Top-down techniques, often referred to as physical methods, are used to fragment bulk materials into nanostructured materials. Several methods include laser ablation and arc discharge, mechanical milling, electrospinning, lithography, sputtering, electrochemical exfoliation and ultrasonic assisted.

In bottom-up approaches, nanostructured particles are formed through the controlled assembly of small molecules and atoms.

CQDs synthesised from green natural precursors or biomass sources have gained significant interest, establishing this approach as a sustainable and environmentally friendly synthesis method. CQDs are mainly structured by the core structural component, carbon, which contributes to their abundant surface functional groups and their overall stability [54]. The carbon-rich precursors, such as green natural and biomass-derived materials being the most prominent contributors [55], acting as the source of carbon when incorporated with top-down or bottom-up synthesis methods. Table 1 summarises the studies reported with the utilisation of various precursor sources with different synthesis method. However, despite the abundance and accessibility of these precursors, CQDs often exhibited variability in particle size, showing the importance of further optimisation for reproducible outcomes. The variation in precursor carbon content and chemical composition also influence the variations in surface groups, structural features, and properties of CQDs synthesised.

After synthesis with various precursors, the yield contains impure CQDs with the presence of by-products, unreacted precursors, and impurities which might interfere with the characterisation, potentially leading to experimental error [79]. This might explain the wide variation in the CQD sizes observed in certain precursors. Hence, purification techniques are essential for the isolation of pure CQDs from other unwanted products to ensure purity, while also narrowing down overall size distribution.

Dialysis is a common technique that utilises membranes with molecular weight cut-offs ranging from 0.5 to 3.5 kDa to purify CQDs. Dipcin et al. [80] demonstrated that three cycles of centrifugation followed by 3 days of dialysis efficiently removed impurities while preserving the surface functional groups of CQDs [80]. Similarly, González-González et al. [81] demonstrated that dialysis effectively eliminates residual contaminants [81], ensuring that the physicochemical properties of the CQDs can be assessed accurately without the interference of the contaminants. Raeispour et al. [82] also reported that the use of a dialysis membrane promotes uniform particle size distribution [82], which may otherwise vary due to batch-to-batch variation of precursors [79].

Although purification effectively removes by-products and ensures the production of CQDs with uniform size and consistent properties, achieving consistency during large-scale synthesis remains a challenge. Scaling-up often results in variation depending on the synthesis method employed, and the reaction conditions, which affects size, properties, and yield of CQDs [81]. This limitation can be addressed through continuous flow synthesis, in which precursors are continuously delivered at a fixed rate via a high-pressure pump into a heated reaction zone, followed by product collection through a condenser. This configuration enables scalable and reproducible production of CQDs in larger quantities [8].

Supajaruwong et al. [84] demonstrated the effectiveness of a continuous hydrothermal synthesis approach, integrating a microreactor system with a liquid chromatography pump that applies pressure [84]. The flow rate was adjusted by optimising the reaction temperature and time, obtaining the most suitable parameters to produce CDs with moderate temperature and pressure within a short reaction duration [84]. This method uses optimised reaction parameters to ensure large-scale production with reduced energy consumption, whilst minimising batch-to-batch variation compared with conventional synthesis techniques [85]. Therefore, issues related to varying size distribution and low production efficiency can be mitigated through the integration of effective purification and scalable synthesis strategies.

CQDs have attracted significant interest as bioimaging agents, largely due to their tuneable PL, which facilitates real-time, minimally invasive cellular imaging through confocal and fluorescence microscopy [2]. Their UV-Vis absorption and emission properties can be tailored through surface passivation and functionalisation strategies, allowing modulation of fluorescence intensity and emission wavelength, including advantageous, red-shifted outputs for bioimaging [86]. Emission across the visible spectrum, spanning blue to red fluorescence, is generally attributed to surface defect states associated with carbon-oxygen functional groups in addition to size-dependent optical properties [87, 88]. Common approaches to modulate these emissive features involve adjusting synthesis parameters such as reaction temperature and precursor composition [89], as well as by heteroatom doping strategies that alter the electronic structure of the carbon core [90].

Beyond imaging alone, CQDs are increasingly explored as multifunctional theranostic platforms that integrate diagnostic and therapeutic capabilities. As highlighted by Daby et al. [91], their tuneable PL enables high contrast fluorescence imaging, however, challenges relating to in vivo stability, biodistribution, and long-term safety remains key barriers to clinical translation [91]. At the subcellular level, Liu et al. [92] demonstrated that organelle-targeting CQDs could preferentially accumulate within mitochondria, the Golgi apparatus, and the endoplasmic reticulum, illustrating how variations in precursor composition and surface functional groups can influence intracellular localisation and enhance organelle-specific imaging contrast [92].

CQDs have also been widely explored in cancer imaging. N-CQDs synthesised from lemon precursors by Tadesse et al. [93] exhibited bright intracellular fluorescence and good cytocompatibility in MCF-7 breast cancer cells, supporting their application in live cell imaging [93]. Similarly, auxin-derived CDs developed by Noorkhajavi et al. [94], also exhibited strong 1 intracellular fluorescence in murine 4T1 cancer cells, with uptake inhibition studies suggesting the involvement of receptor-mediated uptake, thereby highlighting their potential for targeted tumour visualisation [94]. Complementary work using green-derived CQDs synthesised from bread waste demonstrated efficient cellular internalisation and widespread cytoplasmic distribution in HT-29 and CT-26 colon carcinoma cells, enabling clear visualisation of subcellular features under fluorescence microscopy [95].

In addition to imaging, CQDs have shown promise as multifunctional carriers for biomolecular delivery. Several studies report the formation of stable CQD-Ct-DNA complexes capable of intracellular delivery in L929 and MCF-7 cell lines, allowing fluorescent monitoring of gene transport, moreover, suggesting potential applications as non-viral vectors for bioimaging-guided gene transport [96, 97]. More recently, red-emissive CQDs synthesised from Echinophora tenuifolia were shown to selectively stain non-viable cells with high fluorescent intensity and photostability, a property attributed to their negatively charged surface favouring interactions with compromised cellular membranes, sparing viable cells [98]. Taken together, these studies highlight how precise control of surface chemistry and emission properties can be leveraged to direct cellular uptake and intracellular localisation, reinforcing the potential of CQDs as versatile nanomaterials for diagnostic and therapeutic applications.

From a materials design perspective, the suitability of CQDs for bioimaging is closely linked to synthesis-dependent control over particle size, surface passivation, and heteroatom incorporation. Bottom-up synthesis approaches, particularly when combined with N or S doping, tend to produce CQDs with enhanced PL efficiency and improved emission stability. Effective surface passivation minimises non-radiative energy loss and photobleaching, which offers a rationale for the continuous fluorescence behaviour frequently reported in imaging studies. These synthesis-dependent optical features collectively govern the strong imaging performance of CQDs at the cellular and subcellular levels. Figure 4 provides representative examples of CQD bioimaging in both in vitro and in vivo models, demonstrating efficient cellular uptake and selective tumour accumulation.

CQDs in drug and gene delivery demonstrated high effectiveness as carriers due to their non-toxic nature, biocompatibility, PL, small size and large surface area for rapid cellular uptake through binding or adsorption without affecting the drug’s activity, making them a convenient choice [2, 9]. The acidic environment of the targeted site breaks the bond between CQDs and drug, resulting in controlled release [2]. Free or unconjugated CQDs are excreted through hepatobiliary or renal system [2]. Doxorubicin a common Food and Drug Administration (FDA) licensed chemotherapy drug inhibits the growth of cancer cells by utilising enzymes from malignant cells DNA to accelerate DNA base pair intercalation [2]. For example, arginine-glycine-aspartic acid-GQDs used for drug delivery and targeted imaging purposes is an effective conjugation of doxorubicin as GQDs are able to accurately assess cellular absorption and release of doxorubicin [9]. The hydrogen bonding between GQDs and doxorubicin was shown to affect drug release. This conjugation demonstrated that its cytotoxic activity was significantly higher than of free doxorubicin when tested against U251 glioma cells [9].

In addition, Liu et al. (2012) [99] disclosed the encapsulation of PEG-modified nanographene oxide with the aromatic anticancer drug SN38, resulted in approximately 1000-fold greater anticancer efficacy compared with the FDA-approved treatment for colon cancer [9]. In addition, CQDs encapsulated with garlic extract within alginate beads showed enhanced surface loading, with coated CQDs achieving more than 60% higher garlic extract retention compared with uncoated alginate beads [9]. The drug released system was controlled by pH with its therapeutic effect to be found stimulating [9]. In various cancers and metabolic disorders, mitochondria play a crucial role, hence CQDs produced from precursors such as chitosan through hydrothermal synthesis can be utilised for targeted delivery to the nucleus and mitochondria, demonstrating long-term mitochondrial imaging capability [100]. Collectively, these studies suggest that the performance of CQDs in drug delivery is not determined by nanoscale size or intrinsic PL alone but is largely governed by how the particles are synthesised.

Hydrothermal and other bottom-up synthetic routes allow surface chemistry and precursor composition to be tuned during formation, enabling the incorporation of functional groups, heteroatoms, and biomimetics. These synthesis-dependent choices directly influence drug loading capacity, pH-responsive release, organelle targeting, and cellular uptake, highlighting that therapeutic efficacy arises from deliberate surface and structural engineering, rather than from the carbon core alone. In biotechnology, gene therapy plays a significant role by correcting aberrant gene expression and is regarded a long-lasting and curative therapeutic strategy [2, 9]. This technique is able to transfer genetic materials within the cells through carriers, thus attracting significant interest in non-viral nanocarriers [2]. Various classes of nanoparticles have therefore been explored to enable targeted gene transport and intracellular delivery [9].

CQDs are particularly attractive in this context due to their ease of surface functionalisation with components such as cell-penetrating peptides and targeting ligands, simultaneously enabling image-guided tracking [2, 101]. A study demonstrated the formation of stable DNA/polymer carbon dot (PCD) complexes, where DNA compaction within CDs was confirmed by agarose gel electrophoresis [9, 100]. Gene expression was visualised through green fluorescence from enhanced green fluorescent protein, while the blue fluorescence of PCDs confirmed their utility as intrinsic makers for gene delivery and bioimaging [9, 96]. Similarly, multicolour fluorescence was observed from polyethylenimine-functionalised CQDs in COS-7 cells, highlighting their multifunctional potential in gene delivery, bioimaging, and cellular labelling [2, 100]. Pandey et al. (2021) [102] further reported that CDs synthesised from citric acid or β-alanine, carrying a positive surface charge, enabled efficient plasmid DNA delivery into Escherichia coli (E. coli), resulting in improved transfection efficiency and targeted gene delivery.

In summary, CQDs have shown extraordinary potential as carriers for both drug and gene delivery due to their physiological properties, and controlled release behaviour, combined with efficient excretion pathways, making them promising platforms for targeted therapeutic applications. The general mechanism of CQD-facilitated targeted drug delivery is illustrated in Figure 5.

Biosensors can be described as analytical devices that measure changes in biological processes and convert them into a readable signal, such as electrical signals [103]. Biosensors can be used in disease detection and prevention, rehabilitation, and health management [104]. CQDs are a group of fluorescent nanoparticles characterised by their tiny size, under 10 nm, with structures that vary depending on their synthesis methods [21, 105]. Moreover, the numerous functional groups on their surface (hydroxyl, amino, carboxyl) contribute to their water solubility and biocompatibility [21]. They also exhibit excellent polymerisation capabilities, allowing interactions with various biologically active, organic, and inorganic materials [21]. Surface passivation further improves their physical robustness and photoluminescent efficiency [21]. Collectively, these features make CQDs particularly suitable for integration into biosensing platforms and diagnostic systems.

Among the exploited mechanisms in CQD-based biosensing is fluorescence quenching. Fluorescence quenching refers to the loss or reduction in the fluorescence of CQDs after reacting with quenchers [24]. Quenching occurs through non-reactive decays linked to mid-gap states in the CQD band gap energy as they dissipate excitation energy as heat, preventing electron-hole recombination for the generation of fluorescence [24]. Further, photoinduced electron or energy transfer, whereby CQDs function as electron acceptors or donors [24]. For example, iron ions (Fe³⁺) have shown to strongly inhibit CQD fluorescence by accepting photoexcited electrons [24, 106]. Khan et al. [107] have reported a negative linear relationship between CQD fluorescence intensity ratio and Fe³⁺ concentrations [107]. These responses make them beneficial for use in highly sensitive optical biosensors for selective monitoring of metal ions.

Additionally, fluorescence of N-CQDs have shown to be selective fluorescence quenching effects upon interaction with dopamine molecules, due to the electrostatic reaction between CQD surface groups and the amino groups of dopamine molecules, showing potential for dopamine sensing in biological fluids [108]. Surface defects and aggregation are other factors which could contribute to the fluorescence quenching of CQDs due to unwanted electronic states of higher density which trap photo-excited carriers, enhancing non-radiative decays [108]. For instance, solvo-thermally synthesised yellow emissive CQDs have been used to quantify bilirubin in human urine and serum [109].

Furthermore, combining CQDs with rhodamine 6G can be used to monitor glucose levels [110], where a change in colour from blue to green indicates an increase in glucose levels and has good selectivity for glucose over the major components in human blood and can be used with serum [110]. These biosensors increase the accessibility to detect glucose using the naked eye [110]. Yu et al. [111] reported that combining CQDs with glucose oxidase and cellulose acetate complex sensitive film leads to high selectivity, reproducibility and anti-interference ability for real-time detection of low concentrations of glucose. By utilising the fluorescence quenching ability of the film, the biosensor can manifest a rapid fluorescence response to glucose [111].

Moreover, CQDs can act as an electrochemical biosensor, which effectively detects and measures the movement of electrons to determine the analyte [112]. CQDs act as simple immobilising compounds for the development of enzymatic biosensors due to their high surface-to-volume ratio and adaptive nature and can be enhanced using redox-active enzymes [112]. Liu et al. [113] developed an electrogenerated chemiluminescence (ECL) biosensor by combining N-CQDs with DNA to detect microRNA (miRNA)-21 [113]. Whereby an initial DNA sequenced was utilised to conjugate CQDs to aid in the hybridisation of miRNA and the assistance probe [113]. The nicking enzyme identified this complex and cleaved the DNA, releasing the miRNA to engage in more hybridisation cycles [113].

After DNA cleavage, the miRNA hybridises with a detection hairpin immobilised on a graphene oxide electrode, which produces an ECL signal that can be quantitatively measured [112, 114, 115]. Furthermore, mixing nanomaterials with CQDs could improve their properties [112, 114, 115]. Pourmadadi et al. [115], developed an electrochemical aptasensor to determine prostate-specific antigen (PSA) using CQDs and gold nanoparticles [115]. The interaction between the nanomaterials showed enhancement in the electrical conductivity of CQDs following the rise in the current peak of cyclic voltammetry (CV) in COQ-gold nanoparticles compared to CQDs [115]. In addition, CQDs can enhance ECL performance by promoting electron transfer and providing a stable interface for biomolecule immobilisation using TiO₂-CQD medium, where the medium’s pH value could be altered to measure total-PSA and free-PSA separately. CV was used to examine the changes, showing a linear relationship between the concentration of the sample solution and the resistance of the electron transfer on the surface of the electrode [114].

Taken together, these biosensing studies demonstrate that CQD sensing performance is predominantly dictated by synthesis-driven control over surface chemistry and defect states rather than optical properties alone. Heteroatom doping, surface passivation, and the deliberate introduction of oxygen- and nitrogen-containing functional groups during synthesis generate reactive surface sites that govern analyte binding, electron transfer, and signal transduction. Consequently, sensitivity and selectivity across fluorescence, electrochemical, and ECL-based biosensors emerge from rational surface and defect engineering at the synthesis stage, underscoring the tight coupling between synthetic design and biosensing functionality, which is also directly extended to the detection of ROS.

ROS are reactive forms of oxygen which include superoxide anion radical, hydrogen peroxide, singlet oxygen and hydroxyl radical, they are usually produced during normal metabolism of oxygen inside the mitochondria [116]. Peroxynitrite, a kind of ROS in which its overproduction is linked to various diseases. Zhou et al. [117] developed a strategy to prepare CDs with various oxygen-rich surface groups that could selectively detect peroxynitrite in living cells, showing a linear chemiluminescence response and exhibiting efficient cellular uptake, shown by emission of strong green PL in cells after co-incubation. Stimulation with SIN-1 or PMA can enable real-time monitoring of both exogenous and endogenous peroxynitrite production [117]. Since CQDs possess structural and optical features similar to CDs, it is anticipated that they will exhibit comparable performance in detecting ROS.

Photodynamic therapy (PDT) uses specific wavelengths of light to activate photosensitisers (PSs), which can be derived from natural or synthetic compounds [103]. The photosensitisers produce ROS that kill tumour cells. However, oxygen dependence of PDT results in limitation of its efficacy against hypoxic tumours. Further, conventional PSs have low stability, poor solubility and potential damage to neighbouring tissues [103]. To overcome these limitations CQDs can be encapsulated with hematoporphyrin (HP) (HP-CQDs) using HP monomer as a precursor [118]. The product retained optical and chemical properties of HP with significant improvement in water solubility [118]. HP-CQDs can generate ROS effectively under red light and has enhanced effectiveness in PDTC against MCF-7 human breast cancer cells [118]. In comparison with HP alone, HP-CQDs have higher phototoxicity and lower dark toxicity, allowing them to be more effective in targeting cancer cells [118].

Various CQD-based systems have shown potential anticancer effects. Li et al. [119] synthesised CQDs from tender ginger juice which inhibited tumour growth in mice within 14 days [119], whereas He et al. [120] developed Diketopyrrolopyrrole-based CDs which are highly photostable and capable of strong cellular uptake and tumour suppression at low concentrations [120]. CDs containing porphyrin are minuscule in size, have good water solubility and are photostable [121]. They generate singlet oxygen under irradiation which induces cell apoptosis, inhibiting the growth of hepatoma. Beack et al. [122], synthesised CQD-chlorine e6-hyaluronate conjugates that generated singlet oxygen at higher efficacy compared to free chlorine e6, leading to complete suppression of B16F10 murine melanoma cells following laser irradiation [122]. Through these observations, we can conclude that different CQD compositions cam be used as potent anticancer agents upon visible light irradiation.

Other than the generation of ROS from CQDs, the antibacterial mechanism of CQDs have been recorded such as DNA binding, membrane destabilisation, physical and mechanical damage and inhibition of bacterial metabolism. Thus, preventing bacteria from developing resistance [121]. Positively charged or N-CQDs interact electrostatically with the negatively charged bacterial membranes (lipopolysaccharides, lipoteichoic acids), leading to enhanced bacterial adhesion and antibacterial activity [123]. N-CQDs synthesised via a one-step chemical route affected the cell structure of Staphylococcus aureus (S. aureus) and methicillin-resistant S. aureus (MRSA) but were less effective against Escherichia coli (E. coli) [124].

CQDs can also be encapsulated into polymer films and hydrogels to form potent photodynamic antibacterial surfaces [121]. These components demonstrate strong antibacterial activity against S. aureus, E.coli and Klebsiella pneumoniae under blue light irradiation. Gamma ray irradiation of these components induced change in morphology and chemical composition which contributed to increased antibacterial activity towards S. aureus and E. coli [125]. Moreover, CQD thin films exhibited bactericidal effects against MRSA and Pseudomonas aeruginosa [126].

In summary, these findings highlight CQDs as potential light-activated antimicrobial agents, effective at preventing and treating infections. Their modifiable surface properties, biocompatibility and ROS-generating capabilities make them promising candidates for photoactivated disinfection, antibacterial coatings, and wound-healing applications, especially in environments where antibiotic resistance remains a growing concern [121].

Furthermore, CQDs have demonstrated significant potential as multifunctional agents in cancer therapy acting both as a nanocarrier as well as a photosensitiser for PDT [127]. For instance, carbon quantum dots clathrates loaded with methotrexate, conjugated with folic acid receptors, enabled targeted drug delivery to tumour cells. Upon near infrared laser irradiation (1064 nm), the formation of ROS and increased release of medication into cancer cells were observed.

Shahshahanipour et al. [128], developed a method of synthesising CQDs using Lawsonia inermis (henna) via hydrothermal procedure, where the CQDs demonstrated excellent stability and fluorescence characteristics, without surface functionalisation. Their use as biocompatible fluorescent probes for selective methotrexate detection in human serum via Förster resonance energy transfer [128]. Another study demonstrated CQDs loaded with doxorubicin or combined with multifunctional platforms (i.e., FeN@CQDs with folic acid and riboflavin, or SWCNT-PEG-Fe₃O₄-CQDs-doxorubicin conjugates) that demonstrate combined chemotherapy and PDT effects with increased specificity toward cancer cells [127], further elucidating the theranostic potential of CQDs, where their fluorescence allows imaging and tracking, while their capacity to deliver drugs and generate ROS under light exposure enhances their therapeutic efficacy.

Cumulatively, these studies indicate that the therapeutic performance of CQDs in photodynamic and photothermal applications is fundamentally governed by synthesis-driven modulation of photophysical properties rather than by the carbon core alone. Heteroatom co-doping, defect engineering, and hybridisation with photosensitisers or metallic components enhance light absorption, charge separation, and ROS or heat generation, directly influencing treatment efficacy and tissue penetration. However, increasing structural and compositional complexity also introduces challenges related to batch-to-batch reproducibility, long-term biosafety, and regulatory assessment. Consequently, balancing phototherapeutic performance with synthetic simplicity and reproducibility remains a critical consideration for advancing CQD-based systems toward clinical translation.

Emerging evidence indicates that CQDs can exert antiviral effects across diverse viral systems, although translational relevance varies by model. In a plant virology model, Farooq et al. [129] demonstrated that cysteine-functionalised CQDs were shown to reduce begomoviral titres and symptom severity in Nicotiana bethamiana, concurrently modulating immunity-related gene expression, including pathways associated with endocytosis and sphingolipid signalling [129]. Although this work highlights antiviral potential in a botanical host, it is important to note that its applicability to mammalian antiviral systems is limited. In contrast, coal-derived CQDs exhibited potent inhibition of SARS-CoV-2 replication in vitro, achieving more than 95% reduction in viral growth at half-maximal inhibitory concentration (IC₅₀) of approximately 5.47 μg/mL, with minimal cytotoxicity in normal cell lines, indicating genuine antiviral activity in a human-relevant viral system [123].

Similarly, Chen et al. [130] reported that curcumin-derived CQDs demonstrated potent antiviral activity against the Japanese encephalitis virus by selectively binding to its surface envelope protein. Through binding key domains involved in host cell attachment and membrane fusion, these CQDs effectively disrupted early-stage viral infection, resulting in a marked reduction in viral entry and replication, suggesting that surface functionalisation can confer targeted disruption of viral-host interactions [130]. Consistent with these findings, recent reviews have identified recurring antiviral mechanisms among carbon-dot-based nanomaterials, including viral surface protein interference and redox-mediated viral inactivation, however, they also emphasis the lack of standardised synthesis, limited in vivo validation, and incomplete toxicological characterisation, which collectively constraint clinical translation [131].

CQDs have also demonstrated considerable promise as antibacterial agents, particularly when engineered to enhance redox activity or integrate into composite nanostructures. Chen et al. [132] reported that CQD-ZnO nanocomposites containing 7–14 wt% CQDs exhibited significantly enhanced antibacterial effects against Streptococcus mutans, Enterococcus faecalis, and E. coli compared with ZnO alone. This effect was attributed to increased ROS generation and bacterial membrane disruption [132]. Similarly, Shen et al. [133] synthesised N-CQDs from Solanum nigrum extract and reported minimum inhibitory concentrations of approximately 1.1–1.2 mg/mL against S. aureus and E. coli [133]. Mechanistic analyses revealed that bacterial inactivation was associated to hydroxyl radical generation, and electrostatic interactions between positively charged N-CQDs and negatively charged bacterial membranes, leading to morphological changes and cell wall damage [133].

At a broader level, a review by Collins et al. [134] highlighted the antimicrobial versatility of CQDs across foodborne and clinically relevant pathogens, noting common mechanisms such as oxidative stress induction, biofilm inhibition, and interference with bacterial metabolic pathways [134]. In addition, Ghacham et al. [135] reported that green-synthesised CQD-silver nanocomposites exhibited broad spectrum antibacterial activity alongside accelerated wound-healing, attributed to the synergistic effects of silver ion-mediated membrane disruption and CQD-driven oxidative stress [135]. Collectively, these studies support the versatile antibacterial potential of CQDs through various mechanisms, however, most evidence remains derived from in vitro systems, with limited in vivo validation and safety profiling, highlighting the need for further evaluation prior to clinical translation.

Beyond their antiviral and antibacterial applications, CQDs have also been explored for antifungal applications, although evidence remains comparatively limited. Belal et al. [136] reported that N-CQDs significantly reduced fungal counts in both in vitro assays and an in vivo rat model of mucormycosis, with antifungal effects attributed to ROS generation, disruption of fungal cell wall integrity, and increased membrane permeability [136]. While these findings suggest potent fungicidal activity, the study represents a preclinical investigation, and further validation in larger, clinically relevant models are required to establish therapeutic relevance. In relation, Slewa et al. [137] demonstrated that lemon-and onion-derived CQDs incorporated into biodegradable packaging films effectively inhibited the growth of Rhizopus spp., Penicillium spp., Candida albicans, Aspergillus spp., and Botrytis cinerea, highlighting the function of CQDs in surface-associated antifungal applications in addition to their therapeutic uses [137].

Overall, the available evidence supports CQDs as versatile antimicrobial nanomaterials capable of interfering with viral entry, bacterial membrane integrity, and fungal viability, largely through surface-dependent interactions and oxidative mechanisms. However, further work is required to standardise synthesis approaches, define safety profiles, and validate efficacy in physiologically relevant models to support future clinical translation. The antiviral, antibacterial, and antifungal activities of CQDs are strongly governed by synthesis-driven modulation of surface chemistry, charge, and defect density rather than by the carbon core alone.

Heteroatom doping, surface functionalisation, and composite formation enhance electrostatic interactions with microbial membranes, promote ROS generation, and enable targeted disruption of viral surface proteins or microbial cell walls. However, while increasing structural complexity can amplify antimicrobial potency, it also raises challenges related to reproducibility, biosafety, and the lack of standardised toxicological evaluation. Consequently, translating CQD-based antimicrobial systems toward clinical or applied use will require careful balancing of antimicrobial efficacy with controlled synthesis, rigorous safety profiling, and validation in physiologically relevant in vivo models. Figure 6 summarises the proposed antimicrobial mechanisms of CQDs, alongside representative in vitro antibacterial, and in vivo wound-healing evidence.

The blood-brain barrier (BBB) represents a major translational obstacle in the treatment of neurological diseases and remains one of the greatest challenges in the development of effective neurotherapeutics [138]. Formed by tightly apposed endothelial cells supported by pericytes and astrocytic end-feet, the BBB restricts the entry of most small molecules, biologics, and nearly all gene-editing constructs [139]. While essential for maintaining cerebral homeostasis, this barrier concurrently limits pharmacological interventions, often resulting in sub-therapeutic drug concentrations [140]. In response to these limitations, CQDs have emerged as promising candidates for central nervous system (CNS) delivery, owing to their ultrasmall size, tuneable surface chemistry, and potential to engage endogenous transport mechanisms [141].

Seven et al. [142] reported that glucose-derived CDs successfully crossed the BBB transporting small molecular cargo in both murine and zebrafish models, providing evidence of in vivo BBB permeability [142]. While glucose transporter-mediated uptake was experimentally validated in budding yeast, the specific mechanism underlying BBB translocation in vertebrates remains inferred rather than directly demonstrated. The authors noted that glucose transporter expression is upregulated following neurological injury, suggesting potentially enhanced delivery potential [142]. However, transporter involvement in physiologically relevant BBB models remains to be mechanistically confirmed, warranting further validation in mammalian systems. Recent reviews indicate that suitably engineered CQDs can traverse the BBB via receptor-mediated endocytosis and adsorptive-mediated transcytosis, with surface charge and amphiphilicity influencing endothelial interactions and uptake efficiency, although the relative contributions of these pathways are not yet well defined [143]. Consequently, careful surface engineering is critical in achieving reliable and translatable CNS delivery.

Protein misfolding and aggregation underlie major neurodegenerative diseases such as Alzheimer’s (AD), Parkinson’s, and Huntington’s disease, further characterised by neuronal loss, mitochondrial dysfunction, toxic fibril formation, and oxidative stress. Guerrero et al. [144] demonstrated that sodium-citrate CQDs could prevent hen egg-white lysosome (HEWL) from forming mature amyloid fibrils, a common model of amyloidogenic protein aggregation [144]. In addition, these CQDs disaggregated mature fibrils with low cytotoxicity, indicating their potential for both prophylactic and therapeutic intervention in protein-misfolding disorders [144]. Similarly, Mukherjee et al. [145] showed that microwave-synthesised CQDs significantly reduced HEWL amyloidogenesis by delaying nucleation, inhibiting b-sheet conversion, and disrupting protofibril elongation, ultimately yielding fewer and shorter aggregates with reduced cytotoxicity [145].

While these studies highlight the neuroprotective potential of CQDs in controlled assays, HEWL is a model protein with limited structural homology to disease-relevant aggregates, and the in vitro assays do not fully reflect neuronal environments. Therefore, translation to human neurodegenerative diseases remain uncertain. Lim et al. [146] further reported that curcumin-derived CQDs inhibited amyloid-b fibril aggregation and the associated oxidative stress in AD models via interactions with b-sheet domains, preventing peptide self-assembly, and concurrent curcumin-mediated free-radical suppression [146]. Overall, while these findings demonstrate promising neuroprotective effects, the pharmacokinetics, long-term biodistribution, and potential off-target toxicity of CQDs remain largely unexplored, highlighting the need for more relevant studies prior to clinical application.

Crossing the BBB is particularly advantageous in malignancies such as glioblastoma multiforme (GBM), where diffuse infiltration and chemoresistance hinder surgical and pharmacological management. GBM constitutes the most malignant subtype of glioma and the most prevalent primary brain tumour in adults, accounting for 45.2% of cases [147]. Yan et al. [145] developed paclitaxel-derived carbon quantum dots (PTX-CDs) which penetrated an in vitro BBB model, achieving a reported permeability of 19.8%. In an orthotopic glioblastoma mouse model, PTX-CDs accumulated within intracranial tumours and exerted antitumour effects through ROS-mediated oxidative stress [145]. Mechanistic analysis using endocytosis pathway inhibitors indicated that PTX-CDs predominantly traverse the BBB via macropinocytosis, highlighting the involvement of energy-dependent vesicular uptake mechanisms in BBB translocation [148].

At the cellular level, PTX-CDs localised to mitochondria, disrupted mitochondrial membrane potential, and induced oxidative stress-mediated cytotoxicity [148]. However, while this dual mechanism of drug delivery and ROS-generation enhances antitumour efficacy, the potential for off-target oxidative damage to healthy brain cells and associated neurotoxicity was not comprehensively evaluated, which may in-turn limit translational interpretation. Similarly, Algarra et al. [149] showed that 2-acrylamido-2-methylpropanesulfonic-acid-derived CDs, effectively delivered riluzole to GBM cells, significantly reducing cell viability while exhibiting negligible intrinsic toxicity, thereby highlighting their biocompatibility and therapeutic stability as nanocarriers [149]. Complementing this, Hettiarachchi et al. [150] reported that transferrin-conjugated CDs co-loaded with temozolomide and epirubicin produced markedly enhanced cytotoxicity across multiple GBM cell lines compared to single-drug formulations, illustrating the advantage of combined targeting and multi-drug delivery strategies [150]. The ultrasmall particle size of around 3.5 nm following conjugation further supported BBB uptake and intracellular delivery [150]. Collectively, these studies reinforce the versatility of CQDs, however comprehensive in vivo assessments of BBB transport, biodistribution, and long-term neurotoxicity are still required to substantiate clinical translational potential.

Accumulating evidence underscores the emerging neuroprotective potential of CQDs across diverse experimental models. Zhang et al. [151] reported that CDs derived from Crinis carbonisatus significantly reduced infarct volume whilst improving neurological function in a rat middle cerebral artery occlusion/reperfusion model. The effects were attributed to decreased BBB permeability, suppression of pro-inflammatory cytokines (TNF-a and IL-6), and modulation of excitatory and inhibitory neurotransmitter signalling [151]. Extending these findings, Mosalam et al. [152] demonstrated that hyaluronic acid-modified verapamil-loaded CQDs, exhibited enhanced neuronal uptake and neuroprotective efficacy in an in vitro amyloid-induced neurotoxicity model, suggesting transporter-mediated BBB translocation and enhanced intracellular drug distribution [152].

Complementary mechanistic insight is provided by recent work on free-radical-scavenging-nanoparticles, including CQDs, which highlights their ability to attenuate neuroinflammation and oxidative stress through ROS scavenging and suppression of lipid peroxidation, thereby positioning CQDs as both therapeutic nanocarriers and active redox modulators [153]. Collectively, these studies suggest that CQDs can be rationally engineered to traverse the BBB, deliver neuroactive agents, and directly modulate inflammatory and oxidative pathways. However, the predominance of in vitro and acute in vivo models highlights the need for systematic evaluation of long-term safety, biodistribution, and translational relevance in clinically relevant neurological settings [151–153]. A summary of various CQD systems engineered for BBB transport and neurotherapeutic applications is provided in Table 2.

Across these diverse biomedical applications, ranging from bioimaging and biosensing to drug delivery, phototherapy, antimicrobial action, and neurotherapeutics, a common design principle is revealed - the biomedical behaviour of CQDs is governed primarily by precursor choice and synthesis-driven surface engineering, rather than by the carbon core itself. Differences in precursor chemistry, heteroatom doping, and surface functionalisation consistently dictate optical properties, cellular uptake, subcellular localisation, and therapeutic performance. This modularity enables CQDs to be rationally tailored toward various objectives through deliberate design choices during synthesis. To consolidate these design-function relationships across biomedical contexts, Figure 7 provides a schematic overview of representative precursors and engineering strategies discussed in this section.

Recent studies have highlighted the favourable in vitro biocompatibility and functional versatility of plant-derived CQDs across various biomedical applications. Yalshetti et al. [65] reported that CQDs synthesised from Hibiscus rosa-sinensis exhibited moderate cytotoxicity in L929 fibroblasts with approximately 76% cell viability at 100 μg/mL, while showing minimal toxicity in HaCaT keratinocytes, supporting their selective safety profile for nanotherapeutic research [65]. Similarly, Spinacia oleracea-derived CQDs maintained high viability in human mesenchymal stem cells at lower concentrations, and displayed dose-dependent reductions at higher exposure concentrations, indicating their suitability for bioimaging and stem cell-related applications [154]. Nitrogen-doped CQDs (N-CQDs) synthesised from castor seeds demonstrated negligible cytotoxicity in WI-38 lung fibroblasts across a wide concentration range, highlighting their safety profile in prolonged cellular studies and pH-sensing applications [155].

In addition to favourable biocompatibility, several plant-derived CQDs have demonstrated functional activity in disease-relevant cellular models. For instance, artichoke leaf–derived CQDs showed effective dose-dependent anticancer activity in MCF-7 cells, with an IC₅₀ of 96.5 μg/mL, while retaining strong fluorescence properties, thereby demonstrating dual utility in bioimaging and cytotoxicity-based research [156]. In addition, Kamble et al. [157] synthesised red-fluorescent CQDs (R-CQDs) from Lawsonia inermis, which displayed minimal toxicity in both fibroblast and cancer cell lines and exhibited notable antioxidant activity, supporting their application as safe bioimaging probes [157].

Additionally, CQD-based platforms have also been adapted for intracellular sensing, illustrated by cobalt-conjugated N-CQDs, which functioned as a non-toxic fluorescent sensors for dichloroacetic acid in HuH-7 cells, enabling real-time pharmacological monitoring without compromising cell viability [158]. However, it is important to note that the majority of these investigations employ in vitro models, and predominantly plant-derived CQDs, which, although advantageous for initial biocompatibility screening, may not accurately represent their behaviour in the human body.

Other studies have highlighted the capacity of these CQDs to be tailored for targeted biomedical applications. Boron-doped CQDs (B-CQDs), namely Formulation 1 (F1) and Formulation 2 (F2), preserved greater than 86% viability across multiple cell lines, demonstrating organ-specific targeting capabilities relevant to bioimaging and drug delivery strategies [159]. Finally, Syzygium malaccense-derived CQDs (SM-CQDs) exhibited weak cytotoxicity in 3T3-L1 adipocytes, with an IC₅₀ of 444.5 μg/mL, while enhancing glucose uptake, reflecting their promise as biocompatible nanomaterials for metabolic and therapeutic applications [160]. Collectively, these findings indicate that carefully engineered CQDs offer low in vitro toxicity while offering broad functional potential in biomedical research and applications. Thus, while cellular studies indicate low toxicity and functional potential, further in vivo studies are required to assess distribution, metabolism, and clearance, as well as their long-term effects on organs and immune response (Table 3).

CQDs demonstrate organ-specific accumulation influenced by particle size, surface charge, and administration route. In murine models, Spinacia oleracea-derived CQDs showed safe distribution at concentrations up to 100 μg/mL with no histological organ damage, whereas higher doses elicited mild liver lymphocyte infiltration [154]. Additionally, B-CQDs administered intravenously localised in kidneys and lymph nodes, intraperitoneally in intestines, liver, heart, and spinal cord, and orally in liver, kidneys, and lymphatic tissue, with peak tissue accumulation occurring at 30 min, and rapid systemic clearance within 2 hours [159].

Furthermore, cobalt-conjugated N-CQDs demonstrated efficient penetration into neuronal and non-neuronal tissues in zebrafish, a well-established non-mammalian vertebrate model, without eliciting detectable apoptosis or oxidative stress [157]. Smaller (<6 nm), hydrophilic, negatively charged CQDs generally undergo efficient renal clearance, whereas larger or positively charged particles show partial retention in the reticuloendothelial system (RES), especially in the liver and spleen [161, 163]. Tumour-targeting CQDs preferentially accumulate in neoplastic tissues, minimising off-target organ deposition (Table 4).

Neurotoxicity evaluations, although limited, indicate that CQDs are generally non-neurotoxic at standard doses. Henriquez et al. [164] reported that citric acid-derived CQDs effectively mitigated oxidative stress in SH-SY5Y neuroblastoma cells, while also preserving dopaminergic neuronal integrity in paraquat-exposed models [164]. In a related line of investigation, Raut et al. and Güven et al. [158, 159] demonstrated that B-CQDs and cobalt-conjugated N-CQDs elicited no detectable oxidative stress, apoptotic signalling, or microglial activation within neuronal tissues [158, 159]. Extending this safety profile beyond the CNS, Chen et al. and Mishra et al. [163, 165] have reported that off-target toxicity in peripheral organs remained minimal, although they emphasised that excessively high doses or positively charged CQDs may still provoke ROS formation, mitochondrial dysfunction, or the activation of inflammatory pathways [163, 165].

Across these studies, CQD uptake, cytotoxicity, and organ accumulation are strongly influenced by surface chemistry and dose. Positively charged CQDs exhibit enhanced cellular uptake, but greater organ retention and potential oxidative stress, whereas negatively charged or neutral particles show lower tissue retention and minimal mitochondrial or oxidative perturbation. Most CQDs are rapidly cleared renally, limiting long-term retention, although doped or larger particles may persist in liver, spleen, or tumour tissue [161, 166].

The consistent evaluation framework includes in vitro cytotoxicity assays including MTT, CCK-8, and Trypan Blue, apoptosis and necrosis analysis, ROS quantification, haemolysis testing, and fluorescence imaging. In vivo assessments employ histopathology, serum biochemistry, organ fluorescence imaging, and multi-organ biodistribution analysis. Oxidative stress markers (HO-1, ROS), inflammatory markers (NF-κB), mitochondrial integrity, and tight junction proteins provide sensitive indicators of sub-lethal toxicity, guiding safe biomedical translation [162, 166].

Collectively, CQDs exhibit high biocompatibility, low cytotoxicity, and predictable biodistribution when engineered with the appropriate size, surface charge, and doping. Dose-dependence, surface chemistry, and renal clearance or RES clearance critically influence both therapeutic efficacy and potential off-target effects. Positively charged and metal-doped CQDs show heightened oxidative and inflammatory responses, while negatively charged or hydrophilic CQDs remain generally safe. Future investigations should address chronic exposure, long-term retention, neurotoxicity, and standardised reporting to ensure safe clinical translation.