Article Information

Mingxue Su^1,2, Zhenguo Guo³, Ning Li^1,2* and Bing Zhua²

¹Hefei Cement Research and Design Institute Corporation Ltd, Hefei, 230022, Anhui, China

²Anhui Key laboratory of Green and Low-carbon Technology in Cement Manufacturing, Hefei, 230022, Anhui, China

³School of Chemistry and Chemical Engineering, Hefei University of Technology, Hefei, 230009, Anhui, China

^*Corresponding Author: Ning Li, Hefei Cement Research and Design Institute Corporation Ltd, Hefei, 230022, Anhui, China

And Anhui Key laboratory of Green and Low-carbon Technology in Cement Manufacturing, Hefei, 230022, Anhui, China.

Received: 01 February 2023; Accepted: 07 February 2024; Published: 26 February 2024.

Citation: Mingxue Su, Zhenguo Guo, Ning Li and Bing Zhu. Enhanced Oxygen Tolerance in CO2 Electroreduction to Formic Acid on SnO2/CN Catalyst through Alkali-Heat Treatment. Journal of Analytical Techniques and Research 6 (2024): 19-27.

Abstract

Keywords

CO2, HCOOH, g-C3N4, SnO2, Alkali-heat treatment.

Article Details

1. Introduction:

When compared to photo-reduction and thermal-reduction, electro-reduction offers a simpler and more environmentally friendly method for producing high-value chemicals and reducing CO₂ emissions, studied by [1]. The compound formic acid (HCOOH), a product of CO₂ electro-reduction, serves as a versatile feedstock for various useful chemicals and a hydrogen fuel storage medium [2]. Economically, formic acid outperforms various competitors such as CO, CH₃OH, C₂H₄, and CH₃CH₂CH₂OH [3]. Several metallic elements, metal oxides, and composites have been used for CO₂ electro-reduction, with tin dioxide (SnO₂) showing a higher selectivity towards HCOOH [4-6]. An endeavor fostered to improve the stimulating efficiency of SnO₂, and the integration with an N-doped carbon (CN) substrate is considered satisfactory [7, 8]. However, [4] implied the nitrogen doping procedure for CN is commonly complex and involves external nitrogen sources. Graphite-like carbon nitride (g-C₃N₄), a nitrogen-rich material, provides a more regulated procedure for preparation [9, 10]. Current research has presented g-C₃N₄ as a catalyst support in electrocatalytic systems [11-13]. The amino groups and pyridinic N species in g-C₃N₄ can anchor acidic CO₂ to benefit the CO₂ reduction reaction (CO₂RR) [14], making it a cost-effective CN carrier without the need for amino group grafting. Most CO₂ electro-reduction researches have been conducted under pure CO₂ conditions, neglecting challenges posed by low CO₂ concentrations in the industry exhaust gas [15]. Creating a localized enhancement of a CO₂ microenvironment surrounding the electrocatalyst becomes crucial [16]. Amino groups, through Lewis acid-base interactions, provide CO₂ adsorption sites and enhance CO₂ electro-reduction kinetics [17-19]. g-C₃N₄, with its inherent amino functional groups, can potentially improve precise surface area and electron transfer capability through alkali and hydrothermal treatment [20], making it an ideal electrocatalyst support for direct industry exhaust gas electro-reduction. In this study, the g-C₃N₄ support was exposed to metallic and alkali-heat ablation. Our primary focus was on determining their impact on electron transfer capabilities and CO₂ adsorption/activation at active Sn sites. In addition, we investigated the electrochemical performance of the SnO₂/g-C₃N₄ catalyst under pure CO₂ gas flow and simulated industry exhaust gas (SIEG) to assess both oxygen sensitivity and electrocatalytic efficacy.

2. Material and Methods

2.1 Materials

The Sinopharm Chemical Reagent Company of China provided us with melamine (CP), potassium hydroxide (KOH, AR), alcohol (AR), and tin (II) chloride. Toray (Japan) supplied the sheet of carbon paper. Nafion solution (5wt%) and Nafion 117 membrane were obtained from DuPont (USA). All substances have been used as received, with no further modifications.

2.2 Catalyst preparation

The formation of g-C3N4 included a thermal polymer of melamine at 550°C, with a rate of heating of 10°C per minute in a covered container for two hours in an air atmosphere. The resulting yellow substance was then crushed into a powder after cooling and given the designation CN.

1 g of CN was immersed in 20 mL of 0.1 M of KOH solution. The product was obtained by centrifugation after three hours of stirring, followed by a deionized (DI) water wash and overnight drying at 75°C. This material was marked as CNOH.

Similar to CNOH, with an additional step of heating the solid at 550°C in a covered beaker in dynamic air for one hour after centrifugation. Following the cooling process, the solution evaporated overnight at 75°C after being rinsed with dehydrated water. The tag for this substance was CNOHHT. A study by [21] found that SnO₂/g-C₃N₄ composite catalysts has been generated using a new method based on prior research. A standard solution utilized 80 milliliters of ethanol to scatter 0.38 grams of SnCl₂ and 0.5 grams of CN (CNOH or CNOHHT). After one hour of stirring, the mixture was reflux for one hour at 100°C. The final product was collected, let to cool to room temperature, purified with DI water, then left to dry at 70°C instantaneously. SnO₂/CN, SnO₂/CNOH, and SnO₂/CNOHHT were the final product identities.

2.3 Identification of materials

Materials were investigated using a Smart Lab X-ray diffractometer made by Smart Labs (in Japan) with a Cu-Kα source to obtain powder XRD patterns. Materials were analyzed for CO₂-TPD using a ChemStar TPx chemisorption analyzer (Quantachrome Instrument, USA). The functional chemicals in each sample were identified using Fourier transform infrared spectroscopy (FTIR) on a ThermoFisher Nicolet IS20 instrument (United States), using KBr disks for analysis. Materials' X-ray photo electron spectra (XPS) were collected using a Thermo ESCALAB 250 (U.S.A) and a monochromate Al Kα X-ray source. The textural characteristics of the materials were investigated using an Autosorb Q instrument (Quanta chrome).

2.4 Preparation of functional electrodes

Initially, 10 milligrams of SnO₂/CN (SnO₂/CNOH or SnO₂/CNOHHT) were submerged in 1 milliliter of alcohol with 100 μl of Nafion solution. The combo was ultrasonically treated for 30 minutes to ensure appropriate dispersion and homogeneous coating. The resulting mixture was then applied to a carbon paper substrate with a spray cannon. The process of pouring on the carbon paper was repeatedly processed before it reached the desired amount of 1 mg/cm². The result was selected as a functioning electrode.

2.5 Electrochemical experimentations

Electrochemical studies were conducted using a CHI-760E electrochemical computer (Shanghai Chenhua Instruments Company, China). The Nafion 117 membrane separated the cathodic and anodic chambers. The functional and standard electrodes (Ag/AgCl) were in the cathodic area, with the counter electrode (Pt sheet) in the anode chamber. The electrolyte used was a 0.5 M KHCO₃ solution. Before those studies, high-purity CO₂ was given for at least 30 minutes. The linear sweep voltammetry (LSV) studies were performed at the scanning speed of 50 mV/s, spanning -2.1 and zero volts (against Ag/AgCl). Electrochemical impedance spectroscopy (EIS) has been done at low rates (0.01 Hz) and at elevated speeds (100 kHz), with an amplitude of 5 mV at an open circuit voltage. Stability testing was conducted using chronoamperometry with an electrolytic setting of -1.5 V vs. Ag/AgCl. Gas products were detected by gas chromatography (GC, Thermo Trace 1310, USA) with flame ionization detector (FID) and thermal conductivity (TCD). Ion chromatography (IC, Metrohm Eco, Switzerland) was used for analyzing liquid products. The FEHCOOH is determined utilizing a formula (1):

In this context, 'n' indicates the proportion of formic acid per mol, 'F' indicates the Faradaic constant (96485 C/mol), and 'Q' represents the galvanic charge in C.

The FE of CO and H₂ were determined using the formula (2):

The indication "v%" shows the CO concentration or H₂ in the measured gas. 'G' denotes the CO₂ rate of flow (10 ml/min), 'R' the gas constant (8.314 J/mol/K), 'T' the temperature (298 K), and 'I' the constant current.

3. Results and discussion

The crystallographic phases found in each sample were identified by the XRD spectrum. Figure1 shows three different peak angles at 26.6°, 33.8°, and 51.7°, signifying (110), (101), and (211) facets of SnO₂, respectively studied by [22]. The modest but clear spectra at 27.7° are characterized to (002) interlayer showing CN's graphitic-like composition. Significantly, the highest intensities at 27.7° for SnO₂/CNOH and SnO₂/CNOHHT were lower and wider than those for SnO₂/CN. [23] shows a decrease in the in-plane compositional loading pattern of CN following alkali and alkali-heat therapies.

Figure 1: XRD patterns of all the samples

The textural attributes of every sample are summarized in Table 1. The aggregate volume of pores remained constant throughout all samples. SnO₂/CN had a BET surface area of 9.98 m²/g, which was much lower than 11.18 m²/g of SnO₂/CNOH but higher than 9.66 m²/g of SnO₂/CNOHHT. The inclusion of alkali can lead to deterioration of the C layer, raising the specific region of C-containing compounds. Alkali-heat ablation can produce structural collapse in C-based materials, leading to a decline in an exact surface region.

Table 1: Textural characteristics of each material

As displayed in Figure 2, the FTIR analysis displays the molecular framework of all materials. The area at 811 cm-1 corresponds to the triazine unit. The peaks at 1100 cm-1 and 1600 cm-1 are due to the typical resonance of C-N and C≡N patterns. The spectra at 3179 cm-1 and 3269 cm-1 represent the vibratory stretching of -NHx. Each sample contains such distinct peaks, indicating the preservation of CN's distinctive compositional component. SnO₂/CNOHHT exhibits unique bands at 2175 cm-1, similar to metal-doped CN materials [20]. This indicates the potential for potassium to be attached to CN through heat treatment.

Figure 2: FTIR spectrums of all the samples

All materials under investigation were thoroughly analyzed in terms of their compositions and chemical states using profiles obtained from X-ray photoelectron spectroscopy (XPS). Tin 3d3/2 and Tin 3d5/2 have been recognized as two distinct bands in Figure3 at about 487 and 496 electron volts, respectively. This indicates that Sn⁴⁺ of SnO₂ are present across each of materials. The O 1s spectra of O^2- and oxygen absorbed are shown in Figure4 at exactly 530 and 532 electron volts, respectively. Three peak levels, located at approximately 398 eV, 399 eV, and 401 eV in Figure5, are indicative of amino chains, tertiary N groups, and sp2 polarized N atoms in the CN substrate [24]. All species of nitrogen are provided in Table 2. The amino groups' peak region increased to 7% for SnO₂/CNOHHT from 3.7% for SnO₂/CN, showing that alkali and alkali-heat interventions may significantly raise the amount of surface amino bonds in CN. The rise in surface amino groupings improves the catalyst's alkalinity and accelerates its ability to absorb CO₂. Figure6 illustrates that the O 1s peak of SnO₂/CNOHHT changed to low bonding energy, showing that SnO₂/CNOHHT has more oxygen deficit sites. Moreover, the N 1s signal and Sn 3d peak in SnO₂/CNOHHT changed with varying binding energies, showing that electrons might be carried from N to Sn. As a result, the highly electron-rich centers of Sn species in SnO₂/CNOHHT facilitate CO₂ adsorption and activation. Furthermore, the electrostatic field generated speeds up charge transfer [13]. These outcomes show that heating the CN substrate with alkali can increase CO₂ electrically reduced through chemical bonds between its active sites and transporters.

Figure 3: High-resolution Sn 3d XPS spectra of (a) SnO₂/CN, (b) SnO₂/CNOH and (c) SnO₂/CNOHHT

Figure 4: High-resolution O 1s XPS spectra of (a) SnO₂/CN, (b) SnO₂/CNOH and (c) SnO₂/CNOHHT

Figure 5: High-resolution N 1s XPS spectrum of (a) SnO₂/CN, (b) SnO₂/CNOH and (c) SnO₂/CNOHHT

Figure 6: XPS spectra of (a) SnO₂/CN, (b) SnO₂/CNOH and (c) SnO₂/CNOHHT

Table 2: The percentage of all the N variety in each sample

Moreover, CO₂-TPD was used to evaluate the CO₂-adsorption and activating properties of catalysts, as shown in Figure 7. SnO₂/CN and SnO₂/CNOH showed three equivalent CO₂ desorption levels at 95°C, 170°C, and 650°C. Conversely, SnO₂/CNOHHT showed distinct CO₂ desorption levels at 150°C, 470°C, and 620°C. Higher desorption rates often suggest more chemical bonds between the catalyst and CO₂. As a result, SnO₂/CNOHHT exceeded SnO₂/CN and SnO₂/CNOH in terms of both absorption and activity of CO₂. Table 3 highlights the CO₂ adsorption levels of each sample, with SnO₂/CNOHHT having an optimal ability to adsorb at 48.8 mmol/g. In theory, the rate of adsorption should be strongly correlated with the material's particular size and volume of pores. However, Table 1 shows that SnO₂/CNOHHT has the lowest particular surface area and pore size of the samples. The major adsorption ability of SnO₂/CNOHHT is modulated by accessible groups of amino and electron-rich Sn species, consistent with XPS studies.

Figure 7: CO₂-TPD profiles of all samples

Table 3: Rate of CO₂ adsorption in each sample

The electrocatalytic efficiency had originally been assessed using the LSV experiment in a N₂ and CO₂-saturated 0.5 M KHCO₃ solution. Compared to the N₂-saturated solution, Figure8a reveals increased current levels for all samples in the CO₂-saturated solution across the investigated potential range, indicating that CO₂RR was more favorable than H₂ evolution reaction (HER). Notably, SnO₂/CNOHHT had the highest level of current (80 mA/cm²) at -1.87 V (vs. Ag/AgCl). The potential voltages of SnO₂/CN and SnO₂/CNOH have to exceed 2.1 and 1.9 V (vs. Ag/AgCl), respectively, to attain a comparable density of current. Additionally, SnO₂/CNOHHT also showed the lowest onset potential. These results demonstrated the alkali-heat treatment played a positive role in lowering onset potential and increasing current density, and eCO₂RR was more satisfactory to occur on SnO₂/CNOHHT, which was also consistent with the XPS and CO₂-TPD analysis.

Electroreduction processes were carried out utilizing all samples for two hours in a CO₂-saturated 0.5 M KHCO₃ solution at a constant potential (-1.5 V vs. Ag/AgCl) to fully assess catalytic performance. In Figure 8b, after electrolysis, the foremost products for all catalysts were CO, H₂, and HCOOH. Furthermore, compared to 62.7% and 81.8% for SnO₂/CN and SnO₂/CNOH, SnO₂/CNOHHT had the highest FE for formic acid among the three catalysts, reaching 90.5%. Notably, the faradaic efficiency of hydrogen (FEH₂) on SnO₂/CNOHHT was much lower than that of the other two catalysts. The results indicate that after alkali-heat therapy, the displayed surface amino bonds of the CN substrate and highly electron-rich centers of Sn species could inhibit the hydrogen evolution reaction (HER) and improve HCOOH selectivity via enhanced CO₂ adsorption and stimulation features. The FE of CO did not exhibit apparent changes among all the samples, indicating the absence of a competitive relationship between the evolution of carbon monoxide (CO) and the intended production of HCOOH.

Figure 8: (a) LSV curves of the electrodes in N₂- and CO₂-saturated 0.5 M KHCO₃ electrolytes at a scan rate of 50 mV/s and (b) FE of HCOOH, CO, and H₂ at the potential -1.5 V vs. Ag/AgCl after 2 h of eCO₂RR

To evaluate the reaction kinetics for eCO₂RR, Tafel slopes of all samples were studied and shown in Figure 9a. SnO₂/CNOHHT possessed the lowest Tafel slope compared to the SnO₂/CN and SnO₂/CNOH, suggesting the enhancement of reaction kinetics activity for eCO₂RR to HCOOH. The low Tafel slope is useful to reduce CO₂ swiftly with enhancing overpotential. Electrochemical impedance spectroscopy (EIS) was directed at open-circuit voltage to investigate the charge transfer process. As shown in Figure 9b, SnO₂/CNOHHT exhibited the smallest charge transfer resistance (6.2 ohm) compared to the SnO₂/CN (6.8 ohm) and SnO₂/CNOH (8.8 ohm), suggesting its fastest electron transfer rate and reaction kinetics in eCO₂RR [25]. According to the XPS results, alkali-heat treatment effectively enhanced the chemical bonding between SnO₂ and CN, leading to the improvement of overall electronic conductivity, thereby reducing impedance. This ensures improved charge transfer between the active center and CO₂.

Figure 9: (a) Tafel plots and (b) EIS Nyquist plots of all samples in CO₂-saturated 0.5 M KHCO₃ solution

Electrochemical stability, another crucial property of the catalyst, was assessed through an electrolytic experiment conducted in a CO₂-saturated 0.5 M KHCO₃ solution at -1.5 V (vs. Ag/AgCl). The current-time relationship for SnO₂/CNOHHT, depicted in Figure 10, revealed a consistent current value of 23.5 mA/cm² for 8 hours. This result underscores the electrochemical stability of each catalyst.

Figure 10: Stability test of all samples at the potential -1.5 V vs. Ag/AgCl

To further understand the impact of alkali-heat therapy, all catalysts' electroreduction efficiency was evaluated in an SIEG scenario. Figure 11 shows LSV curves for electrodes in SIEG at a scanning pace of 50 mV/s. As seen in Figure 11, when contrasted to pure CO₂ flow, the density of all catalysts decreased with SIEG. SnO₂/CNOHHT showed the smallest decline in current density. Figure 8b shows that the FE of HCOOH for SnO₂/CNOHHT was 69.4%, higher than SnO₂/CNOH (49.3%) and SnO₂/CN (15.4%). Despite a decline in FE of HCOOH for all samples, SnO₂/CNOHHT had the smallest percentage decrease (23.3%) when compared to SnO₂/CN (75.4%) and SnO₂/CNOH (39.7%).

Figure 11: LSV curves of the electrodes in SFG at a scan rate of 50 mV/s

Multi-component adsorption experiment was conducted to assess the effect of alkali and alkali-heat pretreatment on catalyst CO₂ adsorption capability in SIEG. As illustrated in Figure 12, the CO₂/O₂ separation coefficients for SnO₂/CN, SnO₂/CNOH, and SnO₂/CNOHHT were 0.98, 0.95, and 0.90, respectively. A significant divergence from the separation coefficient of one indicates more segregation of the mixed gas particles. This reveals that, under SIEG conditions, the catalyst with alkali-heat treatment showed efficient adsorption of CO₂ from SIEG.

Figure 12: The multi-component competitive adsorption curves of SnO₂/CN (left), SnO₂/CNOH (middle), and SnO₂/CNOHHT (right). Gas condition: 15%CO₂, 8%O₂ and 77%N₂

Conclusion

Both untreated and alkali treated catalysts were outperformed by the alkali-heat treated catalyst, which had a superior faradaic efficiency for formic acid at 90.5% at a -1.5 V voltage (vs. Ag/AgCl). Moreover, the alkali-heat treated catalyst demonstrated electrical stability for up to 8 hours. During electro-reduction of simulated industry exhaust gas, this alkali-heat treated catalyst outperformed the other two catalysts in the present of O₂. The increasing fraction of amino-nitrogen in the carbon nitride carrier confirmed the efficiency of alkali-heat pretreatment in terms of the combined catalyst alkalinity and adsorption of CO₂. Furthermore, the treatment strengthened metal-support contacts, increasing the flow of electrons from CN's nitrogen (N) to Sn and creating Sn species with very electron-rich nuclei. It also improved the association between the Sn active site and CO₂, hence activating and reducing CO₂. The raised alkalinity and interaction of the Sn active site with CO₂ enriched from SIEG, improved the oxygen-induced tolerance and electrolytic activity of the composite catalyst during SIEG electro-reduction.

Acknowledgments

The Chinese Ministry of Sciences and Technology of the PRC provided a grant for this research under its National Key Development and Research Program (No. 2020YFC1908704).

Data and code availability

Not Applicable.

Supplementary information

No supplementary materials available.

Ethical approval

Not Applicable.

Conflict of Interests

The authors do not have any financial or other conflicts of interest to declare.

References

1. PP Yang, MR Gao. Enrichment of reactants and intermediates for electrocatalytic CO₂ Chem Soc Rev 52 (2023): 4343-4380.
2. D Ewis, M Arsalan, M Khaled, et al. Electrochemical reduction of CO₂ into formate/formic acid: A review of cell design and operation. Sep Purif Technol 316 (2023): 123811.
3. A Somoza-Tornos, OJ Guerra, AM Crow, et al. Process modeling, techno-economic assessment, and life cycle assessment of the electrochemical reduction of CO₂: a review. Iscience 24 (2021): 102813.
4. H Cheng, S Liu, JD Zhang, et al. Surface Nitrogen-Injection Engineering for High Formation Rate of CO₂ Reduction to Formate. Nano Lett 20 (2020): 6097-6103.
5. S Zhang, P Kang, TJ Meyer. Nanostructured Tin Catalysts for Selective Electrochemical Reduction of Carbon Dioxide to Formate. J. Am. Chem. Soc 136 (2014): 1734-1737.
6. XS Wang, WH Wang, JQ Zhang, et al. Carbon sustained SnO₂-Bi₂O₃ hollow nanofibers as Janus catalyst for high-efficiency CO₂ Chem Eng J 426 (2021): 131867.
7. D Sun, W Li, RT Guo, et al. Preparation of N-doped biomass C@SnO₂ composites and its electrochemical performance. Diam Relat Mater 120 (2021): 108674.
8. H Wang, Z Liu, Q Liang, et al. A facile method for preparation of doped-N carbon material based on sisal and application for lead carbon battery. J. Clean. Prod 197 (2018a): 332-338.
9. F Li, Y Huang, C Gao, et al. The enhanced photo-catalytic CO₂ reduction performance of g-C₃N₄ with high selectivity by coupling CoNiSx. Mater Res Bull 144 (2021): 111488.
10. W Yang, L Jia, P Wu, et al. Effect of thermal program on structure–activity relationship of g-C₃N₄ prepared by urea pyrolysis and its application for controllable production of g-C₃N₄. J Solid State Chem 304 (2021): 122545.
11. C Hu, MT Liu, A Sakai, et al. Synergistic effect of Cu and Ru decoration on g-C₃N₄ for electrocatalytic CO₂ J Ind Eng Chem 115 (2022): 329-338.
12. BB Mulik, BD Bankar, AV Munde, et al. Electrocatalytic and catalytic CO₂ hydrogenation on ZnO/g-C₃N₄ hybrid nanoelectrodes. Appl Surf Sci 538 (2021): 148120.
13. J Tian, M Wang, M Shen, et al. Highly Efficient and Selective CO(2) Electro-Reduction to HCOOH on Sn Particle-Decorated Polymeric Carbon Nitride. ChemSusChem 13 (2020): 6442-6448.
14. KY Tao, K Yuan, W Yang, et al. A template co-pyrolysis strategy towards the increase of amino/imino content within g-C₃N₄ for efficient CO₂ Chem Eng J 455 (2023): 140630.
15. Y Xu, JP Edwards, J Zhong, et al. Oxygen-tolerant electroproduction of C2 products from simulated flue gas. Energ Environ Sci 13 (2020): 554-561.
16. Y Takeda, S Mizuno, R Iwata, et al. Gas-fed liquid-covered electrodes used for electrochemical reduction of dilute CO₂ in a flue gas. J CO₂ Util 71 (2023): 102472.
17. N Meng, C Liu, Y Liu, et al. Efficient Electrosynthesis of Syngas with Tunable CO/H₂ Ratios over ZnxCd1-xS-Amine Inorganic-Organic Hybrids. Angew Chem Int Edit 58 (2019): 18908-18912.
18. PS Li, X Lu, ZS Wu, et al. Acid-Base Interaction Enhancing Oxygen Tolerance in Electrocatalytic Carbon Dioxide Reduction. Angew Chem Int Edit 59 (2020): 10918-10923.
19. X Chen, J Chen, NM Alghoraibi, et al. Electrochemical CO2-to-ethylene conversion on polyamine-incorporated Cu electrodes. Nat Catal 4 (2021): 20-27.
20. Z Sun, JMTA Fischer, Q Li, et al. Enhanced CO₂ photocatalytic reduction on alkali-decorated graphitic carbon nitride. Appl Catal B-Environ 216 (2017): 146-155.
21. H Zhang, T Ouyang, J Li, et al. Dual 2D CuSe/g-C₃N₄ heterostructure for boosting electrocatalytic reduction of CO₂. Electrochim Acta 390 (2021a): 138766.
22. HJ Wang, GW Jiang, XJ Tan, et al. Simple preparation of SnO₂/C nanocomposites for lithium ion battery anode. Inorg Chem Commun 95 (2018b): 67-72.
23. X Li, X Sun, L Zhang, et al. Efficient photocatalytic fixation of N₂ by KOH-treated g-C₃N₄. J Mater Chem A 6 (2018): 3005-3011.
24. Y Shang, Y Ding, P Zhang, et al. Pyrrolic N or pyridinic N: The active center of N-doped carbon for CO₂ Chinese J Catal 43 (2022): 2405-2413.
25. BH Zhang, S Chen, BR Wulan, et al. Surface modification of SnO₂ nanosheets via ultrathin N-doped carbon layers for improving CO₂ electrocatalytic reduction. Chem Eng J 421 (2021b): 130003.