocatalytic H2 Evolution
The photocatalytic activity of the as-prepared QDs were investigated by using the QDs as absorber and in situ loaded Ni 2+ as cocatalysts. It indicated that without the introduction of Cu, there was no H2 to be detected for the ZnIn2S4 QDs, and only trace amount of H2 was produced for In2S3 QDs under 460 nm LEDs irradiation (Figure 4a). However, the photocatalytic H2 evolution activity dramatically increased after the doping of Cu for both In2S3 and ZnIn2S4 QDs, probably owing to the outstanding visible light harvesting by the narrowed bandgap. Therefore, the photocatalytic H2 evolution activities for different contents of Cu decorated QDs were further explored. As shown in Figure S3, the photocatalytic H2 evolution activity of both Cu-doped In2S3 and ZnIn2S4 QDs increased with the

Photocatalytic H 2 Evolution
The photocatalytic activity of the as-prepared QDs were investigated by using the QDs as absorber and in situ loaded Ni 2+ as cocatalysts. It indicated that without the introduction of Cu, there was no H 2 to be detected for the ZnIn 2 S 4 QDs, and only trace amount of H 2 was produced for In 2 S 3 QDs under 460 nm LEDs irradiation (Figure 4a). However, the photocatalytic H 2 evolution activity dramatically increased after the doping of Cu for both In 2 S 3 and ZnIn 2 S 4 QDs, probably owing to the outstanding visible light harvesting by the narrowed bandgap. Therefore, the photocatalytic H 2 evolution activities for different contents of Cu decorated QDs were further explored. As shown in Figure S3, the photocatalytic H 2 evolution activity of both Cu-doped In 2 S 3 and ZnIn 2 S 4 QDs increased with the increased Cu contents at first and then decreased. This is due that the introduction of Cu would decrease the bandgap of QDs, which resulted in the enhanced photocatalytic activity. While the excess introduction of Cu would lead to the decreased redox driving force as well as the formation of Cu defects for non-radiative carrier recombination centers, which was unfavorable for the photocatalytic H 2 evolution.
increased Cu contents at first and then decreased. This is due that the introduction of Cu would decrease the bandgap of QDs, which resulted in the enhanced photocatalytic activity. While the excess introduction of Cu would lead to the decreased redox driving force as well as the formation of Cu defects for non-radiative carrier recombination centers, which was unfavorable for the photocatalytic H2 evolution. The photocatalytic H2 evolution rate for optimal Cu-doped ZnIn2S4 (Cu0.4-ZnIn2S4) QDs could reach to 144.4 μmol h −1 mg −1 , which was nearly 480-fold and 6-fold higher than that of pristine In2S3 and Cu0.4-In2S3 QDs under identical conditions respectively ( Figure  4a), implying the significant impact of Zn introduction for the enhancement of photocatalytic performance. Apparently, the photocatalytic H2 evolution rate is even comparable to the state-of-the-art cadmium chalcogenides QDs [26,39]. Control experiments also demonstrated that without the Ni 2+ , the photocatalytic activity would decrease by three fourths, indicated the vital role of cocatalysts. Long-time H2 evolution performance of the Cu0.4-ZnIn2S4 QDs was also examined under optimal conditions. As shown in Figure 4b, the rate was decreased with the irradiation time probably owing to the photocurrosion, however, ~1013 μmol of H2 could be obtained from 6.0 mL aqueous solution within 16 h irradiation, giving the TON of 20260 per QD. And the AQY of Cu0.4-ZnIn2S4 QDs could reach 11.8% at 460 nm.

Mechanism for Photocatalytic H2 Production
To shed light on the effect of the introduction of transition metal ions, the UV-visible absorption spectra of various QDs were carried out. As shown in Figure 5a, the In2S3 QDs exhibited faint absorption for visible light, which was corresponding to their negligible photocatalytic activity. As the intrinsic bandgap of bulk In2S3 is 2.3 eV [40][41][42], the faint visible light absorption of the as-prepared In2S3 QDs was mainly owing to broadening bandgap resulted by the quantum confinement effect. Further investigation indicated that the introduction of Zn and Cu elements played the opposite effect for the In2S3 QDs [43]. The absorption onset of ZnIn2S4 QDs was less than 400 nm, implied the broader bandgap The photocatalytic H 2 evolution rate for optimal Cu-doped ZnIn 2 S 4 (Cu 0.4 -ZnIn 2 S 4 ) QDs could reach to 144.4 µmol h −1 mg −1 , which was nearly 480-fold and 6-fold higher than that of pristine In 2 S 3 and Cu 0.4 -In 2 S 3 QDs under identical conditions respectively (Figure 4a), implying the significant impact of Zn introduction for the enhancement of photocatalytic performance. Apparently, the photocatalytic H 2 evolution rate is even comparable to the state-of-the-art cadmium chalcogenides QDs [26,39]. Control experiments also demonstrated that without the Ni 2+ , the photocatalytic activity would decrease by three fourths, indicated the vital role of cocatalysts. Long-time H 2 evolution performance of the Cu 0.4 -ZnIn 2 S 4 QDs was also examined under optimal conditions. As shown in Figure 4b, the rate was decreased with the irradiation time probably owing to the photocurrosion, however,~1013 µmol of H 2 could be obtained from 6.0 mL aqueous solution within 16 h irradiation, giving the TON of 20260 per QD. And the AQY of Cu 0.4 -ZnIn 2 S 4 QDs could reach 11.8% at 460 nm.

Mechanism for Photocatalytic H 2 Production
To shed light on the effect of the introduction of transition metal ions, the UV-visible absorption spectra of various QDs were carried out. As shown in Figure 5a, the In 2 S 3 QDs exhibited faint absorption for visible light, which was corresponding to their negligible photocatalytic activity. As the intrinsic bandgap of bulk In 2 S 3 is 2.3 eV [40][41][42], the faint visible light absorption of the as-prepared In 2 S 3 QDs was mainly owing to broadening bandgap resulted by the quantum confinement effect. Further investigation indicated that the introduction of Zn and Cu elements played the opposite effect for the In 2 S 3 QDs [43]. The absorption onset of ZnIn 2 S 4 QDs was less than 400 nm, implied the broader bandgap after the introduction of Zn. However, the Cu-doped In 2 S 3 QDs displayed the robust visible light harvesting and the absorption tail could rise up to 700 nm, implied the narrower bandgap. What's more, the doping of Cu would also narrow the bandgap of ZnIn 2 S 4 QDs, and the absorption onset of Cu 0.4 -ZnIn 2 S 4 QDs redshifted to 600 nm [44]. Therefore, the introduction of Cu makes the main contribution for the bandgap narrowing.
after the introduction of Zn. However, the Cu-doped In2S3 QDs displayed the robust visible light harvesting and the absorption tail could rise up to 700 nm, implied the narrower bandgap. What's more, the doping of Cu would also narrow the bandgap of ZnIn2S4 QDs, and the absorption onset of Cu0.4-ZnIn2S4 QDs redshifted to 600 nm [44]. Therefore, the introduction of Cu makes the main contribution for the bandgap narrowing. Although the optical density of Cu0.4-ZnIn2S4 was slightly higher than that of Cu0.4-In2S3 at 460 nm (Figure 5a), the difference of absorption ability hardly resulted in the 6fold enhancement of photocatalytic activity. Therefore, the band positions of the four QDs were further investigated by Tauc plot and XPS valence band spectra. Figure 5b revealed that the bandgap of In2S3, ZnIn2S4 and their Cu-doped counterparts were 2.98 eV, 3.32 eV, 1.93 eV and 2.25 eV, respectively, which was consistent with the results of absorption spectra. While the valence band edges of In2S3 and ZnIn2S4 were close, at 1.20 V and 1.22 V (vs. NHE) respectively (Figure 5c), which was owing that the valence band of the chalcogenide semiconductors were mainly composed by S 3p orbitals [45]. However, the introduction of Cu ions would create a Cu dopant level above the pristine valence band edge [3,46], causing their obvious upshift (ca. 0.8 V). Combined with the difference in Although the optical density of Cu 0.4 -ZnIn 2 S 4 was slightly higher than that of Cu 0.4 -In 2 S 3 at 460 nm (Figure 5a), the difference of absorption ability hardly resulted in the 6-fold enhancement of photocatalytic activity. Therefore, the band positions of the four QDs were further investigated by Tauc plot and XPS valence band spectra. Figure 5b revealed that the bandgap of In 2 S 3 , ZnIn 2 S 4 and their Cu-doped counterparts were 2.98 eV, 3.32 eV, 1.93 eV and 2.25 eV, respectively, which was consistent with the results of absorption spectra. While the valence band edges of In 2 S 3 and ZnIn 2 S 4 were close, at 1.20 V and 1.22 V (vs. NHE) respectively (Figure 5c), which was owing that the valence band of the chalcogenide semiconductors were mainly composed by S 3p orbitals [45]. However, the introduction of Cu ions would create a Cu dopant level above the pristine valence band edge [3,46], causing their obvious upshift (ca. 0.8 V). Combined with the difference in bandgap, the conduction band edges of In 2 S 3 and ZnIn 2 S 4 could be calculated at −1.78 V and −2.10 V, indicated the 0.32 V upshift after the introduction of Zn, which could also be observed for the comparison of Cu 0.4 -In 2 S 3 and Cu 0.4 -ZnIn 2 S 4 (Figure 5d). This is because that the conduction band minimum is mainly composed of hybrid d and sp orbitals of the metal cations [47], therefore, the introduction of Zn would result in the change of conduction band from In 5s5p to more negative hybrid orbitals of In 5s5p and Zn 4s4p [46,48]. The upshift of the conduction band edges gave rise to a higher thermodynamic driving force of proton reduction, and hence remarkably accelerating the H 2 evolution rate in water splitting.
Next, we employed the steady-state emission quenching experiments to evaluate the charge transfer process. Take the Cu 0.4 -ZnIn 2 S 4 QDs as example, excitation of QDs at 430 nm would result in strong luminescence at 630 nm (Figure 6a), which roughly corresponded to the bandgap energy of Cu 0.4 -ZnIn 2 S 4 QDs, indicating that this is the band edge emission of Cu 0.4 -ZnIn 2 S 4 QDs. However, the emission intensity was dramatically quenched with the adding of Ni 2+ (Figure 6a), indicating that the binding of Ni 2+ with the QDs would result in the electron transfer from the QDs to Ni 2+ and consequently inhibit the radiative recombination of photogenerated electron-hole pairs [49]. What's more, after adding electron donor, H 2 A, into the QDs aqueous solution, the emission would also be quenched obviously (Figure 6b), demonstrating the hole transfer from QDs to H 2 A [50].
Nanomaterials 2021, 11, x FOR PEER REVIEW 9 of 12 bandgap, the conduction band edges of In2S3 and ZnIn2S4 could be calculated at −1.78 V and −2.10 V, indicated the 0.32 V upshift after the introduction of Zn, which could also be observed for the comparison of Cu0.4-In2S3 and Cu0.4-ZnIn2S4 (Figure 5d). This is because that the conduction band minimum is mainly composed of hybrid d and sp orbitals of the metal cations [47], therefore, the introduction of Zn would result in the change of conduction band from In 5s5p to more negative hybrid orbitals of In 5s5p and Zn 4s4p [46,48]. The upshift of the conduction band edges gave rise to a higher thermodynamic driving force of proton reduction, and hence remarkably accelerating the H2 evolution rate in water splitting. Next, we employed the steady-state emission quenching experiments to evaluate the charge transfer process. Take the Cu0.4-ZnIn2S4 QDs as example, excitation of QDs at 430 nm would result in strong luminescence at 630 nm (Figure 6a), which roughly corresponded to the bandgap energy of Cu0.4-ZnIn2S4 QDs, indicating that this is the band edge emission of Cu0.4-ZnIn2S4 QDs. However, the emission intensity was dramatically quenched with the adding of Ni 2+ (Figure 6a), indicating that the binding of Ni 2+ with the QDs would result in the electron transfer from the QDs to Ni 2+ and consequently inhibit the radiative recombination of photogenerated electron-hole pairs [49]. What's more, after adding electron donor, H2A, into the QDs aqueous solution, the emission would also be quenched obviously (Figure 6b), demonstrating the hole transfer from QDs to H2A [50]. On the basis of the above results, we proposed that the introduction of Zn and Cu ions would regulate the conduction and valence band of indium chalcogenide QDs, respectively. As the upshift of valence band edge was significantly larger than that of conduction band edge, the bandgap of Cu doped QDs was clearly narrowed and gave rise to the robust visible light harvesting. Therefore, the photocatalytic activity of QDs remarkably enhanced after the doping of Cu. On the other hand, with the upshift of the conduction band edge, the thermodynamic driving force of proton reduction raised, thus the Cudoped ZnIn2S4 QDs exhibited the superior photocatalytic activity. Under visible light irradiation, the photogenerated electron of Cu0.4-ZnIn2S4 QDs would transfer to the surface Ni species, the H2 evolution cocatalysts, which would assist the proton reduction and formation of H2. Simultaneously, the hole transferred to the surface of QDs and oxidized the electron donor to accomplish the whole reaction (Figure 7). On the basis of the above results, we proposed that the introduction of Zn and Cu ions would regulate the conduction and valence band of indium chalcogenide QDs, respectively. As the upshift of valence band edge was significantly larger than that of conduction band edge, the bandgap of Cu doped QDs was clearly narrowed and gave rise to the robust visible light harvesting. Therefore, the photocatalytic activity of QDs remarkably enhanced after the doping of Cu. On the other hand, with the upshift of the conduction band edge, the thermodynamic driving force of proton reduction raised, thus the Cu-doped ZnIn 2 S 4 QDs exhibited the superior photocatalytic activity. Under visible light irradiation, the photogenerated electron of Cu 0.4 -ZnIn 2 S 4 QDs would transfer to the surface Ni species, the H 2 evolution cocatalysts, which would assist the proton reduction and formation of H 2 . Simultaneously, the hole transferred to the surface of QDs and oxidized the electron donor to accomplish the whole reaction (Figure 7).

Conclusions
In summary, we have regulated the bandgap and conduction/valence band levels of indium-based chalcogenides QDs simultaneously by introducing transition metal ions to construct efficient and noble metal-free photocatalysts. The bandgap is mainly determined by the doping of Cu, which could remarkably upshift the valence band edge. While the introduction of Zn would slightly enhance the conduction band level and provide higher driving force for proton reduction. Therefore, the as-prepared Cu-doped ZnIn2S4 QDs exhibit outstandingly higher photocatalytic performance. Under the optimal conditions, the Cu0.4-ZnIn2S4 QDs could produce H2 with the rate of 144.4 μmol h −1 mg −1 , which is even comparable to the state-of-the-art cadmium chalcogenide QDs. We believe that this approach has given a much deeper recognition on the band engineering, which can be extended to related systems as an effective strategy for the design of photocatalysts.
Supplementary Materials: The following are available online at www.mdpi.com/xxx/s1, Table S1: Comparison of photocatalytic H2 evolution with reported cadmium chalcogenide QDs-based systems; Figure S1: The XRD patterns of Cu-doped Zn-In-S QDs (the ratio in the graph represents the molar ratio of Cu:Zn:In); Figure S2: The high-resolution XPS spectra of S 2p of In2S3, Cu0.4-In2S3, ZnIn2S4 and Cu0.4-ZnIn2S4 QDs; Figure S3

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflicts of interest.

Conclusions
In summary, we have regulated the bandgap and conduction/valence band levels of indium-based chalcogenides QDs simultaneously by introducing transition metal ions to construct efficient and noble metal-free photocatalysts. The bandgap is mainly determined by the doping of Cu, which could remarkably upshift the valence band edge. While the introduction of Zn would slightly enhance the conduction band level and provide higher driving force for proton reduction. Therefore, the as-prepared Cu-doped ZnIn 2 S 4 QDs exhibit outstandingly higher photocatalytic performance. Under the optimal conditions, the Cu 0.4 -ZnIn 2 S 4 QDs could produce H 2 with the rate of 144.4 µmol h −1 mg −1 , which is even comparable to the state-of-the-art cadmium chalcogenide QDs. We believe that this approach has given a much deeper recognition on the band engineering, which can be extended to related systems as an effective strategy for the design of photocatalysts.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/nano11051115/s1, Table S1: Comparison of photocatalytic H 2 evolution with reported cadmium chalcogenide QDs-based systems; Figure S1: The XRD patterns of Cu-doped Zn-In-S QDs (the ratio in the graph represents the molar ratio of Cu:Zn:In); Figure S2: The high-resolution XPS spectra of S 2p of In 2 S 3 , Cu 0.4 -In 2 S 3 , ZnIn 2 S 4 and Cu 0.4 -ZnIn 2 S 4 QDs; Figure

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.