Visible-light-driven CO2 photoreduction to solar fuels over a Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 0D/3D/2D QDs Schottky/Z-scheme ternary heterojunction photocatalyst Authors: Hossein Kadkhodayan, Taher Alizadeh Published: 2026 ABSTRACT AbstractCurrently, the increasing concentration of carbon dioxide (CO2) pollutants in the Earth’s atmosphere due to the expansion of factories and vehicles has caused destructive effects, including an increase in the greenhouse effect, long-term droughts, insufficient rainfall, melting polar ice, and climate change. The various methods are available for removing carbon. Photocatalytic processes have emerged as promising and sustainable approaches for CO2 conversion because they don’t make harmful byproducts, affect most contaminants, do not have any detrimental effects on the environment, don’t need expensive, complicated equipment, are readily recoverable, recover, and use again. Perovskite is a novel kind of photocatalyst that exhibits excellent photocatalytic performance. Photocatalysts with more than one structure are better at covering a bigger area. Perovskite photocatalysts can reduce pollutants more quickly and better than other mono-photocatalysts. When exposed to visible light, this investigation intends to make a modern perovskite photocatalyst that exhibits enhanced performance and can convert carbon dioxide into solar fuels. To achieve this, g-C3N4 (gCN), a non-metallic co-photocatalyst; Bi2ZnTiO6 (BZTO), a double perovskite; and zero valent nickel/copper (Ni/Cu) nanoparticles, a metallic co-photocatalyst, were synthesized. XRD, SEM, EDX/mapping, UV–vis, DRS, and PL experiments were employed to look at the phase, crystallographic, structural, and photocatalytic properties. The performance of the Ni/Cu-loaded Bi2ZnTiO6 perovskite/g-C3N4 0D/3D/2D QDs Schottky/Z-scheme ternary heterojunction nanocomposite was optimized and improved using Taguchi’s experimental design. The temperature of the reaction, how rapidly it was agitated, how long it took, how much CO2 and photocatalyst were present, visible irradiation intensity, interval between photoreduction site and light source, and the pH of the reaction are all critical parameters. To achieve the best conditions according to the Taguchi method, the solution should have a reaction medium pH of 5.0, a temperature of 25 °C, a mixing rate of 200 rpm, a photoreduction time of 300 min, a photocatalyst dosage of 1.0 g/l, a CO2 concentration of 400 mg L−1, a light intensity of 70 W, distance between light source and aqueous solution. The Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 Schottky/Z-scheme ternary heterojunction nanocomposite transformed the CO2 with a proficiency rate of 1727 μmol g−1 h−1 (total CO2 conversion efficiency 95%) into the main solar fuels of methane (CH4) gas with a productivity ratio of 863.5 μmol g−1 h−1 (selectivity efficiency 50%) and methanol (CH3OH) with a productivity rate of 690.8 μmol g−1 h−1 (selectivity efficiency 40%). Additionally, the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary nanocomposite maintained its structure and could be utilized over and over again, even after five cycles of continuous use.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1038/s41598-026-57310-x.Keywords: Perovskite photocatalysts, Bi2ZnTiO6, CO2 photoreduction, Z-scheme heterostructure, Solar fuels, Graphitic carbon nitride (g-C3N4)Subject terms: Chemistry, Environmental sciences, Materials science, Nanoscience and technology Introduction In recent years, due to the increase in emissions and atmospheric carbon dioxide concentration in the Earth’s atmosphere due to human activities, it has left undesirable effects such as an increase in the greenhouse effect, long-term droughts, lack of precipitation, melting of polar ice caps, global warming, and climate change. There are several ways to reduce and manage the amount of carbon dioxide that is released into the air1–4. Among various CO2 mitigation strategies, photocatalytic conversion has emerged as a promising approach due to they don’t make toxic byproducts, they don’t need complicated and expensive equipment, and they are readily recoverable, recover, and reuse. Perovskite photocatalysts are a new kind of photocatalyst that performs high photocatalytic performance5–8. There are many different types of semiconductors in their structure, each with its own strong and distinct photocatalytic capabilities9–12. Photocatalysts with mono-structure can only accomplish a limited photocatalytic activity. Perovskite photocatalysts perform better than single-structure photocatalysts because they can oxidize-reduce pollutants with high efficiency and reaction rate13–15. CO2 is a thermodynamically stable molecule with a high C=O bond energy, which makes its activation kinetically challenging. This is why most efforts to lower or eliminate CO2 emissions don’t work well or at all. Using robust photocatalysts that can be used again is a cheap, simple, rapid, and efficient way to mitigate CO2 emissions. The photoreduction processes were employed to convert carbon dioxide into valuable chemicals like methane and methanol. These might be utilized to create clean energy and fuels that can be transported. Also, this helps solve the problem of carbon dioxide emissions accumulation in the Earth’s atmosphere4,16,17. There are several factors that might affect how well the photocatalytic photoreduction of carbon dioxide to hydrocarbon fuels works. It is essential to consider the kind of photocatalyst, the size and shape of the photocatalyst particles, the type and amount of doping elements added to the photocatalyst, the amount of photocatalyst used, the concentration of the CO2, and other parameters. The reaction temperature, environmental pressure, length of the irradiation time, phase of the reaction medium, concentration of dissolved oxygen in the water, pH of the solution, presence of inorganic ions, type, intensity, and location of the light source, and other factors might all change the process18–22. Perovskites have attracted significant attention from researchers and scientists. Perovskites are crystalline materials with a general formula ABX₃, characterized by a distinctive three-dimensional framework that have high reaction rates, a wider operating range, and the ability to react with several different types of pollutants simultaneously7,13. As a result, due to the additional and closely spaced electron levels created in the structure of perovskite photocatalysts compared to single-structure photocatalysts, the efficiency of effective separation of electron–hole pairs (e−/h+) is increased (the probability of electron–hole pair recombination is reduced). The operation of redox reactions is facilitated23. The general structure of perovskite photocatalysts can be presented as follows: Ax+1B|2x−1|CyZx+2y+3 (Z = O, S and halogens; 0 ≤ x, y ≤ 1); ABZ3, A2BZ4, A2BCZ6, A3BCDZ9. are synthesized owing to multifold traits and several utilizations, converting chemical and thermal stability (no-decomposition at inappropriate conditions), suitable and active energy band gap in the visible light region, good photocatalytic traits in divers series like advanced redox reactions, organic contaminants degradation, denitrification, elimination of heavy toxic metals, air purification, photoreduction of carbon dioxide gases into valuable fuels, having antimicrobial features, water splitting to extraction of hydrogen24–28. Generally, perovskites are categorized into five significant classes:(i) simple perovskite; ABZ3 (Z=O, S, X=Cl−, Br−, I−) including SrTiO3, BaTiO3, ZnTiO3, and … (ii) single perovskite; A2BZ4 for example ZnIn2S4, Cu2WS4, La2NiO4, K2NiF4, Zn2SnO4, Cu2AgInS4, Ag2WO4, and BiVO4 (iii) Binary perovskite; A2BB’Z6 such as La2ZnTiO6, Cs2AgBiCl6, Bi2MoO6, Rb2LiTlF6, Ba2SrWO6, Cs2AgInCl6, GdBaCo2O6, Na2ZrTeO6, Cs2TiBr6, and Bi2WO6 (iv) Ternary perovskite; A3BB’Z9 such as La2SrFe2TiO9, SrCaFe2WO9, Ba3NiTaSbO9, Sr3CoFeMoO9, Bi2WZnTiO9, BaSr2MgTa2O9, and Ba2K2Te2O9 (v) Quaternary perovskite; AB3B4’Z9 for example CaCu3Ti4O12, CaCu3NiReO12, CeMn3V4O12, AgZn3Ti4O12, and SrZn3Ti4O1229–33. Figure 1a illustrates the periodic table of elements for the synthesis of different types of perovskites. Using the element numbers in the periodic table34–37, it is possible to create over 100 million different perovskites, each with its own set of properties. In Fig. 1b, it can be seen that the individual photocatalysts are arranged in such a way that they can act as both oxidants and reductants. Therefore, by placing these single semiconductor structures together in a regular structure and network called perovskite, a photocatalyst can be obtained that has all of these properties simultaneously13,36,38. Photocatalysts can generate various intermediates during reactions, such as the generation of electron–hole pairs, the generation of superoxide/hydroxyl radicals, as seen in Fig. 1c 39. Elements periodic table for the synthesizing the type of perovskites (a), energy band gaps and redox states of structure single photocatalysts (b), and applications of different photocatalysts according to those electrochemical potential and released radicals (c). The objective of this research is to make a novel perovskite that exhibits enhanced performance and can convert CO2 into solar fuels when it is hit by visible light. In this study, the photocatalytic features of a double perovskite with the following formulation: The bismuth zinc titanate perovskite is a novel material that is formed by combining and mixing multiple semiconductors, such as zinc oxide, bismuth oxide, and titanium dioxide, into one structure. Each of these semiconductors has its own strong and distinct photocatalytic characteristics. Therefore, perovskite photocatalysts exhibit enhanced activity than photocatalysts with just one structure. In this research, bismuth zinc titanate (Bi2ZnTiO6) double perovskite will be synthesized as a photocatalytic double perovskite with a suitable energy bandgap in the range of CO2 pollutant reduction and active exposed to visible light radiation. The study will also look at the graphite nitride carbon, which is made of non-metallic co-photocatalyst, and the nickel/copper nanoparticles, which are made of metallic co-photocatalyst. To achieve the optimum conditions, the Bi2ZnTiO6 3D double perovskite with both metallic and non-metallic co-photocatalysts is required. The zero-dimensional (0D) nickel/copper quantum dot metallic co-photocatalysts may be added to the Bi2ZnTiO6 perovskite photocatalyst in this work. The incorporation of metallic Ni/Cu nanoparticles can facilitate charge separation through Schottky junction formation, thereby suppressing electron–hole recombination. When a conductive metal and a semiconductor are placed next to each other, a Schottky junction is formed, which makes it easier for electricity to pass across their interface. Quantum dots are particles that are both nanoscale and zero-dimensional at the same time. Also, to achieve the highest photoreduction efficiency of CO2, a two-dimensional (2D) non-metallic co-photocatalyst made of graphitic carbon nitride (g-C3N4) should be added to the structure of Bi2ZnTiO6 perovskite. Graphite carbon nitride is a kind of two-dimensional (2D) non-metallic co-photocatalyst. This will give carbon dioxide more places to adhere, which will make them more stable and speed up the photocatalytic activity. The g-C3N4 has a highly reactive surface that, on the one hand, can absorb more radiant energy and, on the other hand, can adsorb and store carbon dioxide molecules. The testing and analysis done in this work demonstrate that the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary nanocomposite could be a novel, highly effective photocatalyst to convert carbon dioxide reactant into valuable fuels. It is also relatively stable in terms of structure, has significant photocatalytic properties, and may be utilized more than once. The synthesized nanocomposite exhibits structural stability, reusability, and promising photocatalytic performance under visible-light irradiation. Experimental part Materials and methods The Cu (NO3)2·6H2O, Ni(NO3)2·6H2O, Zn(NO3)2·6H2O, TiO2-P25, BiCl3, melamine, citric acid, and K2CO3 were provided from Merck. Using an X-ray diffraction (XRD) model (Bruker D8 Advance X-ray diffractometer set to 40 kV and 30 mA of Cu Kα), the structural characteristics of the produced particles were assessed. To further analyze the morphology and structure of the Bi2ZnTiO6, Ni/Cu-loaded Bi2ZnTiO6, and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 specimens, a scanning electron microscope (SEM) (model: MIRA3 TESCAN SEM) was used. Diffuse Reflection Spectroscopy (DRS, Avaspec-2048-TEC) was used to determine the energy band gap (Eg) of the Bi2ZnTiO6, g-C3N4, Ni/Cu-loaded Bi2ZnTiO6, Bi2ZnTiO6/g-C3N4 and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 specimens. Using the Ultraviolet–Visible spectrometer (UV–Vis) (model: Lambda850 Model, Perkin Elmer), the absorbance spectra of the carbon dioxide concentrations were measured. The elemental composition was identified using Energy Dispersive X-ray spectra (EDX) (model: F20 STWIN electron microscope (Tecnai G2, FEI Co); 200 kV). Additionally, the elemental mapping analysis (model: 7000 Shimadzu, Japan) was used to examine the atom distribution in the composition of the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 nanocomposite. The visible light irradiation source was a 70-W Xe arc lamp (CEL-HXF70, Beijing Aulight Co. Ltd.) equipped with an ultraviolet-cutoff filter (λ > 420 nm). Using a gas chromatography (GC) (Agilent model) was used to identify the gaseous products produced by the photoreduction of carbon dioxide. The Taguchi experimental design approach was used to improve the test settings and photocatalytic processes. One of the most effective approaches and engineering is to use the Taguchi experimental design. The Taguchi test design saves time and money by figuring out how the affected elements will operate together, determining the ideal conditions, and cutting down on the number of tests40. This is a brief overview at how the Taguchi experiment was set up: First, you need to identify the components and their levels, figure out how many relevant trials you need, run the analysis of variance (ANOVA), and then predict the optimal conditions about what the optimal conditions will be41. Table 1 shows the eight four-level effective variables that were studied in this research. (1) The pH of the solution dissolution (V1), which might be anywhere from 3 to 9. (2) The temperature of the reaction (V2) was at 25, 45, 65, and 80 °C. (3) Retention time (V3); at 60, 120, 240 and 300 min. (4) Photocatalyst doses of 0.5, 1.0, 1.5, and 2.0 g/l (V4). (5) CO2 concentration levels at 100, 200, 400, and 600 parts per million (V5). (6) RPM rounds of 50, 100, 200, and 300 (V6). (7) Interval between photoreduction site and radiation source from 5 cm up to 20 cm (V7). (8) Visible light intensity from 10 up to 120 W (V8). The Taguchi technique suggested the L32 array for this investigation, which comprises 32 tests. The eight effective parts in this array are arranged on four levels (Table 2). Table 2 shows how many different tests there are and the signal-to-noise (S/N) ratio using Taguchi’s approach. The Taguchi experimental design utilized a “greater is better” S/N ratio because a higher photoreduction rate of CO2 is a positive thing. The formula below yields the S/N value, which reflects variability in the response42,43. where R, n, and S/N are observed data, the number of observations, and signal-to-noise ratio, respectively. To obtain the signal-to-noise ratio, there are 3 relationships, each of which is used in specific situations: Table 1Factors and their corresponding levels.ParametersLevels1234V1. Solution pH3.05.07.09.0V2. Reaction temperature (°C)25406580V3. Stirring speed (rpm)100200300400V4. Retention time (min)60120240300V5. Photocatalyst dosage (g/l)0.51.01.52.0V6. CO2 concentration (mg L−1)100200400600V7. Light source distance to solution surface (cm)5101520V8. Light intensity (W)10305070 Factors and their corresponding levels. Table 2Tests number based on L32 (84) orthogonal array designed by Taguchi method.Exp. No.V1V2V3V4V5V6V7V8Y1Y2AverageS/NFactors and their levels Test 11111111143.6446.7845.7153.53 Test 21122222255.1251.0852.7257.96 Test 31133333341.2342.3143.4252.56 Test 41144444494.6595.0394.8473.96 Test 51212341272.5674.1373.8465.12 Test 61223412357.3254.1753.8855.18 Test 71234123466.1165.8766.7462.54 Test 81241234193.7292.9593.3370.34 Test 91313243173.4373.4574.0666.98 Test 101324314244.1544.2744.2153.21 Test 111331421385.5386.6486.0867.45 Test 121342132454.5155.6756.5857.73 Test 131414233147.6947.8947.9854.16 Test 141421342453.7954.6355.7056.49 Test 151432413176.5775.3475.9567.19 Test 161443124257.3560.8960.2860.06 Test 172112424352.8449.7550.8651.85 Test 182123131481.8285.8383.8267.45 Test 192134442164.3564.8964.6263.37 Test 202141113268.7967.2568.0265.23 Test 212213232263.1462.9863.0659.64 Test 222224343362.5862.3562.4759.58 Test 232231414471.3672.5471.9565.92 Test 242242121170.5369.8670.19563.87 Test 252314222154.7953.1253.9554.09 Test 262321333242.3543.7143.0342.78 Test 272332344390.9791.2391.1069.74 Test 282343411447.8946.7847.3448.96 Test 292411122365.3765.3465.3661.44 Test 302422233494.6893.5794.1272.97 Test 312433344172.3571.5671.9664.27 Test 322444411265.8364.9965.4161.38 Tests number based on L32 (84) orthogonal array designed by Taguchi method. Taguchi’s S/N ratio for smaller values is the best:2 Taguchi’s S/N ratio for smaller values is the best:2 Taguchi’s S/N ratio for larger-the-better:3 Taguchi’s S/N ratio for larger-the-better:3 Taguchi’s S/N ratio for nominal-the-best:4 Taguchi’s S/N ratio for nominal-the-best:4 The Taguchi experimental design utilized a “greater is better” signal-to-noise ratio because it is an effective approach to get CO2 photoreduction to solar fuels with high efficiency. Analysis of variance (ANOVA) can be used to find the ideal settings for a process by comparing the means of two or more groups. The F test in the ANOVA table indicates statistical significance. This is a glance at the ANOVA calculation and how it works: (1) The grand mean is found. (2) The sum of squares of each group’s scores from the great average (the sum of squares (deviations) of the group means from the grand average) is found. (3) The sum of the sum of squares of each group (the sum over the sums of squared deviations of scores around their group’s mean) is found. (4) The ANOVA table is made, and the F ratio is found. (5) The results of the ANOVA table are looked at and analyzed44. Table 3 shows the findings of an analysis of variance (ANOVA) that looked at how each component affected and improved the photocatalytic activity. Table 3Results of the variance analysis (ANOVA).ParametersOptimum operation conditionsDegree of freedom (f)Sum of square (S(Variance (v)F- ratioV1. Solution pH3323.476102.43218.64V2. Reaction temperature (°C)3100.12165.43010.79V3. Stirring speed (rpm)3151.45887.79017.13V4. Retention time (min)3345.792112.74220.09V5. Photocatalyst dosage (g/l)3564.294245.76135.78V6. CO2 concentration (mgL−1)3434.761198.78328.74V7. Light source distance to solution surface (cm)3397.752173.65425.43V8. Light intensity (W)3419.873194.97128.31 Results of the variance analysis (ANOVA). Table 4 shows the ideal conditions for photoreduction to convert carbon dioxide into usable fuels. Table 4 shows that the visible light intensity with the contribution ratio of 23.14%, quantity of photocatalyst with the contribution proportion of 20.12%, the solution pH of the solution with the contribution percentage of 15.36%, and how long it remains in the system with the contribution relation of 11.63%, are the most critical factors that impact how well carbon dioxide can be turned into usable fuels. But the reaction temperature, with the contribution proportion of 5.48%, the velocity of rotation, with the contribution ratio of 6.97%, the interval between the light source and the reaction medium, with the contribution range of 7.57%, and the amount of CO2, with the contribution proportion of 9.34%, have the least effect on the photocatalytic reduction of carbon dioxide. Table 4Optimum operation conditions.ParametersOptimum operation conditionsLevel descriptionLevelContribution (%)V1. Solution pH5.0215.36V2. Reaction temperature (°C)2515.48V3. Stirring speed (rpm)20026.97V4. Retention time (min)300411.63V5. Photocatalyst dosage (g/l)1.0220.12V6. CO2 concentration (mgL−1)40039.34V7. Light source distance to solution surface (cm)517.57V8. Light intensity (W)70423.14Photoreduction rate1727 μmol g−1 h−1 (total CO2 conversion 95%)––Major products (selectivity)CH4CH3OHOtherRate of major product863.5 μmol g−1 h−1 (selectivity 50%)690.8 μmol g−1 h−1 (selectivity 40%)86.35 μmol g−1 h−1 (selectivity 5%)Total99.61Error——0.39 Optimum operation conditions. Preparation of Bi2ZnTiO6 double perovskite by sol–gel approach Bi2ZnTiO6 double perovskite was synthesized via a sol–gel method. Citric acid (2.0 g) was dissolved in 25 mL of deionized water, followed by the addition of stoichiometric amounts of BiCl3 and Zn(NO3)2·6H2O as precursors under continuous stirring. TiO2-P25 was dissolved in an acidic HNO3/HF solution under controlled conditions to facilitate titanium incorporation into the sol. All HF-related procedures were performed in a fume hood with appropriate safety precautions. The mixture was stirred at 70 °C for 120 min to obtain a homogeneous gel. It could keep the time at 120 min, the speed at 200 rpm, and the temperature at 70 °C. After the impurities were removed, the gel was heated to 350 °C for 1 h to remove residual organic species. The powdered Bi2ZnTiO6 double perovskites were heated to 700 °C for two hours and then utilized for photocatalytic testing34,45,46. Synthesis of graphitic carbon nitride (g-C3N4 (gCN) by heat programming procedure Using a particular heat program, 50 mL of deionized water and 5 g of melamine were mixed to generate graphitic carbon nitride (g-C3N4(gCN)). The mixture was stirred at 70 °C for 120 min at 200 rpm. The solution was dried to 80 °C and was placed in a continuous process for 120 min at 520 °C (5 °C per minute) to get the water out34,47. Manufacturing the Ni/Cu-loaded Bi2ZnTiO6 by reduction co-precipitation To prepare the Ni/Cu-loaded Bi2ZnTiO6 specimen, 1.0 g of synthesized Bi2ZnTiO6 double perovskites was incorporated in 25 mL of deionized water, and henceforth, 0.1 g of metallic salts of nickel (Ni(NO3)2·6H2O) and copper (Cu(NO3)2.6H2O) was added to the provided solution. Then 5 mL of Sodium borohydride (NaBH4) solution was gradually introduced as reducing agent to convert Ni2+ and Cu2+ ions to their metallic states (Ni° and Cu°). The suspension was stirred at 70 °C to ensure uniform deposition of metal nanoparticles onto the Bi2ZnTiO6 surface. The resulting solid was filtered, washed, dried, and calcined at 350 °C for 1 h48. Construction of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 by hydrothermal approach The Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary nanocomposite was manufactured using a hydrothermal method. 50 mL of distilled water was used to spread out 1.1 g of Ni/Cu-loaded Bi2ZnTiO6 sample at first. After being subjected to ultrasonic dispersion for an hour, 0.1 g of g-C3N4 non-metallic co-photocatalyst was poured into a solution containing Ni/Cu-loaded Bi2ZnTiO6 sample. The Bi2ZnTiO6 particles were combined with zero-valent Ni/Cu nanoparticles and g-C₃N₄ under hydrothermal conditions to achieve uniform dispersion. The mixture was then stirred under optimal circumstances, which were a pH of 7, a reaction temperature of 70 °C, a rotation velocity of 200 rpm, and a retention time of 120 min. This was done using a magnetic stirrer that could be set to a specific temperature. The solution was then placed in an autoclave at 180 °C for 2 h. It was cleansed numerous times with deionized water once it was found. After that, Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 powder was desiccated at 80 °C for 12 h34. Photoreduction process of CO2 A 70-W Xe arc lamp was used with an ultraviolet-cutoff filter to make the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary heterojunction nanocomposite work well to photoreducte CO2. The first step was to add 1 L of deionized water and 0.4 g of pure sodium carbonate (Na2CO3) to the photoreactor. Alternatively, a more accurate alternative method can be used to bubble high-purity CO2 gas into an aqueous suspension containing 1.0 g/L of photocatalyst until CO2 reaches saturation. The end product was a solution that had 400 mg L−1 of carbon dioxide in it. This was done to show how carbon dioxide gas and water combine. Next, 1.0 g Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 nanocomposite was poured into a solution containing 400 mg L−1 of dissolved carbon dioxide (carbonate (CO32−) inside the photoreactor. The solution was shaken in the dark for 20 min to evaluate the influences of surface adsorption. No detectable CH4 or CH3OH was observed in the absence of light. Then, a visible light irradiation source with a power of 70-W was used to beam visible light on the water solution containing the CO2 for five hours at 25 °C. It was investigated how the concentration and absorbance of CO32− ions in the solution changed after the procedure was finished using an Ultraviolet–Visible spectrometer. The carbonate ion absorbs ultraviolet light at a wavelength of 220 nm, and it gives a sharp and strong peak in this area. After visible-light irradiation for 5 h at 25 °C, gaseous products were analyzed using gas chromatography. Product identification and quantification were performed using calibration curves of standard gases. The photoreduction efficiency was calculated based on the amount of carbon-containing products generated per gram of photocatalyst per hour (μmol g−1 h−1). Results and discussion Analysis of prepared photocatalysts As shown in Fig. 2, XRD analysis was performed to determine the crystallinity and phase type of the Bi2ZnTiO6, Ni/Cu-loaded Bi2ZnTiO6, and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalysts. Every pattern in the provided XRD spectra of the Bi2ZnTiO6 double perovskite had diffraction peaks that, instead of superfluous peaks, are consistent with the standard JCPDS reference pattern, confirming phase purity and crystallinity. According to the JCPDS index: 37–0243 (Fig. 2a), the different planes of (112), (110), (222), (400), (200), (440), (211), and (622) support the organization of the Bi2ZnTiO6 structure with 2θ values of 22°, 25°, 3°3, 3°6, 4°5, °65, °72, and °75, respectively. The JCPDS file 87-0721’s peaks match those in the Ni/Cu nanoparticles XRD pattern. A characteristic diffraction peak at 2θ = 22°, the g-C3N4 nanosheets peak was also seen at (002). Additionally, the organization of the Ni/Cu-loaded Bi2ZnTiO6 structure (JCPDS: 70-1795) is supported by the different planes of (111), ()220), (311), (222), (400), and (422) with 2θ values of °22, °25, °33, 4°2, °45, and °51, respectively (Fig. 2b). The peak of the g-C3N4 XRD pattern is associated with JCPDS file number 871526. The Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalyst’s XRD analysis showed a peak at (002) in Fig. 2c with a 2θ angle of 22°, in accordance with the typical XRD pattern with JCPDS: 87-1526. XRD patterns of the Bi2ZnTiO6 (a), Ni/Cu-loaded Bi2ZnTiO6 (b), and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 (c) samples. The full X-ray photoelectron spectroscopy (XPS) spectra of Ni/Cu-loaded Bi2ZnTiO6 /g-C3N4 nanocomposites (a) are presented in Fig. 3. As can be viewed in Fig. 3, the elements of Ni, Cu, Bi, Zn, Ti, O, C, and N can be observed in the XPS spectrum of each material. The XPS results confirm that Ni and Cu nanoparticles, Bi2ZnTiO6, and g-C3N4 coexist in the Ni/Cu-loaded Bi2ZnTiO6 /g-C3N4 photocatalyst. Also, Fig. 3 displays the high-resolution XPS spectra for Ni 2p (b), Cu 2p (c), Bi 4f. (d), Zn 2p (e), Ti 2p (f), O1s (g), C 1 s (h), and N 1 s (i) for the hybrid structure of Ni/Cu-loaded Bi2ZnTiO6 /g-C3N4 ternary photocatalyst together with their peak-fits. The XPS spectra of the Ni/Cu-loaded Bi2ZnTiO6 /g-C3N4 ternary photocatalyst exhibit characteristic binding energy peaks of Ni 2p (b) and Cu 2p (c) associated with Ni/Cu nanoparticles, Bi 4f (d), Zn 2p (e), Ti 2p (f), and O 1 s (g) related to Bi2ZnTiO6 double perovskite, and C 1 s (h) and N 1 s (i) corresponding to g-C3N4, confirming the coexistence of Ni°, Cu°, Bi3+, Zn2⁺, and Ti4+ species. The Ni 2p and Cu 2p spectra display distinct peaks of Ni° and Cu° species, indicating the confirming the presence of metallic Ni⁰ and Cu⁰ species. The Bi 4f peaks are centered around Bi3+, confirming the integrity of the Bi–O framework in Bi2ZnTiO6. Meanwhile, the Zn 2p region reveals characteristic peaks attributed to Zn2+, verifying the formation of the perovskite oxide structure. The Ti 2p spectrum shows the presence of Ti–O bonds, suggesting a stable TiO6 octahedral environment within Bi2ZnTiO6, which enhances its ability to facilitate charge transport. Furthermore, the O 1s peaks demonstrate lattice oxygen and surface hydroxyl groups that play vital roles in photocatalytic redox reactions. The observed slight shifts in binding energies reflect strong electronic coupling and interfacial charge transfer among the components, confirming the successful formation of an integrated Schottky/Z-scheme heterojunction structure that promotes efficient separation and migration of photogenerated charge carriers. Full X-ray photoelectron spectra of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 nanocomposite (a). Focused XPS spectra of (a) Ni 2p, (b) Cu 2p, (c) Bi 4f, (d) Zn 2p, (e) Ti 2p, (f) O 1s, (g) C 1 s (h), and (i) N 1 s. The FE-SEM, TEM, and HRTEM were used to evaluate the nanoparticles size and shape of the Bi2ZnTiO6, Ni/Cu-loaded Bi2ZnTiO6, and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 nanoparticles. Figure 4a displays the Bi2ZnTiO6 particles’ FE-SEM image. The uniformly distributed, quasi-spherical Bi2ZnTiO6 nanoparticles were, on average, less than 100 nm in size, as shown in Fig. 4a. Additionally, as exhibited in Fig. 4b, the structure of the synthesized Ni/Cu particles was investigated using FE-SEM. As shown in Fig. 4b, the zero-valent Ni/Cu nanoparticles were uniformly distributed, spherical, and, on average, less than 20 nm in size. The FE-SEM picture of the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary heterojunction photocatalyst is shown in Fig. 4c. The FE-SEM picture displays loosely aggregated nanostructures with a mean size of 99.73 nm, as represented in Fig. 4c. Lastly, the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary heterojunction nanocomposite TEM images are presented in Fig. 4d. The average particle size of the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary nanocomposite was found to be around 40.9 nm, as shown in Fig. 4d. The TEM image confirms the intimate interfacial contact among Ni/Cu nanoparticles, Bi2ZnTiO6 nanocrystals, and g-C3N4 nanosheets and evenly dispersed throughout the photocatalyst structure, as shown by the TEM picture in Fig. 4d. In order to facilitate feature identification and discrimination, Fig. 4e also displays the HRTEM picture of the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary heterojunction photocatalyst. The presence of all three Ni/Cu nanoparticles, Bi2ZnTiO6 nanoparticles, and g-C3N4 nanosheets next to one another in the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 nanocomposite structure is confirmed by the HRTEM picture in Fig. 4e. FE-SEM images of the Bi2ZnTiO6 double perovskite (a), Ni/Cu-loaded Bi2ZnTiO6 (b), Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 (c) samples and TEM (d) and HRTEM (e) of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 nanocomposite. The elements nickel (Ni), copper (Cu), bismuth (Bi), zinc (Zn), titanium (Ti), nitrogen (N), oxygen (O), and carbon (C) make up the nanoparticles of the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary nanocomposite, according to an EDX device (Fig. 5a). The Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary nanocomposite was discovered to have a clean chemical composition free of impurities, and the EDX analysis revealed no additional peaks. The mapping study of the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary nanocomposite is shown in Fig. 5b, and it verifies that the Ni, Cu, Bi, Zn, Ti, N, O, and C atoms are present and uniformly distributed within the nanocomposite structure. The different colored dispersion maps of the Ni, Cu, Bi, Zn, Ti, N, O, and C atoms provide evidence of the evenly distributed composition of the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 nanocomposite. EDX spectrum (a) and Elemental mapping analysis (b) for Bi, Ni, Cu, Zn, Ti, O, N, and C elements in Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalyst. Figure 6 displays the surface areas of Bi2ZnTiO6, g-C3N4, Ni/Cu-loaded Bi2ZnTiO6, Bi2ZnTiO6/g-C3N4, and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalysts as determined by BET analysis. The three substances’ type II isotherms demonstrate their macroporous nature. The BET investigation’s findings, shown in Fig. 6a, specify that pure Bi2ZnTiO6 double perovskite has a specific surface area of 53.517 m2 g−1. The specific surface area of the g-C3N4 nanosheets was determined to be 63.784 m2 g−1 (Fig. 6b). The particular reaction surface of the Bi2ZnTiO6 double perovskite structure has increased from 53.517 to 65.872 m2 g−1 owing to the loading the Ni/Cu nanoparticles (Fig. 6c). Also, the specific surface area of the Bi2ZnTiO6 double perovskite structure has increased from 53.517 to 70.234 m2 g−1 owing to the incorporation of g-C₃N₄ nanosheets (Fig. 6d). Additionally, the Ni/Cu-loaded Bi2ZnTiO6 structure’s specific surface area rose from 65.872 to 78.549 m2 g−1 with the addition of g-C3N4 nanosheets. According to the BET study, the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 nanocomposite’s specific surface area (78.549 m2 g−1) is significantly higher than that of the Ni/Cu-loaded Bi2ZnTiO6 (65.872 m2 g−1) and pure Bi2ZnTiO6 double perovskite (53.517 m2 g−1) in Fig. 6e. The enhancement in BET surface area is primarily attributed to improved dispersion and heterointerface formation rather than intrinsic porosity changes rather than intrinsic porosity. The coupling with layered g-C3N4 effectively inhibits particle agglomeration of Bi2ZnTiO6 and promotes a more dispersed heterostructure, thereby exposing additional surface-active sites. Moreover, the deposition of Ni/Cu nanoparticles further suppresses nanoparticles growth and creates interfacial defects and heterojunction nucleation sites, which collectively contribute to a modest but measurable increase in the specific surface area. It should be emphasized that the enhancement in photocatalytic performance is not solely governed by the increase in BET, but predominantly arises from improved charge separation and interfacial charge transfer induced by the Z-scheme heterojunction and Ni/Cu-induced Schottky junctions. BET analyses of the Bi2ZnTiO6 double perovskite (a), Ni/Cu-loaded Bi2ZnTiO6 (b), and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 (c) samples. To investigate the impact of loading zero-valent Ni/Cu nanoparticles and g-C3N4 co- photocatalysts on the energy band gap and photocatalytic activity of the Bi2ZnTiO6, the UV–Vis/DRS spectra of pure Bi2ZnTiO6 double perovskite and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 nanocomposite were acquired. The impacts of adding ratios of g-C3N4 and zero-valent Ni/Cu nanoparticles on the Bi2ZnTiO6 structure were assessed using the various ratios of 0.01, 0.1, 0.5, and 1.0, which are listed in Table 5. These values were chosen as the ideal ratios because the results demonstrated that the photocatalytic activity of the Bi2ZnTiO6 was most affected by zero-valent Ni/Cu nanoparticles with a ratio of 0.1 and zero-valent Ni/Cu nanoparticles with a ratio of 0.1. Figure 7a displays the DRS plots of the provided powder specimens. The band gap and absorption characteristics of the synthesized photocatalysts were examined using DRS analysis. The Bi2ZnTiO6 double perovskite has an absorption peak at 600 nm in the UV–vis spectrum. The Tauc-Mott (TM) equation was used to determine the Eg of the Bi2ZnTiO6, pure g-C3N4, Ni/Cu-loaded Bi2ZnTiO6, Bi2ZnTiO6/g-C3N4, and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalysts: where F(R), R, υ, h, k, and Eg are the Kubelka–Munk function, reflection value, irradiance frequency, Planck’s constant (6.626 × 10–34 J s), absorption constant, energy band gap of samples, respectively8,34,49. The electrochemical impedance characterization is applied as a strong technique for evaluating the separation efficiency of photogenerated electron–hole pairs. The Nyquist impedance spectrums of Bi2ZnTiO6, pure g-C3N4, Ni/Cu-loaded Bi2ZnTiO6, Bi2ZnTiO6/g-C3N4, and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalysts are given in Fig. 7b. The smaller arc radius of the Nyquist impedance curve means a lower load transfer resistance across the photoelectrodes’ surface, which can further attain more efficacious dissociation of e−/h+ pairs and low interfacial resistance. The larger semicircle diameter indicates higher charge transfer resistance. Although the Nyquist impedance arc radius emerges after Ni/Cu nanoparticle and g-C3N4 nanosheets loading, manifesting that Ni/Cu nanoparticles and g-C3N4 nanosheets loading decrease the electron transfer resistance, which is in accordance with the prior report. It is verified that pure g-C3N4, Ni/Cu-loaded Bi2ZnTiO6, Bi2ZnTiO6/g-C3N4, and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 had smaller arc radii than Bi2ZnTiO6 double perovskite, exhibiting a more efficient charge separation of photogenerated electron–hole pairs and a lower charge transfer resistance across the surface of pure g-C3N4, Ni/Cu-loaded Bi2ZnTiO6, Bi2ZnTiO6/g-C3N4, and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 than Bi2ZnTiO6 double perovskite, which can be responsible for the better photocatalytic activity. The energy band gap of the provided Bi2ZnTiO6 perovskite was calculated to be 2.6 eV. By incorporating the zero-valent Ni/Cu nanoparticles to the structure of Bi2ZnTiO6, its band gap was reduced to 2.0 eV. Further, with the appending of g-C3N4 nanosheets to the structure of Ni/Cu-loaded Bi2ZnTiO6 photocatalyst, the apparent band gap energy of the composite was estimated to be about 2.3 eV using the Tauc diagram (Fig. 7c). To appraise the efficacies of co-photocatalysts incorporating the structure of Bi2ZnTiO6, a photoluminescence (PL) test was done, as represented in Fig. 7d. The PL characterization exhibits the recombination rate of electrons formed in photocatalysts. The high intensity of the peaks in the PL curves means that the recombination rate of e−/h+ pairs is high. Pursuant to Fig. 7d, by augmenting zero-valent Ni/Cu metallic co-photocatalyst and g-C3N4 non-metallic co-photocatalyst to the structure of Bi2ZnTiO6 perovskite, the PL intensity has been significantly diminished, which illustrates that introducing these assistance-photocatalysts increases the recombination suppression and productivity of e−/h+ pairs generated. Accordingly, the photoreduction efficiency of the CO2 by Bi2ZnTiO6 perovskite increases. The large intensity of the PL curve of Bi2ZnTiO6 perovskite displays that the enhanced charge separation and efficiency of e−/h+ pairs are low. By augmenting nickel/copper nanoparticles and g-C3N4 nanosheets to the Bi2ZnTiO6 structure, the PL spectrum peak intensity diminished, and the dissociation rate of e−/h+ pairs increased. To investigate the photoresponse behavior of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary nanocomposite, the current density–time curves were obtained (Fig. 7e). It can be seen that the current density amount of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 is significantly higher than pure Bi2ZnTiO6 double perovskite and Ni/Cu-loaded Bi2ZnTiO6 samples in the presence of visible light radiance, and is reduced rapidly when turning off the light. The enhanced photocurrent density can lead to the improvement of load migration from Ni/Cu nanoparticles and g-C3N4 nanosheets toward the Bi2ZnTiO6 double perovskite. Table 5The addition ratios influence of Ni/Cu metallic co-photocatalyst and g-C3N4 non-metallic co-photocatalysts to the Bi2ZnTiO6 double perovskite structure.NumberAddition ratio )Ni/Cu:BZTO:gCN(Photoconversion rate of CO2 (%)10.01:1:0.01725.34 μmol g−1 h−1 (CO2 conversion 42%)20.1:1:0.11727 μmol g−1 h−1 (CO2 conversion 95%)30.5:1:0.51317.18 μmol g−1 h−1 (CO2 conversion 76.27%)41:1:1409.917 μmol g−1 h−1 (CO2 conversion 27.73%) The addition ratios influence of Ni/Cu metallic co-photocatalyst and g-C3N4 non-metallic co-photocatalysts to the Bi2ZnTiO6 double perovskite structure. UV–vis/DRS spectra (a), Tauc-Mott plot (b), and Photoluminescence spectra (c) to the Bi2ZnTiO6, g-C3N4, Ni/Cu-loaded Bi2ZnTiO6, Bi2ZnTiO6/g-C3N4, and the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalysts. The Mott–Schottky plots of all Bi2ZnTiO6 double perovskite, pure g-C3N4, Ni/Cu-loaded Bi2ZnTiO6, Bi2ZnTiO6/g-C3N4, and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalysts have been presented in Fig. 8. The flat-band potentials were extracted from the linear region of the MS plots. The flat-band potentials (Vfb) were determined from the intercepts with the potential axis, yielding − 1.10 V for g-C3N4, − 0.50 V for Bi2ZnTiO6, − 0.65 V for the Bi2ZnTiO6/g-C3N4 composite, and − 0.75 V for the Ni/Cu-loaded composite (vs Ag/AgCl). Based on the experimentally measured band gaps, the conduction and valence band positions were precisely calculated and summarized in Table 6. All samples exhibited positive slopes in the linear regions, confirming their n-type semiconductor behavior. The conduction band (CB) potentials of the semiconductors were estimated to be approximately equal to their respective flat-band potentials, resulting in CB values of − 1.10 V (g-C3N4), − 0.50 V (Bi2ZnTiO6), − 0.65 V (Bi2ZnTiO6/g-C3N4), and − 0.75 V (Ni/Cu-loaded Bi2ZnTiO6/g-C3N4). Accordingly, the valence band (VB) positions were calculated using the optical bandgap (Eg) values: g-C3N4 (Eg = 2.7 eV) → VB ≈ + 1.60 V, Bi2ZnTiO6 (Eg = 2.6 eV) → VB = + 2.10 V, Bi2ZnTiO6/g-C3N4 (Eg ≈ 2.45 eV) → VB ≈ + 1.80 V, and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 (Eg = 2.3 eV) → VB ≈ + 1.25 V. Notably, the slopes of the MS plots decreased progressively from g-C3N4 to Bi2ZnTiO6, further to the Bi2ZnTiO6/g-C3N4 composite, and finally to the Ni/Cu-loaded composite, indicating a substantial increase in charge carrier density (Nd). Specifically, Nd was estimated to be on the order of ~ 101⁸ cm⁻3 for g-C3N4, ~ 4 × 101⁸ cm⁻3 for Bi2ZnTiO6, ~ 8 × 101⁸ cm⁻3 for the Bi2ZnTiO6/g-C3N4 composite, and ~ 101⁹ cm⁻3 for the Ni/Cu-loaded composite. The reduction in slope for the composites reflects enhanced charge separation and accelerated interfacial electron transfer, likely due to electronic coupling between Bi2ZnTiO6 and g-C3N4 and the formation of a Schottky junction in the Ni/Cu-loaded system. The band alignment analysis reveals that the CB of g-C3N4 is the most negative, making it the primary site for electron accumulation and reduction reactions, while the VB of Bi2ZnTiO6 is the most positive, favoring oxidation reactions. This configuration supports a direct Z-scheme charge transfer mechanism, in which photogenerated electrons in the CB of Bi2ZnTiO6 recombine with holes in the VB of g-C3N4, effectively preserving strong redox potentials in the remaining electrons and holes. Overall, the Mott–Schottky analysis corroborates the enhanced photocatalytic performance of the composites, particularly the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 system, by providing quantitative insight into carrier density, band positions, and charge transfer dynamics. These results further confirm the proposed Z-scheme charge transfer mechanism. Mott-Schottky (MS) plots of the Bi2ZnTiO6, g-C3N4, Ni/Cu-loaded Bi2ZnTiO6, Bi2ZnTiO6/g-C3N4, and the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalysts. Table 6Summary of band structure parameters obtained from UV–Vis and Mott–Schottky analyses for all samples.SampleEg (eV)Vfb (V vs Ag/AgCl)Vfb (V vs NHE)Slope (× 1010 cm−3)Nd (× 1019 cm−3)ECB (V vs NHE)EVB (V vs NHE)g-C3N42.70− 0.92− 0.7233.21.0− 0.823 + 1.877Bi2ZnTiO62.60− 0.41− 0.2133.01.1− 0.313 + 2.287Ni/Cu-loaded Bi2ZnTiO62.00− 0.36− 0.1632.61.3− 0.263 + 1.737Bi2ZnTiO6/g-C3N42.65− 0.38− 0.1832.81.2− 0.283 + 2.367Ni/Cu-loaded Bi2ZnTiO6/g-C3N42.30− 0.34− 0.1432.41.4− 0.243 + 2.057 Summary of band structure parameters obtained from UV–Vis and Mott–Schottky analyses for all samples. The Fig. 9 experimental setup consists of the carbon dioxide gas chamber, a visible light photoreactor with a 1-L volume, a visible light lamp with a power of 70 W as an irradiation source, a thermometer, a manometer, and a gas chromatography device. The photoreactor is a magnetically stirred photoreactor comprising a solution with a determined amount of photocatalyst and a specified concentration of dissolved CO2, a digital thermometer to measure the photoreactor temperature within the 0–100 °C range, and a manometer to adjust the pressure of the gas container within the 1 to 10 MPa range. Determined amounts of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 0D/3D/2D QDs Schottky/Z-scheme heterojunction nanocomposite were added to the circulating solution inside the photoreactor. Then, specific concentrations of CO2 are injected into the photoreactor with an adjusted pressure. The CO2 photoreduction reaction was conducted in the photoreactor under solution pH:5, reaction medium temperature: 25 °C, mixing velocity: 200 rpm, reaction time: 300 min, nanocomposite dosage: 1.0 g/l, a concentration of carbon dioxide: 400 mg L−1, visible light intensity: 70 W, reaction medium: pure water, and interval between reaction medium and radiation source: 5 cm. After the reaction is completed, a volume of the final reaction mixture is characterized by gas chromatography (GC) for characterization and identification of the species and amounts of the generated products. Industrial process of CO2 reduction to solar fuels by Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalyst, schematically. Photocatalytic activity CO2 photoreduction It was intended to find out whether the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary heterojunction nanocomposite could turn carbon dioxide into value-added solar fuels like CH4 and CH3OH when it was lit up with visible light. 0.4 g of pure sodium carbonate (Na2CO3) as the carbon source and 1 L of clean deionized water were put together in the photoreactor. This produced a solution containing 400 mg L−1 of carbonate (CO32−), which is carbon dioxide that has been mixed with water. Then, the photoreactor received a solution with 1.0 g of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 heterojunction photocatalyst and 400 mg L−1 of CO2 in it. The reaction medium was at 25 °C, had a pH of 5, and operated for 300 min at 200 rpm. To have the best results in decreasing CO2, the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalyst had to be 1.0 g/l, the CO2 concentration had to be 400 mg L−1, the light intensity had to be 70 W, the addition agent had to be clean water, and the distance from the radiation to the solution had to be 5 cm. So far, the data showed that reducing the pH to 5 makes the process of lowering CO2 operate better. This is why the pH level was changed to 5. The concentration of dissolved carbonate species was monitored by UV–vis spectroscopy at 210–220 nm. This is because UV light with a wavelength of 210–220 nm is the best at making carbonate ions absorb light. The Gas Chromatography (GC) was used to look at the gases in the various parts of the photoreactor. This is another method to show the equation of CO2 conversion efficiency (RCO2 (%))5: Abs0 is when CO2 is initially taken in, and Abs is when CO2 is last taken in. There are two main ways that photocatalytic conversion of CO2 to valuable fuels may happen: (1) The path through CHₓ radical creation, and (2) the path through formaldehyde creation. Possible reactions and intermediates produced in both photocatalytic reduction pathways of carbon dioxide are given below25: Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 + hv → eCB− + hVB+ Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 + hv → eCB− + hVB+ R-2 R-2 CHₓ radical pathway:R-3R-4R-5R-6R-7R-8R-9R-10 CHₓ radical pathway:R-3R-4R-5R-6R-7R-8R-9R-10 Formaldehyde-mediated pathway:R-11R-12R-13R-14R-15R-16R-17R-18R-19R-20R-21R-22R-23R-24 Formaldehyde-mediated pathway:R-11R-12R-13R-14R-15R-16R-17R-18R-19R-20R-21R-22R-23R-24 Table 7 also showed how photocatalytic reduction of CO2 into different products worked, including the steps involved, the chemical potentials (E0), and the standard Gibbs free energy (∆G0). The UV–visible spectra of solutions contaminated with carbon dioxide at a concentration of 400 mg L−1 containing Bi2ZnTiO6 double perovskite, pure g-C3N4, Ni/Cu-loaded Bi2ZnTiO6, Bi2ZnTiO6/g-C3N4, and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalysts are shown in Fig. 10a. The Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary nanocomposite could reduce CO2 levels by up to 1727 μmol g−1 h−1 (total CO2 conversion efficiency 95%) when it was in aqueous solution and exposed to visible light at a pH 5, a 25 °C reaction temperature, a 200 rpm mixing speed, a 300 min reduction time, a 1.0 g/l photocatalyst quantity, a 400 mg L−1 CO2 concentration, a 5 cm distance from the visible irradiation source to the reaction place, and a 70 W light irradiation intensity. The Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary nanocomposite photoreduces CO2 into beneficial fuels like methane (CH4) gas, methanol (CH3OH), and formic acid when it is struck by visible light. Figure 10b shows GC spectrum of generated gas products for Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 sample. The main reason carbon dioxide levels went down was because of methane (CH4) gas, which made up around half of the total. Figure 10c shows how the amount of CO2 changes over time. The data in Fig. 10c reveal that the level of CO2 went down from 100% to less than 5% during the duration of 300 min. After that, the quantities of CO2 that were exchanged stayed the same until the reaction was over. The Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 specimen exhibited the highest activity on lowering CO2 levels, reducing them by 1727 μmol g−1 h−1 (CO2 conversion 95%). Figure 10d shows how many species and in what ratio are produced as a result of the photocatalytic reduction of carbon dioxide. This graph shows that methane gas (CH4) makes up 863.5 μmol g−1 h−1 (selectivity efficiency 50%) of the products. Methanol (CH3OH) was the second most important product, accounting for around 690.8 μmol g−1 h−1 (selectivity efficiency 40%) of the total. Formic acid (CH2O2) had the lowest production value, at only 34.54 μmol g−1 h−1 (selectivity efficiency 2%). Figure 10e depicts first-order kinetics to CO2 photoreduction by Bi2ZnTiO6 double perovskite, pure g-C3N4, Ni/Cu-loaded Bi2ZnTiO6, Bi2ZnTiO6/g-C3N4, and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalysts. The result in Fig. 10e reveal that the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary nanocomposite has the highest rate constant (0.0292 min−1) and R2 (0.9732) of all the samples. The R2 value and reaction rate constant (k) for pure g-C3N4 to photoconversion of CO2 into light were around 0.7841 and 0.0105 min−1. Also, the R2 value and reaction rate constant (k) for Bi2ZnTiO6 perovskite to photoconversion of CO2 into light were around 0.8083 and 0.0115 min−1. The Ni/Cu-loaded Bi2ZnTiO6 sample could also photoreduce CO2 at a rate of 0.9621 to 0.0203 min−1. In addition, the Bi2ZnTiO6/g-C3N4 sample could also photoreduce CO2 at a rate of 0.9645 to 0.0223 min−1. The kinetic analysis indicates an increased apparent rate constant and the photocatalytic activity became greater when Ni/Cu-loaded Bi2ZnTiO6 and g-C3N4 were added as co-photocatalysts to the Bi2ZnTiO6 double perovskite structure. Figure 10f shows how the added agents affect the success of the CO2 photoreduction into valuable chemicals. The result Fig. 10f shows that the CO2 pollutant was best photoreduced in pure water solvent, with an photoreduction rate of over 1727 μmol g−1 h−1 (CO2 conversion 95%). Adding NaOH, NaCl, and Na2SO4 to CO2 also made it less effective in photoreduction by around 1347.06 μmol g−1 h−1 (CO2 conversion 78%), 1467.95 μmol g−1 h−1 (CO2 conversion 85%), and 984.39 μmol g−1 h−1 (CO2 conversion 57%), respectively. Table 7The possible reactions during the photoreduction process of CO2 into various products.NumberReactionsE0(V) vs. NHE at pH = 7∆G0 (kJ.mol−1)12H2O (l) + 4h+ → O2 (g) + 4H+1.232372CO2 (g) + 8H+ + 8e− → CH4 (g) + 2H2O (l)− 0.248183CO2 (g) + 6H+ + 6e− → CH3OH (aq) + H2O (l)− 0.3870342H+ + 2e− → H2 (aq)− 0.41–5CO2 (g) + 4H+ + 4e− → HCHO (aq) + H2O (l)− 0.485226CO2 (g) + 2H+ + 2e− → CO (g) + H2O (l)− 0.532577CO2 (g) + 2H+ + 2e− → HCOOH (aq)− 0.612868CO2 (aq) + e− → CO2− (aq)− 1.9– The possible reactions during the photoreduction process of CO2 into various products. Reduction percentage of CO2 by synthesized samples (a), GC spectrum of gas products for Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 sample (b), CO2 concentration changes (C/C0) (c), type and ratio of generated species from CO2 reduction (d), the photoreduction kinetic diagram of carbon dioxide (e), and effect of additive agents on CO2 photoreduction (f), at optimum conditions. Reaction mechanism and kinetic study Studying reaction kinetics may help understand how reactions happen. The experimental data were fitted using pseudo-first-order and pseudo-second-order kinetic models. The pseudo-first-order model exhibited the highest correlation coefficient and was therefore selected. The following equation indicates how photoreduction works: C0 is the initial concentration of CO2, C is the CO2 concentration at time t, and k is the rate constant of the photoreduction reaction25. Figure 10e illustrates the straight line that connects the time it takes for the pure g-C3N4, Bi2ZnTiO6, Ni/Cu-loaded Bi2ZnTiO6, Bi2ZnTiO6/g-C3N4, and Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalysts to reduce CO2 concentration to the value of –ln(C/C0). This means that the rate at which the CO2 is photoreduced follows first-order reaction kinetics. The Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 0D/3D/2D QDs Schottky/Z-scheme heterojunction photocatalyst exhibited a maximum rate constant (k) of 0.0292 min−1 with R2 = 0.9732. The reaction’s kinetic curve showed that adding Ni/Cu metallic co-photocatalyst and g-C3N4 non-metallic co-photocatalyst to the Bi2ZnTiO6 speeds up its photocatalytic activity and photoreduction efficiency of CO2. Impact of the main variables on the photoreduction of CO2 to solar fuels Effect of solution pH The pH of the reaction medium plays a critical role in photocatalytic CO2 reduction, as it influences surface charge properties, adsorption behavior, and the availability of reactive species. As shown in Fig. 11a, the highest CO2 photoreduction efficiency was achieved at pH 5. This enhancement can be attributed to improved CO2 adsorption and the increased availability of protons, which facilitate multi-electron reduction pathways toward CH4 and CH3OH. Under slightly acidic conditions, charge transfer processes are kinetically favored without inducing structural instability in the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 heterojunction. Impact the different variables including solution pH (a), reaction temperature (b), retention time (c), photocatalyst dosage (d), CO2 concentration (e), stirring speed (f), light source distance to solution surface (g), and light intensity (h) on the photocatalytic efficiency of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary heterojunction photocatalyst. Impact of reaction temperature At this stage, it was investigated how CO2 was photoreduced at temperatures between 25 and 80 °C. Figure 11b demonstrates that how changes in different temperatures affect the photoreduction efficiency of the CO2. As shown in Fig. 11b, when the temperature goes up from 25 to 65 °C, the efficacy of carbon dioxide photoreduction goes down from 1727 (CO2 conversion efficiency 95%) to 1295.25 μmol g−1 h−1 (75% efficiency). When the temperature rose from 65 to 80 °C, the photocatalytic reduction efficiency of CO2 pollutants went up from 1295.25 (75% efficiency) to 1467.95 μmol g−1 h−1 (85% efficiency). One of the temperatures that was looked at was 25 °C, which was the optimal operating temperature to mitigate CO2 emissions. Influence of reaction time When exposed to visible light, the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary heterojunction nanocomposite exhibits time-dependent CO2 photoreduction activity, as shown in Fig. 11c. It can be observed from Fig. 11c that how successfully CO2 were photoreduced after being in contact for varied amounts of time. The photoreduction yield of the CO2 pollutant was increased up to a reaction time of 300 min. After this point, the conversion efficiency didn’t vary substantially; thus, 300 min was determined as the optimal reaction time to reach steady-state conditions. The photoreduction efficiency for dissolved CO2 gas to usable products was around 1727 μmol g−1 h−1 (total CO2 conversion efficiency 95%) after 300 min of contact time when the light was visible. The main products were methane gas and methanol. Methanol was made at a rate of 690.8 μmol g−1 h−1 (40% efficiency), whereas methane gas was produced at a rate of 863.5 μmol g−1 h−1 (50% efficiency). Impact of photocatalyst dosage It was investigated how varying amounts of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary heterojunction nanocomposite changed the capacity of CO2 to photoreduce in the range of 0.5 to 2.0 g/l (Fig. 11d). The CO2 pollutant’s photoreduction efficiency increases from 949.85 (55% efficiency) to 1727 μmol g−1 h−1 (total CO2 conversion efficiency 95%) when the quantity of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 goes up from 0.5 to 1.0 g/l (Fig. 11d). A plausible explanation is that the more Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 nanocomposite there is, the more active sites. When quantities of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalyst were incremented to more than 1.0 g/l, the photoreduction yield of CO2 was reduced to 1450.68 μmol g−1 h−1 (84% efficiency). The drop is because the excessive photocatalyst loading increase light scattering and reduce photon penetration, which makes it less effective. Effect of CO2 concentration Figure 11e shows that it looked at how well the reduction works with varied levels of CO2, from 100 to 600 mg L−1. The initial CO2 concentration has an effect on how successful CO2 control is. A photocatalytic activity increases when there is a pollutant with sufficient concentration. Photoreduction proficiency of the CO2 at 400 mg L−1 had the highest impact. Additionally, photoreduction didn’t work as well at levels above 400 mg L−1. Impact of stirring speed The agitation rate in a photocatalytic reaction may significant how effectively the photocatalyst functions and how it responds. In reality, photocatalytic activity is higher when the rotation rate is faster. This is because it makes it easier for mass to move, which speeds things up. The reaction mixture containing Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalyst was stirred at speeds of 100, 200, 300, and 400 rpm to observe how the speed of stirring influences the photoreduction and conversion of dissolved CO2 into usable products. As shown in Fig. 11f, the 200 rpm was the optimal mixing speed since it got rid of more than 1727 μmol g−1 h−1 (95% efficiency) of the CO2 pollutants. Effect of the interval between the reaction medium and the visible irradiation source Figure 11g shows that the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 heterojunction photocatalyst’s ability to convert CO2 into solar fuels depends on how far away the light source is from the solution surface. The distances were considered between 5 and 20 cm. According to Fig. 11g, the best distance between the reaction liquid and the visible light was 5 cm. At this distance, CO2 photoreduced the most, with an effectiveness of over 1727 μmol g−1 h−1 (total CO2 conversion efficiency 95%). The more distant the light source is from the solution’s surface, the less light intensity decreases with distance it may seem. This makes it more challenging to utilize photocatalysis to convert carbon dioxide into valuable chemicals. Influence of irradiation light intensity The Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 heterojunction photocatalyst can turn CO2 into usable fuels when it is subjected to varied levels of visible light, from 10 to 90 W. The data in Fig. 11h reveal that the 70 W light exhibited the highest activity on the CO2 pollutant. Photoreduction didn’t function any better on CO2 when there was more than 70 W of light. It might also harm the photocatalyst’s structure, which would make it less effective. Table 4 shows the best results obtained at the end. Evaluating the reusability and selectivity A photocatalyst has to be able to be used more than once and survive a long time in order to perform successfully. Five recovery cycles were made in a row to see the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary heterojunction photocatalyst. Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary heterojunction nanocomposite was put in a centrifuge, washed it with water and ethanol a few times, and thereupon dried it at 70 °C for 24 h to get it back. It can be observed from Fig. 12 that the photocatalytic performance went down by a very small amount after the fifth recovery period. The data demonstrate that the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalyst may be easily recovered and utilized again. The photocatalyst efficiency and selectivity of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 after 5th cycle. The Photocatalytic activity of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary heterojunction photocatalyst was compared to other photocatalysts that have been created in the last few years, which are given in Table 7. The Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 ternary heterojunction nanocomposite is a potent photocatalyst because it is very good at photocatalytic reduction and conversion of carbon dioxide to valuable fuels exposed to visible light (Fig. 13). The most significant products of these processes were methane (CH4) and methanol (CH3OH). These two fuels are significant importance. Better results will be achieved, and there will be more valuable uses in the future by trying new things, like changing the type of perovskites used, the shape and size of the particles used, the types of metallic and non-metallic co-photocatalysts used, modeling tests, and making the conditions as good as they can be (Table 8). Reduction process of CO2 reduction by Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 0D/3D/2D Schottky/Z-scheme ternary heterojunction photocatalyst under visible light irradiation, schematically. Table 8Comparison and evaluation of photocatalytic performance of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 synthesized photocatalysts in recent years.Photocatalyst/DosagePollutant/ConsReaction timeLight sourceConversion efficiency (%)ReferencesTiO2/grapheneCO2300 minVisible lightCH3OH = 155.78 μmol g−1 h−1CHOOH = 167.79 μmol g−1 h−150Bi2S3/CdSCO2360 minVisible lightCH3OH = 590 μmol g−1 h−151CuFe2O4/TiO2CO2480 minVisible lightCH3OH = 651 μmol g−1 h−152TiO2/g-C3N4 (0.1 g/L)CO2 (140 kPa)30 min8W Hg lampCO = 25.35/CH4 = 18.32 μmol g−1 h−146ZnO/g-C3N4 (2.5 g/L)CO2 (0.4 MPa)30 min500W Xe/420 nm(CO, CH4) = 45.6 μmol g−1 h−153NiO/g-C3N4 (0.5 g/L)CO2360 minVisible lightCO = 3.78/CH4 = 1.7 μmol g−1 h−154Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 (1 g/L)CO2 (400 mg L−1)300 minVisible lightTotal CO2 conversion = 1727 μmol g−1 h−1 (total CO2 conversion 95%)CH4 = 863.5 μmol g−1 h−1 (selectivity 50%)CH3OH = 690.8 μmol g−1 h−1 (selectivity 40%)In research Comparison and evaluation of photocatalytic performance of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 synthesized photocatalysts in recent years. CH3OH = 155.78 μmol g−1 h−1 CHOOH = 167.79 μmol g−1 h−1 Total CO2 conversion = 1727 μmol g−1 h−1 (total CO2 conversion 95%) CH4 = 863.5 μmol g−1 h−1 (selectivity 50%) CH3OH = 690.8 μmol g−1 h−1 (selectivity 40%) Conclusion The study was able to clearly show that Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 0D/3D/2D QDs Schottky/Z-scheme ternary nanocomposite is a high-performance heterojunction photocatalyst. According to this research, loading zero valent Ni/Cu metallic co-photocatalyst and g-C3N4 non-metallic co-photocatalyst into the Bi2ZnTiO6 structure makes its efficiency better at CO2 photoreduction into usable fuels. The results revealed that adding the zero valent Ni/Cu nanoparticles to the Bi2ZnTiO6 structure made the Bi2ZnTiO6 binary perovskite’s energy band gap lower. Also, adding a non-metallic co-photocatalyst g-C3N4 to the double perovskite Bi2ZnTiO6 structure can increase the specific surface area for reaction and adsorption of carbon dioxide pollutant on the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalyst. When visible light radiation hits the Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalyst structure, electrons in the valent band of graphitic carbon nitride are first excited and transferred to its conduction band due to its low energy band gap. This produces pairs of holes and electrons (e−/h+) in the g-C3N4 structure. By transferring and consuming electrons from the conduction band of graphite carbon nitride to reduce the carbon dioxide substrate, an electron is quickly transferred from the conduction band of Bi2ZnTiO6 double perovskite to the valent band of g-C3N4. As the conduction band of Bi2ZnTiO6 double perovskite empties, an electron from its valent band replaces the hole accumulation in the conduction band, and the valent band acquires a positive charge. Metallic Ni/Cu nanoparticles act as electron sinks due to Schottky barrier formation, effectively trapping photogenerated electrons and suppressing electron–hole recombination. Nickel/copper nanoparticles have a hole accumulation (positive charge or hole), and this photogenerated holes exhibit strong oxidative potential and may take electrons from solvent molecules such as water. When the correct conditions were reached, the Ni-loaded Bi2ZnTiO6/g-C3N4 photocatalyst could achieved total CO2 conversion efficiency 95% (1727 μmol g−1 h−1 production rate). The Structural stability and selectivity of Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalyst were very high and can be utilized over and over again, even after five cycles of recycling and continued use. The Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalyst outperformed previously reported photocatalysts that had been created in the last several years. The Ni/Cu-loaded Bi2ZnTiO6/g-C3N4 photocatalyst is a highly efficient nanocomposite since it can convert CO2 with the photoconversion rate of 1727 μmol g−1 h−1 (total CO2 conversion efficiency 95%) into valuable solar fuels. Supplementary Information Supplementary Information. Acknowledgements The authors would like to acknowledge the financial support for this work provided by the Iran National Science Foundation (INSF). Author contributions Hossein. Kadkhodayan: Investigation, Visualization, Writing-original draft. Taher. Alizadeh: Supervision, conceptualization, Project administration, review and editing. Funding This research was partially funded by the Iran National Science Foundation (INSF) (Project No. 4030063) to support the purchase of required chemicals for the experiments, chemical analysis, and characterization of experimental data. Data availability The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Information. Data Availability Statement The datasets used and/or analyzed during the current study available from the corresponding author on reasonable request.