CHAPTER 1: Recent Advances in the Controlled Design of One-dimensional Carbon Nitrides for Thermal CO Oxidation Reaction

The excessive utilization of nonrenewable fossils has caused the emission of huge amounts of hazardous gasses (i.e., CO, CO₂, and CH₄) that have harmful effects on humans (i.e., headache, nausea, coma, fainting, oxidative stress/endothelial dysfunction, and kidney calcification) and the environment (i.e., climate change, global warming, smog/acid rain, and melting ice), so it is critical to reduce these gas emissions. The new generations of green, renewable, and zero-emission energy conversion technologies (i.e., solar cells, water splitting, oxygen reduction, alcohol oxidation fuel cells) driven by metal-based catalysts are among the most promising ways to reduce CO₂/CO emission.^1–15  Another effective way is gas conversion reactions, including CO₂ reduction reaction (CO₂RR) and CO oxidation reaction (CO OR) on graphitic carbon nitrides (gCNs) that are of great importance in industrial, environmental, and fundamental applications.^16,17  The materials used are polymeric materials comprising C, N, and H, heterocycles of heptazine or triazine rings linked via carbon sp²-bonded nitrogen atoms (N(C)₃ units), or –NH– groups. In 1834, J. Liebig¹⁸  prepared a melon for the first time, which is a polymer that contains tri-s-triazines via secondary nitrogen; however, the breakthrough in the fabrication and utilization of gCNs was just attained in the last two decades.^{16,17,19–26}  gCNs have electron-rich properties, basic surface functionalities, and H-bonding motifs in addition to great thermal stability, chemical stability in different solvents, and mechanical properties, active and sites of gCNs that are significant merits in multidisciplinary catalytic, energy, and environmental applications.^{16,17,19–24}  Also, gCNs are semiconductors with a band gap energy of 2.7 eV and great visible light absorption < 450 nm that enhances the photocatalytic performance towards water splitting, dye degradation, and solar cells.^17,27,28  Gas capture or conversion reactions, including CO₂RR and CO ORs on gCNs are of great importance in industrial, environmental, and fundamental applications due to the hazardous effects of the use of fossil fuels on humans and the environment.^29–34 

The unique composition of gCNs contains an electronegative nitrogen atom bonded to positive carbon, resulting in electron-deficiency on the carbon atom in the form of carbon⁺–nitrogen⁻ that is favored for the electrophilic and nucleophilic attack alongside promoting the charge transfer and adsorption of CO molecules.^{21–24,35–38}  One-dimensional gCN nanostructures (i.e., nanotubes, nanorods, nanofibers, nanowires, and needles) possess various advantages over other shapes (i.e., zero-dimensional and two-dimensional). One-dimensional nanostructures have a greater surface area, multiple adsorption sites, quicker charge migration, short electron-diffusion, and less-tendency for aggregation.^24,39–43  There are various methods for the controlled fabrication of 1D gCN nanostructures such as template-based, solvothermal, self-assembly, and thermal condensation. Notably, CO oxidation on 1D gCN nanostructures is not emphasized enough relative to other applications.^34,44,45  There are some recently published reviews about the fabrication of 1D gCN nanostructures for various catalytic applications, but not for CO oxidation.^34,46–50 

This chapter emphasizes the controlled synthesis of 1D gCN nanostructures for CO oxidation. This includes the fabrication of nanotubes, nanorods, nanofibers, and needles and their related mechanisms. This is in addition to the challenges and outlooks of gCNs for gas conversion reactions.

Template-based methods (i.e., hard, soft, and hybrid templates) are the most effective for the rational synthesis of uniform 1D gCN nanostructures in high yield, with tunable shape composition driven by proper selection of a template (i.e., foam, tubes, and rods).^51–53  The reaction parameters (i.e., template type, reactant concentrations, pH, solvent, and temperature) are not only the most critical factor determining the uniformity and morphology, but also tuning the physicochemical (i.e., stability, surface area, and mechanical properties), catalytic, photocatalytic properties of the final 1D gCN product.^28,51–53 

The hard templating method is widely used for 1D gCNs, due to its simplicity and low level of precautionary measures compared with the soft-template method. The process is based on the initial filling of a rigid template (i.e., ZnO, SiO₂, anodic aluminum oxide membranes (Al₂O₃), calcium carbonate, or any other inorganic material) with a carbon–nitrogen precursor, followed by pyrolysis to allow condensation and polymerization of the precursor inside the pore or channel of the template and finally erosion of the template by acidic/alkaline etchants (Figure 1.1A).^7,54–56  However, AAO and SiO₂ are the most common templates. They are widely used in the fabrication of 1D gCNs due to their high thermal stability, ease of handling, availability, low cost, uniformity, and stability against side reactions with the gCN monomers during the condensation process. Meanwhile, the fabrication of 1D gCN nanostructures using hard templating is not emphasized enough compared to other 3D or 2D structures.^54–57 

Nickel-modified AAO enabled the preparation of uniform small gCN nanorods (10–20 nm) with unique catalytic performance and includes the initial deposition of nickel on AAO using an impregnation approach to form NiO/Al₂O₃ filled with ethylenediamine and tetrachloride. This was followed by annealing at 600 °C in Ar to allow the conversion process as well as reducing NiO to Ni nanoparticles that nearly distributed over three to four layers of the gCN nanorods.⁵⁸  This unique composition has great electronic effect, with the electron-donating merits of Ni nanoparticles, the rich electron density of the gCN nanorods enhanced the catalytic hydrogenation of p-nitrobenzoic acid at various pH with an impressive durability.⁵⁸  Silica ‘SBA-15’ was used as a template for the preparation of mesoporous gCN rod-like shaped gCN nanostructures (but not 1D shapes) with a high surface area up to 517 m² g⁻¹, based on the acidification of SBA-15 by hydrochloric acid at 80 °C. It was then dried and mixed with cyanamide under sonication, vacuumed, mixed with water, then dried and annealed at 550°, finally eroding SBA-15 by NH₄HF₂.⁵⁹  Likewise, a mesoporous rod-like shaped g-C₃N₄ nanocomposite decorated with Ag nanoparticles with a high surface area of 563.4 m² g⁻¹ was prepared using SBA-15 as a template, based on the dissolving of dicyandiamide in water solution containing AgNO₃. It was mixed with g SBA-15 while stirring at 80 °C until the evaporation of water, and the resultant powder was annealed at 550 °C under nitrogen.⁶⁰  Notably, the SiO₂-based template is usually used to prepare 2D or 3D gCNs, with hardly any reports on gCNs with 1D shapes.^55,57,61  However, SiO₂-based templates are feasible for different modifications (i.e., cross-linking, doping, and co-polymerization) and easily combined with other methods for tailoring shapes of gCN nanostructures.^55,61–64  Porous 1D gCN nanotubes were prepared via the polymerization of ethylenediamine and carbon tetrachloride in the presence of an AAO template at 600 °C under nitrogen, then etching of AAO by NaOH solution.⁶⁵  The as-obtained gCN nanotubes were used as a substrate for supporting the deposition of Pt nanoparticles for the catalytic hydrogenation of cyclohexene.⁶⁵  Confined 1D gCN nanorods, free from defects or free NH₂ group, were prepared by the polymerization of cyanamide using an AAO template 600 °C under nitrogen and then by etching AAO by hydrochloric acid.⁶⁶  Using AAO as a template enhanced the crystallinity, reduced the highest occupied molecular orbital (HOMO) position, and enhanced the photocatalytic properties of the thus obtained gCN nanorods for a full water-splitting reaction (i.e., hydrogen evolution and oxygen evolution).⁶⁶  This indicates the possibility of modulating the optical and photo properties of gCN materials using an AAO membrane with various dimensions in the presence of different carbon–nitrogen monomers. Despite great progress in the hard template for the synthesis of 1D gCNs, using hazardous etchants, multiple steps, and unavoidable defects remain great challenges. Additionally, filling the template with carbon–nitrogen monomers is still difficult and requires high precautionary measures.

Unlike the hard template, the soft template is based on the self-assembly of a nitrogen-carbon monomer with surfactants or ionic liquids or block polymers including ionic co-polymers (i.e., cetyltrimethylammonium bromide (CTAB) and benzylhexadecyldimethylammonium chloride (BDAC)), non-ionic co-polymers (i.e., polyvinylpyrrolidone (PVP), Pluronic F127, Plurong P123, Brig-58), or their hybrids to build block shapes (i.e., rods, tubes, or 3D micelles) enabling the formation of gCN nanoarchitectonics in different shapes.⁵¹  The built blocks are then decomposed during the condensation process due to their lower thermal stability and subsequently remove the need for an additional etching step as in the hard templating method.⁵¹ 

The soft template's concentration should be above the crucial micelle's concentration to allow the self-assembly process; meanwhile, the solvent type and pH are critical factors determining the final shape and properties of the gCNs.⁵¹  Wang et al. used various soft templates (i.e., Pluronic P123 and Triton X-100, Pluronic F12Brij-x) for the rational synthesis of a mesoporous gCN nanostructure using dicyandiamide as a monomer while carrying out the condensation at 550 °C, but not for 1D gCNs.⁶⁷  Both Triton and Pluronic F127 were preferred for the production of gCNs with a high surface area. This study opens the way for the soft-templating approach in the preparation of 1D gCN nanostructures.

For instance, Pluronic F127 was used as a template for the synthesis of g-C₃N₄ nanotubes via the co-polymerization of urea and thiourea at two annealing steps at 350 °C and then at 600 °C (Figure 1.1B).⁶⁸  Following that, g-C₃N₄ nanotubes were used as a substrate to support PtNi nanoparticles' growth by the chemical reduction method using NaBH₄. Although there was a low Brunauer–Emmett–Teller (BET) surface area of those formed g-C₃N₄ nanotubes (20 m² g⁻¹), the photocatalytic performance for hydrogen evolution reaction was outstanding. Likewise, g-C₃N₄ nanotubes co-doped with Na and S were prepared by the co-polymerization of urea and thiourea in the presence of Pluronic F127 as a template and NaHCO₃ as Na source and then annealed.⁶⁹  The surface area of the doped g-C₃N₄ nanotubes was significantly higher than the bulk g-C₃N₄; also, the surface area depended on the Na content.⁶⁹  This method is flexible and can allow mono doping with either Na or S via selecting the initial type or concentration of the used reactants. Although there has been great progress in the soft template, it is not emphasized enough to synthesize 1D gCNs. The template's self-assembly mechanism with the carbon-monomer is still ambiguous, and it is not easy to decipher.⁵¹  Meanwhile, the precautions, including concentrations of the template and reaction conditions, remain unreproducible and not feasible for large-scale applications.⁵¹ 

One-dimensional gCN nanotubes are of great importance in various catalytic applications due to their high surface area, accessible sites, low density, quick charge mobility, and abundant adsorption sites for the reactant molecules, so massive effort has been dedicated to their fabrication.⁵⁵  An early attempt at the rational design of g-C₃N₄ nanotubes was made by Guo et al. They prepared g-C₃N₄ nanotubes with an average size of 50–100 nm and wall thickness of 20–50 nm, synthesized using the facile benzene-thermal process (Figure 1.2A and B).²⁵  This included mixing 1,3,5-trichlorotriazine and sodium azide in benzene, then autoclaving at 220 °C for 15 h; after cooling, the obtained powder was washed with benzene and water and then dried under vacuum.²⁵  g-C₃N₄ nanotubes were prepared by the direct annealing of a melamine-cyanuric acid complex at 400 °C under nitrogen to form nanorods, then calcination was performed at 550 °C under nitrogen to form nanotubes.⁷¹  The same mechanism was found in the preparation of g-C₃N₄ microtubes by the polymerization of melamine−cyanuric acid under autoclaving at 180 °C to form complex melamine−cyanuric acid (MCA) microrods with an average length of 20–50 µm and width of 1–2 µm average diameter, which converted to nanotubes with a wall thickness of 50 nm after annealing under nitrogen at 550 °C (Figure 1.2C).⁷²  The mechanism was attributed to the initial formation of a MCA hexagonal plate shape that consequently converted to hexagonal microrods upon hydrothermal treatment based on the longitudinal growth direction with the assistance of water, meanwhile during annealing under nitrogen, this allowed the quick sublimation for the triazine molecules in the internal part or microrod to grow outward to form g-C₃N₄ tube morphology.⁷²  The surface area of the g-C₃N₄ microtubes was 2.2 times higher than that of bulk g-C₃N₄ along with a superior photocatalytic activity by 3.1-fold.⁷²  Porous gCN nanotubes decorated with Rh nanoparticles were prepared by the initial preparation of gCN nanosheets by the direct pyrolysis of melamine at 550 °C in air, and then gCN nanotubes were mixed with ethanol/water and RhCl3 followed by the addition of NaBH₄ as a reducing agent to facilitate the reduction of RhCl₃ with the assistance of sonication.⁷³  The use of sonication and ethanol/water solvents and the high reduction power of NaBH₄ under sonication may be the main reason for the formation of porous nanotubes.⁷³ 

Millimeter-long g-C₃N₄ nanotubes with an exterior size of 300–500 nm were prepared in a simple one-pot method, including protonation of melamine in ethylene glycol solution by nitric acid, then thermal sintering at 350 °C (Figure 1.2D).²⁴  The formation mechanism includes the protonation of the abundant active amino groups of melamine by nitric acid.²⁴  The protonated melamine was polymerized into an s-triazine ring-based nanofiber, and then the ribbons rolled up to the nanotubes upon heating to reduce the surface free energy of the intermediate fiber ribbons.²⁴  The presence of nitric acid is necessary to direct nanotube formation while using NaNO₃ and HCl drove the formation of the g-C₃N₄ nanofiber. Additionally, using other acids like sulfuric or phosphoric acid did not form 1D nanostructures.²⁴  The as-obtained g-C₃N₄ nanotubes and nanofibers revealed a great luminescent emission, quick photoresponse, and high photoconductivity.²⁴  Notably, most of the reported 1D gCN nanotubes were undoped, and it is known that doping can enhance the catalytic properties and improve the physicochemical properties of gCN materials. In this context, hierarchical porous g-C₃N₄ nanotubes doped with oxygen (Og-C₃N₄NTs) composed of multi-walled nanotubes with a size of 20–30 nm were synthesized using thermal oxidation exfoliation and curling‐condensation of bulk g-C₃N₄.⁷⁴  In particular, bulk g-C₃N₄ was obtained by direct condensation of melamine at 550 °C that consequently annealed at 600 °C for 4 h under both nitrogen and air.⁷⁴  The formation mechanism is based on the exfoliation of g-C₃N₄ nanosheets to individual sheets driven by breaking the hydrogen bonds that turn into nanotubes under the driving force of airflow.⁷⁴  With airflow assistance, g-C₃N₄ nanotubes condense and self‐assemble into a hierarchical porous nanotube on the tubular furnace.⁷⁴  The BET surface area of Og-C₃N₄NTs (36 m² g⁻¹) was superior to bulk g-C₃N₄NTs (9 m² g⁻¹) by four times in addition to the higher photocatalytic activity for CO₂ reduction, due to the excellent CO₂ adsorption, light absorption, delay electron–hole recombination, and quick charge mobility. Likewise, phosphorus-doped g-C₃N₄ nanotubes with mesoporous < 25 nm were produced by the direct one-step annealing of melamine and sodium hypophosphite monohydrate at 550 °C after cooling. The obtained powder was ground, washed with water, and dried at 70 °C.⁷⁵  The phosphine gas generated from the thermal decomposition of NaH₂PO₂·H₂O induces the formation of P-g-C₃N₄ nanotubes from g-C₃N₄ nanosheets.⁷⁵  The BET surface area of phosphorus-doped g-C₃N₄ nanotubes (13.38 m² g⁻¹) was 4.4 times higher than undoped g-C₃N₄ (3.02 m² g⁻¹) in addition to higher optical and photo properties along with higher CO₂ absorption, lower zeta potential, and charge separation.⁷⁵ 

Porous C₃N₄ nanotubes with controllable nitrogen (N-g-C₃N₄) and tunable nitrogen vacancies were prepared by the thermal etching method, including the initial preparation of C₃N₄ via autoclaving of cyanuric chloride and melamine in acetonitrile at 180 °C, that were subsequently pyrolyzed at 550 °C under nitrogen to form C₃N₄ nanotubes.⁷⁶  This was followed by annealing under air at 550 °C for 15, 30, 45, and 60 min to integrate nitrogen-deficiency into the C₃N₄ nanotubes.⁷⁶  Increasing the annealing time from 15 to 60 min not only increases the nitrogen vacancies but also increases the BET surface area from 23 m² g⁻¹ to 207 m² g⁻¹, in addition to maximizing the utilization of visible light, providing more active sites for capturing photoexcited electrons that preclude the electron–hole recombination, and increasing the CO₂ adsorption.⁷⁶ 

Likewise, nitrogen doped (g-C₃N₄) with adjustable nitrogen content was prepared by the supramolecular self-assembly strategy, including mixing hydroxylammonium chloride and melamine in water, then autoclaving at 120 °C to form a supramolecular intermediate that dried at 60 °C after washing, and then annealing at 520 °C under ammonia gas with different flow rates of 50, 100, 200, and 400 mL min⁻¹.⁷⁷  The nitrogen content of the N-g-C₃N₄ nanotubes increased with increasing flow rate but almost collapsed the nanotube shape at the highest rate, 400 mL min⁻¹, due to the etching properties of ammonia.⁷⁷  Interestingly, the g-N-g-C₃N₄ nanotubes' surface area reached 135.8 m² g⁻¹, which was 15.7 times higher than that of bulk g-C₃N₄.⁷⁷  Meanwhile, annealing under air, argon, and nitrogen atmospheres instead of ammonia also results in g-C₃N₄ nanotubes but with less porosity.⁷⁷  Owing to the porous nanotube shape, Lewis basicity, high surface area, and rich nitrogen content of N-g-C₃N₄ nanotubes, its photocatalytic properties, and CO₂ adsorption/desorption increased significantly.⁷⁷  Long aligned g-C₃N₄ nanotubes with lengths over 4 µm with the open end of 60 nm, were prepared by dispersion of melamine in water at 80 °C, followed by addition of salicylic acid with different amounts (0.92, 0.69, 0.55, 0.46, and 0.39 g). After drying, the obtained powder was subjected to a low annealing step at 550 °C under nitrogen and then under air to allow the thermal-oxidative etching process. Increasing the amount of salicylic acid leads to increasing pore volume due to its ability to act as a pore modulator; however, it decreases the surface area. The thermal etching process and the decomposition of salicylic acid during annealing possibly allow the rolling mechanism to form nanotubes.⁷⁸  The photocatalytic properties, charge carrier mobility, and light absorption were enhanced using a higher amount of salicylic acid. This study opens new avenues for utilizing organic acids as a pore generator for C₃N₄ 1D shapes.

Similarly, porous g-C₃N₄ microtubes were prepared by a liquid–liquid interfacial self-assembly method, including mixing dicyandiamide in a mixture of solvents of water and isopropanol under stirring, then autoclaving at 180 °C to form the initial supramolecular precursors that were annealed a 550 °C in air after being purified and dried⁷⁰  (Figure 1.2E). The surface area of 82.84 m² g⁻¹ was 13.3 times higher than that of bulk g-C₃N₄ prepared by direct annealing of dicyandiamide along with superior photocatalytic activities and higher redox-active sites.⁷⁰  Intriguingly, using double solvent water with isopropanol is a prerequisite to allow nanotube formation driven by the formation of 1D hexagonal supramolecular pillars afterwards. Comparatively, using individual water or isopropanol forms an irregular flake-like or sheet-like shape.⁷⁰  Water with isopropanol creates 1D hexagonal supramolecular pillars after autoclaving, as water probably creates the inner part of the 1D pillar. In contrast, isopropanol generates the hexagonal framework of the 1D pillar as an outer layer.⁷⁰  g-C₃N₄ nanotubes were prepared using a melamine sponge as a template that is soaked in urea solution and then frozen at −20 °C, and then freeze-dried before being finally annealed at 550 °C under a N₂ atmosphere (Figure 1.3A).⁷⁹  The BET surface area of the g-C₃N₄ nanotubes was 1.6 times higher than that of bulk g-C₃N₄ prepared without a melamine sponge.⁷⁹  This study may allow the utilization of melamine-based foam or sponge for the tailored fabrication of 1D gCN nanostructures with tunable shapes and compositions.

3D g-C₃N₄ nanotubes composed of 1D thick tubes decorated with multiple thin tubes were prepared by the thermal polymerization of cyanuric acid and melamine in water with the assistance of phosphorous acid under an autoclave and heated at 180 °C to form nanorods and then a second annealing process at 550 °C was performed under nitrogen to form 3D nanotubes (Figure 1.3B).⁸⁰  The BET surface area 3D g-C₃N₄ nanotubes (71 m² g⁻¹) was superior to bulk g-C₃N₄ by 11-fold alongside enhancement of the photocatalytic properties.

Although the formation of tabular gCNs is the most reported, other morphologies were also reported using different chemical methods. Table 1.1 summarizes the fabrication process of various 1D gCN nanostructures for different applications that clearly showed rare reports on CO₂RR and CO oxidation relative to other applications. As shown in Table 1.1, the template-based method is the most common in the rational synthesis of 1D gCN nanostructures or microstructures. Also, melamine is usually used as an initial precursor for 1D gCNs, due to its ease of handling and polymerization under mild conditions, as well as its stability against side reactions with the template. Additionally, melamine is feasible for coupling with other precursors like urea and cyanuric acid to tailor uniform, high-yield, and well-defined 1D gCN nanostructures.

Compared to bulk gCNs, 1D gCN nanorods possess various inimitable inherent catalytic/photocatalytic properties due to their superior surface area, light absorption, quick charge-carrier migration with a higher magnitude, and proper aspect ratio. Also, 1D gCN nanorods enhance the interaction between electrolyte and electrode, owing to the short diffusion length for the electrolyte's electrons and ions. However, the fabrication of 1D gCN nanorods has not been studied enough relative to bulk gCNs or 2D gCN, or 1D gCN nanotubes. To this end, Atonietti's group synthesized 1D gCN nanorods with a length up to 5 µm and width of 260 nm via the thermal condensation of cyanamide in the presence of AAO membrane as a template. The photocatalytic water-splitting performance of nanorods was superior to bulk gCNs.⁶⁶  Likewise, Wang's group obtained conjugated C₃N₄ nanorods via the supercritical fluid technique, including autoclaving of cyanuric chloride and melamine in acetonitrile solvent at 180 °C.¹⁴⁸  This developed solution‐processing approach allows the facile bottom‐up design of 1D gCN nanorods at low temperature driven by the one-pot solution assembly, crystallization, and cross‐linking chemistry. The as-obtained 1D gCN nanorods showed a lower decreased band gap energy and greater light‐harvesting compared with bulk gCNs. Following that, various efforts were dedicated to rational fabrication of gCN nanorods. For instance, g-C₃N₄ nanorods were prepared by the simple reflux of g-C₃N₄ nanoplates in an aqueous methanol–water solution based on the exfoliation and regrowth followed by a subsequent rolling mechanism.¹⁴⁹  This includes the initial preparation of g-C₃N₄ nanoplates by the polycondensation of dicyandiamide at 550 °C, followed by dispersion in methanol/H₂O solution and refluxing at 68 °C at different times 1–6 h (Figure 1.4A).¹⁴⁹  The BET surface area, uniformity/size, crystallinity, and catalytic properties of the thus obtained g-C₃N₄ nanorods increased with increasing the refluxing time and methanol concentration.¹⁴⁹  Similarly, nanoporous g-C₃N₄ microrods were synthesized via exfoliation and etching of bulk gC₃N₄ using concentrated H₂SO₄ (Figure 1.4B).¹⁵⁰  Initially, bulk g-C₃N₄ nanosheets were obtained by the direct annealing of melamine at 520 °C and then dispersed in concentrated H₂SO₄ in an ice bath container for 12 h to allow the exfoliation and delamination process. Water was added under stirring before centrifugation and then annealing at 550 °C in air.¹⁵⁰  The synthetic mechanism is attributed to the strong oxidative properties of sulfuric acid that delaminate the g-C₃N₄ structure into separated nanosheets resulting from the internal stresses due to the intercalation of sulfate ions with g-C₃N₄ sheets that can be rolled up into microrods.¹⁵⁰  The intercalation of sulfate ions also results in the formation of interconnected nanopores throughout the microrods.¹⁵⁰  Sulfuric acid is necessary for the fabrication of nanoporous microrods. In contrast, using concentrated hydrochloric and nitric acids instead of sulfuric acid did not exfoliate or form g-C₃N₄ microrods due to their inferior oxidation properties relative to the sulfuric acid.¹⁵⁰  Although sulfuric acid can have hazardous effects and is difficult to handle, this study opens new ways to utilize strong acids for the controlled preparation of porous 1D gCNs without templates. However, the template-based method is still the most favored for the tailored design of 1D gCN nanorods with uniform size, shape, and controllable pore size/volume. To this end, mesoporous g-C₃N₄ nanorods with hexagonal mesostructured pores were prepared using chiral mesostructured SiO₂ nanorods, obtained initially using F127 and CTAB as templates.¹⁵¹  The final nanorod product should be condensed at 550 °C in air, followed by the removal of SiO₂ by ammonium hydrogen NH₄HF₂. Another attempt for the green synthesis of 1D gCN nanostructures includes the fabrication of hierarchical g-C₃N₄ porous nanorods by the simple sonication of melem in water solution to allow for molecular rearrangement, followed by annealing to reduce the defects.¹⁵² 

This includes the initial preparation of melem by direct annealing of melamine at 430 °C under air, then grinding before being mixed with water under ultrasonic irradiation to form melem hydrate followed by purification, drying, and finally annealing at 550 °C under air (Figure 1.5A).¹⁵²  The obtained hierarchical g-C₃N₄ porous nanorods possess a high surface area of 88.6 m² g⁻¹, porosity, greater electric conductivity, better crystallinity with fewer defects, higher photocatalytic properties when compared with g-C₃N₄ nanorods prepared using melamine instead of melem hydrate.¹⁵²  This method eliminates the need for using hazardous chemicals or multiple complicated reaction steps. However, the effect of sonication is not yet understood in the preparation process and needs further study. Likewise, 1D gCN rod-like nanostructures were prepared by a simple one-pot and template-free approach, including the polymerization of melamine in ethylene glycol solution using nitric acid and then drying and annealing at 450 °C (Figure 1.5B and C).¹⁵³  This is based on protonation of melamine and then deamination to form heptazine and then melon that self-assembled into a rod-like shape morphology with the assistance of possible migration of carbonated species from inside to outwards during annealing.¹⁵³  This method was developed based on the initial attempts made by Tahir et al.¹⁵⁴  High crystalline and ordered g-C₃N₄ nanorods were prepared via molten salts and reflux methods of protonated melamine obtained after mixing with HCl solution melamine followed by leaching concentrated H₂SO₄.¹⁵⁵  The same method is feasible for the preparation of Cu single atoms impeded g-C₃N₄ nanorods (Figure 1.6A and B).¹⁵⁵ 

Particularly, in the presence of HCl, CuCl₂ complexed to form [CuCl₅]_n³ⁿ⁻, that directly complexed with the protonated melamine that directly polymerized hepatize-based g-C₃N₄ followed by molten salt at 550 °C and then refluxed in H₂SO₄ for leaching to form Cu−CCN nanorods.¹⁵⁵  The photocatalytic properties of g-C₃N₄ nanorods enhanced significantly after integration of a Cu single atom.¹⁵⁵  This may pave the way for the preparation of g-C₃N₄ nanorods with various metal-based single atoms for various catalytic and photocatalytic reactions.

The fabrication of porous 1D gCNs not only involves multiple reaction steps and using templates besides hazardous etchants to remove the template, but also the obtained structures possess lower electrical conductivity and electrochemical surface area.¹⁵⁶ 

To this end, 3D branched g-C₃N₄ nanorods were prepared by a modified ionothermal method, including the dispersion of grinding of bulk g-C₃N₄ (obtained by the direct condensation of melamine) with eutectic salts (KCl and LiCl), then degassing using argon, pyrolysis at 550 °C, while purging argon (Figure 1.7A).¹⁵⁷  Following cooling to room temperature, the powder was directly washed with ice water to allow the quenching process to remove any impurities and dried at 60 °C.¹⁵⁷  The eutectic salts promote the conversion of melon, the formation of nanorods, and induce crystallinity; meanwhile, LiCl acts as an etchant and erodes unstable parts.¹⁵⁷  The quenching provides dissimilar thermal stress at a different site, resulting from the inferior thermal conductance of water.¹⁵⁷  The same method was used to prepare 3D g-C₃N₄-N nanoneedles but with quenching of the annealed powder in the freezing deionized water.¹⁵⁷  The BET surface areas of 3D g-C₃N₄-N nanorods and nanorods were higher than that of bulk g-C₃N₄. Also, the surface area, photocatalytic properties, and charge mobility of nanoneedles and nanorods were superior to bulk g-C₃N₄.¹⁵⁷ 

Tahir et al. developed a facile one-pot and template-free method for the preparation of g-C₃N₄ dense microfibers with an average length of 20 µm and width of 100 nm. This is based on the polymerization of melamine ethanol solution then dried at 450 °C in a chemical vapor deposition (CVD) furnace (Figure 1.7B).¹⁵⁸ 

The same approach was used for the preparation of g-C₃N₄ nanotubes¹⁶⁰  as well as g-C₃N₄ micro-strings,¹⁵⁹  but using ethylene glycol instead of ethanol along with increasing the annealing time. Mainly melamine was polymerized in ethylene glycol solution using nitric acid, followed by drying and annealing at 400 °C, to result in ultra-long micro-strings of 2–4 µm length (Figure 1.7C).¹⁵⁹ 

Screening the growth mechanism at different reaction times showed the formation of nanorods after 60 min annealing, that destructed to produce small beads, which subsequently formed a small string-like shape after fusion and annealing for 120 min and finally grew to a long micro-string after 150 min.¹⁵⁹  The BET surface area of the thus obtained micro-string (290 m² g⁻¹) was 29 times higher than that of bulk g-C₃N₄ (10 m² g⁻¹) and had better photocatalytic properties.¹⁵⁹  Another interesting example is the fabrication of a g-C₃N₄ ultra-long fiber with a diameter width of 0.8 mm and length up to 1 mm prepared by two annealing steps of melamine at 680 °C and then at 720 °C under nitrogen driven by the vapor–liquid–solid mechanism.⁹⁹  Mainly melamine is initially evaporated in the form of CN_x (x > 1) small nucleus vapor phase to the downstream region of a quartz tube with the assistance of a nitrogen carrier deposited on the inner wall of the quartz tube that acts as a substrate for supporting the growth of the CN_x nucleus to the short fiber. Upon increasing the annealing temperature and time, the micrometer-scale nuclei are produced and grew longer due to the greater condensation of vapor around the nuclei.⁹⁹  This study opens new avenues for the scalable fabrication of g-C₃N₄ fibers at the millimeter scale.

Recently, Eid et al. developed a facile, one-pot, and template-free approach for the rational design of various 1D gCN nanostructures in situ doped with mono or binary metals under ambient conditions inspired by the template-free approach mentioned above.¹⁶¹  To this end, porous 1D gCN nanowires co-doped with Pd and Cu (Pd/Cu/gCN NWs) were prepared by the activation of melamine in an ethanol solution containing K₂PdC_l4 and CuCl₂ with the assistance of nitric acid followed by drying and then annealing at 673 K under nitrogen.¹⁶¹  Nitric acid allowed melamine activation to initiate the polymerization process to melon, which provides abundant adsorption sites for the Pd/Cu precursor, and finally, the carbonization process occurred at high temperature under nitrogen. The as-synthesized Pd/Cu/gCN NWs had uniform nanowire structures with an average width of 80 nm and length up to 2000 nm; meanwhile, the nanowires are assembled in network-like porous structures. Elemental mapping showed the homogenous distribution of C, N, Pd, and Cu in the thus obtained Pd/Cu/gCN NWs. The ratio of C/N was about 41/57.8, which is close to the C₃N₄ phase, while the loading of Pd/CU was about 1.2 wt%. Interestingly, co-doping with Pd and Cu slightly increased the BET surface area of Pd/Cu/gCN NWs (120 m² g⁻¹) compared to that of the metal-free gCN NWs (110 m² g⁻¹).¹⁶¹  The thermal CO oxidation activity of Pd/Cu/gCN NWs was tested relative to Cu/gCN NWs, Pd/gCN NWs, and metal-free gCN NWs under (4% CO and 20% O₂) while heating from room temperature to 723 K. The pretreatment for activation of all catalysts at 250 °C under air, and then under H₂, is needed prior to the CO oxidation tests.

The complete CO oxidation to CO₂ (100% conversion) was achieved on Pd/Cu/gCN NWs at (149 °C) compared to Pd/gCNs (283 °C), and Cu/gCN NWs (329 °C), whereas metal-free gCN NWs only showed inferior activity, 7% conversion at 450 °C. Co-doping with binary Pd/Cu plausibly enhances the CO-adsorption and O₂-adsorption/dissociation, accelerating the CO oxidation kinetics. Moreover, Pd/Cu/gCN NWs reserved their activity for 20 h while Pd/gCN NWs and Cu/gCN NWs revealed a slight loss in the activity. These results indicate the substantial effect of doping with either Pd/Cu or Pd or Cu on the CO oxidation activity of gCNs. Inspired by these results, this approach is extended to fabricate gC₃N₄ nanofibers co-doped atomically with Au and Pd (Au/Pd/gC₃N₄NFs), using isopropanol instead of ethanol.¹⁶²  The preparation process includes the polymerization of melamine in isopropanol solution involving HAuCl₄ and Na₂PdCl₄ by nitric acid and then drying before thermal pyrolysis at 480 °C under nitrogen (Figure 1.8).^19,162  Using isopropanol as a solvent and/or source for carbon allowed the fabrication of uniform Au/Pd/gC₃N₄ long nanofibers with a length up to 10 ± 1 µm and width of 80 ± 2 nm as confirmed by TEM and SEM images (Figure 1.8A and B). The long aliphatic carbon chain of isopropanol plausibly increased the length of the gC₃N₄ nanofiber relative to the nanowires obtained using ethanol. The EDX and element mapping analysis revealed the presence of C/N with an atomic ratio of 40, 59, 0.55, and 0.45, respectively, and thus formed Au/Pd/gC₃N₄NFs. The BET surface area of doped Au/Pd/gC₃N₄NFs (85 m² g⁻¹) was slightly larger than the undoped gC₃N₄NFs (72 m² g⁻¹).^19,162  The CO oxidation to CO₂ on Au/Pd/gC₃N₄NFs was significantly higher than that of Pd/gC₃N₄NFs, Au/gC₃N₄NFs, and gC₃N₄NFs, due to the electronic effect of co-doping. Mainly, the complete CO conversion on Au/Pd/gC₃N₄NFs was achieved at 144 °C, which is lower than that of Pd/gC₃N₄NFs (191 °C) by 47 °C and Au/gC₃N₄NFs (205 °C) by 61 °C (Figure 1.8C). Meanwhile, undoped gC₃N₄NFs did not exhibit any significant conversion rate.^19,162  Intriguingly, the CO oxidation performance on Au/Pd/gC₃N₄NFs was superior to the previously reported metal and metal oxide-based catalysts (Table 1.2).^{70,155,163–174}  After a durability test for 20 h, the as-formed Au/Pd/gC₃N₄NFs did not exhibit any loss in activity, while Pd/gC₃N₄NFs lost 10% and Au/gC₃N₄NFs lost 15%. Thereby, co-doping with both Au and Pd enhances activity due to the electronic effect and catalytic merits of binary dopants and also boosts the durability.^19,162 

Later, gC₃N₄ nanotubes in situ co-doped with Au and Pd at the atomic level (Au/Pd/gC₃N₄NTs) were formed by the polymerization of melamine in ethylene glycol solution containing HAuCL₄ and K₂PdCl₄ using nitric acid before drying and then carbonization under nitrogen at 450 °C (Figure 1.9).¹⁷⁵  That drove the preparation of uniform gC₃N₄ nanotubes in situ co-doped with Au and Pd at the atomic level with a length of 1.3 µm and inner pore of 70 nm with a wall thickness of 8 nm (Figure 1.9A and B).¹⁷⁵  The as-obtained Au/Pd/gC₃N₄NTs nanotubes had an exterior layer of polycrystalline graphitic layer, and the inner core of amorphous carbon comprises various defects and lattice distortion due to the co-doping effect.¹⁷⁵  The surface area of Au/Pd/gC₃N₄NTs (320.6 m² g⁻¹) was higher than that of undoped gC₃N₄NTs (275.7 m² g⁻¹). Au/Pd/gC₃N₄NTs with a high surface can afford abundant active catalytic sites for the adsorption/activation/dissociation of reactant, while pores promote the mass transfer and electron mobility during the catalytic applications.¹⁷⁵  Therefore, the CO conversion on Au/Pd/gC₃N₄NTs was completed at 165 °C compared to 1% only on metal-free gC₃N₄NTs, due to the electronic and synergetic effect of Au/Pd (Figure 1.9C).¹⁷⁵  The CO oxidation kinetics on Au/PdgC₃N₄NTs was quicker than metal-free as inferred in its complete conversion at 30 min under 165 °C relative to only 2% for gC₃N₄NTs (Figure 1.9D).¹⁷⁵  Furthermore, Au/PdgC₃N₄NTs maintained activity after 10 durability cycles without any significant loss (Figure 1.9E). Likewise, Pd/Cu/gC₃N₄NTs were formed using nitric acid to polymerize melamine in an ethylene glycol solution involving K₂PdCl₄ and CuCl₂, then dry annealing at 550 °C under nitrogen.^20,176 

Raising the annealing temperature and using Pd/Cu salts instead of Au/Pd salts drove the formation of shorter (1.5 µm) Pd/Cu/gC₃N₄NTs nanotubes with a smaller hollow interior of 60 nm. Interestingly, Pd/Cu/gC₃N₄NTs showed a great catalytic activity towards CO oxidation and CO₂ reduction compared to undoped gC₃N₄NTs.^20,176  Particularly, the complete CO oxidation to CO₂ was achieved at 154 °C, which was superior to Pd/gC₃N₄NTs (210 °C) by 56 °C and Cu/gC₃N₄NTs (250 °C) by 96 °C.^20,176 

Pd/Cu/gC₃N₄NTs allowed the electrochemical and photoelectrochemical CO₂RR to form formic acid as the main product with an inferior amount of ethanol at room temperature in an aqueous sol NaHCO₃. The electrochemical CO₂RR current of Pd/Cu/gC₃N₄NTs was higher than that of the metal-free gC₃N₄NTs by four-fold; meanwhile, the visible light enhanced the current density of Pd/Cu/gC₃N₄NTs by 2.92 times compared to dark conditions.^20,176  It should be noted that the CO oxidation performances of our newly developed Au/Pd/gC₃N₄NFs, Pd/Cu/gC₃N₄NTs, and Au/Pd/gC₃N₄NTs, and Pd/Cu/gCN NWs were superior to metal nanoparticles (i.e., Au/Pd, Pd, and Pt) and metal oxides (i.e., Co₃O_4, MnOx, and Cu₂O) over various supports (Table 1.2).^{21–23,35–37,162,177}  These results open new ways to utilize 1D gCNs in thermal CO oxidation. In pursuit of this aim, 1D gCN nanorods were co-doped with Pd and Pt (PtPd/gCNs) for electrochemical and photoelectrochemical CO ORs.¹⁷⁸  The synthesis process including using HCl and NaNO₃ for the polymerization of melamine in ethylene glycol solution contains K₂PdCl₄ and Na₂PtCl₄, under sonication, then drying and annealing at 550 °C under nitrogen (Figure 1.10A).¹⁷⁸  The mechanism is proposed according to the in situ formation of HNO₃ (H⁺ from HCl and NO₃ from NaNO₃) to allow the amino group's protonation (–NH₂) of melamine, to form melem and melon sheets similar to other reports.^24,158–160  So, HCl and NaNO₃ should be used together; also, the reaction parameters like addition and solvent should be considered to avoid the formation of undesired shapes like sheets, flakes, or wires.¹⁷⁸ 

Melon sheets with various interlayers provide massive adsorption sites for Pt/Pd atoms. Finally, melon sheets are rolled up to form nanorod shapes after the carbonization process. The nanorods had a rough surface with a twined end surface and an average length and width of 94 ± 2 nm and 11 ± 1 nm, respectively (Figure 1.10B and C).¹⁷⁸  The BET surface area of PtPd/CNs (155.2 m² g⁻¹) was superior to metal-free nanorods gCN nanorods (149. 2 m² g⁻¹).

The electrochemical CO oxidation activity and stability of PtPd/CNs nanorods (containing PtPd 1.5 wt%) measured in KOH solution were significantly superior to the commercial Pt/C catalyst (containing 20 wt% Pt) and undoped gCNs.¹⁷⁸  This is obvious in the higher CO oxidation current density of PtPd/CNs (14.75 mA cm⁻²) than Pt/C by 2.01 times and gCN nanorods by 23.41 times under the same potential (Figure 1.10D).¹⁷⁸  Additionally, the CO oxidation current of PtPd/CNs under any applied potential was higher than that of either the Pt/C or gCN nanorods (Figure 1.10E). This is due to the great electrochemical active surface area of the PtPd/CNs nanorods (75 m² g⁻¹) compared to the gCN nanorods (68 m² g⁻¹) and Pt/C (64 m² g⁻¹), as well as the electronic and synergetic effect of PtPd.¹⁷⁸  Furthermore, the photoelectrochemical CO oxidation current of PtPd/CNs increased 1.48-fold than under dark conditions (Figure 1.10F), originating from the unique photocatalytic properties of gCN nanorods, and catalytic properties of PtPd.¹⁷⁸ 

This chapter summarized the recent advances in the fabrication process of 1D gCN nanostructures, including self-assembly and template-based methods (i.e., soft template and hard template). The template-based method is based on the simple filling of the carbon–nitrogen precursors (i.e., melamine, urea, cyanuric acid) into the hard template (i.e., AAO and SBA-15) or the soft template (non-ionic copolymers P123 and F127) followed by solidification or sintering and then template removal by acid (hydrogen fluoride or ammonium difluoride). These methods allowed the fabrication of uniform and well-aligned 1D nanorods, nanowires, nanofibers, nanotubes, and nanoneedles. The diameter, shape, compositions, surface area, conductivity, catalytic, photocatalytic, physiochemical merits, and crystallinity of the thus obtained 1D nanostructures are shaped by the type of precursors, template, and reaction conditions (i.e., solvents, annealing temperature, and reactant concatenations). However, the multiple reaction steps, using hazardous chemicals or solvents, and high-cost templates are the main drawbacks of template-based methods. Also, the unavoidable reaction between templates and carbon–nitrogen precursors may lead to undesired shapes or phases. There are also newly developed methods for preparing one-dimensional nanostructures like hydrothermal, sol–gel, sonication, and molten salts. Although these methods are eco-friendlier and low-cost, they could not produce uniform or aligned 1D nanostructures as in the template methods. These methods also still comprise multi-step reactions. Recent advances include the combination of different methods to reduce the cost and improve the quality and yield of the products, like thermal condensation with templates or sonication, or acid etching/exfoliation, or molten salts. Our group developed a novel one-pot, green, and template-free approach for the rational fabrication of 1D gCN nanostructures doped with mono and binary metal dopants (i.e., PdPt, Pd, Cu, Au, AuPd) including nanowires, nanofibers, nanorods, and nanotubes, via the simple polymerization of melamine in organic solvents (i.e., ethanol, methanol, isopropanol, and ethylene glycol) with the assistance of mild nitric acid followed by annealing. The catalytic performance of 1D gCNs towards CO oxidation and CO₂RR was discussed in detail, supported by various examples. The CO₂RR allowed the production of various hydrocarbons (i.e., formic acid, methanol, methane, and CO), but most reported articles were photoelectrochemical without enough emphasis on the electrocatalytic activities. The thermal CO oxidation on 1D gCNs formed CO₂ at mild temperatures 140–190 °C due to their unique adsorption/dissociation of O₂ along with retarding the intermediates. However, the CO oxidation and CO₂RR on 1D gCNs remain impractical and need further improvements (i.e., ease of preparation methods, reduction of cost, enhancement of the production yield, reduction of the thermal oxidation temperature).

The electrocatalytic CO oxidation and CO₂RR on 1D gCNs have not been studied enough compared to other applications. There is a critical need for future theoretical and experimental studies to tailor the catalytic properties of 1D gCN (i.e., activate the surface area, electrical conductivity, and charge dynamic) and photocatalytic properties (i.e., reduce band gap energy, delay the electron–hole recombination, and enhance the visible light absorption > 450 nm) to allow scalable CO oxidation and CO₂RR on 1D gCNs. These targets could be achieved by combining 1D gCNs with various catalysts such as porous metal catalysts, especially those that are multimetallic, due to their unique redox properties at room temperature;^193,194  likewise, carbon materials (i.e., fullerenes, carbon dots, graphene oxide, carbon nanotubes), metal carbide or nitride, transition metal oxides (i.e., Cu₂O, TiO₂, Ce₂O), due to their great catalytic performance for gas conversion reactions under ambient conditions.^{116–121,124–126,195–198}  There is also a need to obtain more insight into the structural-related fundamentals of 1D gCNs for various gas conversion reactions rather than CO and CO₂, such as methane and ammonia.