CoAl2O4-g-C3N4 Nanocomposite Photocatalysts for Powerful Visible-Light-Driven Hydrogen Production.

There is no doubt that the rate of hydrogen production via the water splitting reaction is profoundly affected to a remarkable degree based on the isolation of photogenerated electrons from holes. The precipitation of any cocatalysts on the substrate surfaces (including semiconductor materials) provides significant hindrance to such reincorporation. In this regard, a graphite-like structure in the form of mesoporous g-C3N4 formed in the presence of a template of mesoporous silica has been synthesized via the known combustion method. Hence, the resulting g-C3N4 nanosheets were decorated with varying amounts of mesoporous CoAl2O4 nanoparticles (1.0-4.0%). The efficiencies of the photocatalytic H2 production by CoAl2O4-doped g-C3N4 nanocomposites were studied and compared with those of pure CoAl2O4 and g-C3N4. Visible light irradiation was carried out in the presence of glycerol as a scavenger. The results showed that the noticeable photocatalytic enhancement rate was due to the presence of CoAl2O4 nanoparticles distributed on the g-C3N4 surface. The 3.0% CoAl2O4-g-C3N4 nanocomposite had the optimum concentration. This photocatalyst showed extremely high photocatalytic activities that were up to 22 and 45 times greater than those of CoAl2O4 and g-C3N4, respectively. This photocatalyst also showed 5 times higher photocatalytic stability than that of CoAl2O4 or g-C3N4. The presence of CoAl2O4 nanoparticles as a cocatalyst increased both the efficiency and productivity of the CoAl2O4-g-C3N4 photocatalyst. This outcome was attributed to the mesostructures being efficient charge separation carriers with narrow band gaps and high surface areas, which were due to the presence of CoAl2O4.

Introduction

Currently, fossil fuels have been used to yield approximately 80% of energy used in the world; hence, an increasing number of environmental problems and crises have been declared. To overcome these problems, many ecofriendly sources of renewable energy have been classified as important for potential evolution and progress. Starting from this point, many researchers have developed the use of another type of fuel, that is, hydrogen produced from water splitting photocatalytic systems. These cells used are functional designs used to obtain clean energy.[1−3] In the past decade, a photoelectrochemical water splitting process for the production of hydrogen and oxygen in the presence of TiO2 was reported by Fujishima and Honda.[4] This type of conversion involves the transformation of solar energy to another form of energy, chemical energy. This conversion can be carried out using different photocatalysts and constitutes an efficient and appropriate solution to overcome the most obvious energy and environmental problems.

In addition, two-dimensional semiconductor photocatalysts have received considerable attention as a result of the photocatalytic response that they exhibit to visible light.[5,6] Among these materials, graphitic carbon nitride, g-C3N4, is a polymeric metal-free semiconductor with a band gap energy (Eg) of approximately 2.7 eV, enabling it to absorb the visible light. It also exhibits many other characteristics, including nontoxicity, excellent stability, low cost, and versatile structural properties.[7,8] The drawbacks of g-C3N4 nanosheets have been widely reported in various studies.[7−9] Metal or nonmetal doping, semiconductor coupling, and construction of porous structures are some of the solutions used to avoid these drawbacks.[10−13] However, these proposed structures still suffer from a significant number of disadvantages, including the fast recombination of electron–hole pairs and insufficient absorption of visible light. Pristine g-C3N4 still displays a considerably limited performance with respect to photocatalytic activity.[14] Heterostructures containing g-C3N4 have been calcified to produce the best g-C3N4 compositions in the photocatalysis field, which noticeably promoted photocatalytic achievements among all previously known types. This outcome may result from the development of charge carrier separation, which occurs with each of the catalysts g-C3N4/Ag2MoO4, g-C3N4/Bi2O4, g-C3N4/perovskite oxide, and g-C3N4/TiO2.[14−19] Combined semiconductors containing CoAl2O4 have been utilized for photocatalytic decomposition. CoAl2O4 also has a narrow band gap of 1.80 eV and exhibits a strong response to visible light.[20−30] To the best of our knowledge, hydrogen production using CoAl2O4–g-C3N4 photocatalysts has not been reported. In this regard, heterostructure-based CoAl2O4 and mesoporous g-C3N4 for the formation of CoAl2O4–g-C3N4 were synthesized by simple sol–gel procedures. The chemical structures of the resulting products were confirmed using various techniques. The photocatalytic activities were evaluated for hydrogen production under visible light. Finally, a likely hydrogen production mechanism for the mesoporous CoAl2O4–g-C3N4 heterostructured nanocomposites was also proposed.

Experimental Section

Materials

EO106-PO70EO106 surfactant was used as a triblock copolymer with an average MW of 12,600 g/mol (F-127). Co(NO3)2·6H2O, Al(NO3)3·9H2O, acetic acid, hydrochloric acid, and ethanol were all purchased from Sigma-Aldrich.

Preparation of Mesoporous CoAl2O4

A sol–gel procedure was used to prepare mesoporous CoAl2O4 using a structure-directing agent, namely, the F127 triblock copolymer. The required material was synthesized using molar ratios on the order of 1:0.02:50:2.25:3.75 for CoAl2O4/F127/C2H5OH/HCl/CH3COOH, respectively. For example, a solution of 1.6 g of F127 in 30 mL of ethanol was stirred for 60 min. Next, 0.74 mL of HCl and 2.3 mL of CH3COOH were added to the previous solution, and magnetic stirring was continued for 30 min. Co and Al precursors were weighed out in a 1:2 ratio and added to the F127–CH3COOH mesophase with additional stirring for 60 min. A humidity chamber (40%) was used to hold the prepared mesophase at 40 °C for 12 h to reduce the amount of ethanol, leading to the formation of a gel. Further aging at 65 °C for 24 h was carried out in the resulting gel. Finally, the samples were calcined at 600 °C at a heating rate of 1°C/min in air for 4 h and then cooled at a rate of 2°C/min in order to eliminate the F127 surfactant and obtain the mesoporous CoAl2O4 as a final product.

Synthesis of Mesoporous g-C3N4

Urea and dicyandiamide were purchased from Sigma-Aldrich. High-surface-area mesoporous silica (HMS) (∼500–1000 m2 g–1) was used to prepare g-C3N4 with a large surface area. Furthermore, pyrolysis of dicyandiamide and urea in air was performed. The detailed HMS preparation was easily executed as reported in the literature.[31] Approximately 50 mL of distilled water and 1 g of HMS were dispersed for 30 min. A mixture of dicyandiamide (3 g) and urea (5 g) was carefully added to the abovementioned solution. Continuous stirring at 80 °C was done to enhance the dissolution of both components. The sample was dried overnight at approximately 80 °C to remove the excess water. Calcination was performed at 550 °C for 4 h. Next, the obtained material was immersed in a solution of NH4HF2 (2 M, 50 mL) with vigorous stirring for 24 h to drive out the HMS template. To release any contaminants adsorbed by the produced g-C3N4 nanoparticles, they were easily cleaned by washing several times with water. Thereafter, the synthesized pure material was dried by heating for 12 h at 100 °C.

Synthesis of Mesoporous CoAl2O4–g-C3N4 Nanocomposites

A water exfoliation method was used to synthesize CoAl2O4–g-C3N4 nanocomposites. The samples were synthesized as follows: 0.2 g of the as-prepared g-C3N4 was mixed with the required amount of mesoporous CoAl2O4, and the mixture was then sonicated in 400 mL of deionized water for 3 h at a power of 40 kHz. This procedure allowed the formation of thin-layered CoAl2O4–g-C3N4 products. A centrifugation process was used to collect the final products with the general abbreviation xCoAl2O4–g-C3N4, where the nominal molar content of CoAl2O4 was represented by “x” in this formulation (x = 1, 2, 3, and 4%).

Characterization

A JEOL JEM-1230 transmission electron microscope was used to determine the images of the prepared samples at 200 kV. Phase identification of the prepared materials was carried out using a Bruker AXS D8 Endeavor X-ray diffractometer. A Nova 2000 series Chromatech apparatus was used to determine the texture properties of the prepared photocatalysts. A Shimadzu system (RF-5301, Japan) was applied for the determination of the photoluminescence (PL) spectra of the prepared photocatalysts. The photocurrent intensity of the prepared photocatalysts was determined using a Zahner Zennium electrochemical workstation. The Fourier transform infrared (FT-IR) spectrum was measured in a KBr dispersion in the range of 400–4000 cm–1 using a PerkinElmer spectrometer. A V-570 spectrophotometer (Jasco, Japan) was used to obtain the UV–vis–NIR spectra. The band gap values were determined by UV–vis diffuse reflectance spectroscopy.

Photocatalytic Tests

A certain quantity of the photocatalyst was suspended in 450 mL of H2O in the presence of a glycerol scavenger (10% vol) prior to the production of hydrogen. The required experiments were carried out under normal conditions at room temperature and atmospheric pressure. To overcome the effect of lamp heating on the reaction, a cooler made from quartz was used. Before photocatalysis began, nitrogen gas was bubbled for 30 min to eliminate oxygen dissolved in water. The area above the photoreactor was fixed with a 500 W xenon lamp producing visible light. The photocatalytic process for H2 production started when the lamp was switched on. An Agilent GC 7890A gas chromatograph with nitrogen carrier gas was used to examine the quantity of H2 produced over separate periods of time throughout the photocatalytic process. Further reactions, as additional confirmations of the optimized parameters, were carried out without a lighting source and without the desired photocatalyst.

Results and Discussion

Investigation of the Product Samples

The X-ray diffraction (XRD) patterns for the pure g-C3N4 and CoAl2O4–g-C3N4 nanocomposites are illustrated in Figure A. The XRD diffraction patterns for pure CoAl2O4 are illustrated in Figure A,B. All the diffraction patterns obtained confirm the suggested structures. g-C3N4 was indicated by the diffraction peak observed at 27.4° in Figure A, according to card number JCPDS 87-1526. On the other hand, the XRD diffractogram assigned to pure CoAl2O4 corresponded to that in card number JCPDS 044-0160, as all essential peaks have been mentioned. These peaks are attributed to the CoAl2O4 phase, as shown in Figure B. The diffractograms also show that the g-C3N4 peak intensities showed considerable decreases as the CoAl2O4 content increased (1.0–4.0%). All CoAl2O4–g-C3N4 diffractograms show that no additional peaks related to pure CoAl2O4 were still present, which is attributed to the strong CoAl2O4 adhesion to the surface of g-C3N4 nanosheets. Additionally, this result was attributed to the lower CoAl2O4 content present in each composition. The XRD diffraction patterns also showed no additional equivocal peaks in any samples. This observation provides good evidence for the formation of the heterojunction nanocomposite between CoAl2O4 and g-C3N4 nanosheets.

Figure 1

(A): XRD patterns of g-C3N4 and CoAl2O4–g-C3N4 samples. (B)XRD diffraction pattern of the prepared pure CoAl2O4 sample.

The FT-IR spectra of the prepared pure CoAl2O4, g-C3N4, and CoAl2O4–g-C3N4 samples are illustrated in Figure . The triazine stretching mode present in both the pure g-C3N4 and CoAl2O4–g-C3N4 nanocomposites was observed at 808 cm–1. Along with the peaks for the typical CN-heterocyclic stretching modes, five additional peaks were observed at 1633, 158, 1408, 1322, and 1243 cm–1.[32−34] The FT-IR spectra also revealed that the intensity of the peak for pure g-C3N4 was significantly reduced as the CoAl2O4 content increased. An absorption peak at approximately 664 cm–1 was also present in the FT-IR spectrum of pure CoAl2O4 nanoparticles.

Figure 2

FT-IR spectra of pure CoAl2O4, g-C3N4, and CoAl2O4–g-C3N4 samples.

The X-ray photoelectron spectroscopy analysis for the 3.0% CoAl2O4–g-C3N4 nanocomposite is shown in Figure . The high-resolution spectra of Co, Al, O, C, and N are shown in Figure A–E. The presence of Co2+ and Co3+ ions in the prepared nanocomposites was confirmed by the presence of the major peaks assigned to Co 2p1/2 at ∼794.7 and 804.2 eV and Co 2p3/2 at ∼779 eV and 783.6 eV (Figure A). It is easily determined that the obtained values are very similar to those reported in the literature.[35] Furthermore, Figure B displays one peak for Al 2p at 73.5 eV, confirming the presence of Al as Al oxide.[36] Furthermore, Figure C shows that the O 1s spectrum consists of two peaks at 531 and 530 eV that could be related to the adsorbed oxygen species and CoAl2O4 lattice oxygen, respectively.[37,38] Two main C 1s peaks at ∼287.9 and ∼284.6 eV were also detected, as shown in Figure D. These peaks indicate the presence of sp2 C connected to N in the N-containing aromatic rings and sp2 C–C bonds. Figure E shows that the N 1s peak appears at 398.3 eV, which reveals the presence of sp2-hybridized N atoms. The structure of graphitic carbon nitride g-C3N4 was confirmed by all the abovementioned information.[39]

Figure 3

High-resolution spectra of 3.0 wt % CoAl2O4–g-C3N4 for Co 2p (A), Al 2p (B), O 1s (C), C 1s (D), and N 1s (E) species.

The transmission electron microscopy (TEM) images of the CoAl2O4, g-C3N4, and 3.0% CoAl2O4–g-C3N4 samples are displayed in Figure . The average particle sizes of the prepared CoAl2O4 nanoparticles were in the 5–8 nm range (Figure A). The typical nanosheet structure of g-C3N4 is shown in Figure B. The TEM images of the CoAl2O4–g-C3N4 nanocomposite are shown in Figure C and exhibit a considerable dispersion of CoAl2O4, in the form of spherical particles, over the g-C3N4 nanosheet. In addition, a significant decoration of CoAl2O4 has been noted. Figure D shows the high-resolution transmission electron microscopy (HRTEM) image of the 3.0% CoAl2O4–g-C3N4 nanocomposite product. Examination of the image confirms the higher distribution of CoAl2O4 on the g-C3N4 surface. The existence of g-C3N4 and CoAl2O4 was also confirmed by the determination of lattice spacings of 0.320 and 0.460 nm for the (002) and (111) planes, respectively.[39] Hence, a strong interfacial interaction between g-C3N4 and CoAl2O4 is clearly revealed by the HRTEM image.

Figure 4

TEM images of CoAl2O4 (A), g-C3N4 (B), and 3.0 wt % CoAl2O4–g-C3N4 (C) samples, and (D) HRTEM image of the 3.0 wt % CoAl2O4–g-C3N4 sample.

The surface properties of the obtained nanocomposites were explored, and Figure demonstrates the N2 adsorption–desorption isotherms for the g-C3N4, CoAl2O4, and 3.0% CoAl2O4–g-C3N4 samples. As per the IUPAC convention, the obtained isotherms are classified as IV-type isotherms, which are indicative of mesostructured materials. This observation indicates that after the dispersion of CoAl2O4 nanoparticles over the g-C3N4 nanosheets, the mesoporous characteristics remain without any changes. The surface areas of g-C3N4, pure CoAl2O4, and various loadings of CoAl2O4 on g-C3N4 are listed in Table . As the results show, the surface area of pure g-C3N4 is 175 m2/g, which is considerably larger than the reported literature value.[40] This increase in the obtained surface area is mainly the result of the presence of the initial HMS precursor, as previously highlighted in the Experimental Section. The slight decrease due to the presence of CoAl2O4 could be attributed to the pore filling of C3N4 with homogeneously dispersed particles on the surface.

Figure 5

N2 adsorption–desorption isotherms for the g-C3N4, CoAl2O4, and 3.0% CoAl2O4–g-C3N4 samples.

Table 1

BET Surface Areas of g-C3N4 and CoAl2O4@g-C3N4 Samples

samples	SBET (m2/g)
g-C3N4	175.00
1.0 wt % CoAl2O4@g-C3N4	184.00
2.0 wt % CoAl2O4@g-C3N4	188.00
3.0 wt % CoAl2O4@g-C3N4	192.00
4.0 wt % CoAl2O4@g-C3N4	193.00
CoAl2O4	210.00

Figure illustrates the UV–vis spectra of the pure CoAl2O4, g-C3N4, and CoAl2O4–g-C3N4 photocatalysts with varying CoAl2O4 contents. The results confirm the absorption of visible light by all samples. The presence of CoAl2O4 also enhanced the width of both the absorption bands and band edges (Figure ). UV–vis spectra were also used to determine the band gaps in all cases, and the outcomes are listed in Table . The calculated values of the band gaps of g-C3N4 were heavily affected by the loading percentage of CoAl2O4 in the nanocomposites. Consistently, the band gap was reduced as the integrated weight percentage of CoAl2O4 on the surface of the g-C3N4 nanosheets was increased.

Figure 6

UV–vis spectra of pure CoAl2O4, g-C3N4, and CoAl2O4–g-C3N4 samples with various CoAl2O4 contents.

Table 2

Band Gaps of g-C3N4 and CoAl2O4@g-C3N4 Samples

samples	band gap, eV
g-C3N4	2.70
1.0 wt % CoAl2O4@g-C3N4	2.20
2.0 wt % CoAl2O4@g-C3N4	2.10
3.0 wt % CoAl2O4@g-C3N4	1.94
4.0 wt % CoAl2O4@g-C3N4	1.92
CoAl2O4	1.80

Evolution of H2 via Visible Light Irradiation with the Obtained Catalysts

The targeted CoAl2O4–g-C3N4 nanocomposite photocatalysts were examined and compared with pure CoAl2O4 and g-C3N4 for hydrogen production upon irradiation with visible light. The initial reaction conditions included a photocatalyst content of 1.2 g/L, a reaction solution volume of 450 mL, the presence of glycerol (10 vol %), a Xe lamp (500 W) light source, and 9 h of irradiation at room temperature. The effect of different CoAl2O4 loadings from 1.0 to 4.0 wt % in the CoAl2O4–g-C3N4 nanocomposite on the quantity of hydrogen produced was studied and compared with the volumes obtained with both pure CoAl2O4 and g-C3N4, as illustrated in Figure A. The results revealed that the quantities of hydrogen produced were 810 and 400 μmol g–1 for CoAl2O4 nanoparticles and g-C3N4 nanosheets, respectively. The use of various weight percentages in the CoAl2O4–g-C3N4 samples (1.0, 2.0, 3.0, and 4.0 wt %) used for the generation of hydrogen resulted in 1912, 9450, 13050, and 13,095 μmol g–1 of hydrogen, respectively. The values are greater than those in some published works[40,41] and less than those in other published works.[42−44] Thus, the results obtained indicated that the addition of CoAl2O4 nanoparticles significantly increased the extent of charge carrier separation and the surface area and decreased the band gap energy. Therefore, the CoAl2O4 content in the original photocatalyst showed a direct and positive effect on the H2 yield until a certain loading weight (3.0%) was reached. Above this weight percentage, there was no additional effect on the yield, which did not respond to the addition of any extra photocatalyst in the reaction mixture. The production of hydrogen was increased to 8775, 10,125, 13,050, 16,875, and 18,225 μmol g–1 as a result of the gradual increase in the content of photocatalyst from 0.4 to 2.0 g/L, as shown in Figure B. These results may have occurred because the total number of active sites over the 3.0% CoAl2O4–g-C3N4 photocatalyst surface showed a noticeable increase. The level of hydrogen production was at least 15,120 μmol g–1 when the photocatalyst content was greater than 2.4 g/L. This result may be due to an effective reduction in light penetration during the illumination process in the presence of a higher particle content in the reaction solution.[45−49]

Figure 7

(A) Effect of CoAl2O4 content on hydrogen evolved using the g-C3N4 photocatalyst. (B) Effect of the amount of 3.0% CoAl2O4–g-C3N4 photocatalyst used for hydrogen evolution.

The measurements of both PL and transient photocurrent responses emphasize the results obtained in this study. As seen in Figure A, the PL spectrum of g-C3N4 shows the highest PL emission intensity among all samples. However, upon increasing the content of CoAl2O4 nanoparticles adsorbed over the g-C3N4 nanosheet surface, the PL emission intensity noticeably decreased, as illustrated. The observed PL emission intensities decreased as follows: g-C3N4 > CoAl2O4 > 1.0% CoAl2O4–g-C3N4 > 2.0% CoAl2O4–g-C3N4 > 3.0% CoAl2O4–g-C3N4 ≈ 4.0% CoAl2O4–g-C3N4. The CoAl2O4 nanoparticles have a high PL emission intensity and show a lower band gap energy (1.80 eV). Therefore, CoAl2O4 displays a low photocatalytic activity, and the recombination rate of the charge carriers in the presence of CoAl2O4 is very high. However, the photocatalyst effectiveness remains clear and apparent from the standpoint of photocatalytic activity. The photocurrent transient responses are given in Figure B. The results indicate that a lower photocurrent density was observed for g-C3N4, while a substantial increase occurred as the content of CoAl2O4 deposited on the surface of g-C3N4 increased. The photocurrent densities of the designed nanocomposites increased in the following order: g-C3N4 < CoAl2O4 < 1.0% CoAl2O4–g-C3N4 < 2.0% CoAl2O4–g-C3N4 < 3.0% CoAl2O4–g-C3N4 ≈ 3.0% CoAl2O4–g-C3N4. These outcomes also show that the success of the photocatalytic process for the CoAl2O4–g-C3N4 nanocomposites coincides closely with, and is proportional to, the results of the PL measurements.

Figure 8

(A) PL spectra of pure CoAl2O4, g-C3N4, and CoAl2O4–g-C3N4 samples with various CoAl2O4 contents. (B) Photocurrent transient responses of pure CoAl2O4, g-C3N4, and CoAl2O4–g-C3N4 samples with various CoAl2O4 contents.

Figure shows the photocatalytic reproducibility of reused photocatalysts. As previously mentioned, the 3.0% CoAl2O4–g-C3N4 photocatalyst contains the optimum composition and shows substantial recycling potential. CoAl2O4–g-C3N4 may be recycled five times without exhibiting any significant defects. The fifth round affords 99.7% of the hydrogen evolution efficiency observed in the first use. From the above results, the optimized photocatalyst, 3.0% CoAl2O4–g-C3N4, demonstrated high stability, representing a highly applicable and valuable photocatalyst for the evolution of hydrogen. The XRD, UV–vis, and PL characterizations of the photocatalysts used also confirmed that the photocatalysts are stable. Additionally, inductively coupled plasma analysis of the solution remaining after catalysis confirmed that there were no Co or Al ions present, which confirmed the stability of the photocatalyst.

Figure 9

Reuse and reproducibility of 3.0 wt % CoAl2O4–g-C3N4 photocatalyst used five times.

Suggested Mechanism for the CoAl2O4–g-C3N4 Nanocomposite

The separation of photoelectrons and holes in g-C3N4 nanosheet-reinforced CoAl2O4 nanoparticles has been explained by using the proposed mechanism below (Scheme ). The following equations have been used to calculate the band energy levelswhere the valence and conduction bands are designated EVB and ECB, respectively; the band gap value is given as Eg and is determined from optical measurements; the absolute electronegativity of the semiconductor is represented as X; and the normal hydrogen electrode versus the redox-level measurement on the absolute vacuum scale is given as E0 (E0 = −4.5 eV). A narrow band gap value for g-C3N4 nanosheets has been previously reported. Hence, a lower energy is required to excite the system. As a result of the photocatalytic irradiation, the photogenerated electrons from pure g-C3N4 originate from the valance band and are promoted to the conduction band. However, in the CoAl2O4–g-C3N4 nanocomposite, CoAl2O4 accepts the excited electrons, thereby realizing the desired charge carrier separation. The energy of the CoAl2O4 conduction band (+0.06 eV) exhibits a more positive value than that of g-C3N4 (−1.13 eV). Additionally, the distribution of CoAl2O4 nanoparticles across the g-C3N4 nanosheets in the nanocomposites provides a noticeable increase in the number of active sites on the CoAl2O4–g-C3N4 photocatalyst surface; the photocatalytic activity is enhanced and hydrogen production is considerably accelerated relative to either CoAl2O4 or g-C3N4. In total, the gross efficiency of the hole-scavenging action is greatly increased because the reaction solution contains glycerol as a scavenger. Protons are readily produced by this process and can additionally react with charge carriers to create more H2. Therefore, according to Scheme , water splitting can occur at 1.23 eV according to ref (50), which lies within the band gap of g-C3N4. CO2 formation is an obvious product of hole transfer from p-type CoAl2O4 to the attached g-C3N4, exhibiting an energy of +1.57 eV. These holes could produce protons and CO2 from the obvious decomposition of glycerol, as seen in previous literature reports.[51] With the assistance of the separated electrons in the CB of supported CoAl2O4, hydrogen generation is made possible by the combination of two protons with electrons.

Scheme 1

Photocatalytic Process of the 3.0 wt % CoAl2O4–g-C3N4 Photocatalyst for Hydrogen Production

Conclusions

It is easy to prepare g-C3N4 nanosheets via a combustion process using a template material of mesoporous silica. Various CoAl2O4 nanoparticle contents (1.0–4.0%) were used as adsorbents on the g-C3N4 nanosheets as a result of the preparation process. The g-C3N4 sheets were affected by the dispersion of CoAl2O4 on the surface of the nanocomposites. Prevention of electron–hole reincorporation was significantly enhanced by the decrease in the band gap energy. The photocatalyst CoAl2O4–g-C3N4 (3.0 wt %) produced 18,225 μmol g–1 of hydrogen, the maximum amount produced by the catalysts prepared with various compositions. In addition, a maximum photocatalyst weight of up to 2.0 g/L was used, with irradiation carried out for 9 h at room temperature. The synergetic effect of CoAl2O4 and g-C3N4 enhances the production of hydrogen. The CoAl2O4–g-C3N4 composites produce a significantly greater amount of hydrogen than either the g-C3N4 sheets or pure CoAl2O4 nanoparticles. A highly efficient, stable product has been developed in the form of CoAl2O4–g-C3N4. A maximum of five repeated cycles was also studied, without any loss of hydrogen evolution in any of the cycles.