Unveiling Versatile Applications and Toxicity Considerations of Graphitic Carbon Nitride

Abstract

Metal-free, low-cost, organic photocatalytic graphitic carbon nitride (g-C₃N₄) has become a promising and impressive material in numerous scientific fields due to its unique physical and chemical properties. As a semiconductor with a suitable band gap of ~2.7 eV, g-C₃N₄ is an active photocatalytic material even after irradiation with visible light. However, information regarding the toxicity of g-C₃N₄ is not extensively documented and there is not a comprehensive understanding of its potential adverse effects on human health or the environment. In this context, the term “toxicity” can be perceived in both a positive and a negative light, depending on whether it serves as a benefit or poses a potential risk. This review shows the applications of g-C₃N₄ in sensorics, electrochemistry, photocatalysis, and biomedical approaches while pointing out the potential risks of its toxicity, especially in human and environmental health. Finally, the future perspective of g-C₃N₄ research is addressed, highlighting the need for a comprehensive understanding of the toxicity of this material to provide safe and effective applications in various fields.

1. Introduction

Graphitic carbon nitride (g-C₃N₄), as the most stable and resistant allotrope out of five polymorphs of carbon nitrides (α-C₃N₄, β-C₃N₄, cubic-C₃N₄, pseudocubic-C₃N₄, and graphitic-C₃N₄), is a novel metal-free organic semiconductor with several notable physiochemical properties. In the last decade, g-C₃N₄ has attracted remarkable attention in research due to its prospective application in numerous fields such as photocatalysis, energy storage, electrochemical sensors, and biomedical engineering. The attribute “graphitic” refers to the layered structure with individual heterocyclic layers bound together by weak van der Waals forces like graphite. The hexagonal network consisting of C-N covalent bonds is mainly formed by s-triazine (triazine) or tri-s-triazine (heptazine) units. The general formula C₃N₄ could be misleading, as the theoretical C/N ratio of 0.75 is hardly obtained by synthesis and the g-C₃N₄ also contains a considerable amount of hydrogen from amino uncondensed groups. With the convenient band gap energy of ~2.7 eV, g-C₃N₄ is an attractive photocatalyst active under the irradiation of visible light, suitable for systems utilising sunlight [1,2,3,4].

To date, many reviews have already been published on the topic of g-C₃N₄, covering synthesis methods, strategies for performance improvement, and applications in various fields [5,6,7,8]. There are considerably fewer reviews summarising the current knowledge and research concerning the toxicity and biocompatibility of this material. In this review, we first describe the fundamental methods of synthesis of both bulk and exfoliated g-C₃N₄ nanomaterials. Subsequently, we aim to provide a brief overview of the latest research on diverse applications of g-C₃N₄, mainly on electrochemical and photocatalytic domains. Lastly, we focus on the summary of the toxicity of this material, discovered so far through various toxicity assays, namely, in vitro toxicity assessment on human and animal cells and ecotoxicity tests on animals, microorganisms, and plants.

2. Preparation Methods of g-C₃N₄

Evaluating the resource efficiency of g-C₃N₄ synthesis methods is essential for assessing its overall sustainability. Green synthesis routes that minimise energy consumption, waste generation, and the use of hazardous materials are preferred. The synthesis of g-C₃N₄ is influenced by several key factors that determine its structural, morphological, and functional properties. These factors include the choice of precursors, synthesis method, temperature, and reaction atmosphere (Figure 1). By carefully controlling these factors, researchers can tailor the properties of g-C₃N₄ to suit specific applications, ranging from photocatalysis and water splitting to sensors and biomedical applications.

2.1. General Conditions and Precursors of g-C₃N₄ Green Synthesis

One of the most commonly used methods of g-C₃N₄ synthesis is the calcination of various nitrogen- and carbon-rich precursors, generally melamine, dicyandiamide, thiourea, and urea, under the O₂ or N₂ atmosphere [4,9]. These precursors undergo thermal polymerisation or condensation reactions, occurring between 400 °C and 600 °C, while the building units and g-C₃N₄ polymer network fully form above 500 °C. When passing the temperature of 600 °C, the polymer becomes unstable, and beyond 700 °C it is decomposed completely [10].

During the preparation of g-C₃N₄, the most important factors that influence the composition and properties of the end product are precursor selection and polymerisation conditions, i.e., the temperature of calcination and heating rate. Zheng et al. conducted a study in which melamine and urea were chosen and compared as reagents for g-C₃N₄ formation solely through thermal polymerisation at varying temperatures (500, 550, 600, and 650 °C). It was found that the surface area of g-C₃N₄ gradually increased with an increasing pyrolysis temperature in both cases; however, urea-derived g-C₃N₄ was overall a lot less dense and lighter, with a highly porous, “sponge-like” structure in comparison to melamine-derived g-C₃N₄ with a high order of crystallinity. This phenomenon is attributed to a much larger atmosphere of gases (NH₃, H₂O, CO₂) generated from the pyrolysis of urea, creating bubbles that may function as gas templates, which increases the distance between the layers. Samples derived from urea also showed better adsorption and photocatalytic abilities [11].

More than the calcination temperature itself, a major factor in controlling the outcome of g-C₃N₄ products may be the heating rate during polymerisation. Sharma et al. compared g-C₃N₄ prepared from urea and melamine at varying heating rates. They found that g-C₃N₄ synthesised from both urea and melamine showed enhanced photocatalytic and surface properties when produced at a slower heating rate of 2 °C/min in comparison with g-C₃N₄ prepared at a high heating rate of 10 °C/min [12]. In addition to the calcination described above, i.e., the thermal polymerisation of nitrogen- and carbon-rich precursors, other methods for the preparation of g-C₃N₄ are known, including solvothermal and hydrothermal methods, molten salt synthesis, and chemical vapour deposition (CVD). In solvothermal synthesis, g-C₃N₄ is prepared in a solvent (either water or an organic solvent) under high pressure and temperature, which can result in various morphologies of g-C₃N₄, such as nanosheets or g-C₃N₄ with a porous structure [13]. Using the molten salt method, g-C₃N₄ with a high degree of crystallinity and a large surface area can be prepared. In this technique, molten salts at high temperatures, such as chlorides, nitrates, and carbonates, are utilised as the reaction medium. This method is also often used in the preparation of g-C₃N₄ nanosheets [14]. CVD can be utilised to form a thin film of g-C₃N₄ on a suitable substrate using gas-phase precursors at elevated temperatures. This method yields thin g-C₃N₄ films with controlled thickness and excellent uniformity [15].

2.2. Preparation of g-C₃N₄ Nanomaterials

Bulk g-C₃N₄ without further treatment, activation, or modification exhibits inadequate photocatalytic properties due to a relatively low surface area and few active sites, low solar-light adsorption, and fast recombination of electron–hole pairs [16]. There is an important emphasis on the preparations of g-C₃N₄ nanomaterials mainly in the form of nanosheets or even nanotubes, which exhibit noticeably enhanced physiochemical properties compared to bulk g-C₃N₄. Furthermore, for improved properties of g-C₃N₄, elemental doping with nonmetals (B, S, O) or metals (Ni, Co, Mg) or doping with various molecules and materials (TiO₂, carbon nanotubes) to introduce certain functional groups into the structure of g-C₃N₄ and many other methods are also being explored as means of g-C₃N₄ functionalisation for technological applications [17,18,19].

Nanosheets of g-C₃N₄ with a hexagonal lattice similar to graphene are typical examples of 2D nanomaterials, and their formation can be achieved through two different approaches, regarded as the “top-down” and “bottom-up” methods (Figure 2). When utilising a “top-down” method, the layered bulk material is delaminated into nanosheets. Nanosheets of g-C₃N₄ with a thickness of around 2 nm can be obtained by simple thermal oxidation of bulk g-C₃N₄ in the air atmosphere [20]. The other most commonly used top-down methods are chemical, ultrasound, or microwave-assisted exfoliation [21,22]. In contrast, a “bottom-up” method typically relies on the use of various templates to build the desired materials out of the most elementary units—atoms and molecules. In the hard-template method, a rigid template of pre-synthesised material, e.g., silica, alumina, or clays, acts as a mould to give shape to the precursor material. On the other hand, soft templating relies on the use of surfactants, block copolymers, and ionic liquids to act as templating agents to assist in the assembly of the precursor material into a porous structure [23]. Gas-template methods could also be included in bottom-up soft-template methods. Easy one-step synthesis with the use of NH₄Cl as a gas-template agent could be used for the preparation of 3.1 nm g-C₃N₄ nanosheets directly from dicyandiamide as a precursor [24]. The main advantages of the top-down approach are scalability and reproducibility, while the bottom-up approach enables the direct synthesis of nanosheets with uniform size and crystallinity [25]. In addition to 2D g-C₃N₄ nanosheets, 1D nanotubes of g-C₃N₄ have also shown significant performance enhancements. Mo et al. successfully prepared nitrogen-rich g-C₃N₄ with a tubular structure by calcination of a supramolecular precursor under an NH₃ atmosphere. The precursor for g-C₃N₄ nanotubes was prepared using the hydrothermal method through the supramolecular self-assembly of melamine with hydroxylammonium chloride. The resulting g-C₃N₄ nanotubes exhibited a significantly increased specific surface area, approximately 16 times higher than bulk g-C₃N₄ [26].

As has been shown, the synthesis of g-C₃N₄ can be achieved through various methodologies, each affecting the material’s efficiency, morphology, and purity. Each synthesis method offers distinct advantages and trade-offs, impacting the suitability of g-C₃N₄ for specific applications such as photocatalysis, water splitting, and sensor technologies. These are summarised in

Table 1. Synthesis methodologies of g-C₃N₄ and their outcomes [5,6,25,27].

Synthesis Type	Result	Efficiency and Purity
Thermal polymerisation	Produces bulk g-C3N4 with high crystallinity and decent photocatalytic properties. However, this method often results in a limited surface area and poor dispersion in solutions.	High purity with moderate efficiency due to limited surface area and electron–hole recombination issues.
Solvothermal and hydrothermal methods	Creates g-C3N4 with various morphologies, such as nanosheets or nanospheres, and can introduce porosity.	Higher surface area and enhanced photocatalytic activity compared to bulk g-C3N4. Purity depends on the solvent and reaction conditions but is generally good.
Template-assisted synthesis	Produces porous or structured g-C3N4 with high surface areas and tailored morphologies.	High surface area leads to improved catalytic performance. Template removal can impact purity if not performed thoroughly.
Chemical vapour deposition (CVD)	Yields thin films of g-C3N4 with controlled thickness and high uniformity.	Very high purity and good control over material properties, leading to efficient catalytic and electronic applications.
Molten salt synthesis	Produces g-C3N4 with a high degree of crystallinity and large surface area.	High efficiency due to better dispersion and surface properties. Purity is generally high but depends on the post-synthesis washing to remove residual salts.
Doping and composite formation	Enhances electronic properties, photocatalytic activity, and stability. Creates composites with synergistic properties.	Improved efficiency due to enhanced charge separation and increased active sites. Purity can be impacted by the choice of dopants and composite materials.
Exfoliation techniques	Produces ultrathin g-C3N4 nanosheets with high surface areas and excellent catalytic properties.	High efficiency due to increased active sites and better light absorption. Purity is generally high if exfoliation is performed without introducing impurities.

.

3. Applications of g-C₃N₄

g-C₃N₄ showcases outstanding multifunctionality and enhanced performance in various applications, positioning it as a leading material in photocatalysis and related fields (Figure 3).

In electrochemical sensing, g-C₃N₄-modified electrodes exhibit high sensitivity and selectivity for detecting organic compounds and heavy metals. The enhanced detection limits are attributed to its high surface area and conductivity, making g-C₃N₄-modified electrodes more effective than traditional ones. This performance improvement ensures faster, cheaper, and more reliable in situ analysis, critical for applications in environmental monitoring and healthcare diagnostics.

Photoelectrochemical (PEC) sensing benefits from g-C₃N₄’s role as a photoactive electrode. It enhances PEC system efficiency by generating effective photogenerated electron–hole pairs. This results in higher sensitivity and specificity compared to many conventional materials. g-C₃N₄’s ability to improve PEC sensing performance is particularly advantageous in detecting low concentrations of analytes, making it suitable for applications in food safety testing and pollution monitoring.

In photocatalytic applications, g-C₃N₄ is effective in pollutant degradation and water disinfection, offering an eco-friendly alternative to traditional methods. While bulk g-C₃N₄ has limitations, its nanostructured forms exhibit enhanced photocatalytic activity, often surpassing other photocatalysts in efficiency. This makes g-C₃N₄ an excellent choice for environmental remediation, where sustainable and efficient solutions are needed.

For water splitting, g-C₃N₄ nanocomposites demonstrate high efficiency in hydrogen production. Their tunable bandgap and stability make them competitive with other photocatalytic materials. The ability to efficiently split water into hydrogen and oxygen positions g-C₃N₄ as a key material in the development of the hydrogen economy, which is crucial for addressing climate change and reducing reliance on fossil fuels.

In drug delivery systems, g-C₃N₄ nanoparticles are valued for their biocompatibility and functionalisation capability. These properties enable targeted drug delivery with controlled release, offering improved therapeutic efficacy over some traditional drug delivery systems. The use of g-C₃N₄ in biomedical applications highlights its potential to enhance the effectiveness and precision of treatments.

Phototherapy is another area where g-C₃N₄ excels due to its superior light absorption and photothermal conversion efficiency. This material provides a non-invasive treatment option with potential advantages over conventional photothermal agents, particularly in cancer treatment and other medical applications. g-C₃N₄’s properties ensure effective and safe phototherapy, promoting better patient outcomes.

3.1. Electrochemical Sensing

The basic principle of any electrochemical sensor or biosensor is the ability to convert chemical signals into easy-to-interpret electrical signals, therefore offering a fast, cheap, and selective in situ analysis. To reach even lower limits of detection, elevated sensitivity, and selectivity, extensive research is being put into the development of modified commercially available electrodes. Properties such as high specific surface area and conductivity, facile surface functionalisation, tunable bandgap, and exceptional thermal and chemical stability are only some that are typically highlighted within the topic of g-C₃N₄ as an electrode modifier. Several formats of electrodes with g-C₃N₄ modifications have recently been prepared for the detection of a plethora of substances [28].

A carbon paste electrode (CPE) consisting of graphite and paraffin wax, along with g-C₃N₄ as an attractive modifier for the heightened signal response, was used for the detection of tryptophan [29] and acyclovir [30]. Highly accurate determination of heavy metals (Cr, Ni) in tap water was accomplished by potentiometric titration with g-C₃N₄ and multiwalled carbon nanotube (MWCNT)-modified CPE [31].

Screen-printed electrodes (SPE) modified solely with nanoscale g-C₃N₄ doped with additional oxygen-containing functional groups were utilised for dibenzofuran analysis via impedimetric detention to achieve an exceptionally low LOD (pM range) [32]. In another study, Cu, Zn, and Fe metal–organic frameworks (MOFs) enriched with melamine-derived g-C₃N₄ were evaluated towards 2,4-dichlorophenol sensing on a modified SPE. The Fe-MOF/g-C₃N₄ nanocomposite showed the most fitting electrochemical behaviour and was used for real sample analysis, achieving nanomolar detection limits [33]. Niaz et al. used cathodic stripping voltammetry for the detection of iodide ions (I⁻) on a chitosan/g-C₃N₄ composite-modified SPE. After applying a cleaning potential, the modifying layer remained unchanged for several repeated measurements, which stands out among many single-use sensing methods utilising SPEs [34].

Glassy carbon electrodes (GCE) are arguably the most widely used electrodes in electroanalytical sensing, with endless potential for surface modification to suit the specific choice of analyte. Metronidazole was successfully detected with appropriate selectivity with no interference from other electroactive compounds, using a MoS₂/g-C₃N₄ composite anchored with Nafion on the electrode’s surface [35]. A GCE modified with a nanocomposite of Bi₂Te₃/g-C₃N₄ achieved some of the best limits of detection (calculated as 0.7 μM) reported to date for the electrochemical analysis of epinephrine, also known as adrenaline [36]. Impressive results for perphenazine determination were also achieved via a LaCoO₃/g-C₃N₄-modified GCE. In this work, the nanocomposite resembled “belts” of individual sheet-like layers of g-C₃N₄ dotted with circular nanoparticles of the lanthanum compound [37].

Each electrochemical sensing application has unique requirements and challenges. Optimising g-C₃N₄-modified electrodes for specific analytes, matrix compositions, and environmental conditions remains a challenge. Tailoring the properties of g-C₃N₄ to address these specific application needs requires further investigation. One of the primary challenges associated with g-C₃N₄-modified electrodes is the complexity and variability of synthesis methods. Achieving consistent and reproducible g-C₃N₄ structures with desired properties can be challenging, impacting the reliability and reproducibility of electrode performance. While g-C₃N₄ shows promise for electrochemical sensing applications in research settings, scaling up production and commercialisation remain hurdles. Developing cost-effective synthesis methods and ensuring scalability while maintaining performance is crucial for transitioning g-C₃N₄-modified electrodes from the lab to practical applications.

3.2. Photoelectrochemical Sensing

Photoelectrochemical (PEC) detection utilises light for excitation and measures current or voltage as the output signal. In the construction of PEC platforms, g-C₃N₄ can have different roles, such as a photoactive electrode material, sensitiser, and either electron donor or acceptor. Additionally, owing to diverse surface functional groups, g-C₃N₄ can act as an immobiliser for bioreceptors such as aptamers [38,39].

A PEC system was developed on an F-doped SnO₂ (FTO) glass platform first modified with g-C₃N₄ and BiOI on top for accurate detection of ascorbic acid. The continuously decreasing cathodic photocurrent was attributed to an increasing ascorbic acid concentration. The research states that this combination of electrode modifiers leads to the formation of abundant charge transfer pathways. Their reordering or subsequent layering creates a template for task-specific modulation of these pathways based on the intended use [40].

Surface-modified g-C₃N₄ nanosheets decorated with clusters of Cu (2%) provided an accurate photoelectrochemical sensing platform for the detection of 4-nitrotoluene. Cu (II) ions effectively worked in synergy with the exfoliated g-C₃N₄ nanosheets to promote separation and migration of the photogenerated electron–hole pairs, therefore greatly enhancing the photocurrent response to the analyte. The copper ions in this system served as an electron-trapping medium to facilitate the reduction of p-nitrotoluene. The surface modification proved to be crucial in this system, as g-C₃N₄ nanosheets themselves showed minimal change in photocurrent in response to increasing 4-nitrotoluene concentration [41].

A photoelectrochemical aptasensing platform for the detection of extremely low concentrations of atrazine was developed by Yan et al. In this study, additional cyano functional groups, as well as nitrogen deficiencies, were introduced into the bulk g-C₃N₄ material to trap photoinduced electrons, improve the adsorption capacity of visible light, and therefore improve PEC performance. Atrazine-specific aptamer as the identification element was further used to modify g-C₃N₄, and atrazine concentration was relative to the decrease in detected photocurrent [42]. Tian et al. prepared an aptasensor based on the electrochemiluminescence (ECL) of g-C₃N₄ for the accurate detection of aflatoxin B1. The biosensor was formed by deposition of g-C₃N₄ modified with aflatoxin-specific aptamer on GCE. By itself, g-C₃N₄ exhibited intense ECL signals; therefore, growing aflatoxin concentration was directly proportional to a decrease in ECL signal response, as the aflatoxin–aptamer complex blocked active sites on the material [43].

Achieving precise control over surface functionalisation and maintaining the stability of modified g-C₃N₄ structures are significant challenges that warrant attention in PEC sensing applications. Potential interference from other compounds and optimised selectivity for real-world applications are critical considerations for the practical implementation of g-C₃N₄-based PEC sensors. As with electrochemical sensing, scaling up the production and commercialisation of g-C₃N₄-modified PEC platforms presents challenges. Developing cost-effective synthesis methods and ensuring scalability while maintaining performance and reproducibility are essential for translating laboratory-scale research into practical sensing devices.

Table 2. Sensing applications of g-C₃N₄ (mentioned in Section 3.1 and Section 3.2) and their performance comparison to other photocatalytic materials.

Analyte	g-C3N4-Based Sensing Platform	Ref.	LOD (μM)	Comparison Material	LOD (μM)	Ref.
Tryptophan	g-C3N4-modified CPE	[29]	0.085	Graphene-modified CGE	0.3	[44]
Acyclovir	g-C3N4-modified CPE	[30]	3 × 10−3	Magnetic CdO NPs-modified GCE	0.3	[45]
Heavy metals (Cr, Ni)	g-C3N4, MWCNT-modified CPE	[31]	6.7 × 10−3 and 0.012	Zeolite and chlorinated MWCNT-modified CPE	0.06 (Cr)	[46]
Dibenzofuran	Oxygenated g-C3N4-modified SPE	[32]	1.58 × 10−6	Silver electrode modified with MnO2 nanofibers	1.2 × 10−3	[47]
2,4-dichlorophenol	Fe-MOF/g-C3N4-modified SPE	[33]	1.2 × 10−3	Cu-MOF/rGO composite-modified GCE	0.083	[48]
Iodide ions (I−)	g-C3N4/chitosan-modified SPE	[34]	0.01	Silver oxide microparticles PAA/PVA-modified GCE	0.3	[49]
Metronidazole	MoS2/g-C3N4-modified GCE	[35]	0.09	Graphene nanosheets/Fe3O4 modified-GCE	0.23 × 10−3	[50]
Epinephrine	Bi2Te3 /g-C3N4-modified GCE	[36]	0.71	Zeolite imidazole framework on GCE	2.1	[51]
Perphenazine	LaCoO3/g-C3N4-modified GCE	[37]	4.3 × 10−3	Graphene oxide nanosheets on GCE	38.4	[52]
Ascorbic acid	FTO glass/g-C3N4/BiOI PEC platform	[40]	3.3	Cu-porphyrin MOF	0.023	[53]
P-nitroroluene	g-C3N4 nanosheets/Cu(2%) PEC platform	[41]	0.13	-	-	-
Atrazine	g-C3N4 nanosheets doped with cyano groups and N deficiencies on ITO glass PEC aptasensor	[42]	3.33 × 10−11	BiOI nanoflowers/TiO2 nanotubes PEC platform	0.5 × 10−6	[54]
Aflatoxin B1	g-C3N4 + aptamer on GCE	[43]	0.005 ng/mL	Methylamine perovskite quantum dots, encapsulated by ZIF-8 MOF on GCE	3.5 fg/mL	[55]

summarises the aforementioned applications of several g-C₃N₄-based sensors including their comparison to methods using different modifiers on the electrode surface. In most cases, sensors utilising g-C₃N₄ were able to achieve higher sensitivity and significantly lower limits of detection to the same analyte. Additionally, in several examples, the modifier containing g-C₃N₄ requires fewer steps and simpler synthesis approaches, proving the practicality of this material in electrochemical and photoelectrochemical sensing.

3.3. Photocatalytic Applications

Among the first discovered and most prominently researched attributes of g-C₃N₄ is its photocatalytic properties; however, as has been mentioned, g-C₃N₄ in its bulk form is severely hindered for these applications, showing a relatively small specific surface area and therefore insufficient active sites as well as the rapid recombination of excited electron–hole pairs. Numerous studies over the years show that the limitations of bulk g-C₃N₄ can be easily worked around with the introduction of g-C₃N₄-based nanostructures or by engineering electron redistribution by introducing vacancies or doping g-C₃N₄ with other elements. It has been noted that many approaches to modifying g-C₃N₄ for enhanced photocatalytic properties rely on creating structures with various metals, which may be detrimental to the environment in the long run due to the low availability of some metals in nature and their inherent toxicity. Special attention is being paid to developing g-C₃N₄ catalysts using methods and materials with sustainability in mind [56,57].

A novelty approach to 1,2 amino alcohol synthesis has been developed through ultrathin porous g-C₃N₄ nanosheets-enabled photocatalysis under visible light by Xu et al. The 1,2-amino alcohol component serves as a crucial chemical backbone in numerous natural products, biologically active compounds, and ligands. This approach enables simultaneous reductive and oxidative reactions, leading to the formation of two radical intermediates that couple to produce the final product with almost 100% yields. Most notably, the synthesised photocatalyst exhibited satisfactory recyclability, remaining effective for up to seven cycles. Additionally, the reaction mechanism was explored for its applicability to other complex molecules, suggesting potential industrial applications for ultrathin porous g-C₃N₄ nanosheets in the synthesis of valuable pharmaceutical intermediates [58].

Ozonation refers to the production of reactive oxygen species (ROS) from ozone (O₃) molecules, capable of attacking and degrading a wide range of organic compounds and microorganisms. Nanoclustered g-C₃N₄ functionalised with additional oxygen-containing (mainly C=O and C–O) groups was found to be an exceptional catalyst, capable of generating primarily ^•OH and O₂^•− radicals in the presence of O₃, as confirmed by EPR analysis. When tested, atrazine removal of a system containing oxygen-functionalised g-C₃N₄ and O₃ reached nearly 93% compared to O₃ alone (63%). The material’s catalytic abilities remained constant under both acidic and alkaline conditions, suggesting a wide applicability [59]. Another study reports significantly lower (around 30%) usage of O₃ and complete inactivation of Escherichia coli bacteria and JC virus in water, utilising a dual system of g-C₃N₄/O₃, greatly reducing the cost of such water treatment. The main ROS generated were identified to be ^•OH [60].

Xu et al. prepared g-C₃N₄ via a multiple-step calcination route resulting in a material with a combined crystalline and amorphous structure and a narrower bandgap than its bulk g-C₃N₄ counterpart, making it more conductive. Upon stimulation with xenon lamplight in an environment abundant in O₂, g-C₃N₄ with a combined crystalline and amorphous structure was capable of efficient H₂O₂ production and nearly 100% removal of methylene blue under simulated conditions. The key ROS generated and important in the mechanism of pollutant degradation reaction were identified to be O₂^•− radicals and singlet oxygen ¹O₂ [61].

Photocatalytic disinfection strategies against microbial and bacterial pollution have been considered recently as a preferable solution in contrast to traditional bacterial inactivation methods because of their ecological sustainability. Pure g-C₃N₄ without any additives shows certain antibacterial properties under visible light illumination on account of its unique photocatalytic properties. As previously mentioned, the photoexcited free radicals (^•OH, O₂^•−) can attack cell walls and membranes, causing them to rupture, leak intracellular materials, and kill bacteria [62].

These approaches demonstrate innovative solutions to enhance the photocatalytic properties of g-C₃N₄, enabling its application in diverse environmental remediation and synthesis processes. However, concerns about the environmental impact of using metal-based modifications to enhance g-C₃N₄ photocatalytic properties are rising. It is important to develop sustainable g-C₃N₄ catalysts using methods and materials that minimise environmental harm. This reflects a growing awareness within the scientific community regarding the need for eco-friendly approaches to material synthesis and catalysis. Developing scalable and reproducible synthesis routes for g-C₃N₄-based nanostructures or doped materials is essential for practical applications but often presents challenges in terms of cost and complexity. Assessing the stability of these materials over extended operation periods and under different environmental conditions is crucial for evaluating their practical viability for industrial applications. Further research is needed to elucidate the underlying processes governing photocatalytic reactions on g-C₃N₄ surfaces and to identify key factors influencing catalyst activity and selectivity.

3.4. Watter Splitting

Another possible application of a photoelectrochemical system based on g-C₃N₄ is utilising the photogenerated electron–hole pairs to drive redox processes associated with water splitting. Simply put, water splitting is a chemical reaction in which hydrogen (H₂) and oxygen (O₂) are generated from water molecules. The development of economical and efficient means of water splitting is regarded to be one of the major building blocks for the hydrogen economy, considered to be one of the most promising ways to mitigate ongoing climate change caused by the use of fossil fuels [63]. Water splitting is divided into three main routes, photocatalytic, electrocatalytic, and photoelectrocatalytic. Photocatalytically, water splitting occurs through the formation of bound electron–hole pairs (excitrons) by adsorption of light. For electrocatalytic water splitting, O₂ is generated on the anode and H₂ is generated on the cathode at an electrode potential of 1.23 V, assuming the electrochemical cell has been assembled with appropriate materials. When these two principles combine, water splitting can also occur via a photoelectrocatalytic route, meaning a photocatalyst material is bound to an electrode and potential is applied to the system in the presence of light [64].

Wang et al. presented an interesting new approach to photoelectrocatalytic water splitting with the use of g-C₃N₄ nanofibers decorated with black phosphorous nanoparticles. Black phosphorous has the properties of a p-type semiconductor with notable optoelectrical qualities, with an adjustable bandgap ranging from 0.3 eV to 2 eV, depending on the specific morphology of the material, while g-C₃N₄ functions as a n-type semiconductor with a defined bandgap of 2.7 eV. The combination of these two materials as a p-n heterojunction led to the formation of a highly functional photoresponsive material, achieving a high evolution rate of hydrogen and being stable for over 8 h [65].

Sun et al. achieved exceptional H₂ production rates through photocatalytically driven water-splitting reactions on a complex of g-C₃N₄ nanosheets deposited with single-atom cerium sites. In addition to the water splitting reaction, the prepared nanocomposite was able to oxidize amines after irradiation with visible light. The lifetime of photoinduced electrons was improved by 5 ns in comparison to bare g-C₃N₄, and the production of superoxide radicals was proven to be the driving force of the reaction mechanisms. The addition of cerium also improved charge transfer resistance and showed significant photoluminescence quenching compared to the bare g-C₃N₄ material [66].

The reaction of O₂ evolution has been remarked on as highly energy-demanding ad not particularly fast due to its mechanism requiring the transfer of four electrons. Iridium oxide (IrO₂) is the most commonly used catalyst in water electrolysis, but its main disadvantage is the low availability of the rare metal and, therefore, an exceptionally high price. Aiming to improve on these facts, Torres-Pinto et al. developed an electrocatalytic system consisting of g-C₃N₄ deposited onto Ni foam substrates. Upon extensive performance evaluation of this system, for the first time it was noted that a g-C₃N₄-based catalyst was able to surpass the point-of-reference catalyst IrO₂ in terms of rates of generated O₂ in an alkaline medium [67].

Table 3. Properties of common water-splitting catalysts compared to g-C₃N₄ [68,69,70,71].

Material Type	Bandgap (eV)	Stability	Cost and Scalability	Environmental Impact	Other Drawbacks
g-C3N4	2.7—easily excited by visible light.	Thermally and chemically stable.	Cheap and abundant precursors, relatively simple synthesis, effective for large-scale applications.	Environmentally compatible.	Prone to rapid electron–hole recombination, functionalisation is necessary to overcome drawbacks.
Metal oxides	Several catalysts such as TiO2, SnO2, ZnO, and NiO are limited by larger bandgaps (>3.2)—not able to adsorb visible light.	Extremely stable.	Limited availability and high cost of several noble metals (Pt, Ru, Ir, Pd). Difficult for large-scale production.	Can be environmentally hazardous due to the potential release of toxic heavy metals.	Prone to deactivation due to adsorption of impurities or changes in pH. In nanoparticle form, prone to agglomeration.
Metal chalcogenides	Typically a much narrower bandgap (1.2–2.4).	Susceptible to photocorrosion and environmental degradation.	Several metals are expensive and low in abundance. Difficult for large-scale production.	Can be environmentally hazardous due to the potential release of toxic heavy metals.	Limited activity, low conductivity, low synthesis yield.
Perovskite materials	Adjustable, depending on the composition.	Prone to degradation from oxygen, moisture, UV exposure, and high temperatures (above 80).	Low cost of materials, faces challenges in large-scale processing.	Some perovskites contain lead. Additionally, toxic solvents are employed in processing.	Low catalytic performance when used alone.
Porous organic polymers (POPs)	Adjustable, depending on composition.	Generally stable.	Inexpensive precursors, scalable. Estimated production cost lower than 10 USD/kg.	Generally nontoxic, energy-efficient, and recyclable, though case-specific.	Relatively large pore size (>1 nm), which may be a limiting factor in some applications.

summarises some advantages and disadvantages of g-C₃N₄ materials compared to other common catalysts. However, it should be noted that even though significant progress has already been made in this research, the use of g-C₃N₄ in photoelectrocatalysis is still in its beginning stages. Current trends focus on environmental friendliness and cost efficiency in technological materials suggesting great potential for g-C₃N₄-based catalysts.

3.5. Drug Delivery Systems and Phototherapy

Low biotoxicity and good biocompatibility make g-C₃N₄ a favourable material for biomedical applications, especially in the field of cancer treatment. It can be used as a nanocarrier by loading the therapeutic agent or drug into the cavities to make a complex. The drug molecules or therapeutic agents are loaded on the nanocarrier by weak forces of attraction, e.g., London dispersion force [72]. Assuming the acidic microenvironment of the tumour (a pH value around 5), both the anticancer drug and its carrier (g-C₃N₄) become negatively charged once absorbed by the tumour. Hence, the electrostatic repulsion facilitates the release of the anticancer drug. Recent insights into g-C₃N₄ in drug delivery testing have been summarised in more detail by Pourmadadi et al. [17].

Among many types of water-splitting materials, g-C₃N₄ has attracted considerable attention for its adjustable bandgap and band position. After the modification, water splitting can be driven under a high penetrable laser, which makes g-C₃N₄ suitable for phototherapy. Unfortunately, some disadvantages still restrict the application of g-C₃N₄ in water splitting, including a high recombination rate of photogenerated electron–hole pairs and lower light utilisation. Therefore, the modification design of g-C₃N₄ for further widening the light absorption range and improving stability is necessary to meet the requirements of practical water splitting [73].

4. Toxicity Considerations of g-C₃N₄

The mechanism of any phenol-like compound degradation using g-C₃N₄ typically involves a photocatalytic process wherein g-C₃N₄ serves as a catalyst to facilitate the decomposition of phenolic compounds under light irradiation [74]. When exposed to light, g-C₃N₄ absorbs photons, promoting electrons from the valence band to the conduction band, creating electron–hole pairs. The photoexcited electrons (e⁻) and holes (h⁺) can react with water and oxygen molecules adsorbed on the surface of g-C₃N₄ to generate already mentioned ROS. Phenol molecules present in the solution are adsorbed onto the surface of g-C₃N₄ through π–π interactions or hydrogen bonding. The ROS generated on the surface of g-C₃N₄ reacts with the adsorbed phenol molecules, leading to the oxidation of phenolic compounds into intermediate products such as hydroquinone, benzoquinone, and eventually, smaller organic acids, aldehydes, and carbon dioxide. During this process, several risks that can negatively impact the environment or human health need to be considered. These are summarised in

Table 4. g-C₃N₄ mechanisms and toxicity considerations during photocatalytic processes.

Risk	Impact
Toxic intermediates	As phenol undergoes degradation, various intermediate products are formed. Some of these intermediates may exhibit toxicity to humans and aquatic organisms, potentially posing risks to environmental and human health.
Incomplete degradation	Incomplete degradation of phenol or its intermediates could result in the accumulation of persistent organic pollutants in the environment, leading to long-term ecological impacts and potential bioaccumulation in the food chain.
Unwanted ROS	Uncontrolled generation of ROS during the photocatalytic process may cause oxidative stress in aquatic organisms and disrupt ecosystems, particularly in sensitive aquatic environments.
Release of nanoparticles	While g-C3N4 is generally considered to be biocompatible, the long-term effects of nanoparticle exposure on human health and the environment are still not fully understood.

.

To mitigate these risks, thorough research and risk assessments are necessary to understand the fate and toxicity of intermediate products, optimise photocatalytic conditions to minimise the formation of toxic by-products, and implement appropriate wastewater treatment strategies to ensure the safe disposal of treated effluents. Additionally, proper monitoring and regulation of g-C₃N₄-based photocatalytic processes are essential to prevent potential adverse impacts on human health and the environment.

Various experimental data show that g-C₃N₄-based materials are biocompatible only under the investigated doses. Biodegradability, chronic toxicity, and larger animal model testing have not been systematically and deeply studied yet. Therefore, the biocompatibility of g-C₃N₄-based photocatalysts still requires detailed biosafety assessments. Only when the biocompatibility of g-C₃N₄-based photocatalysts is fully demonstrated will it be possible to proceed with subsequent clinical applications [75]. For example, nanosheets of g-C₃N₄ exhibit promising applications as efficient photosensitizers for photodynamic tumour therapy and as pH-responsive nanocarriers for drug delivery. As photosensitizers, these nanosheets demonstrate the capability to generate ROS, effectively targeting and eliminating cancer cells under low-intensity light irradiation. Despite this mechanism holding significant potential for enhancing cancer treatment outcomes, the adverse effects on surrounding healthy tissues and cells need to be considered. However, when this material undergoes unintended processes such as the spontaneous release of certain components (e.g., metals) and the generation of ROS, its toxicity can be deemed a negative outcome [76].

Table 5. g-C₃N₄ mechanisms and toxicity considerations in biomedical applications.

Consideration	Impact
Particle size and shape	Nanoparticles of g-C3N4 may exhibit different properties and behaviours compared to bulk materials. Their small size and high surface area-to-volume ratio could increase interactions with biological systems, potentially leading to adverse effects such as cellular uptake, oxidative stress, and inflammation.
Chemical composition	The chemical composition of g-C3N4, including any surface functionalisation or impurities, could influence its toxicity profile. For example, surface groups or contaminants may enhance cellular uptake or trigger immune responses, leading to cytotoxic or immunotoxic effects.
Biological interactions	When introduced into biological systems, g-C3N4 nanoparticles may interact with cellular components such as proteins, lipids, and nucleic acids. These interactions could disrupt cellular processes, interfere with signalling pathways, or induce cellular damage, ultimately leading to cytotoxicity or genotoxicity.
Oxidative stress	Nanoparticles of g-C3N4 have the potential to generate ROS through photoactivation or chemical reactions. Excessive ROS production can overwhelm cellular antioxidant defences, leading to oxidative stress and cellular damage.
Biodegradation and clearance	The biodegradation and clearance of g-C3N4 nanoparticles from the body are critical factors in determining their long-term toxicity. If nanoparticles persist in biological tissues or accumulate in organs over time, they may elicit adverse effects such as chronic inflammation, fibrosis, or organ damage.
Aggregation and agglomeration	Nanoparticles of g-C3N4 may agglomerate or aggregate in biological fluids or tissues, altering their physicochemical properties and biological interactions. Aggregated nanoparticles could lead to localised toxicity, impaired cellular uptake, or obstruction of biological pathways.

summarises possible mechanisms through which g-C₃N₄ could exhibit toxicity during biomedical applications (i.e., cancer therapy).

5. Assessment of g-C₃N₄ Toxicity

As mentioned above, in research for future practical use, the more attractive form of g-C₃N₄ has become the g-C₃N₄ nanosheets due to its improved properties such as higher photocatalytic activity or enhanced photoluminescence compared to bulk g-C₃N₄ material. Until now, g-C₃N₄ has been considered a nontoxic material, according to the following studies.

5.1. In Vitro Cytotoxicity Assessment on Human Cells

Thin nanosheets of g-C₃N₄ are considered a promising candidate within the field of bioimaging based on their various favourable attributes including enhanced photoluminescence, metal-free nature, and relatively low-cost preparation. An important step in researching the practical use of materials in bioimaging involves assessing cell toxicity. A summary of biocompatibility tests for g-C₃N₄ can be found in Table 6. Zhang et al. reported that ultrathin g-C₃N₄ nanosheets, prepared by ultrasound-assisted liquid exfoliation of bulk g-C₃N₄ in water, showed no evident reduction of HeLa cell viability even at a high incubation concentration of 600 μg/mL. The size distribution of tested exfoliated g-C₃N₄ was assessed by atomic force microscopy imaging, revealing a range from 70 to 160 nm, centred around 100 nm, and a thickness of nanosheets around 2.5 nm. The biocompatibility of prepared ultrathin nanosheets was further validated through a bioimaging test using an excitation wavelength λ = 405 nm, while the activity and morphology of HeLa cells were not disrupted [77].

A comparison of bulk material toxicity with the toxicity of nanosheets was tested on A549 cells, a lung epithelial carcinoma cell line, by Duan et al. The thickness of g-C₃N₄ nanosheets with a porous structure prepared by thermal exfoliation at 500 °C was determined as 1.0 nm. As the concentration of tested photocatalyst increased from 5 up to 100 μg/mL, the percentage of cell viability decreased after a 24-h incubation period irrespective of whether bulk or nanosheet material was used. Interestingly, g-C₃N₄ nanosheets were less destructive to A549 cells as the decrease in cell viability with increasing concentration was lower than bulk material. Even at the highest tested concentration of g-C₃N₄ nanosheets, the decrease in the A549 cell viability after 24 h was relatively low, which suggests good biocompatibility of g-C₃N₄ nanosheets toward tested cells within the tests [78].

Another study investigating the toxicity of bulk and g-C₃N₄ nanosheets was conducted by Davardoostmanesh et al. Nanosheets of a size under 100 nm were prepared by electrophoresis fractionation of bulk g-C₃N₄ synthesised from urea. The cytotoxicity tests of prepared materials were performed on Saos-2, a bone carcinoma cell line, and foreskin fibroblast HFF cells. Bulk g-C₃N₄ exhibited lower toxicity against Saos-2 cells after 24 and 48 h of incubation compared to g-C₃N₄ nanosheets. Determined half maximal inhibitory concentrations (IC₅₀) were 66.8 ± 3.3 μg/mL and 27.0 ± 4.2 μg/mL for nanosheets and 214.4 ± 13.7 μg/mL and 104.0 ± 8.5 μg/mL for bulk after 24 and 48 h of incubation. Fluorescence microscopy revealed that g-C₃N₄ nanosheets were able to penetrate the tumour cells, distribute widely within the cytoplasm, and accumulate near the nuclear membrane. This finding was not observed with bulk material. There was no significant cell viability difference observed after treating normal HFF skin cells with 48-h IC₅₀ concentrations of g-C₃N₄ in bulk or nanosheet form. Based on these discoveries, g-C₃N₄ nanosheets were proved to be potential candidates in bioimaging, drug delivery, and also anticancer therapy [79].

Table 6. Summary of g-C₃N₄ biocompatibility tests on human and animal cells.

Material Type	Cell Line	Methods	Concentration	Results	Ref.
g-C3N4 nanosheets	HeLa	MTT assay	600 μg/mL	No cytotoxic effect even at a high concentration of g-C3N4.	[77]
Bulk g-C3N4	A549	MTT assay	5–100 μg/mL	Dose-dependent decrease in cell viability.	[78]
g-C3N4 nanosheets				Better biocompatibility compared to bulk material.
t-C3N4		MTT, WST-8 assay	25–500 μg/mL	t-C3N4 exhibited higher toxicity compared to h-C3N4.	[80]
h-C3N4
Bulk g-C3N4	Saos-2	MTT assay, fluorescence microscopy	12.5–200 μg/mL	48-h IC50 = 104.0 ± 8.5 μg/mL.	[79]
g-C3N4 nanosheets				48-h IC50 = 27.0 ± 4.2 μg/mL.
Bulk g-C3N4	HFF		48-h IC50	No cytotoxic effect.
g-C3N4 nanosheets
COOH-rich g-C3N4 nanosheets	MCF-7	CCK-8 assay, fluorescence microscopy	up to 400 μg/mL	No cytotoxic effect.	[81]
NH3-rich g-C3N4 nanosheets
g-C3N4 nanosheets	HaCaT	MTS assay, optical microscopy	3.125–500 μg/mL	No cytotoxic effect.	[82]
Ni-doped g-C3N4 nanosheets				24-h IC50 = 53.93 μg/mL.
Cu-doped g-C3N4 nanosheets				24-h IC50 = 157.00 μg/mL.
Cu-Ni-doped g-C3N4 nanosheets				24-h IC50 = 40.10 μg/mL.
g-C3N4 nanosheets	PC12	MTT assay	1–100 μg/mL	No significant cytotoxic effect under dark or white LED irradiation conditions.	[83]
Metal (Fe, Cu, Zn)-doped g-C3N4 nanosheets
Cu-doped g-C3N4 nanosheets combined with upconversion nanoparticles	4T1	MTT assay	15.63–500 μg/mL	High cytotoxicity toward tumour 4T1 cells after NIR laser irradiation.	[84]
	L929			No obvious cytotoxic effect.

Table 6. Summary of g-C₃N₄ biocompatibility tests on human and animal cells.

Material Type	Cell Line	Methods	Concentration	Results	Ref.
g-C3N4 nanosheets	HeLa	MTT assay	600 μg/mL	No cytotoxic effect even at a high concentration of g-C3N4.	[77]
Bulk g-C3N4	A549	MTT assay	5–100 μg/mL	Dose-dependent decrease in cell viability.	[78]
g-C3N4 nanosheets				Better biocompatibility compared to bulk material.
t-C3N4		MTT, WST-8 assay	25–500 μg/mL	t-C3N4 exhibited higher toxicity compared to h-C3N4.	[80]
h-C3N4
Bulk g-C3N4	Saos-2	MTT assay, fluorescence microscopy	12.5–200 μg/mL	48-h IC50 = 104.0 ± 8.5 μg/mL.	[79]
g-C3N4 nanosheets				48-h IC50 = 27.0 ± 4.2 μg/mL.
Bulk g-C3N4	HFF		48-h IC50	No cytotoxic effect.
g-C3N4 nanosheets
COOH-rich g-C3N4 nanosheets	MCF-7	CCK-8 assay, fluorescence microscopy	up to 400 μg/mL	No cytotoxic effect.	[81]
NH3-rich g-C3N4 nanosheets
g-C3N4 nanosheets	HaCaT	MTS assay, optical microscopy	3.125–500 μg/mL	No cytotoxic effect.	[82]
Ni-doped g-C3N4 nanosheets				24-h IC50 = 53.93 μg/mL.
Cu-doped g-C3N4 nanosheets				24-h IC50 = 157.00 μg/mL.
Cu-Ni-doped g-C3N4 nanosheets				24-h IC50 = 40.10 μg/mL.
g-C3N4 nanosheets	PC12	MTT assay	1–100 μg/mL	No significant cytotoxic effect under dark or white LED irradiation conditions.	[83]
Metal (Fe, Cu, Zn)-doped g-C3N4 nanosheets
Cu-doped g-C3N4 nanosheets combined with upconversion nanoparticles	4T1	MTT assay	15.63–500 μg/mL	High cytotoxicity toward tumour 4T1 cells after NIR laser irradiation.	[84]
	L929			No obvious cytotoxic effect.

In another study, Dong et al. examined and compared the cytotoxicity of g-C₃N₄ with triazine (t-C₃N₄) and heptazine (h-C₃N₄) structure moieties on the A549 cell line. According to the results obtained by MTT assay, t-C₃N₄ exhibits higher toxicity toward A549 cells compared to h-C₃N₄ across all tested concentrations ranging from 25 to 500 μg/mL. The same result was also achieved using the WST-8 assay after the same incubation time of 24 h as in the MTT assay. The suggested mechanism of t-C₃N₄ and h-C₃N₄ toxicity toward A549 cells via oxidative stress caused by ROS formation was confirmed by measurements of total glutathione in cells after incubation with both t-C₃N₄ and h-C₃N₄ at two different concentrations (50 and 100 μg/mL). It was proved that t-C₃N₄ induced higher levels of oxidative stress in A549 cells compared to h-C₃N₄, but both t-C₃N₄ and h-C₃N₄ showed significantly lower toxicity when compared to graphene oxide under similar experimental conditions [80].

Surface charge represents another important factor affecting the biological properties of g-C₃N₄. The relation between g-C₃N₄ surface charge and toxicity towards MCF-7 breast adenocarcinoma cells was investigated by Huang et al. Alteration in surface charge can be achieved through functionalisation via chemical oxidation to produce g-C₃N₄ rich in carboxyl groups (Cg-C₃N₄) or by chemical hydrolysis to create a nitrogen-rich derivate with amino groups (Ag-C₃N₄). Despite using the different synthesis methods, the obtained nanosheets were similar in size with 40–50 nm diameters. Neither of the prepared materials showed any toxicity effect and thus a significant reduction in cell viability after 24 h up to the concentration of 400 μg/mL. Cellular uptake examination by fluorescence spectroscopy and field emission scanning electron microscopy (FE-SEM) revealed that at equal concentrations, the Ag-C₃N₄ exhibited faster rates of cellular entry, which resulted in a higher accumulation of Ag-C₃N₄ within the cells. Additionally, at the lower incubation concentrations of g-C₃N₄, a larger contrast in cellular uptake was observed, suggesting that cationic Ag-C₃N₄ displays a greater affinity for cellular interactions because of the existing electrostatic interaction with the cell membrane charged negatively [81].

One of the other methods used for modifying the g-C₃N₄ structure is metal doping or codoping. The effect on the toxicity of g-C₃N₄ catalyst after metal nanoparticle doping was studied by Pieta et al. Cytotoxicity of prepared g-C₃N₄ and Ni-, Cu-, and Cu-Ni-doped g-C₃N₄ nanosheets, with sizes ranging from 1.1 to 1.5 nm, was investigated on the HaCaT human skin cell line. Microscopic analysis revealed that HaCaT cells incubated for 24 h with metal-doped g-C₃N₄ catalyst samples exhibited dose-dependent changes in morphology, while there was no evident change in morphology of cells observed after incubation with pure g-C₃N₄ nanosheets, even at the highest concentration of 500 μg/mL. In addition to the examination of HaCaT cell morphology, the impact of catalysts on the cell viability/proliferation was also determined using the MTS assay. A statistically significant reduction in cell viability was noted at concentrations of 25 μg/mL in the Ni-doped sample, 50 μg/mL in the Cu-doped sample, and 12.5 μg/mL in the Cu-Ni-codoped sample. In contrast, g-C₃N₄ nanosheets without modifications did not show any cytotoxic effect on HaCaT cells. The calculated IC₅₀ values were 53.93 μg/mL for Ni-doped catalyst, 157.00 μg/mL for Cu-doped catalyst, and 40.10 μg/mL for Cu-Ni samples [82].

5.2. In Vitro Cytotoxicity Assessment on Animal Cells

In addition to the toxicity studies on human cell lines, a few studies have also examined the cytotoxic effects of g-C₃N₄ and its modifications on animal cells. Chung et al. studied the toxicity of g-C₃N₄ and metal (Fe, Cu, Zn)-doped g-C₃N₄ on rat pheochromocytoma PC12 cells. It was found that neither g-C₃N₄ nor its metal derivates caused a notable decrease in cell viability across the broad range of tested concentrations (0–100 μg/mL) under dark or under white LED irradiation conditions. Achieved results confirm good biocompatibility of g-C₃N₄ and metal-doped g-C₃N₄, as well as the negligible impact of ROS generated after irradiation of photocatalysts on animal cells [83].

The cytotoxicity of pyrrolic nitrogen-rich Cu-doped g-C₃N₄ nanosheets, in conjunction with upconversion nanoparticles (UCNPSs) acting as artificial enzymes, was tested on mouse cancerogenic 4 T1 and healthy L929 fibroblast cells. No apparent toxic effects of the prepared enzyme-like nanocomposite were observed in L929 cells after 24 h of incubation across a wide range of concentrations (0–500 μg/mL). Due to its good biocompatibility and ability to be taken up by 4 T1 cells, the nanocomposite with g-C₃N₄ was further investigated as a promising photosensitive drug against cancerous 4 T1 cells. Following irradiation with a 980 nm laser at varying concentrations of g-C₃N₄ nanocomposite, the 4 T1 cells were eliminated, while NIR laser irradiation alone resulted in a high survival of cells, indicating that 980 nm laser irradiation alone was ineffective in killing the cells [84].

5.3. Ecotoxicity Tests on Microorganisms, Fishes, and Plant Seeds

Besides the tests evaluating toxicity on various types of cells, it is crucial to investigate the ecotoxicity of g-C₃N₄ on various organisms, especially considering that a significant amount of g-C₃N₄ application research is focused on water treatment and environmental remediation. Rosa et al. evaluated the acute toxicity of g-C₃N₄ and magnetic nanocomposite consisting of g-C₃N₄ and Fe₃O₄ nanoparticles on zebrafish (Danio rerio) embryos. Bulk g-C₃N₄ used in the study was prepared by calcination of the mixture of melamine and urea, and the magnetic nanocomposites were prepared by the coprecipitation method. In the case of prepared samples of g-C₃N₄ at all tested concentrations (0–100 mg/L), no statistically significant changes were observed in all monitored parameters indicating lethality, developmental variations, or malformations during the development of the zebrafish embryo–larva 24, 48, 72, and 96 h after fertilisation. The results indicate that both g-C₃N₄ alone and in combination with magnetic nanoparticles do not cause teratogenic or acute toxic effects on the embryos of zebrafish [85].

The acute toxicity of ternary nanocomposite consisting of ZnO, CdO, and g-C₃N₄ was evaluated in Nile tilapia (Oreochromis niloticus) by Berhanu et al. Based on the observed results, g-C₃N₄ in combination with ZnO and CdO semiconductors does not cause any change in experimental toxicological parameters at low exposure concentrations (10 mg/L). However, at the higher concentrations (500 and 1000 mg/L) a few hours after exposure, tested fish started to show signs of respiratory distress such as dyspnoea, swimming to the water surface, or rapid movement of the operculum. The determined 50% lethal concentration (LC₅₀) was at 12 h 113 mg/L [86].

Abdel-Moniem et al. analysed the toxicity of g-C₃N₄ and a nanocomposite consisting of Bi₂S₃ and g-C₃N₄ nanosheets using the Microtox model 500 analyser, capable of conducting in vitro tests of acute toxicity on the nonpathogenic bacterium Vibrio fisheri. The calculated effective 50% concentrations (EC₅₀-15 min) of g-C₃N₄ nanosheets and Bi₂S₃-x@C₃N₄ nanosheets were ≥100, which indicates that these materials are non-toxic, environmentally friendly, and suitable for water treatment processes. Prepared samples were also tested on various microorganisms, specifically Escherichia coli, Staphylococcus aureus, and Candida albicans, for disinfecting properties under dark conditions. Out of all tested materials, g-C₃N₄ nanosheets displayed the weakest antimicrobial properties against all selected microorganisms. Its higher antimicrobial activity was observed against E. coli, with the value almost reaching 80%. However, the antimicrobial activity against S. aureus and C. albicans was negligible, not even reaching 20%. Meanwhile, the antimicrobial activity of the g-C₃N₄ nanosheets modified with Bi₂S₃ nanosheets significantly improved, with the values almost reaching up to 100% toward all tested microorganisms [87].

The possible phytotoxicity of g-C₃N₄ toward rice seedlings (Oryza sativa L.) was evaluated by Hao et al. Tested g-C₃N₄, with an average size spanning several hundred nanometres and thickness of about 3–7 nm, was synthesised by urea calcination followed by freeze-drying. The obtained weights of rice biomass, treated with g-C₃N₄ (250 mg/kg), after 30 days of exposure showed increases of 29.9% (rice roots), and 22.4% (rice shoots) compared to untreated control rice seedlings. The utilisation of a low dose of g-C₃N₄ did not affect the rice seedlings’ growth and it even mitigated the phytotoxicity induced by heavy metals (As, Cd) present in the soil through adsorption of metals on g-C₃N₄, which caused inhibition of As and Cd distribution and accumulation in rice plants [88].

6. Conclusions and Future Perspectives

The prospects of g-C₃N₄ lie in its further development and improvement for a variety of applications such as photocatalysis, water splitting, environmental remediation, and even sensor applications. The sensitivity of g-C₃N₄ towards various analytes, including organic compounds and heavy metals, has been demonstrated, making g-C₃N₄ a suitable material useful in processes such as environmental monitoring, healthcare diagnostics, and food safety testing. Overall, while g-C₃N₄ is considered to be relatively safe for many applications, it is also essential to carefully evaluate the potential toxicity risks associated with its use (Figure 4) in sensing and electronics applications. The potential toxicity of these nanocomposites may depend on factors such as the size, shape, surface chemistry, and dose of the g-C₃N₄ nanoparticles, as well as the specific application and exposure scenario. Understanding whether g-C₃N₄-containing materials can be recycled or safely disposed of at the end of their lifecycle is important for sustainable product stewardship. Proper waste management practices should be followed to minimise the release of g-C₃N₄ nanoparticles into the environment and mitigate potential ecological risks. In biomedical applications, such as drug delivery systems or biomedical imaging, the potential toxicity of g-C₃N₄ nanoparticles to human cells or tissues needs to be thoroughly evaluated. This includes assessing cytotoxicity, genotoxicity, immunotoxicity, and any potential long-term health effects associated with exposure to g-C₃N₄ materials.

g-C₃N₄ is often used in photocatalytic coatings for environmental remediation purposes, such as air purification or water treatment. While g-C₃N₄ itself is considered to be relatively inert, there may be concerns about the release of nanoparticles, metals, or degradation products from these coatings into the environment, which could potentially impact human health or ecosystems. Once g-C₃N₄ is doped with heavy metals such as cadmium to enhance its photocatalytic activity for water purification, there is a risk of leaching of toxic metal ions into the environment during the catalytic process. These metal ions can pose significant environmental and health hazards due to their toxicity and potential for bioaccumulation in organisms. If released into water bodies, the metal ions could contaminate the aquatic environment, posing risks to aquatic organisms and potentially entering the food chain.

Understanding the mechanical properties of g-C₃N₄ is essential for ensuring the robustness, proper design, and optimal functioning of coatings, devices, and systems based on this material. The mechanical characterisation of g-C₃N₄ reveals a material with high stiffness, moderate hardness, and good wear resistance, suitable for a range of applications from coatings to structural components. However, its brittle nature and anisotropic properties need careful consideration in design and application. Researchers and engineers can optimise g-C₃N₄ for specific uses by understanding and leveraging these mechanical properties, ensuring robustness and longevity in practical applications. Exploring modifications, heterojunctions, novel composite structures, and nanostructures for enhancing g-C₃N₄ efficiency will allow the expansion of its practical applications in sustainable energy and environmental management. Overall, the perspectives of g-C₃N₄ are vast and multidisciplinary. Continued research efforts aimed at the examination of its properties and synthesis methods as well as toxicity investigations are necessary for further progress.