Sustainable Hydrogen Production from Starch Aqueous Suspensions over a Cd0.7Zn0.3S-Based Photocatalyst

We explored the photoreforming of rice and corn starch with simultaneous hydrogen production over a Cd0.7Zn0.3S-based photocatalyst under visible light irradiation. The photocatalyst was characterized by UV–vis diffuse reflectance spectroscopy, X-ray diffraction, and X-ray photoelectron spectroscopy. The influence of starch pretreatment conditions, such as hydrolysis temperature and alkaline concentration, on the reaction rate was studied. The maximum rate of H2 evolution was 730 μmol·h−1·g−1, with AQE = 1.8% at 450 nm, in the solution obtained after starch hydrolysis in 5 M NaOH at 70 °C. The composition of the aqueous phase of the suspension before and after the photocatalytic reaction was studied via high-performance liquid chromatography, and such products as glucose and sodium gluconate, acetate, formate, glycolate, and lactate were found after the photocatalytic reaction.


Introduction
One of the problems facing humanity is the rapid increase in greenhouse gas concentrations in the atmosphere. The main cause is fossil-fuel combustion, which is responsible for the release of the most significant greenhouse gas-carbon dioxide [1]. To overcome this problem, the development of alternative energy sources is necessary. Hydrogen is considered a prospective energy source for the future, as it combines the following benefits: high heating value, environmentally friendly combustion product, and the abundance of hydrogen on the Earth. Most of the hydrogen being produced is obtained from fossil fuels, which cannot be considered a sustainable method because of their limited abundance and because of CO 2 production [1,2]. Biomass also contains great amounts of hydrogen, which can be extracted in several ways. Nowadays, the most used methods are thermochemical processes, such as the pyrolysis and gasification of biomass [3,4]. Both of these require a high temperature, but more importantly, they produce greenhouse gases as byproducts. Therefore, the development of alternative energy sources to produce pure hydrogen under ambient conditions is necessary.
One of the more promising ways to produce hydrogen from biomass is the photocatalytic reforming of plant biomass. Soluble biomass derivatives, such as sugars, alcohols, and carboxylic acids, have been widely studied as substrates in hydrogen evolution reaction [5][6][7][8][9][10][11]. A wide range of valuable organic compounds have been found in reaction solutions after photocatalytic reforming, such as formic acid [12], acetic acid [13][14][15], acetaldehyde, ethanol, acetol, methanol, and glycolaldehyde [16][17][18]. It has been concluded that the nature of the photocatalyst, its activity, and the light source all affect product formation [9]. However, raw biomass components, such as cellulose, lignin, and starch, look more promising from the economical and practical points of view. In these cases, solid sulfide solution with a lattice parameter equal to 5.70 Å and Zn0.95Cd0.05S with a lattice parameter equal to 5.43 Å. The average crystalline size for both phases was 3.8 nm.
(a) (b) Further, we will denote the sample in the text as Cd0.7Zn0.3S/ZnS. The characteristic peaks of platinum were not detected in the XRD pattern, probably because of the low metal content (0.5 wt.%).
To study the optical properties of the photocatalyst, the UV-vis spectrum was obtained ( Figure 1b). The band gap energy was calculated using the Tauc function for direct semiconductors F(R) 2 (hν) 2 and the value was 2.55 eV. The adsorption edge of the photocatalyst was about 480 nm, which means the photocatalyst is able to absorb visible light irradiation with λ = 450 nm.
The surface chemical composition was studied via the XPS technique ( Figure 2). The Pt4f binding energy was 72.8 eV, which corresponds to platinum in the oxidized state (Pt 2+ in PtO oxide) [42,43]. The binding energies of Cd 3d5/2, Zn 2p3/2, and S 2p3/2 were 405.4, 1022.3, and 161.8 eV, respectively, which are typical of cadmium and zinc sulfides. No other ions were identified by the XPS method [44][45][46][47].  Further, we will denote the sample in the text as Cd 0.7 Zn 0.3 S/ZnS. The characteristic peaks of platinum were not detected in the XRD pattern, probably because of the low metal content (0.5 wt.%).
To study the optical properties of the photocatalyst, the UV-vis spectrum was obtained ( Figure 1b). The band gap energy was calculated using the Tauc function for direct semiconductors F(R) 2 (hν) 2 and the value was 2.55 eV. The adsorption edge of the photocatalyst was about 480 nm, which means the photocatalyst is able to absorb visible light irradiation with λ = 450 nm.
The surface chemical composition was studied via the XPS technique ( Figure 2). The Pt4f binding energy was 72.8 eV, which corresponds to platinum in the oxidized state (Pt 2+ in PtO oxide) [42,43]. The binding energies of Cd 3d 5/2 , Zn 2p 3/2 , and S 2p 3/2 were 405.4, 1022.3, and 161.8 eV, respectively, which are typical of cadmium and zinc sulfides. No other ions were identified by the XPS method [44][45][46][47]. solid sulfide solution with a lattice parameter equal to 5.70 Å and Zn0.95Cd0.05S with a lattice parameter equal to 5.43 Å. The average crystalline size for both phases was 3.8 nm.
(a) (b) Further, we will denote the sample in the text as Cd0.7Zn0.3S/ZnS. The characteristic peaks of platinum were not detected in the XRD pattern, probably because of the low metal content (0.5 wt.%).
To study the optical properties of the photocatalyst, the UV-vis spectrum was obtained ( Figure 1b). The band gap energy was calculated using the Tauc function for direct semiconductors F(R) 2 (hν) 2 and the value was 2.55 eV. The adsorption edge of the photocatalyst was about 480 nm, which means the photocatalyst is able to absorb visible light irradiation with λ = 450 nm.
The surface chemical composition was studied via the XPS technique ( Figure 2). The Pt4f binding energy was 72.8 eV, which corresponds to platinum in the oxidized state (Pt 2+ in PtO oxide) [42,43]. The binding energies of Cd 3d5/2, Zn 2p3/2, and S 2p3/2 were 405.4, 1022.3, and 161.8 eV, respectively, which are typical of cadmium and zinc sulfides. No other ions were identified by the XPS method [44][45][46][47].   To elucidate the structure of the composites, we obtained TEM images as well as HAADF-STEM images with EDX elemental mapping of the synthesized photocatalyst ( Figure 3). Figure 3a,b shows that the sample has a disordered structure with coherently scattering domains about 5 nm, which is in good accordance with the XRD results. Note that platinum nanoparticles are not visible in TEM images, likely because of a very small size. However, elemental mapping (Figure 3c) shows the presence of platinum evenly distributed over the surface of the composite sample. The elemental distribution of cadmium and zinc shows that aggregates of cadmium-enriched and zinc-enriched phases with a size of 100-500 nm are formed. These aggregates are in close contact, which should ensure the effective formation of interfacial heterojunctions that provide high photocatalytic activity. To elucidate the structure of the composites, we obtained TEM images as well as HAADF-STEM images with EDX elemental mapping of the synthesized photocatalyst ( Figure 3). Figure 3a,b shows that the sample has a disordered structure with coherently scattering domains about 5 nm, which is in good accordance with the XRD results. Note that platinum nanoparticles are not visible in TEM images, likely because of a very small size. However, elemental mapping (Figure 3c) shows the presence of platinum evenly distributed over the surface of the composite sample. The elemental distribution of cadmium and zinc shows that aggregates of cadmium-enriched and zinc-enriched phases with a size of 100-500 nm are formed. These aggregates are in close contact, which should ensure the effective formation of interfacial heterojunctions that provide high photocatalytic activity.   Following the photocatalyst's characterization, we can conclude that the photocatalyst has a multiphase structure PtO x /Cd 0.7 Zn 0.3 S/ZnS, which benefits photocatalytic hydrogen production via enhanced charge separation [48].

Photocatalytic Activity
At the beginning, we tested rice and corn starch as substrates to find out the most active substrate for hydrogen evolution. In preliminary experiments, it was shown that the amount of hydrogen evolved without using a substrate is negligible. The dependence of the yield of H 2 (t = 2 h) on starch concentration is shown in Figure 4a. The dependence has a weakly expressed maximum equal to 50.8 µmol at a concentration of starch of 5 g·L −1 . The decrease in activity at high starch contents is related to the obstructed light penetration caused by a high content of starch particles in the reaction suspension. Although the yields of H 2 with different substrates were similar, corn starch was more active at 5 g·L −1 and was chosen for further experiments. Following the photocatalyst's characterization, we can conclude that the photocatalyst has a multiphase structure PtOx/Cd0.7Zn0.3S/ZnS, which benefits photocatalytic hydrogen production via enhanced charge separation [48].

Photocatalytic Activity
At the beginning, we tested rice and corn starch as substrates to find out the most active substrate for hydrogen evolution. In preliminary experiments, it was shown that the amount of hydrogen evolved without using a substrate is negligible. The dependence of the yield of H2 (t = 2 h) on starch concentration is shown in Figure 4a. The dependence has a weakly expressed maximum equal to 50.8 μmol at a concentration of starch of 5 g·L −1 . The decrease in activity at high starch contents is related to the obstructed light penetration caused by a high content of starch particles in the reaction suspension. Although the yields of H2 with different substrates were similar, corn starch was more active at 5 g·L −1 and was chosen for further experiments. Another potential reason for the obstructed light penetration may be the high concentration of photocatalyst [49]. On the other hand, the increase in photocatalyst should lead to the enhancement of activity, given the increased amount of active centers for H2 evolution. As such, it is important to establish the optimum catalyst loading [50]. Figure   Figure 4. (a) The dependence of total H 2 amount on starch content; (b) the dependence of H 2 evolution rate on catalyst concentration, corn starch; (c) the dependence of specific activity on the catalyst concentration, corn starch; and (d) comparison of the rate of H 2 evolution from different biomass substrates over a Cd 1-x Zn x S-based photocatalyst. Conditions: C 0 (NaOH) = 5 M, C 0 (starch) = 5 g·L −1 , C(cat) = 2.5 g·L −1 unless otherwise stated, V = 20 mL, λ = 450 nm, t(reaction) = 120 min.
Another potential reason for the obstructed light penetration may be the high concentration of photocatalyst [49]. On the other hand, the increase in photocatalyst should lead to the enhancement of activity, given the increased amount of active centers for H 2 evolution. As such, it is important to establish the optimum catalyst loading [50]. Figure 4b shows the dependence of the rate of H 2 formation on photocatalyst concentration. The rate increases linearly, reaching the maximum at the concentration of 3.75 g·L −1 , and then falls. The specific rate of hydrogen evolution per gram of photocatalyst, presented in Figure 4c, reaches its maximum at the lowest catalyst loading of 1.25 g·L −1 . For further experiments, the concentration of 2.5 g·L −1 was chosen, as it provides the optimum combination of absolute and specific rates of H 2 formation.
It is known the plant biomass undergoes hydrolysis under alkaline conditions [51]. This effect is beneficial for photocatalysis as soluble organic compounds can be easily absorbed onto the photocatalyst's surface and thus react with the photogenerated charge carriers [52]. Therefore, at the next stage, we studied the influence of sodium hydroxide concentration on the reaction rate ( Figure 5a). The increase in NaOH concentration led to enhanced H 2 evolution. The highest reaction rate achieved in the 5 M solution of NaOH was equal to 535 ± 50 µmol·h −1 ·g −1 , with an AQE = 1.3%. It should be noted that this kinetic curve reached a plateau after 2 h of irradiation, and the total amount of H 2 produced was similar to the amount of H 2 that evolved from 1 M NaOH solution after 6 h of irradiation. In the absence of NaOH, the amount of H 2 detected was close to zero. We can thus conclude that starch can be hydrolyzed to a limited extent, even in a high-alkaline environment, and long-term effective H 2 evolution cannot be guaranteed.
4b shows the dependence of the rate of H2 formation on photocatalyst concentration. The rate increases linearly, reaching the maximum at the concentration of 3.75 g·L −1 , and then falls. The specific rate of hydrogen evolution per gram of photocatalyst, presented in Figure 4c, reaches its maximum at the lowest catalyst loading of 1.25 g·L −1 . For further experiments, the concentration of 2.5 g·L −1 was chosen, as it provides the optimum combination of absolute and specific rates of H2 formation.
It is known the plant biomass undergoes hydrolysis under alkaline conditions [51]. This effect is beneficial for photocatalysis as soluble organic compounds can be easily absorbed onto the photocatalyst's surface and thus react with the photogenerated charge carriers [52]. Therefore, at the next stage, we studied the influence of sodium hydroxide concentration on the reaction rate ( Figure 5a). The increase in NaOH concentration led to enhanced H2 evolution. The highest reaction rate achieved in the 5 M solution of NaOH was equal to 535 ± 50 μmol·h −1 ·g −1 , with an AQE = 1.3%. It should be noted that this kinetic curve reached a plateau after 2 h of irradiation, and the total amount of H2 produced was similar to the amount of H2 that evolved from 1 M NaOH solution after 6 h of irradiation. In the absence of NaOH, the amount of H2 detected was close to zero. We can thus conclude that starch can be hydrolyzed to a limited extent, even in a high-alkaline environment, and long-term effective H2 evolution cannot be guaranteed.
Another method described in the literature for starch dissolution is the thermal treatment of the starch aqueous solution [23].  Figure 5a demonstrates the kinetic curves of hydrogen evolution in two series of experiments-without preliminary heating treatment and with heating treatment at 70 °C. When we used a higher temperature (105 °C and 140 °C) for starch pretreatment, only a small amount of H2 was produced after 6 h of irradiation, and these results are not shown. Additionally, no hydrogen evolved from the starch aqueous suspension after heating at  Figure 5a demonstrates the kinetic curves of hydrogen evolution in two series of experiments-without preliminary heating treatment and with heating treatment at 70 • C. When we used a higher temperature (105 • C and 140 • C) for starch pretreatment, only a small amount of H 2 was produced after 6 h of irradiation, and these results are not shown. Additionally, no hydrogen evolved from the starch aqueous suspension after heating at 70 • C without NaOH. On the contrary, heat-pretreating starch at 70 • C with a high alkaline concentration (1 or 5 M) led to an increase in the hydrogen evolution rate. We should note that the kinetic curve of H 2 evolution from starch treated at 5 M NaOH and 70 • C does not reach a plateau even after 6 h of irradiation. The rate of H 2 evolution for the first 2 h was 730 h −1 ·g −1 , with AQE = 1.8% (450 nm); for the next 4 h, these values were 240 µmol·h −1 ·g −1 and 0.6%, respectively. The comparison of these results with data in the literature is given in Table 1. The rate of H 2 evolution and the AQE obtained in this study in most cases are higher in comparison with the activity of other systems, including that of raw biomass substrate. Moreover, this is one of the first studies on the photocatalytic production of H 2 from insoluble starch. Additionally, the formation of H 2 from starch was compared to the results obtained using glucose solutions, or α-cellulose suspension, as substrates for H 2 production over the platinized Cd 1-x Zn x S photocatalysts, under the same conditions (Figure 4d) [34,35]. The rates of H 2 evolution from different polymers of glucose-cellulose and corn starch-are the same, whereas the rate in the case of the use of glucose is 3.5 times higher. This effect is likely related to the high solubility of glucose. Despite the high complexity of the raw biomass substrate, the developed photocatalyst achieved comparable rates of H 2 production. Sulfide-based photocatalysts are prone to photocorrosion [34]. The stability of the synthesized photocatalyst was investigated during eight 3 h runs (Figure 5b). After each run, the photocatalytic reactor was purged with argon. The dependence of H 2 yield on time is clearly complicated; this may be caused by the complexity of the substrate. However, the rate of H 2 evolution was reasonably high during the first 18 h, then the rate decreased in every run, but stayed constant for the last 9 h of the experiment. The total yield of H 2 over 24 h was 460 µmol.
Taking into account the results of the kinetic experiments, the combination of an alkaline environment with heating clearly contributes to efficient starch hydrolysis, with the release of soluble organic compounds. These compounds can act as electron donors and react with photogenerated charge carriers to form oxidized derivatives [54]. On the other hand, the dependence of H 2 evolution rate on hydrolysis temperature reached a maximum at 70 • C, and it was shown that, after starch pretreatment at 105 • C and 140 • C, the obtained hydrolysates were not effective substrates for H 2 formation.
To clarify the reasons for this effect, we studied the composition of the aqueous phase of the reaction mixture at different pretreatment temperatures. The reaction suspension was filtered out and analyzed via HPLC. We analyzed the composition of the soluble organic compounds in suspensions of starch heated in 0.1 M and 1 M NaOH at different temperatures (70 • C, 105 • C, and 140 • C) before and after the photocatalytic reaction. The aqueous phase of the suspension with 5 M NaOH was not studied, as a solution with such a high concentration of NaOH is not possible for HPLC. Table 2 outlines the composition of the aqueous phases of the reaction suspensions before and after photocatalytic H 2 evolution. Only a trace amount of glucose was found in the alkaline (0.1 M or 1 M NaOH) suspension before the reaction without heat treatment owing to the negligible dissolution of the starch. After the photocatalytic production of hydrogen from untreated starch suspensions, glucose and other organic substances were found in the solution, indicating the simultaneous hydrolysis of starch and the oxidation of the hydrolysis products through photocatalytic reaction. Table 2 shows that the heating of a starch alkaline suspension at 70 • C resulted in the hydrolysis of the starch, with the formation of glucose as a product. After carrying out the photocatalytic reaction of hydrogen evolution, the glucose concentration increased; moreover, additional valuable organic products, including lactate and gluconate, were detected. In the case of hydrolysis at 105 • C and 140 • C, no glucose was detected in the solutions either before or after the reaction; the solutions contain mostly formate, lactate, and acetate anions. Additionally, the rate of hydrogen evolution was close to zero, and the concentration of the detected organic compounds remained almost constant before and after the reaction. It should be noted that the formate concentration increased with the hydrolysis temperature (Table 2), while the content of more complex organic compounds decreased. Thus, the increase in temperature facilitates the further alkaline hydrolysis of starch, and results in the formation of simple organic compounds, which are not effective substrates for H 2 evolution over a sulfide-based photocatalyst. As such, the alkaline hydrolysis of starch at 70 • C facilitates the best composition of a reaction solution for H 2 evolution. With the increase in the pretreatment temperature, the content of simple organic compounds increased. In contrast, the relative concentration of glucose decreased. Figure 6 shows the relative distributions of the organic compounds detected in the solutions after the H 2 evolution reaction. One can see that two factors affect the scope of organic compound transformation: alkali concentration and hydrolysis temperature. During the photocatalytic reaction, glucose transforms into valuable organic products, such as gluconic and glycolic acids.
Glucose is a well-known substrate used for photocatalytic hydrogen evolution [53][54][55][56][57]. As such, in this research, the glucose molecules should act as electron donors in the photocatalytic system, and should be oxidized by the photogenerated holes to form formate, gluconate, lactate, glycolate, or acetate anions. However, in such a case, the glucose should be consumed, and its concentration should decrease. Cellulose and starch have been shown to undergo depolymerization under the conditions of photocatalysis [27]. This process can provide additional glucose during photocatalytic hydrogen evolution. Additionally, in alkaline solutions, glucose is present mostly in its deprotonated form, C 6  − species are adsorbed on the photocata-lyst's surface through hydroxyl groups; more details are described elsewhere [58]. Once adsorbed, the glucose ions react with h+ to produce oxidized products, while simultaneously, the electrons reduce the number of protons, leading to hydrogen formation [34,53]. Additionally, total organic carbon (TOC) was measured for all reaction solutions (Figure 7). The high concentration of alkaline (1 M) led to the highest value of TOC at 20 °C and 70 °C, and the TOC content was lower following high-temperature (105 °C and 140 °C) hydrolysis. In contrast, for 0.1 M solutions of NaOH, the TOC value increased with the increase in temperature, but the highest value obtained at 140 °C was lower than that obtained in 1 M NaOH at 20 °C. These results confirm that a high alkaline concentration facilitates more extended starch hydrolysis. After the photocatalytic reaction, when the hydrolysates were obtained without thermal treatment or with treatment at 70 °C at a high alkali concentration, the TOC remained almost unchanged. In the case of the treatment at a low alkali concentration, the TOC remarkably increased. This can be explained by the fact that glucose is oxidized when hydrogen is released, but starch is also hydrolyzed with the release of glucose. In this way, the simultaneous production of hydrogen and valuable organic compounds is achieved by means of starch photoreforming under visible light irradiation. For hydrolysates obtained at high temperatures, the TOC content fell slightly Additionally, total organic carbon (TOC) was measured for all reaction solutions (Figure 7). The high concentration of alkaline (1 M) led to the highest value of TOC at 20 • C and 70 • C, and the TOC content was lower following high-temperature (105 • C and 140 • C) hydrolysis. In contrast, for 0.1 M solutions of NaOH, the TOC value increased with the increase in temperature, but the highest value obtained at 140 • C was lower than that obtained in 1 M NaOH at 20 • C. These results confirm that a high alkaline concentration facilitates more extended starch hydrolysis. After the photocatalytic reaction, when the hydrolysates were obtained without thermal treatment or with treatment at 70 • C at a high alkali concentration, the TOC remained almost unchanged. In the case of the treatment at a low alkali concentration, the TOC remarkably increased. This can be explained by the fact that glucose is oxidized when hydrogen is released, but starch is also hydrolyzed with the release of glucose. In this way, the simultaneous production of hydrogen and valuable organic compounds is achieved by means of starch photoreforming under visible light irradiation. For hydrolysates obtained at high temperatures, the TOC content fell slightly after the photocatalytic reaction as the photocatalytic oxidation of light organic compounds into CO 2 likely occurred, given that dissolved oxygen was present in the reaction suspension in trace amounts. after the photocatalytic reaction as the photocatalytic oxidation of light organic compounds into CO2 likely occurred, given that dissolved oxygen was present in the reaction suspension in trace amounts. Thus, a simple pretreatment in combination with photocatalysis makes it possible to obtain hydrogen and valuable organic compounds from practically insoluble corn starch.

Photocatalyst Characterization
The obtained catalysts were characterized by various techniques, including UV-vis diffuse reflectance spectroscopy, X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS). The diffuse reflectance UV-vis spectra were obtained using a Shimadzu UV-2501 PC spectrophotometer with an ISR-240A diffuse reflectance unit.
The phase composition of the synthesized samples was determined by X-ray diffraction (XRD). The XRD patterns were recorded on a Bruker D8 diffractometer (Bruker, Germany) using CuKα radiation. Diffraction patterns were performed in the 2Ѳ range from 20 to 70° with a scanning step of 0.05° and an accumulation time of each point of 10 s. The Thus, a simple pretreatment in combination with photocatalysis makes it possible to obtain hydrogen and valuable organic compounds from practically insoluble corn starch.

Photocatalyst Synthesis
The following reagents were used for the synthesis and photocatalytic tests without purification: CdCl 2 ·2.5H 2 O (Vekton, Russia, 98%), Zn(NO 3 ) 2 •6H 2 O (Acros Organics, Belgium, 98%), Na 2 S•6H 2 O (Sigma-Aldrich, USA, 60%), H 2 PtCl 6 (Reakhim, Russia, 98%), NaBH 4 (Acros Organics, Belgium, 98%), sodium hydroxide (Sigma-Aldrich, 98%), corn starch (Sigma-Aldrich, USA), and rice starch (Sigma-Aldrich, USA). The photocatalyst 0.5 wt.%Pt/Cd 0.7 Zn 0.3 S/ZnS was synthesized via a simple two-stage technique including the deposition of a sulfide of cadmium and zinc throughout the intermediate stage of hydroxide formation. At the first stage, mixture of 0.1 M solutions of CdCl 2 · and Zn(NO 3 ) 2 with a total volume of 100 mL (Cd/Zn = 3:7) interacted with 0.1 M NaOH solution (100 mL). Then, 250 mL of 0.1 M Na 2 S was added to the hydroxide suspension under continuous stirring and a yellow-orange precipitate of Cd and Zn sulfide solid solution (~1 g) was obtained. After the washing and drying of the obtained sample, platinum (0.5 wt.%) was deposited via the reduction of the metal from a H 2 PtCl 6 solution using NaBH 4 . The synthesis procedure was described in detail in our previous work [35].

Photocatalyst Characterization
The obtained catalysts were characterized by various techniques, including UV-vis diffuse reflectance spectroscopy, X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS). The diffuse reflectance UV-vis spectra were obtained using a Shimadzu UV-2501 PC spectrophotometer with an ISR-240A diffuse reflectance unit.
The phase composition of the synthesized samples was determined by X-ray diffraction (XRD). The XRD patterns were recorded on a Bruker D8 diffractometer (Bruker, Germany) using CuKα radiation. Diffraction patterns were performed in the 2 via the reduction of the metal from a H2PtCl6 solution using NaBH4. The synthesis proce-dure was described in detail in our previous work [35].

Photocatalyst Characterization
The obtained catalysts were characterized by various techniques, including UV-vis diffuse reflectance spectroscopy, X-ray diffraction (XRD), and X-ray photoelectron spectroscopy (XPS). The diffuse reflectance UV-vis spectra were obtained using a Shimadzu UV-2501 PC spectrophotometer with an ISR-240A diffuse reflectance unit.
The phase composition of the synthesized samples was determined by X-ray diffraction (XRD). The XRD patterns were recorded on a Bruker D8 diffractometer (Bruker, Germany) using CuKα radiation. Diffraction patterns were performed in the 2Ѳ range from 20 to 70° with a scanning step of 0.05° and an accumulation time of each point of 10 s. The range from 20 to 70 • with a scanning step of 0.05 • and an accumulation time of each point of 10 s. The calculation of the lattice constants and average crystallite sizes was carried out using the TOPAS software. The value of the parameter x in Cd1−xZnxS was determined using Vegard's law.
The X-ray photoelectron spectra were recorded on a SPECS spectrometer (SPECS, Germany) using AlKα radiation (h * = 1486.6 eV, 200 W). The binding energy of measured spectra was calibrated by the position of the Au4f 7/2 (BE = 84.0 eV) and Cu2p 3/2 (BE = 932.67) peaks.
The structure and microstructure of the photocatalysts were studied by HRTEM using a ThemisZ electron microscope (Thermo Fisher Scientific, TFS, Waltham, MA, USA) operated at an accelerating voltage of 200 kV. The microscope was equipped with a corrector of spherical aberrations, which provided a maximum lattice resolution of 0.06 nm, and a SuperX spectrometer (Thermo Fisher Scientific, TFS, USA). Images were recorded using a Ceta 16 CCD sensor (Thermo Fisher Scientific, TFS, USA). For electron microscopy studies, samples were deposited on perforated carbon substrates attached to aluminum grids using an ultrasonic dispersant.

Photocatalytic Activity Measurements
The method used for the activity test was described elsewhere [34]. Briefly, the reaction of hydrogen evolution was carried out in a sealed batch reactor containing an aqueous suspension of a photocatalyst and starch under irradiation. In this study, we used a 450 nm LED with a light intensity of 14 mW cm −2 and a light spot area of 19.6 cm 2 . For some experiments, the starch was previously heated in a NaOH solution at 70, 105, or 140 • C for 1 h. After cooling to room temperature, the obtained mixture was placed in the reactor with the photocatalyst and photocatalytic activity tests were carried out. The amount of hydrogen evolved was measured by means of a gas chromatograph (Khromos, Russia).
The following equation was used to calculate the apparent quantum efficiency [34]: where W is the rate of hydrogen formation in µmol·min −1 and N phot is the calculated photon flux equal to 69 µE min −1 .
The study of organic products in the reaction solution was conducted by highperformance liquid chromatography (HPLC) using a Prominence LC-20 chromatograph (Shimadzu, Japan). The total organic carbon (TOC) was determined using a Multi N/C 2100S carbon and nitrogen analyzer (Analytik Jena, Germany). More details are described in our previous work [59].

Conclusions
In this research, a PtO x /Cd 0.7 Zn 0.3 S/ZnS nanocomposite photocatalyst was synthesized and tested in the photoreforming of corn starch with simultaneous hydrogen evolution under visible light irradiation. The influence of the starch's pretreatment conditions (sodium hydroxide concentration and temperature) on the reaction rate and organic products produced was studied. It was shown that the thermal treatment of an alkaline starch suspension leads to an increase in the hydrogen production rate owing to starch hydrolysis with glucose formation. The highest rate of H 2 evolution was obtained in a starch suspension in a 5 M solution of NaOH treated at 70 • C, yielding 730 µmol•h −1 •g −1 with AQE = 1.8% at λ = 450 nm. However, further increasing the pretreatment temperature to 105 and 140 • C led to the near-complete inhibition of the formation of hydrogen from the hydrolysate.
For the first time, we have carried out a systematic investigation of the products of starch photoreforming accompanied by H 2 evolution. Glucose and gluconate, acetate, formate, glycolate, and lactate anions were found in the aqueous phase of the reaction mixture after the photocatalytic reaction. Thus, we have shown for the first time that the photocatalytic evolution of hydrogen from starch suspensions over a Cd 1-x Zn x S-based photocatalyst under visible light is accompanied by the formation of valuable organic compounds.

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