Experimental and theoretical overview on the origin of photoluminescence in carbon nanodots

Abstract

Carbon nanodots (CNDs) are extensively explored for bioimaging, sensing, and optoelectronic applications because of their tunable photoluminescence (PL), elevated quantum yield (QY), facile synthesis, and biocompatibility. Nevertheless, despite their discovery in 2004, the origin of their optical characteristics remains ambiguous. Their structural variability results in several emission pathways, such as quantum confinement, surface states, and molecular fluorophores (MFs), which complicates the establishment of a unified structure–property relationship. Inspired by these emerging insights, this review provides a recent comprehensive experimental and theoretical overview of the origin of PL in CNDs over the past two decades. Special emphasis is placed on core-state emission, surface-state emission, MF contributions, and defect-mediated pathways and understanding how variations in environmental conditions modulate PL properties in CNDs. Furthermore, the growing integration of theoretical approaches, especially density functional theory (DFT), with experimental validation is highlighted to elucidate the structure–emission relationship. This review demonstrates that a synergistic theoretical and experimental framework is necessary for uncovering the PL mechanisms and advancing the controlled design of CNDs.

---

1. Introduction

The chemistry of carbonaceous materials has engrossed the scientific domain due to the essential role played by carbon in the origin and evolution of life on Earth. With the advancement of research, various allotropic forms of carbon have been identified, including fullerenes, carbon nanotubes (CNTs), and graphene. Among these, the accidental discovery of CNDs in 2004 marked a significant milestone in nanoscience.¹ However, the pioneering study by Sun et al. in 2006 laid the foundation for modern CND chemistry, igniting widespread interest and extensive research into these nanoscale materials.² Thereafter, CNDs have gained significant attention owing to their impressive resistance to photobleaching,^3,4 high QY,^5,6 tunable fluorescence emission,^7,8 outstanding water solubility,⁹ and excellent biocompatibility.^4,10 CNDs serve as an umbrella term for several classes of carbon-based nanomaterials, including graphene quantum dots (GQDs), and polymer dots (PDs). Carbon nanoparticles (CNPs), typically less than 10 nm in diameter, consist of a mostly sp²-hybridized graphene carbon core, frequently encircled by organic functional groups such as hydroxyl, carboxyl, carbonyl, thiol, and amine. In some cases, CNDs possess MFs on their surface, primarily responsible for their fluorescence. However, it is difficult to precisely understand the origin of emission in CNDs due to their structural diversity arising from the wide range of carbon precursors, solvents, temperature, and synthetic methodologies employed. These factors strongly influence their physicochemical and optical properties.

Therefore, numerous theoretical investigations have been conducted to support the experimental findings and elucidate the physicochemical and optical characteristics of CNDs. To modify the optical characteristics in accordance with the application, functional group manipulation and heteroatom doping are being carried out.^11,12 These modifications are sometimes crucial for tuning the optical behavior of CNDs and enhancing their biocompatibility with biological systems. Consequently, CNDs have been found in various applications across numerous domains, including bioimaging,¹³ sensing,¹⁴ photocatalysis,¹⁵ light-emitting diodes (LEDs),¹⁶ and nanotheranostics.¹⁷ The photoluminescent features of CNDs are unique, and the underlying mechanisms behind PL have been widely investigated, encompassing the influences of surface passivation, heteroatom doping, quantum confinement, structural homogeneity and heterogeneity, surface states, and molecular states.

Nevertheless, accurately delineating the parameters that influence the emission is a challenging task. Consequently, a considerable debate has emerged over the influence of these parameters on the optical characteristics of CNDs. Additionally, the origin of excitation-independent emission in homogeneous CNDs remains a topic of scientific discussion. Along with this, some properties, particularly their high biocompatibility and robust optical performance, have made CNDs promising candidates for live-cell imaging and real-time monitoring of cellular processes.^18–20 Despite these promising characteristics, several fundamental aspects of CND photophysics remain poorly understood. Although slow solvent relaxation and red-edge effects have been cited as contributing factors, these mechanisms fail to explain the large Stokes shifts and broad emission profiles typically seen in CNDs. Understanding the structural and photophysical intricacies of CNDs is crucial for refining their synthetic control and optimizing their performance in practical applications. From 2016 onward, chiral CNDs have emerged as a significant subclass that combines the intrinsic PL of CNDs with molecular or structurally induced chirality.²¹ The current increase in interest at the intersection of chirality and CNDs has propelled chiral CNDs to the forefront of study. The strong fluorescence and chirality arise from chiral precursors, asymmetric surface groups, or distorted carbon frameworks. However, obtaining true intrinsic chiral CNDs remains difficult, as numerous products are more accurately categorized as chiral carbonized polymer dots (chiral CPDs).²² This review emphasizes the current advancements in chiral CNDs, concentrating on their production, structural characteristics, and chiroptical capabilities. Overall, as the field moves forward, unravelling the precise origin of their optical properties, particularly their PL behavior, will be vital to harnessing their full potential in emerging technologies such as nanotheranostics and advanced bioimaging platforms. In the last ten years, multiple reviews have substantially improved the understanding of CNDs and their luminescence characteristics.^23–26 However, the present review fills a crucial gap by integrating experimental results and computational models into a cohesive framework. It improves the mechanistic comprehension of fluorescence in CNDs and provides an umbrella for the fabrication of next-generation multi-responsive CNDs. Therefore, there is an evident opportunity for a further review, and ours is uniquely positioned to advance the field.

This review summarizes the recent advances in CNDs, focusing on understanding the origin of PL. The review also covers synthetic strategies, structural classification, and recent advances in high QY CNDs ranging from blue to near-infrared emission. In addition, it highlights the importance of post-synthetic modulation (surface passivation and purification) and reaction parameters such as precursor type, solvent, temperature, pH, and heteroatom doping in influencing emissive centers and PL behavior. It discusses how different factors such as core structure, surface states, MFs, and defects contribute to the emission behavior of CNDs. Finally, theoretical studies using DFT and TD-DFT are discussed to explain the relationship between CND structure and optical properties, providing guidance for the rational design of CNDs for applications in bioimaging, sensing, optoelectronics, and catalysis.

2. Optical origin of PL in CNDs

The story of PL in CNDs began in 2004, when Xu and coworkers accidentally observed a fluorescent band while purifying single-walled CNTs using gel electrophoresis (Fig. 1).¹ This band moved faster than expected and emitted multiple colors under UV light. These fluorescent carbon fragments raised a fundamental question: why does carbon glow? The first mechanistic answer came from Sun and coworkers in 2006, who found that CNDs produced by laser ablation showed no PL until their surfaces were passivated with organic species.²⁷ This led to the first proposed model suggesting that the emission originates from surface energy traps stabilized by surface passivation, while quantum confinement plays a supporting role. The excitation-dependent broad emission in CNDs was attributed to a distribution of particle sizes and emissive surface sites. In these CNDs, surface passivation combined with a quantum confinement framework became the dominant paradigm for several years. Inconsistencies in this model emerged when Bao and coworkers in 2011 showed that increasing surface oxidation caused a red shift in emission, even when the particle size remained unchanged.²⁸ This decoupled size effects from surface effects and showed that oxygen-containing groups, such as carbonyl and hydroxyl, directly introduce emissive trap states independent of the carbon core. The field consequently adopted a dual core–surface model, assigning excitation-dependent emission to quantum-confined sp² carbon core states and excitation-independent surface state-mediated emission to oxygen-related surface states. This dual model was further systematically investigated in 2012 when Krysmann and coworkers performed temperature-dependent pyrolysis of citric acid (CA) and ethanolamine and found that at low temperatures no nanoparticles formed at all.²⁹ The highly photoluminescent product was a single amide-containing molecular condensation product with a QY of approximately 50%. In contrast, at elevated temperatures, well-defined carbogenic cores formed, which led to a significant decrease in QY to only a few percent. This showed that the high QY was actually due to MFs, not CNDs, and that low-temperature synthesis mainly produces molecular emission rather than true nanoparticle effects. Parallel work on GQDs provided contrasting evidence that genuine quantum-confined excitonic states do exist in properly synthesized CNDs. Suzuki and coworkers in 2016 demonstrated that GQDs of 2 to 7 nm exhibited chiral excitonic transitions arising from quantum confinement in the graphene sheet, with chiroptical activity that disappeared in larger dots that could thermally racemize.²¹ This confirmed that when CNDs are properly prepared and characterized, their optical properties genuinely reflect quantum-confined electronic states of the carbon core. The MF issue was clearly demonstrated by Soni and coworkers in 2021.² They purified red-emissive CNDs synthesized from o-phenylenediamine (OPDA) and found that about 80% of the product was a single MF (QXPDA), showing molecular size, complete degradation below 200 °C, and optical spectra matching theoretical predictions. In contrast, the actual CND content was only ∼7%. This study confirmed that, without appropriate purification, the observed optical properties are mainly due to molecular byproducts rather than true CNDs. Therefore, understanding PL in CNDs requires first unambiguously determining their composition. In 2023, Zhang and coworkers demonstrated that post-oxidation introduces oxygen-related defects in CNDs, generating near-infrared (∼760 nm) emission from defect states formed by substitution of graphitic nitrogen with oxygen.³⁰ Selective passivation of these defects suppressed NIR emission while retaining visible emission, confirming defect states as an independent and tunable source of PL. The understanding of PL in CNDs has progressively evolved through distinct stages. Initially, quantum confinement and surface states were considered the primary contributors. This was followed by the recognition that MFs often dominate emission in bottom-up synthesis. Subsequently, true semiconductor-like core behavior was confirmed in CNDs. Finally, defect state engineering emerged as a deliberate strategy for accessing new emission windows. The intrinsic quantum properties of truly carbonized nanomaterials were clearly shown in 2025 by Wei and coworkers.³¹ They observed semiconductor-like PL blinking in carbonized polymer dots, including multiple intensity levels caused by charged excitons and Auger recombination. The occurrence of these blinking events increased with higher heteroatom doping, confirming that the carbon core exhibits real quantum-confined exciton behavior rather than emission from surface molecular species. Today, the PL of CNDs is understood to arise from multiple coexisting contributions. These include MF emission, surface state emission, quantum-confined core emission, and defect state emission. The relative dominance of each contribution depends on factors such as synthesis temperature, degree of carbonization, heteroatom doping, and the extent of purification prior to optical characterization. This review represents a transitional point, urging the community to move from broad theories to mechanism-specific design.

Fig. 1 Schematic overview of the evolution of PL origin in CNDs including core, edge, surface, and molecular states. (a) The first discovery of fluorescent CNDs. Reproduced from ref. 1 with permission from American Chemical Society, Copyright 2016. (b) First report on CNDs demonstrating that polymer surface passivation enhances fluorescence and attributing excitation-dependent emission to quantum confinement effects. Reproduced from ref. 27 with permission from American Chemical Society, Copyright 2006. (c) Role played by oxygen-related surface states in mediating PL via degree of oxidation in CNDs. Reproduced from ref. 28 with permission from Wiley-VCH, Copyright 2011. (d) Temperature-dependent evolution of core, and molecular state emission in CNDs. Reproduced from ref. 29 with permission from American Chemical Society, Copyright 2012. (e) The first report illustrating chirality in GQDs was obtained via cysteine-induced helical buckling of graphene sheet. Reproduced from ref. 21 with permission from American Chemical Society, Copyright 2016. https://pubs.acs.org/doi/10.1021/acsnano.5b06369. (f) Red emission in bottom-up CNDs arises from an organic fluorophore rather than graphitic cores, highlighting role played by molecular states and purification in PL. Reproduced from ref. 2 with permission from Royal Society of Chemistry, Copyright 2021. (g) Oxygen-related defect engineering: surface-state tuning for NIR emission. Reproduced from ref. 30 with permission from Wiley-VCH, Copyright 2023. (h) The exciton behavior of single carbonized polymer dots, providing insights into their PL properties under different surface chemicals. Reproduced from ref. 31 with permission from American Chemical Society, Copyright 2025.

3. Structural engineering and synthetic strategies of CNDs

3.1 Synthetic routes

Recent mainstream preparation approaches are categorized into two approaches: top-down and bottom-up approaches (Fig. 2a).^32,33 There are many reports that demonstrate that both approaches influence the structural and optical characteristics of CNDs. Some reaction mechanisms or structures have been proposed based on the reaction conditions. Diverse synthetic methodologies have been investigated for the preparation of CNDs, based on the requirements of the target application. The selection of the synthesis approach is mainly based on the following objectives: (i) simplicity and reproducibility, (ii) attainment of high yield or scalability, (iii) reduction of byproduct formation, (iv) cost-effectiveness, and (v) advanced optical characteristics.

Fig. 2 (a) General synthesis methods of CNDs: bottom-up approach, reproduced from ref. 51with permission from Springer Nature, Copyright 2019 and top-down approach. (b) Four types of fluorescent CNPs discussed in this review: GQDs, CQDs, reproduced from ref. 2 with permission from Royal Society of Chemistry, Copyright 2021; CNDs, and PDs. Three categories of proposed CND core: (c) crystalline core. Reproduced from ref. 46 with permission from American Chemical Society, Copyright 2023; (d) amorphous core. Reproduced from ref. 44 with permission from Wiley-VCH, Copyright 2021; (e) both crystalline and amorphous core. Reproduced from ref. 45 with permission from American Chemical Society, Copyright 2023.

The top-down approach involves bulk carbonaceous materials that are broken into small fragments. The carbon sources used for the top-down approach are mainly coal, candle soot, graphite, graphene oxide (GO), and CNTs. Substantial carbonaceous materials can be decomposed into nanostructures using methods such as acidic oxidation, arc discharge, laser ablation, electrochemical exfoliation, and ultrasonication. The top-down method is an efficient and straightforward approach for the synthesis of GQDs. Carbon precursors, including graphite and graphene, exhibit almost a single composition and a well-defined core structure, which is easier to further modify. Therefore, mostly CNDs can be further modified through the incorporation of many functionalities, including drug molecules, fluorophores, targeting ligands, and antibodies, to impart the necessary properties and functions. The bottom-up approach involves synthesizing CNDs from chemical precursors. This approach employs small organic molecules as precursors to synthesize CNDs via chemical reaction, due to their commercial availability and efficient carbonization. After carbonization under specific conditions, graphitic or amorphous cores are formed along with functional groups remaining on the surface of CNDs. The bottom-up approach encompasses hydrothermal methods, combustion, electrochemical processes, and microwave synthesis. Some commonly used precursors for bottom-up approaches are CA, phenylenediamine, glucose, urea, and ethylenediamine (EDA).^34–36

3.2 Structural classification

Based on their crystallinity and structural characteristics, carbon-based nanomaterials are generally categorized into four types (Fig. 2b): (1) CNDs: these typically exhibit an amorphous and crystalline structure and do not show quantum confinement.^37,38 In some cases, CNDs may show a combination of amorphous and crystalline features. (2) Carbon Quantum Dots (CQDs): These nanodots possess a crystalline structure and exhibit quantum confinement effects.³⁹ (3) GQDs: These are π-conjugated, single- or few-layer graphene with well-defined crystalline structures whose edges are functionalised with chemical groups.⁴⁰ (4) PDs: PDs are the cross-linked and aggregated polymer prepared from monomer units, as well as polymeric precursors.⁴¹ Moreover, the carbon core and the associated polymer chains can self-assemble to form PDs and small, incompletely carbonized polymer chains adsorbed onto the surface of the carbon core. Although these carbon-based nanomaterials are synthesized using different carbon precursors, they are distinguished by their intrinsic structure and surface chemical functionalities.

3.3 Structure of CNDs and synthesis of high-QY blue–NIR-emissive CNDs

The CND structure mainly contains core, surface, edge, and molecular states which collectively influence their PL behaviour. The core is mainly composed of sp²/sp³-hybridized carbon domains that may be amorphous or graphitic and provide structural stability and intrinsic electronic properties. The surface states arise from the synergistic interaction between various functional groups such as –OH, –CO, –SH, –NH₂, and –COOH present on the surface and the carbon core, which introduce surface energy levels and strongly affect the PL characteristics. Edge states originate from defects or dangling bonds located at the edges of sp² carbon domains and can act as trap sites that influence the emission wavelength and intensity. In contrast, the molecule state also acts as a PL center formed solely by an MF, which is either attached to the surface or embedded within the carbon backbone and can emit PL directly. In addition, these MFs are formed during the dehydration and carbonization of precursor molecules. These are distinct emissive organic entities formed during synthesis, potentially embedded within or adsorbed onto the carbon matrix or graphene framework.^2,42 In the literature, CND cores are reported to exhibit either amorphous, crystalline structures, or in some cases, a coexistence of both (Fig. 2c–e).^43–50 The diversity in synthesis routes leads to varying degrees of crystallinity and morphological features. Amorphous CNDs are typically characterized by the lack of regular, ordered structure, or by having no clear crystal lattice. These include sp²-hybridised carbon–carbon bonding along with the presence of sp³-hybridised carbon atoms within CNDs core, which contribute to structural disorder. In contrast, crystalline CNDs possess a regular, ordered crystal lattice composed predominantly of sp²-hybridised carbon–carbon bonding within the CND core. Therefore, comprehensive structural characterization (Table 1) is necessary to distinguish whether the synthesized product represents pure crystalline CNDs, amorphous carbon, quasi-CNDs, or aggregated MF. As the PL properties of CNDs are highly structure dependent, precise identification of their structural nature is critical for correctly elucidating the PL origin and establishing a reliable structure–emission correlation.⁵¹

Table 1 Characterization methodologies for identification of pure crystalline CNDs, amorphous CNDs, crystalline quasi CNDs, and aggregated fluorophores. Reproduced from ref. 51 with permission from Springer Nature, Copyright 2019

Techniques	Crystalline CNDs	Amorphous CNDs	Crystalline quasi CNDs	Aggregated fluorophores
HR-TEM	(a) Size 2–5 nm	(a) Size 2–5 nm	(a) Size 2–5 nm	Almost similar to CNDs. A little variation may be observed based on the fluorophore structure
	(b) Lattice fringes	(b) No lattice fringes	(b) Lattice fringes
	(c) Interlayer spacing of 0.34 nm for 002 plane	(c) No interlayer spacing	(c) Interlayer spacing of 0.34 nm for 002 plane
	(d) Interlayer spacing of 0.21 nm for 001 plane		(d) Interlayer spacing of 0.21 nm for 001 plane
XRD	Sharp peak at ∼24° for 002 plane	Broad peaks as compared with CND/Quasi CND	Sharp peak at ∼24° for 002 plane	Sharp peak at ∼24° for 002 plane
AFM	The height profile will confirm the size of 2–5 nm, topology and number of graphene layers in CNDs	The height profile will confirm the size of 2–5 nm and topology	The height profile will confirm the size of 3–5 nm, topology and number of graphene layers in CNDs	The height profile will confirm the size of 2–5 nm and topology
Raman Spectroscopy	(a) Sharp G band at ∼1580 cm^−1	Predominant broad D bands or high D/G ratio	Same as CNDs as it contains the same crystalline core	Almost similar to CNDs. A little variation may be observed based on the fluorophore structure
	(b) 2D band at ∼2680 cm^−1
	(c) less intense D band ∼1360 cm^−1 due to edge defect by surface functional groups
TGA	No mass loss and mostly stable up to 800 °C	No mass loss and mostly stable up to 800 °C	Due to less thermal stability of molecular fluorophore extensive mass loss within 300 °C	Due to less thermal stability of molecular fluorophore extensive mass loss within 300 °C
Confocal Raman PL	Due to high thermal stability of carbon core no change in PL intensity at lower temperature ∼300 °C	Due to high thermal stability of carbon core no change in PL intensity at lower temperature ∼300 °C	Due to less thermal stability of molecular fluorophore huge reduction in PL intensity within 300 °C	Due to less thermal stability of molecular fluorophore complete reduction in PL intensity within 300 °C

---

In this context, rational design of synthesis strategies becomes crucial, as controlled variation in precursors and reaction conditions enables tuning of core structure, surface states, and molecular emissive species. Generally, CNDs obtained via top-down approaches exhibit relatively low QYs, with graphene quantum dots (GQDs) produced through acid-assisted cutting often showing QYs below 1%. In contrast, bottom-up routes such as hydrothermal, solvothermal, and microwave-assisted methods enable better control over carbonization, surface states, molecular states and heteroatom incorporation, resulting in significantly enhanced emission efficiency. To address the limitations of low QY, researchers have utilized advanced molecular engineering, heteroatom doping (e.g., Si, B, N, S, and P), and surface passivation. Zhang et al. reported a one-step microwave synthesis of blue-emissive, N-doped, hydroxyl-functionalized CNDs with an ultra-high fluorescence QY of 99%.¹¹ These Tris-CNDs were prepared using CA and tris(hydroxymethyl)aminomethane, shown in Fig. 3a. The PL of Tris-CNDs originates from the synergistic effect of graphitic nitrogen and surface hydroxyl groups, leading to high QY. Similarly, Chen et al. reported the preparation of ultrasmall, green-emissive organosilica nanodots (OSiNDs) with an average size of ∼2.0 nm and a QY of ∼100% via a one-step hydrothermal treatment (Fig. 3b) using a silane molecule and rose Bengal.⁵² The structural reorganization and halide loss of rose Bengal during the hydrothermal procedure specifically enhance the ultrahigh QY and minimize phototoxicity of the OSiNDs. Guo et al. developed CNDs that show bright yellow fluorescence. These CNDs were synthesized via solvothermal treatment of CA and urea in toluene solvent, followed by silica column chromatography.⁵³ The resulting yellow-emissive CNDs exhibit a large Stokes shift, ultra-high QY exceeding 90%, and good PL stability in both solution and solid states (Fig. 3c). The origin of fluorescence of these CNDs was attributed to conjugated sp²-carbon domains (fused rings) with edge functional groups. Initially, the solvothermal reaction yielded crude yellow-emissive CNDs, showing white fluorescence with a solution QY of 53%. However, after purification through column chromatography, these CNDs achieved a stable solution QY of up to 92%. Zishan Sun et al. reported silicon-doped, orange-emitting CNDs using an economical and straightforward one-pot hydrothermal approach, with OPDA and ethyl orthosilicate used as raw materials.⁵⁴ Silicon doping enhanced the emission properties of orange-emitting CNDs, increasing QY from 39.2% to 64.1%, while maintaining excellent fluorescence stability (Fig. 3d). Silicon doping creates new emissive states and enhances radiative recombination, while TEOS passivates the surface and reduces non-radiative losses, leading to improved QY. Haoran Jia et al. demonstrated that an electron-donating group passivation strategy imparts red bandgap emission to CNDs.⁵⁵ Three red-emissive, electron-donating group-passivated CNDs R-EGP-CNDs-NMe₂, –NEt₂, and –NPr₂ were synthesized and exhibited strong red emissions at 637, 642, and 645 nm, respectively. These CNDs showed excellent solubility in common organic solvents and achieved a high QY of up to 86% in ethanol (Fig. 3e). Developing efficient red-emissive CNDs is a significant advancement toward realizing high-performance electroluminescent warm white LEDs. In these CNDs, theoretical studies show that red emission arises from the π-conjugated core, while –NR₂ groups tune the electronic structure through internal charge redistribution, resulting in high QY. Additionally, NIR-emissive CNDs exhibiting QYs up to 40% have been attained by employing conjugated aromatic precursors, optimized solvothermal synthesis parameters, and implementing targeted post-synthetic modifications. Recent research has shown that chiral CNDs were made using a microwave-assisted hydrothermal method with CA as the carbon source and D-/L-cysteine as the chiral ligand (Fig. 3f).⁹⁸ When heated to 180 °C, CA undergoes carbonization to form a graphitic core. At the same time, cysteine facilitates surface passivation and imparts chirality, forming covalent bonds (such as –COSR/amide-type bonds). This process results in water-soluble, chiral CNDs with functionalized surfaces. The emission in these CNDs originates from coupled core–surface states and chirality which actively influences electronic transitions through spin–orbit coupling and spin-selective charge transfer processes. As a result of these photophysical properties, electron transfer and photocurrent generation are significantly enhanced, which are crucial for the development of next-generation optoelectronics. Ding et al. achieved a novel solvent-controlled synthetic route to produce highly luminescent CNDs exhibiting NIR emission at about 715 nm, achieving a QY of 43% in ethanol and a QY of 36% in water (Fig. 3g), among the highest reported QY values for the NIR-emitting CNDs.⁷⁸ PL in these NIR-emitting CNDs arises from surface states with bandgap narrowing arising from increased sp² conjugation and graphitic nitrogen incorporation, promoted by enhanced carbonization and larger particle size. Table 2 summarizes optical properties of high-QY blue-to-NIR-emissive CNDs, highlighting key parameters including precursors, absorbance, PL, and QY. The data emphasize the role played by conjugated precursors, heteroatom doping, and post-synthetic modifications in achieving enhanced QY and red/NIR emission. Overall, the rational design of high-QY blue-to-NIR-emissive CNDs relies on the synergistic control of precursor selection, reaction conditions, heteroatom doping, surface passivation, and purification strategies, enabling precise tuning of their structural and optical properties. These approaches facilitate efficient and tunable PL across the UV to NIR region, making CNDs highly promising for applications in optoelectronics, bioimaging, sensing, and phototherapy. In conclusion, the PL of CNDs originates from a synergistic interplay of core, surface, and MF states, with the dominant emission pathway governed by the specific synthesis conditions and structural characteristics. In most cases, surface states play a key role in emission, while the sp²-conjugated carbon core regulates the bandgap, leading to red/NIR shifts, and molecular emissive species contribute to enhanced QY.

Fig. 3 (a) Schematic illustration showing the synthesis of blue-emissive Tris-CNDs having QY of 99% utilizing a microwave approach. Reproduced from ref. 11 with permission from Royal Society of Chemistry, Copyright 2016. (b) Schematic depicting the synthesis of green-emissive OSiNDs with exceptionally high QY∼ 100% using a straightforward one-step hydrothermal reaction involving rose Bengal (RB) and silane molecule. Reproduced from ref. 52 with permission from American Chemical Society, Copyright 2018. (c) Schematic representation of yellow-emissive CNDs synthesized via solvothermal treatment of CA and urea in toluene demonstrating QY of 53%, further purified by column chromatography, resulting in Y-CNDs with QY up to 92%. Reproduced from ref. 53 with permission from Wiley-VCH, Copyright 2022. (d) Schematic illustrating that silicon-doped orange-emitting CNDs (OCNDs) are synthesized via a one-pot hydrothermal process, utilizing OPDA and ethyl orthosilicate as precursors. A notable enhancement on the QY with Si4+ doping was reported, i.e. 39.2% for O-CNDs and 64.1% for Si4+-doped O-CNDs at the same concentration. Reproduced from ref. 54 with permission from Elsevier, Copyright 2022. (e) Synthesis of R-EGP-CNDs-NMe2, –NEt2, and –NPr2 using solvothermal treatment of N,N-dimethyl-, N,N-diethyl-, and N,N-dipropyl-p-PD, respectively, with the maximum QY up to 86.0% in ethanol. Reproduced from ref. 55 with permission from Wiley-VCH, Copyright 2019. (f) Synthesis of chiral CNDs using CA as a source of the carbon core and D/L-cysteine as the chiral ligands. Reproduced from ref. 98 with permission, from American Chemical Society, Copyright 2024. (g) Schematic representation of synthesis of NIR-emissive CNDs utilizing L-glutamic acid and the QY in ethanol solution was as high as 43.20% and 12.80%, respectively. Reproduced from ref. 78 with permission from Wiley-VCH, Copyright 2018.

Table 2 Precursors and corresponding optical properties (absorption, emission, and QY) of CNDs reported spanning the blue to NIR spectral region

Sr. no.	Precursor	Absorbance (nm)	PL (nm)	QY (%)	Ref.
1	OPDA, BmimPF6	≈580, ≈630	714	37.90	56
2	PTCDA, urea	741	765	18.80	57
3	PPDA, MnSO4	242 and 294	610	73.80	58
4	CA, Tris, PEG400	330, 570	608	65.50	44
5	Naphthalene tetracarboxylic dianhydride, L-cysteine	361, 380, 552–600	610	60.50	59
6	OPDA and 3APBA	435	540	59.00	12
7	m-phenylenediamine (MPDA) and Resorcinol	283, 293, and 347	540	61.00	60
8	OPDA, methyl red	≈470	552	95.00	61
9	N-Phenyl-ophenylenediamine	490–510	561–618	60.80	62
10	RB, FeCl3	510	530	97.00	63
11	OPDA, L-Try	≈400	535	71.00	64
12	CA, urea, and CaCl2	406	≈515	73.00	65
13	CA and urea	413	516	73.00	66
14	NaOH, CA, and urea	330 and 400	518	75.00	67
15	PDA, Urea, BA	400, 420, 445	470–520	97.20	68
16	CA and cyanamide	410	521	73.00	69
17	m-PDA	263, 468	516/520	80.00/76.00	70
18	Perylene and potassium persulfate	≈450	522	96.00	71
19	RhB and sodium metasilicate	240 and 490	510	92.30	72
20	CA and OPDA	254 and 373	431	87.07	73
21	CA and OPDA	247 and 384	425	94.40	74
22	CA and phenylalanine	220–230, 300	425	90.00	75
23	Pyrene	535 nm	610	98.85	76
24	Dopamine, OPDA, HCl	277, 450–631	710	26.28	77
25	L-Glutamic acid, OPDA, H2SO4	634	715	43.20	78
26	CA, urea	724	770	11.00	79
27	CA, EDA	564	627	53.00	47
28	N,N-Dimethyl-p-PDA	540	637	77.90	55
29	Resorcinol	595	610	72.00	80
30	CA and urea	540	580	46.00	35
31	Phloroglucin	582	598	54.00	81
32	H2SO4, ethylene glycol	232.5, 299.5, 502.5	519	62.90	82
33	Phloroglucinol	498	507	72.00	81
34	p-PDA	287, 507	600	52.46	83
35	CA and DETA	240, 340	456	88.60	84
36	CA and Tris	333	417	93.30	85
37	Folic acid (FA)	280 and 320	400	94.50	86
38	FA	235, 280, and 335	440	85.00	87
39	CA and cyclen	300–400, 400–500	544	48.00	88
40	CA, urea	715	760	10.00	89
41	CA, urea, ammonium fluoride	556, 624	658, 777	9.80	90
42	1,3-Dihydroxynaphthalene, KIO4	530	628	53.00	91
43	OPDA	430	495	88.90	92
44	CA, urea, and manganese carbonate	405	510	61.00	93
45	RhB, NaOH	≈490	526	91.17	56
46	PPDA, maleic anhydride	306	508	76.50	94
47	CA and Nile Blue A sulphate	255 and 523	615	54.40	95
48	OPDA, MnSO4	242 and 294	550	85.40	58
49	OPDA and CPBA	230, 284, and 450	356, 700	86.58	96
50	Nitrated pyrenes and boric acid	490	595	65.78	97

---

4. Effect of post-synthetic modulation on PL properties

Post-synthetic modulation plays an important role in tuning the PL properties of CNDs. After synthesis, surface passivation and purification significantly influence the emission characteristics of CNDs. Surface passivation alters the surface functional groups and electronic states, which can shift emission wavelength, fluorescence QY, and quenching behaviour. In contrast, appropriate purification separates residual MF, unreacted precursors, and other by-products that may contribute to or dominate the observed emission. Therefore, careful control of these post-synthetic processes is essential for accurately interpreting the PL origin and ensuring reliable optical properties of CNDs.

4.1 Effect of surface passivation on PL

Surface functional groups and surface modification substantially influence key properties, including fluorescence QY, complexation capacity, PL, and quenching behavior. Among these, amino, carboxy, and hydroxy groups are the most commonly observed and can be introduced via both covalent and noncovalent modifications. Noncovalent modification of CNDs includes pi interactions, electrostatic interactions, and complexations, which can be modified based on the different functional groups and structure of CNDs. These interactions always depend upon the surface chemistry of CNDs as well as aromatic guest molecules. For example, a Silicon-CNDs@dopamine (Si-CNDs@DA) composite was synthesized in two steps by Jiang et al.⁹⁹ First, Si-CNDs were rapidly prepared via a microwave-assisted approach using glycerol and (3-aminopropyl)triethoxysilane (APTES) as the carbon and silicon sources. Then, dopamine (DA) was introduced by simply mixing the Si-CNDs with dopamine in water and stirring overnight, forming a stable Si-CNDs@DA composite shown in Fig. 4a. This nanomaterial showed strong blue fluorescence and excellent stability. It was selectively quenched by silver ions (Ag⁺), making it useful as a sensitive probe for intracellular Ag⁺ detection and cell imaging due to its low toxicity and good photostability. Furthermore, Chen et al. synthesized CNDs via a one-step hydrothermal approach using ethylenediaminetetraacetic acid (EDTA) as the carbon source and triethylenetetramine (TETA) as the passivating agent, shown in Fig. 4b.¹⁰⁰ The resulting CNDs had –COOH and –NH₂ groups on their surface, enabling direct coordination with Tb³⁺ ions to form the CNDs-Tb nanoprobe. This probe detects dipicolinic acid (DPA) through a ratiometric fluorescence mechanism, where DPA binds to Tb³⁺ and enhances its green emission, while the CND's blue fluorescence serves as a stable reference. Similarly, in pi–pi interactions, the core of CNDs, which contains the graphene sheets having an extended pi system, can undergo pi interactions with incoming aromatic molecules. Several reports have demonstrated that pi–pi interactions between CNDs and exogenous ligands enable detection of nucleic acids, Hg⁺, and Ag⁺. Sarkar et al. synthesized blue- and green-emitting cationic CNDs (CCNDb and CCNDg) and noncovalently modified them with 17β-estradiol hemisuccinate (E2) via electrostatic interactions to target estrogen receptor-positive (ER⁺) cancer cells.¹⁰¹ These E2-functionalized CNDs (CCNDb-E2 and CCNDg-E2) were further loaded with doxorubicin (Dox) via pi–pi stacking, forming CCNDb-E2-Dox and CCNDg-E2-Dox (Fig. 4c). The hybrids showed enhanced fluorescence, selective labeling, and targeted killing of ER⁺ MCF7 cells via a late apoptotic pathway, demonstrating about 2-fold higher efficacy than in ER⁻ and noncancerous cells. Hailong Li et al. used CNPs for nucleic acid detection. The detection of DNA consists of two steps.¹⁰² Firstly, the CNP binds dye-ssDNA through pi–pi interactions between DNA nucleobases, nucleosides with the pi-rich CNP, resulting in the fluorescence quenching of the fluorophore FAM (Fig. 4d). In the subsequent step, the hybridization of dye-ssDNA with its target yields the formation of dsDNA. This causes the release of dye-ssDNA from CNP and hence restores dye fluorescence. Similarly, Minhuan Lan et al. (2014) contributed to the stabilization of the CNP-RhB nanohybrid system through non-covalent stacking between the aromatic rings of CNPs and RhB. While the primary sensing mechanism for Hg²⁺ detection involves electrostatic interaction and metal–ligand coordination, leading to fluorescence quenching, the π–π stacking supports close molecular proximity, enhancing the efficiency of energy or electron transfer (Fig. 4e).¹⁰³ These interactions, although secondary, play a supportive role in maintaining hybrid stability and effective ratiometric fluorescence response. Electrostatic interaction serves as a versatile and efficient strategy for noncovalent surface modification of CNDs, enabling the attachment of various functional molecules such as fluorophores, targeting agents, and therapeutic drugs.^101,103,104 This approach allows for enhanced fluorescence properties, improved solubility, and selective detection of biologically relevant ions such as S²⁻, Hg²⁺, and Cu²⁺ with high sensitivity and low detection limits. Moreover, electrostatically modified CNDs have demonstrated promising applications in real biological systems, including bioimaging and targeted cancer therapy. These findings indicate the potential of electrostatic interactions as a powerful tool for the development of multifunctional nanoprobes and theranostic platforms.

Fig. 4 (a) Schematic illustration of Si-CNDs@DA synthesis and Ag⁺ sensing mechanism. Red dots represent Ag⁺ ions that are selectively adsorbed and reduced by dopamine moieties on the Si-CNDs@DA surface, leading to aggregation and fluorescence quenching. Reproduced from ref. 99 with permission from American Chemical Society, Copyright 2016. (b) Schematic of a ratiometric nanoprobe based on terbium-functionalized CNDs (CNDs-Tb) for DPA detection. Reproduced from ref. 100 with permission from Royal Society of Chemistry, Copyright 2015. (c) Schematic showing synthesis of blue- and green-emitting cationic CNDs (CCNDb and CCNDg), and their noncovalent electrostatic conjugation with estradiol hemisuccinate (E2) to form CCD-E2 hybrids. Reproduced from ref. 101 with permission from American Chemical Society, Copyright 2017. (d) Schematic illustration of CNP-based fluorescence sensing via π–π interaction with ssDNA. Reproduced from ref. 102 with permission from Royal Society of Chemistry, Copyright 2011. (e) Schematic illustration of a CNP–RhB nanohybrid-based ratiometric fluorescence sensor for Hg²⁺ detection. Negatively charged CNPs and positively charged RhB form a stable hybrid via electrostatic interaction, while Hg²⁺ selectively quenches CNP emission through electron or energy transfer. Reproduced from ref. 103 with permission from American Chemical Society, Copyright 2014.

Covalent modification involves introducing functional groups or particular compounds onto the surface of purified CNDs via covalent bonding. This approach offers precise control over morphology, and specific biological targeting, and provides a reliable method for conjugating CNDs with biomolecules for targeted applications. These modifications affect the fluorescence characteristics of the CNDs. Covalent modification offers a robust method for linking nanoparticles to biological systems or conjugating them with other molecules for targeted applications. Common reaction types include amide coupling,¹⁰⁵ sialylation, esterification, sulfonylation, and copolymerization. Liu et al. synthesized multicolor solid-state emissive CNDs by functionalizing CNDs with chlorosalicylaldehyde using Schiff base reactions.¹⁰⁶ For that, pristine CNDs were first synthesized using OPDA and sulfuric acid as precursors via hydrothermal treatment, yielding amine-rich CNDs (Fig. 5a). These amine-rich CNDs were then modified by reacting with different isomers of chlorosalicylaldehyde (ortho-, meta-, and para-substituted). The aldehyde group of the salicylaldehyde formed imine (C=N) bonds with the amine groups on the CNDs, resulting in covalently bound surface ligands. Similarly, Li et al. synthesized thermoresponsive fluorescent CNDs and modified their surface through a copolymerization approach to enhance their water solubility and cellular imaging performance.¹⁰⁷ Initially, hydrophobic luminescent CNDs were prepared by thermal oxidation of CA and L-tyrosine. These CNDs were then passivated via amidation with methacryloyl chloride to introduce polymerizable methacrylate groups, forming CND-MAA monomers. Subsequently, a free radical copolymerization of these CNDs-MAAs with N-isopropylacrylamide (NIPAM) using AIBN as the initiator yielded poly(NIPAM)-grafted CNDs (CNDs-g-PNIPAM), shown in Fig. 5b. This polymer shell significantly improved the aqueous dispersibility of the hydrophobic CNDs and imparted thermoresponsive behavior due to the lower critical solution temperature properties of PNIPAM. The CNDs-g-PNIPAM composites demonstrated excitation-independent blue fluorescence and a progressive reduction in fluorescence intensity as temperature increased, due to the collapse of the PNIPAM chain and aggregation of nanoparticles. Importantly, cytotoxicity assays demonstrated excellent biocompatibility with over 85% cellular viability at high concentrations (up to 2000 μg mL⁻¹), and time-dependent fluorescence microscopy confirmed successful and stable cellular internalization. This study emphasizes the potential of polymer-grafted CNDs as smart, responsive nanoprobes for long-term bioimaging and thermally controlled applications. In another work, Wang et al. synthesized CNDs via a hydrothermal method utilizing CA and urea. After synthesis, the surface of CNDs was further functionalized with glutathione (GSH) through carbodiimide-activated coupling chemistry.¹⁰⁸ This modification was performed using EDC/NHS coupling, which activated the -COOH groups present on the CNDs surface (Fig. 5c). The reaction was performed in phosphate-buffered saline (PBS, pH 7.4) at ambient temperature under dark conditions, followed by purification using filtration and dialysis. The presence of GSH on the CNDs’ surface introduced multiple functional groups, such as –SH, –COOH, and –NH₂ groups, enhancing the aqueous dispersibility, biocompatibility, and PL properties of functionalized CNDs. In a related work, Gao et al. synthesized quaternary ammonium-functionalized CNDs (CTA-CNDs) for the selective and sensitive detection of 2,4,6-trinitrophenol (TNP) in aqueous medium.¹⁰⁹ After preparing passivated CNDs from carboxymethyl cellulose (CMC) and 4,7,10-trioxa-1,13-tridecanediamine (TTDDA) through a hydrothermal approach, the surface of these CNDs was modified via esterification reaction with (3-chloro-2-hydroxypropyl)trimethylammonium chloride (CTA) in the presence of DCC/DMAP as catalysts, as shown in Fig. 5d. This reaction covalently grafted quaternary ammonium groups onto the surface of the CNDs. The modification was aimed at enhancing electrostatic interactions between the positively charged CTA-CNDs and the negatively charged TNP (picrate anion) in water. The successful functionalization was confirmed by XPS and FTIR, showing peaks corresponding to amide, ester, and quaternary ammonium functionalities. Zeta potential analysis showed a charge reversal from negative (−33.35 mV) to positive (+11.49 mV), indicating successful surface modification. These CTA-CNDs showed enhanced fluorescence quenching in the presence of TNP due to static quenching via π–π, hydrogen bonding, and electrostatic interactions, allowing detection of TNP with a low detection limit of 7.04 × 10⁻⁷ M, which is significantly better than unmodified CNDs. Lai et al. modified CNDs through direct thermal coupling with targeting molecules (mannose or folic acid), eliminating the need for EDC/NHS-based wet chemistry (Fig. 5e), yielding biocompatible and selective fluorescent probes for bioimaging and biosensing.¹¹⁰ Sha Li et al. synthesized CNDs from α-cyclodextrin via a hydrothermal approach, resulting in CNDs rich in -OH groups.¹¹¹ To enhance their functionality and biocompatibility, the synthesized CNDs were further modified by grafting hyperbranched polyglycerol (HPG) through an anionic ring-opening polymerization of glycidol (Fig. 5f). The resulting CNDs-g-HPG nanohybrids exhibited improved water solubility, high biocompatibility, and significantly enhanced hemocompatibility compared with unmodified CNDs. Structural confirmation of HPG grafting was carried out through FTIR and ¹H NMR spectroscopy, showing characteristic peaks for polyether groups. The modified CNDs also retained strong photoluminescent properties and demonstrated effective cellular uptake with minimal toxicity, highlighting their potential as bioimaging agents. Yuan et al. synthesized amine-functionalized CNDs utilizing a pyrolytic reaction of CA and a diamine compound 4,7,10-trioxa-1,13-tridecanediamine (TTDDA) in glycerol.¹¹² These CNDs were further functionalized through a multi-step coordination-based surface modification approach to enable selective detection of mercury ions (Hg²⁺). The amine groups on the CND surface first reacted with carbon disulfide (CS₂) in the presence of NaOH to form dithiocarbamate (DTC) groups (Fig. 5g). These newly formed DTC moieties were then coordinated with Cu²⁺ ions to generate a bis(dithiocarbamato)copper(II) complex (CuDTC₂) on the surface of the CNDs. Subsequently, an additional ligand, ammonium N-(dithiocarboxy)sarcosine (DTCS), was coordinated to the copper center to complete the functionalization. This surface-bound CuDTC₂ complex effectively quenched the native fluorescence of the CNDs via electron and energy transfer mechanisms. However, when Hg²⁺ ions were introduced, they displaced the Cu²⁺ ions in the surface complex due to their stronger affinity toward DTC groups, leading to a disruption of the quenching mechanism and a marked “turn-on” fluorescence response.

Fig. 5 (a) Schematic depicts the condensation reaction between amine groups on the surface of CNDs and the aldehyde group of chlorosalicylaldehyde, leading to the formation of an imine (Schiff base) bond which enables their subsequent self-assembly and influences their solid-state luminescence properties. Reproduced from ref. 106 with permission from American Chemical Society, Copyright 2023. (b) Schematic illustrating the sequential preparation process (i) synthesis of CNDs via thermal oxidation, (ii) covalent functionalization through amidation with methacryloyl chloride to yield polymerizable CNDs, and (iii) copolymerization with N-isopropylacrylamide (NIPAM) employing radical polymerization. Reproduced from ref. 107 with permission from Wiley, Copyright 2017. (c) Schematic illustrates the covalent conjugation of GSH to the surface of CNDs through a carbodiimide-activated coupling reaction (using EDC/NHS). Reproduced from ref. 108 with permission from Royal Society of Chemistry, Copyright 2016. (d) Schematics illustrates the covalent grafting of CTA onto passivated CNDs through an esterification reaction for the selective detection of TNP. Reproduced from ref. 109 with permission from Elsevier, Copyright 2018. (e) Schematic depiction of the dehydration-coupling mechanisms employed for mannose-functionalized CNDs (Man-CNDs) and folic acid-functionalized CNDs (FA-CNDs). Reproduced from ref. 110 with permission from Elsevier, Copyright 2016. (f) Schematic illustration of surface passivation of CNDs by grafting hyperbranched polyglycerol (HPG) via surface-initiated anionic ring-opening polymerization. Reproduced from ref. 111 with permission from Royal Society of Chemistry, Copyright 2017. (g) Diagram illustrating the functionalization of amine-coated CNDs with CuDTC2 for the highly sensitive and selective detection of aqueous mercury ions (Hg2+). Reproduced from ref. 112 with permission from American Chemical Society, Copyright 2014.

4.2 Influence of purification on the PL origin of CNDs

Crude CNDs contain molecular impurities and fluorophores that can change their PL, sometimes causing misleading emission behaviour and overestimated QY. Therefore, purification is important to reveal the true emission behaviour in CNDs. CNDs often show complex and varied PL for several reasons, but a main cause is that they are often not purified properly. As a result, many researchers believe they are working with a mixture which includes CNDs and other impurities rather than with pure CNDs. Due to multiple reaction pathways and diverse synthesis approaches, crude CNDs typically contain molecular side products, unwanted impurities, unreacted precursors, MF, and carbon-based nano or microstructures. When top-down approaches are used for CND synthesis, some large-sized derivatives that are still intact or partially degraded often exist as impurities along with CNDs. In contrast, CNDs synthesized by bottom-up processes frequently exhibit heterogeneous surface states and structures, leading to diverse optical characteristics, even within the same batch. During bottom-up methods, organic molecules undergo sequential stages of dehydration, condensation, and carbonization to form CNDs. However, during bottom-up synthesis, not all molecules react completely, leaving some organic molecules and low-molecular-weight oligomers in the crude CNDs solution. The commonly utilized characterization techniques are not able to identify and distinguish the presence of residual molecular entities. The solution, comprising CNDs and the molecular byproducts, will always produce emission. At that point, it is easy to believe that the synthesis of fluorescent CNDs was successful; nonetheless, inadequate purification has resulted in erroneous conclusions. Meanwhile, TEM/AFM images will provide additional proof of CND formation. Moreover, it is essential to acknowledge that both TEM and AFM cannot always differentiate between aggregation of organic molecules and carbon nanomaterials (CNPs). For example, Khan et al. showed that in hydrothermal treatment, CA yields methylene succinic acid, resulting in hydrogen-bonded nano-assemblies, which, upon TEM inspection, appear as spherically shaped nanoparticles with an average diameter of 3.6 nm and a lattice spacing of 0.21 nm, which are common characteristics associated with CNDs.¹¹³ Incomplete purification can therefore significantly alter intrinsic PL properties of CNDs. For instance, impurities like organic molecule fluorophores with intense emission may coordinate, complex, adsorb, or covalently bind to the functional groups on the surface of CNDs. This can cause electron or energy transfer that alters the emission efficiency of the CNDs. Several methods used for purification, such as centrifugation, filtration, dialysis, chromatography, and electrophoresis, have been employed to purify CNDs since 2004. Purification methods employed since 2004 fall into two major categories: removal methods (centrifugation, filtration, and dialysis) and separation methods (electrophoresis and chromatography). Xu et al. employed a preparative electrophoretic purification technique to isolate and purify fluorescent carbon constituents from single-walled carbon nanotube (SWNT) arc-discharge soot, as illustrated in Fig. 6a.¹ Liu et al. isolated pure fluorescent CNPs from a neutralized candle-soot dispersion using denaturing polyacrylamide gel electrophoresis (Fig. 6b).¹¹⁴ Zhang et al. synthesized CNDs via a 100 ml hydrothermal approach using Rosa roxburghii as a carbon source. The formed brown CNDs were then centrifuged and purified using a 0.22 μm filter membrane (Fig. 6c).¹¹⁵ Fan et al. functionalized CNDs using previously reported method with some modifications.¹¹⁶ Initially, they dissolved 2 g of CA in 30 mL of water and stirred the mixture for 10 minutes. After that, EDA was added, and the solution was subjected to ultrasonication for five minutes. The flask was placed in a muffle furnace and heated at a rate of 4 °C min⁻¹ for 5 h, until it attained a maximum temperature of 180 °C. After the reaction, the furnace was allowed to cool down naturally to room temperature. The deep yellowish-brown solution was further centrifuged for 15 minutes at 10 000 rpm to isolate the supernatant. The solution was dialyzed against ultrapure water in a dialysis tube (MWCO: 1000 Da) for 24 h to remove small molecules (Fig. 6d) and unreacted reactant. The N-CNDs were then concentrated to 5 mg ml⁻¹ in DI water. Uthirakumar et al. used a facile and convenient method to isolate CNDs from the reaction mixture via a solvent extraction method (Fig. 6e).¹¹⁷ Soni et al. made red-emitting CNDs using OPDA and 10 ml of water under vigorous conditions, then adding 5 ml of a 5% hydrochloric acid (HCl) solution.² The systematic purification of the solution was carried out via column chromatography (Fig. 6f). For chromatographic purification, one solvent was insufficient for complete separation, therefore a solvent system was employed including 10% ethyl alcohol in 90% ethyl acetate (most polar), a mixture of 50% ethyl acetate and 50% hexane, and 10% ethyl acetate in 90% hexane (least polar). As a result, elution of three primary components exhibiting distinct red, green, and blue colors from the column was achieved. Notably, the red component had the highest product yield, about 80%, while the green component had about 13% and the blue component had the lowest product yield, about 7%. It was observed that the red emission in the CNDs comes from an MF (QXPDA). The blue and green fractions, on the other hand, likely possess quasi-CNDs structures with a varying degree of core and surface states. Ding et al. synthesized CNDs under hydrothermal conditions using urea and p-phenylenediamine (PPDA) as starting precursors. The synthesized products were actually mixtures of CNDs exhibiting various fluorescence colors, which could be effectively separated using silica column chromatography because they had different polarities. Ding et al. synthesized full-color CNDs via using a one-pot hydrothermal method using urea and PPDA (1 : 1 mass ratio) at 160 °C for 10 hours (Fig. 6g).¹¹⁸ The as-prepared products were complex mixtures of CNDs with different fluorescence colors. To separate the complex mixture of CNDs, silica column chromatography was employed for purification. This process yielded eight distinct fractions with PL emission maxima ranging from 440 to 625 nm, which were subsequently freeze-dried and redispersed in water for further application. Yuan et al. synthesized NBE-T-CNDs by utilizing three-fold symmetric phloroglucinol and solvothermal conditions at 200 °C at varying reaction time (Fig. 6h).¹¹⁹ Then, silica column chromatography was then used to purify the product, with dichloromethane (DCM) and methanol as the eluent. There is another report, where Wu et al. synthesized green-emissive CNDs using OPDA and L-tryptophan (L-Try) in DMF with catalytic HCl at 200 °C.⁶⁴ After cooling, the product was purified via repeated DCM extraction until the aqueous phase became clear, and evaporation of the organic layer yielded purified green-emissive CNDs. Overall, systematic purification is essential for unambiguously identifying the true PL origin in CNDs and ensuring reproducibility of their optical characterization.

Fig. 6 (a) Purification of SWCNTs allowing the discovery of CNDs. Reproduced from ref. 1 with permission from American Chemical Society, Copyright 2004. (b) Electrophoretic separation of CNDs synthesized from candle soot revealed the multiple species present within one batch. Reproduced from ref. 114 with permission from Wiley-VCH, Copyright 2007. (c) Scheme illustrating the synthesis of CNDs from Rosa roxburghi and their purification through centrifugation filtration. Reproduced from ref. 115 with permission from Frontiers, Copyright 2020. (d) Schematic illustration showing synthesis and purification of CNDs through dialysis. Reproduced from ref. 116 with permission from World Scientific, Copyright 2018. (e) A schematic representation for isolating CNDs via a solvent extraction technique. Reproduced from ref. 117 with permission from Royal Society of Chemistry, Copyright 2018. (f) Schematic showing that red-emissive CNDs were purified through column chromatography. Reproduced from ref. 2 with permission from Royal Society of Chemistry, Copyright 2021. (g) One-pot synthesis and purification route for CNDs with distinct fluorescence characteristics. Eight CND samples under 365 nm UV light and their corresponding PL emission spectra, with maxima at 440, 458, 517, 553, 566, 580, 594, and 625 nm. Reproduced from ref. 118 with permission from American Chemical Society, Copyright 2016. (h) Schematic illustration of silica column chromatography that was used for the purification of narrow-bandwidth emission triangular CNDs (NBE-T-CNDs). Reproduced from ref. 119 with permission from Springer Nature, Copyright 2018.

5. Influence of environmental conditions on PL properties

The reaction conditions significantly influence the optical characteristics and emission wavelengths of CNDs, even with identical precursors.^76,120 Fluctuations in parameters, including temperature, pressure, solvent, pH, reaction media, and energy source, can result in CNDs emitting across a wide spectral range from blue to NIR with markedly distinct QYs. For instance, various synthetic methodologies, including hydrothermal, microwave, solvothermal, or hybrid approaches, can produce CNDs from carbonaceous precursors exhibiting emission peaks from blue to the NIR range, with QY reaching up to 92%. More severe reaction conditions, such as elevated temperatures or strong acid catalysis, generally promote the synthesis of longer-wavelength-emitting CNDs by facilitating carbonization and conjugation. Moreover, within bottom-up synthetic methodologies, solid-phase techniques such as vacuum heating, microwave-assisted heating, and magnetic hyperthermia (MHT) have demonstrated efficacy for generating highly luminous CNDs with elevated QY and superior product yields. These findings collectively highlight the significance of optimizing reaction conditions for scalable, high-performance CND synthesis tailored to specific applications. Han et al. reported the synthesis of CNDs by tuning several reaction parameters such as the type of precursor, precursor ratio, reaction temperature, solvent environment, and heteroatom doping.¹²¹ By systematically modifying these parameters, they demonstrated that the electronic structure and surface chemistry of CNDs can be controlled, which directly influences their PL behavior (Fig. 7). Their study showed that depending on the synthesis conditions, the emission of CNDs can originate from intrinsic carbon core states (π–π transitions), dopant-induced energy levels within the band gap, or surface functional group-related states, thereby explaining the different PL mechanisms observed in CNDs.

Fig. 7 Structural models of (a) crystalline, (b) doped, and (c) surface-functionalized CNDs and their corresponding energy band structures showing defect-free, dopant-induced, and surface state energy levels. Reproduced from ref. 121 with permission from Royal Society of Chemistry, Copyright 2022.

5.1 Precursor-dependent emission

The predominant precursors for CND synthesis are categorized into natural sources and chemical molecules. The natural compounds include banana juice,^122–124 orange and lemon juice,^125–127 potato,^128,129 egg,^130,131 beer, coffee,^132–135 meat, beverage, soy milk, sugar,^136,137Punica granatum fruit, bread,¹³⁸ sucrose,¹³⁹ jaggery,^140,141 lysozyme, grass,^142,143 silk from Bombyx mori, and starch. Organic molecular precursors include OPDA,^25,144 boric acid, benzene, CA,¹⁴⁵ polyethylene and ethylene glycol,¹⁴⁶ proteins, amino acids,^147–149 glycerol,^150,151 glucose,¹²⁰ diamine,^152–154 and N-acetylcysteine.¹⁵⁵ The precursor is selected based on its compatibility with the synthesis technique of CNDs. Each precursor confers distinct physical and chemical characteristics to CNDs. The optical properties of CNDs are significantly affected by the level of conjugation in the precursor molecules.¹⁵⁶ Precursors exhibiting a greater degree of conjugation facilitate the development of larger π-conjugated domains within the CND core, so efficiently reducing the bandgap and enhancing red to NIR emission. Conversely, CNDs derived from nonconjugated or weakly conjugated compounds frequently exhibit inadequate π-conjugation, resulting in elevated band gaps and consequently blue-shifted visible emission. This tendency underscores the necessity of carefully selecting precursor materials to control the optical characteristics of CNDs. The utilization of large and extended conjugated precursors offers a potential approach for engineering CNDs with preferred long-wavelength luminescence for advanced optical and bioimaging applications. On the basis of precursor selection, longer-wavelength emissive CNDs with ultrahigh QY can be achieved by using aromatic or heteroatom-rich precursors that promote extended π-conjugation and tune the PL properties.

5.2 Emission dependency on solvothermal conditions

Solvothermal conditions also play a key role in determining the structure and emission color of CNDs.^78,157–160 Recently, CNDs have been directly synthesized from the same precursor using different solvents to obtain longer wavelength or multicolour-emissive CNDs. The extent of dehydration and carbonization, which governs the size of conjugated pi domains, can be adjusted using different solvents, thereby modulating the emission of CNDs. In the study by Ru et al., blue, yellow, and red-emissive chiral carbonized polymer dots (CPDs) were synthesized via a solvothermal method using CA and different chiral diamines (such as (R,R)-1,2-diphenylethylenediamine) in DMF (Fig. 8a).¹⁵⁸ The reaction mixtures were sealed in Teflon-lined stainless-steel autoclaves and heated at 180 °C for 12 h. After natural cooling, the crude products were purified by sequential washing with HCl, ethanol, deionized water, and acidified water to remove impurities. Final purification was done by dialysis using a 1000 Da cutoff membrane. By selecting different chiral amines and tuning the carbonization process, CPDs with blue, yellow, and red PL were obtained, demonstrating excellent tunability in emission color and circularly polarized luminescence properties. In another study by Guo et al. (2024), multicolor-emissive CNDs were synthesized via a solvent-engineered solvothermal method using CA and cyanamide as precursors.¹⁵⁷ The synthesis was carried out in various solvent systems, including water, DMF, toluene, and formamide to tailor the emission color (Fig. 8b). The precursor mixtures were placed in Teflon-lined stainless-steel autoclaves and heated at 200 °C for 8 h to form B-CNDs, C-CNDs, Y-CNDs, and W-CNDs. After cooling to room temperature, the crude products were dialyzed and purified. The choice of solvent significantly influenced the reaction pathway, carbonization degree, and surface states, leading to CNDs with tunable PL. Ding et al. synthesized multicolour fluorescent CNDs using L-glutamic acid and OPDA as starting materials through a solvent-engineered strategy (Fig. 8c).⁷⁸ The selection of solvents in the CND synthesis controlled the dehydration and carbonization processes of the precursors and directly influenced the dimensions of the sp²-conjugated domains and the graphitic N content in each CND sample, ultimately leading to emission from blue to NIR. Zhan et al. developed a solvent-engineered solvothermal synthesis strategy for CNDs using 1,3,6-trinitropyrene (TNP) as the precursor.¹⁶¹ Under mild solvothermal conditions (230 °C for 12 h) in different solvent systems of DMF, ethanol, water, and acetic acid, they produced CNDs with tunable fluorescence from blue to red (Fig. 8d). This method allowed precise control over size and bandgap, yielding excitation-independent emission, high QY (up to 59%), and excellent scalability (50–92% yield). The scalable strategy represents a promising route for fabricating multicolor CND synthesis suitable for bioimaging and optoelectronics. In a related study by Wang et al. (2020), full-color CNDs were synthesized using a one-pot solvothermal method.¹⁶² The process involved OPDA as the carbon precursor and various acid reagents such as 4-aminobenzenesulfonic acid (4-ABSA), folic acid (FA), boric acid (BA), acetic acid (AA), terephthalic acid (TPA), and tartaric acid (TA) in ethanol solvent (Fig. 8e). The reaction mixture was sealed in a Teflon-lined stainless-steel autoclave and heated at 180 °C for 12 h. This solvothermal environment facilitated the incorporation of electron-donating and electron-withdrawing groups into the CNDs, leading to tunable PL emission from blue to red and even white light. The acid reagents modulated the particle size and surface functionalization of the CNDs, resulting in distinct emission profiles without requiring complex post-synthesis treatment.

Fig. 8 (a) Diagram illustration of the preparation procedure for multicolor-emitting chiral carbonized PDs with full-color circularly polarized luminescence. Reproduced from ref. 158 with permission from Wiley-VCH, Copyright 2021. (b) Schematic diagram of the preparation process of multicolor-emissive CNDs using CA and cyanamide as starting materials. Reproduced from ref. 157 with permission from American Chemical Society, Copyright 2024. (c) A solvent-engineered approach for synthesis of multicolored fluorescent CNDs utilizing L-glutamic acid and OPDA as starting precursors. Reproduced from ref. 78 with permission from Wiley-VCH, Copyright 2018. (d) A solvent-engineered molecular fusion approach for high-yield production of multicolor fluorescent CNDs using TNP as active monomers, DMF or EtOH as the single or main solvent, and H₂O or CH₃COOH as the auxiliary solvent. Reproduced from ref. 161 with permission from Elsevier, Copyright 2018. (e) An acid reagent engineering strategy for the synthesis of full-color fluorescent CNDs using OPDA as the precursor, along with fluorescence photographs of CNDs under UV light (excited at 365 nm). Reproduced from ref. 162 with permission from American Association for the Advancement of Science, Copyright 2020.

5.3 Reaction time-dependent emission

CND synthesis involves pyrolysis or carbonization, along with condensation, dehydration, and polymerization, all of which depend on reaction time.¹⁶³ In the initial stage, organic molecules rapidly form intermediates through radical reactions, Schiff base condensation, and aldol condensation. Over moderate reaction times, these intermediates undergo further polymerization or aggregation via covalent and noncovalent interactions. With extended reaction time, the resulting aggregates and polymerized structures undergo carbonization to form the carbonaceous core of the CNDs. Papaioannou et al. synthesized CNDs via a hydrothermal approach from glucose as a carbon precursor with reaction times ranging from 2 to 12 h.¹⁶⁴ TEM indicated a reduction in particle size, accompanied by increased monodispersity across the reaction duration. The hypothesized formation mechanism of CNDs comprises four stages: precursor breakdown, polymerization/aromatization, nucleation, and growth. HRTEM and diffraction patterns revealed a decrease in crystalline phases with increasing reaction durations.

Subsequently, Kalita et al. synthesized GQDs with various sizes (2, 4, 5.2, 6.5 nm) utilizing rice grains as a precursor via a one-pot pyrolysis approach at distinct heating durations (3, 5, 7, and 10 min).¹⁶⁵ The synthesis mechanism of the GQDs involves thermal decomposition of starch into glucose oligomers, followed by nucleation and carbonization of these glucose oligomers at elevated temperature and the subsequent growth of the conjugated domains through pyrolysis (Fig. 9a). This approach produced GQDs exhibiting a redshift in emission as particle size increased from 2 nm to 6.5 nm, but their QY diminished from 24% to 16% with the increase in particle size. Ehrat et al. arrested the synthesis at specific intervals and found that CNDs grew rapidly within 30 min and maintained their size throughout the remaining synthesis duration. Rather than increasing in size, the CNDs underwent a significant alteration in their internal structure.⁴ Initially, MF resembling citrazinic acid structures prevailed, producing intense blue emission with elevated QY and extended lifetimes. With increasing reaction time, these MF developed into aromatic domains, resulting in a red shift in emission and changes in photophysical properties. Fig. 9b illustrates the PL behavior over time. Early stage CNDs showed strong, stable emission with slower decay due to MF. As aromatic domains grew, emission became faster decaying, and a shoulder at longer wavelengths appeared, indicating a more graphitic character. The blue line represents the decrease in the fluorophore-related slow PL decay component, while growth of aromatic domains is depicted in a red circle. Fig. 9c shows how CNDs evolved during synthesis: initially, MF such as citrazinic acid formed quickly and dominated the PL.⁴ With prolonged synthesis, these fluorophores facilitated the growth of aromatic domains and drove a structural transition from N-rich to more carbon-rich cores. This temporal progression highlights the capacity to tune the optical properties of CNDs by controlling synthesis duration, providing a simple yet effective strategy for tailoring PL emission toward targeted applications. Zhao et al. showed that reaction time significantly affects the emission color of the CNDs. Blue and green CNDs were synthesized at 200 °C for 8 h, resulting in smaller, less carbonized structures with higher band gaps, leading to blue and green emissions (Fig. 9d).¹⁶⁶ In contrast, red CNDs required a longer reaction time of 12 h, allowing greater carbonization and extended conjugation, which lowered the bandgap and produced red-shifted emission. Thus, longer reaction times favor red emission, while shorter times yield blue or green fluorescence. Yuan et al. demonstrated that a solvothermal reaction time of 6–12 h yields high crystallinity, sharp emission peaks, and well-defined graphene-like structures of the PR- and PG-NBE-T-CQDs (Fig. 9e).⁸⁰ This time duration allows for suitable carbonization, crystal growth, and the development of their excellent optical properties.

Fig. 9 (a) A plausible mechanism for the synthesis of GQDs from rice powder. Reproduced from ref. 165 with permission from Royal Society of Chemistry, Copyright 2016. (b) Carbon content in the CNDs (left) and ratio between long and short decay components of the fluorescence lifetime (right) throughout the synthesis of the CNDs. These characteristics represent the relative abundance of small organic fluorophores similar to citrazinic acid (shown in a blue circle) and aromatic domains (depicted in a red circle). Reproduced from ref. 4 with permission from American Chemical Society, Copyright 2017. (c) Schematic illustration of the possible formation of sp2-hybridized aromatic domains or molecular fluorophores during the formation of CNDs from CA and EDA. Reproduced from ref. 4 with permission from American Chemical Society, Copyright 2017. (d) Time-dependent synthesis of blue, green, and red-emitting CNDs. The photographs of dispersed blue, green, and red-emitting CNDs dissolved in ethanol in daylight (left), and under excitation of 365 nm (right). Reproduced from ref. 166 with permission from Elsevier, Copyright 2019. (e) Time-dependent synthesis of PR and PG-NBE-T-CNDs by solvothermal treatment of resorcinol in two different durations. The photographs of PR and PG-NBE-T-CNDs under daylight (left) and fluorescence images (right) under UV light (excited at 365 nm). Reproduced from ref. 80 with permission from Springer Nature, Copyright 2019.

5.4 Temperature-dependent PL

CNDs show temperature-dependent optical properties.^167–170 The hydrothermal approach is frequently employed and involves a temperature-dependent polymerization and carbonization process.^171,172 The temperature variation during carbonization leads to a change in the size and PL properties of the CNDs, hence enhancing the polymer density over the CND's core. The particle size, QY, and characteristic fluorescent peaks of CNDs are controlled by adjusting the reaction temperature. Zhang et al. synthesized six types of CNDs using different hydrothermal temperature conditions, using CA and ammonium solution (Fig. 10a).¹⁶⁷ At reduced temperatures (120–160 °C), non-fluorescent precursors condensed to form short fluorescent polymer chains that expanded and crosslinked into polymer-like CNDs devoid of a carbon core, resulting in an enhancement of QY. As the temperature increased to 160–180 °C, these polymer-like CNDs underwent carbonization, yielding a carbon core while surface fluorescent polymer chains continued to form, culminating in the optimal QY due to the equilibrium between polymer chain growth and carbon core expansion. Above 180 °C, additional carbon core development depleted these fluorescent chains more rapidly than they could be generated, resulting in a reduction in their quantity and a decrease in QY. Tang et al. reported the dual PL bands of CNDs, which were ascribed to emissions from core and surface states.¹⁷⁰ The two emission bands had exhibited similar temperature dependency, with temperature fluctuations illustrated in Fig. 10b.

Fig. 10 Molecule and core-state formation during temperature variation. (a) Schematic illustration of the temperature-dependent possible formation mechanism of CNDs. Reproduced from ref. 167 with permission from Elsevier, Copyright 2016. (b) Temperature-dependent fluorescence of CNDs. Reproduced from ref. 170 with permission from American Chemical Society, Copyright 2012. (c) A schematic of the CNDs obtained at different hydrothermal temperatures. The absorbance, QY and dependence value of CND solution synthesis at different temperatures. Reproduced from ref. 173 with permission from Royal Society of Chemistry, Copyright 2015.

Moreover, elevated temperature facilitates an increased degree of carbonization as well as enhanced graphitization, leading to an expanded conjugation structure, which includes subdomains with a larger size and molecular state-related structure with extended conjugation, etc. Consequently, the bandgap can become narrower, resulting in potential red-shifted emission. The carbonization temperature can induce a shift in emission wavelength and a variation in PL intensity. Furthermore, Krysmann et al. demonstrated the emission characteristics of CPDs synthesized via thermal treatment using the mixture of CA and ethanolamine. The molecular state is transformed into the carbon core state, resulting in a reduction of the PL contribution from the molecular state and an increase in the PL contribution from the carbon core state.²⁹ Song et al. demonstrated that the highly photoluminescent IPCA molecular state transforms into a highly carbonized carbon core state at elevated synthesis temperature, leading to a reduction in QY, shown in Fig. 10c.¹⁷³ Shamsipur et al. conducted a systematic analysis of the possible formation mechanism of CPDs from CA and EDA, as well as the change of PL centers under different carbonization degrees.¹⁷⁴ At low carbonization levels, fluorescence predominantly originated from the molecular state 2,3-dihydroimidazo[1,2-a]pyridine-5(1H)-one. Subsequent carbonization led to the formation of larger conjugated structures inside polymer/carbon hybrid structures, which contributed to CPD fluorescence via conjugated subdomain states. Further study indicated that the conjugated subdomain state can be not only the planar conjugated structure, but also the curved polycyclic aromatic hydrocarbon structures, which was experimentally supported by the mass spectrum and absorbance spectrum. Ultimately, the carbonization leads to the formation of CNDs exhibiting PL emission from surface states or carbon core states (Fig. 11). Vercelli et al. reported the hydrothermal synthesis of nitrogen-doped CNDs using CA or glucose as carbon sources and urea as the nitrogen precursor.¹²⁰ The study demonstrated that reaction temperature plays a crucial role in determining the PL origin and emission behavior of the CNDs. At 160 °C, the PL emission mainly arises from nitrogen-related surface states, resulting in blue emission with a single emissive center. In contrast, synthesis at 200 °C promotes the formation of graphitic nitrogen-oxide centers, which introduce additional emissive states and lead to dual emissive centers with modified PL decay dynamics and improved stability, shown in Fig. 12.

Fig. 11 Schematic representation of the synthesis of CNDs from CA citric acid and EDA through the pyrolysis. Reproduced from ref. 174 with permission from American Chemical Society, Copyright 2018.

Fig. 12 Proposed PL mechanisms in CNDs synthesized at different reaction temperatures (a) 160 °C and (b) 200 °C. Reproduced from ref. 120 with permission from Wiley-VCH, Copyright 2025.

5.5 pH-dependent PL

CNDs exhibit superior properties, including surface functionalization, intense fluorescence emission, monodispersity, biocompatibility, photostability, and enhanced water solubility. However, the influence of pH during CND synthesis remains a relatively novel and underexplored aspect in this field. Previous studies have rarely addressed this variable, which motivates our investigation into how pH affects the formation and PL properties of CNDs. Several studies indicate that pH variations from 1 to 14 can influence the PL behavior of CNDs through protonation–deprotonation processes, changes in energy levels, aggregation of CNDs, protective shell formation, and proton transfer.^175,176 Polymerization, crosslinking, and carbonization of carbon precursors are key steps in the synthesis of multicolour CNDs. These processes are strongly influenced by pH, which controls the size of sp² domains and the graphitic N content, thereby tuning the photoluminescent emission colors. Dai et al. use 2,3-diaminopyridine as a single precursor to synthesize different color CNDs under varying pH conditions (Fig. 13a).¹⁷⁶ In addition, Lu et al. reported that small organic molecules, such as sugars or amino acids, are carbonized and thermally broken down in a bottom-up approach, frequently with the help of hydrothermal or microwave processes.¹⁷⁷ For instance, CNDs with tunable PL based on synthesis pH are produced by heating L-serine and L-tryptophan in alkaline conditions at 200 °C (Fig. 13b). A smooth carbon sheet with embedded dots appears at alkaline pH (∼8.1), indicating significant PL. Layered nanosheets with increased PL and better regularity appear at somewhat acidic pH values (3–7). The formation of small, amorphous, N-rich CNPs with the highest QY (46.83%) occurs at strongly acidic pH values (1–3). Dehydration and N doping are improved by the higher acidity, which results in smaller particles with more surface defects that increase luminescence. Therefore, pH regulates both optical performance and structural evolution. Wang et al. developed lipid droplets and mitochondria-specific CNDs via a solvothermal approach as well as an acid-controlled synthetic strategy (Fig. 13c).¹⁷⁵ They found that distinct MF show that the precursors experience distinct reaction pathways in acidic and neutral environments. Overall, these studies demonstrate that pH plays a crucial role in controlling the structure and surface chemistry of CNDs by influencing carbonization, doping, and functionalization, thereby modulating the sp²-conjugated core size, surface characteristics, dispersibility, degree of functionalization, PL properties, and organelle-targeting capability.

Fig. 13 (a) pH-dependent synthesis of CNDs influencing the sp² domain, carbonization, and red shift in emission. Reproduced from ref. 176 with permission from Elsevier, Copyright 2022. (b) A schematic illustration of the preparation process for various structural CNDs by hydrothermal carbonization of L-serine and L-tryptophan at several pH values and temperatures. Reproduced from ref. 177 with permission from American Chemical Society, Copyright 2016. (c) pH-controlled synthesis of green and orange CNDs. Reproduced from ref. 175 with permission from American Chemical Society, Copyright 2025.

5.6 Heteroatom doping-dependent emission

One effective method to enhance the PL efficiency of CNDs is heteroatom doping. Generally, organic molecules serving as carbon sources are carbonized in the presence of N-, Si-, B-, O- or S-containing compounds, as well as co-doping with Mg/N, S/N, N/P, resulting in heteroatom-doped CNDs.^{11,12,78,178–182} By enhancing effective radiative electron transitions and inhibiting nonradiative routes, N doping, specifically, the addition of graphitic, pyrrolic, and pyridinic N, is essential for increasing QY. By adding structural flaws, altering surface states, and raising the likelihood of radiative transitions, S doping further enhances QY. When S-containing precursors are employed, S-doped CNDs frequently exhibit redshifted emission and significant QY enhancements, reaching values as high as ∼90%. In conclusion, elemental doping effectively modifies the electronic structure and surface chemistry of CNDs, allowing for notable increases in QY, emission tunability, and luminescence intensity.

6. Origin and mechanistic insights of PL in CNDs

The PL mechanism of CNDs has been widely investigated, yet it remains complex due to the presence of multiple emissive centers. Generally, the PL behavior of CNDs is attributed to contributions from the sp² carbon core, surface functional groups, edge defects, and MFs formed during synthesis. Variations in size, surface chemistry, heteroatom doping, and structural defects can significantly influence the emission properties. As a result, different mechanisms such as core-state emission, surface-state emission, defect-mediated emission, and molecular-state emission have been proposed to explain the PL origin in CNDs. Additionally, purification studies have shown that MF, quasi-CNDs, and fully carbonized CNDs can coexist, and each contributes differently to the observed emission behavior. Understanding these mechanisms is essential for tuning the optical properties of CNDs for applications in bioimaging, sensing, and optoelectronics.

6.1 Core-state emission

Core-state emission in CNDs originates from the sp²-conjugated carbon core, where the electronic structure and bandgap determine PL properties. Factors such as size of the sp² domains, heteroatom doping, and structural defects can modify the HOMO–LUMO gap, leading to tunable emission wavelengths. Thus, the intrinsic electronic transitions within the carbon core play a key role in governing PL behavior. Song et al. synthesized deep-ultraviolet (DUV) emissive CNDs via a solvothermal reaction of phenylenediamine (o-, m-, or p-) and methyl red in DMF.¹⁸³ The CNDs exhibited strong DUV emission centered at ∼290 nm with a QY of 31.6%. The emission originated from intrinsic carbon core states through π–π* carrier radiative recombination, while surface functional groups mainly passivated the carbon core and suppressed defect emission. Lin et al. initially investigated CNDs’ chemiluminescence (CL) features in the presence of oxidizing agents, including KMnO₄ and cerium(IV).¹⁸⁴ Electron paramagnetic resonance confirmed these oxidants introduce holes into CNDs. This mechanism enhances the density of holes in the CNDs and accelerates electron–hole recombination, resulting in energy release in CL emissions. Furthermore, the intensity of the CL depends on the concentration of the CNDs within a specific range. An increase in temperature was shown to positively influence the CL, as the electron distribution in the CNDs achieved thermal equilibrium, as illustrated in Fig. 14a. The CL of CNDs is believed to result from the formation of positively and negatively charged CNDs inside the system. These then recombine to generate CNDs in the excited state, which emit CL upon transitioning to the ground state. Similarly, Holà et al. synthesized CNDs via a two-step process using CA and urea as starting materials.¹⁸⁵ The various luminescent fractions were then separated according to charge using column chromatography. Red-emitting CNDs with the highest negative charge were found in the final fraction (Fig. 14b). XPS, FT-IR, Raman spectroscopy, and DFT calculations demonstrated that the growing amount of graphitic N in the CNDs’ structure was the source of the observed trend in PL red shift, surface charge, and column chromatography separation. Light is absorbed at a significantly lower energy level than usual in undoped systems due to the creation of mid-gap states by graphitic N in the HOMO–LUMO gap. According to the findings, graphitic N is another crucial element that can red-shift the PL of CNDs. Hu et al. demonstrated that silane-functionalized CNDs synthesized from CA, p-phenylenediamine (PPDA), and APTES via a one-step microwave-assisted method exhibit tunable full-color solid-state fluorescence spanning blue (438 nm) to red (633 nm), primarily governed by carbon core states rather than surface states (Fig. 14c).¹⁸⁶ The multicolor emission was attributed to the synergistic increase in sp²-conjugated domain size and graphitic nitrogen content within carbon cores, both of which progressively narrow the HOMO–LUMO energy gap, resulting in a progressive red shift in PL. Time-resolved PL decay analysis revealed two recombination pathways: a fast component (τ₁ ≈ 1.0 ns) associated with core-state recombination and a slow component (τ₂ ≈ 7.0 ns) linked to surface-state recombination. With increasing emission from blue to red, core-state recombination becomes progressively dominant due to the growth of sp²-conjugated domains. This clearly confirms the central mechanistic role played by the carbon core.

Fig. 14 Proposed PL mechanisms in CNDs (a) Schematic illustration of FL and CL mechanisms in CNDs-KMnO4 and CNDs-cerium(IV) systems. CL1 and CL2 represent two CL routes in the system. Reproduced from ref. 184 with permission from Royal Society of Chemistry, Copyright 2012. (b) Schematic diagram showing graphitic N induces midgap states within HOMO–LUMO gap of undoped system, leading to red shift in absorption, which subsequently produces fluorescence at a lower energy of the visible spectrum. Reproduced from ref. 185 with permission from American Chemical Society, Copyright 2017. (c) Formation mechanism and carbon core-driven PL mechanism of full-color solid-state silane-functionalized CNDs. Reproduced from ref. 186 with permission from Elsevier, Copyright 2023.

6.2 Surface-state emission

Surface-state emission in CNDs originates from surface functional groups and defects, which create additional energy levels that influence PL behavior. Changes in surface oxidation and chemical environment can modify these states, resulting in tunable emission properties. Qu et al. reported that the CNDs synthesized from CA and urea exhibit a distinct absorption band at 540 nm and excitation-independent orange PL.³⁵ This behavior indicates that the emission mainly originates from the band-gap transition of conjugated sp² domains in the nitrogen-rich graphene-like carbon core. Ding et al. demonstrated that fluorescence is predominantly associated with the surfaces of the CNDs, and that the surface states govern the PL characteristics.¹¹⁸ The photoluminescent features of the reported CNDs resemble certain forms of molecular fluorescence, leading scientists to hypothesize that the luminescent centers on the CNDs’ surfaces predominantly consist of conjugated carbon and oxygen atoms that are chemically bonded, as illustrated in Fig. 15. The HOMO and LUMO gap of this structure depends upon the integrated O species. The band gap diminishes with increased O atoms within the structure. This occurs due to increased surface oxidation, resulting in a red shift in PL. The pH-dependent behavior of the red-emitting CNDs was investigated to further show that the tunable PL originated from the surface states. Lei et al. systematically investigated the surface state modulation mechanism governing QY and product yield in blue-emitting CNDs synthesized from different carbon sources and silane coupling agents with varying oxygen and nitrogen-containing functional groups (Fig. 16a).¹⁸⁷ The PL origin was unambiguously attributed to surface-state luminescence rather than quantum size effects, as confirmed by the absence of correlation between particle size, graphitization degree, and optical properties. Specifically, high content of C=O bonds and pyrrolic N on the CND surface were identified as the two key structural determinants of high QY, as both promote radiative electron–hole recombination by enhancing surface passivation and increasing exciton capture center density. CNDs with excitation-independent emission and longer fluorescence lifetimes, indicating a well-passivated surface energy level, exhibited high QY values of up to 97.32%. Time-resolved PL decay analysis further confirmed that the long-lifetime component (τ₂), attributed to surface-state recombination, dominates the PL of high-QY CNDs, while the short-lifetime component (τ₁) arising from carbon core states, contributes minimally. These findings establish a clear mechanistic principle: the PL of these CNDs originates predominantly from surface states, and rational control of surface functional group composition, particularly the C=O and pyrrolic N content, is the primary strategy for engineering high-QY CNDs. Zhao et al. demonstrated that the PL origin in CNDs can be exclusively attributed to surface states arising from -NH₂ functional groups (Fig. 16b).¹⁸⁸ To achieve this, electrochemical synthesis from graphite rods was used, which modifies only the surface of CNDs without introducing nitrogen into the carbon core. Initial CNDs (I-CNDs) containing only oxygen-related surface states emitted green light with an energy level gap of approximately 2.3 eV. Progressive introduction of amino groups via 1,8-diaminonaphthalene, OPDA, and a combination of OPDA and PPDA systematically shifted the emission from blue (450 nm, 2.8 eV) to green (558 nm, 2.2 eV) to red (608 nm, 2.0 eV), confirming that –NH₂ surface states directly narrow the HOMO–LUMO energy gap through electron donation and elevation of HOMO levels. Quantum size effects and carbon core modification were rigorously excluded as contributing factors, as all four CND variants exhibited nearly identical particle sizes (1–2 nm) and comparable I_D/I_G ratios in Raman spectroscopy. Zhang et al. demonstrated that panchromatic luminescent CNDs synthesized from 4,4-bipyridine and PPDA via a one-step solvothermal method exhibit full-spectrum emission spanning from purple (441 nm) to red (627 nm) through a surface-state-based PL mechanism (Fig. 16c).¹⁸⁹ The surface functional group C=N and surface defect states are primarily responsible for long-wavelength red-shifted emission, while C–O–C and O–H groups are responsible for blue-shifted emission. By systematically tuning the chain length, polarity, and protonation of the polyol solvent, the surface-state composition of the CNDs was precisely modulated without altering the core structure, thereby enabling fine spectral control through a unified surface-state regulation mechanism. In this study, quantum size effects, graphitic nitrogen content, and bandgap transitions were systematically excluded as dominant contributors to the multicolor emission, confirming the dominant role played by surface states in governing the multicolor PL mechanism. The surface-state-dominated CNDs further demonstrated selective detection of Fe³⁺ through dynamic fluorescence quenching and multicolor cell-labeling imaging, highlighting the practical utility of surface-state engineering in CND design.

Fig. 15 Schematics illustrating the tunable PL of CNDs with varying degrees of oxidation. Reproduced from ref. 118 with permission from American Chemical Society, Copyright 2016.

Fig. 16 (a) Schematic of the synergistic influence mechanism of surface states arising from oxygen- and nitrogen-containing functional groups on the origin of PL in CNDs. Reproduced from ref. 187 with permission from Royal Society of Chemistry, Copyright 2022. (b) Schematic structure and energy level diagram of oxygen- and nitrogen-related surface states governing the origin of PL in CNDs. Reproduced from ref. 188 with permission from Elsevier, Copyright 2022. (c) Schematic illustration of the synthesis strategy and surface state-based PL mechanism of panchromatic CNDs, where C=N surface defect states govern long-wavelength red-shifted emission and C–O–C and O–H surface states drive blue-shifted emission. Reproduced from ref. 189 with permission from Elsevier, Copyright 2025.

6.3 Core and surface emission

In many CNDs, PL arises from the combined contribution of carbon core states and surface-related emissive sites.¹⁹⁰ The radiative and nonradiative processes of photoexcited electrons involve transitions between sp²-related electronic states and defect or trap states present on the surface. Interactions between CND core and surface states govern the relaxation pathways of excited electrons, thereby influencing the fluorescence and phosphorescence behavior of CNDs. There are several reports which demonstrated that the PL heterogeneity of CNDs arises from the distinct contributions of the core and surface states.¹²⁰ The core state, rich in graphitic nitrogen, exhibits strong and efficient PL, whereas the surface state, containing oxygen-functional groups, shows weaker and less efficient PL due to electron-scavenging effects. Their findings highlight that controlling the core and surface composition is critical for tuning PL properties and designing CNDs with tailored emissive behavior for metal-free light-harvesting and other photonic applications. Jiang et al. elucidated the mechanism by which fluorescent CNDs (F-CNDs) converted into phosphorescent CNDs (P-CNDs) upon thermal activation (Fig. 17a).¹⁹¹ In F-CNDs, fluorescence originates from surface-bound fluorophore units within loosely organized polymer chains, generating transient emission through singlet excited states. Following thermal treatment, the F-CNDs yield compact carbonized cores (P-CNDs), immobilizing triplet excitons via intraparticle hydrogen bonding and inhibiting nonradiative decay. The structural rearrangement and N and phosphorus doping that promote intersystem crossover (ISC) permit long-lived room-temperature phosphorescence (RTP) from P-CNDs. Xu et al. proposed energy level structures to explain the behavior of photoexcited electrons in GO sheets and GQDs.¹⁹² The proposed model accounts for nonradiative relaxation of electron from higher-to-lower-lying defect states, thermal decay into nonradiative traps, and their radiative recombination from discrete sp²-related and continuous defect states (Fig. 17b).

Fig. 17 Proposed PL mechanisms in CNDs (a) Schematic representation of the FL and RTP mechanisms for the F-CNDs and P-CNDs. Reproduced from ref. 191 with permission from Wiley-VCH, Copyright 2018. (b) Energy level structures to elucidate the optical characteristics of photoexcited electrons in both GO sheets and GQDs. Reproduced from ref. 192 with permission from American Chemical Society, Copyright 2013.

6.4 Edge states of the sp² carbon core and molecular-state emission

PL in CNDs can also originate from the edge states of the sp² carbon core along with molecular fluorophores present on the surface. These emissive centers contribute to dual emission, where both edge-related electronic states and surface fluorophores influence the fluorescence behavior. Their emission properties are often sensitive to the surrounding environment, such as solvent polarity and hydrogen bonding, which can modulate the PL characteristics of CNDs. Nidhisha V. et al. reported that PL of PPDA-derived CNDs originates primarily from edge states/mid-gap states and surface fluorophores rather than the carbon core.¹⁹³ The emission observed in the 400–500 nm region is attributed to edge-state excitation, while the longer-wavelength emission (≥500 nm) arises from surface fluorophores, confirming the dominant roles they play in the PL mechanism of these CNDs.

Basu et al. presented a simplified Jablonski diagram illustrating the proposed mechanism of fluorescence arising from distinct emissive regions of the CNDs.¹⁹⁴ The emission spectra in all solvents display two main components, corresponding to two distinct populations of emission centers: the edge states at the periphery of sp²-hybridized CNDs core and the surface fluorophores (Fig. 18). Both emission centers exhibit sensitivity to solvent polarity and hydrogen bonding capacity. They exhibit significant red shift in relation to the solvent ET(30) polarity parameter, indicating the same polar characteristics of these centers and their strong interactions with solvent molecules. Nonetheless, between these two emission centers, surface fluorophore emissions are significantly suppressed in high ET(30) solvents, indicating the emergence of new nonradiative relaxation pathways. The differential response of these two emission centers to solvents allows the CNDs to function as ratiometric sensors for solvent polarity. Singaravelu et al. reported that the PL of nitrogen-doped CNDs arises from multiple emissive states including the graphitic core, edge states, and surface functional states, where photoexcitation initially occurs in the carbon core followed by relaxation through edge and surface states, ultimately resulting in green fluorescence emission.¹⁹⁵

Fig. 18 A simplified Jablonski diagram illustrating the most probable mechanism of fluorescence arising from different regions of the CNDs. Reproduced from ref. 2 with permission from American Chemical Society, Copyright 2018.

6.5 Core and molecular-state emission

The PL of CNDs can also arise from the combined contribution of carbon core states, surface trap states, and MF formed during synthesis. In many systems, CNDs contain molecular species, polymer structures, and carbonized cores that act as multiple emissive centers. As a result, dual-emission behavior is often observed, where MF produce excitation-independent emission, while surface traps on the carbon core generate excitation-dependent PL. Understanding the interplay between these emissive centers is therefore essential for rationally designing CNDs with controlled emission.

Comprehensive structural and molecular studies have gradually elucidated the fundamental mechanisms of PL in CNDs. Fig. 19 depicts a representative dual-emission process emphasizing the roles played by both molecular and carbon-core-based emissive centers. During synthesis, CNDs form through multiple molecular fusion and carbonization pathways. For example, Song et al. reported that CNDs consist of a type of fluorescent molecule (imidazo[1,2-a]pyridine-7-carboxylic acid, 1,2,3,5-tetrahydro-5-oxo-, IPCA), polymers, and carbon cores, featuring two primary PL centers (Fig. 19a).¹⁷³ IPCA exhibits robust blue emission that is independent of excitation, whereas the carbon core induces excitation-dependent emission attributed to surface traps. This dual-emission model elucidates the PL behavior of CA-based CNDs and underscores the significance of both molecular and carbon core constituents. Despite the ambiguity surrounding the PL mechanisms of CNDs, their remarkable photoluminescent spectra persistently inspire researchers to pursue further investigations. Macairan et al. demonstrated that CNDs synthesized from glutathione and formamide exhibit dual fluorescence originating from two physically distinct emissive centers: carbon-core state and the molecular state (Fig. 19b).¹⁹⁶ The carbon-core state arises from the sp²-conjugated carbon network, which is responsible for blue emission, while the molecular state originates from fluorophore-like surface moieties, giving rise to red emission. Femtosecond transient absorption spectroscopy and electrochemical analysis confirmed this two-state model, revealing that fluorescence proceeds either through direct electron–hole radiative recombination within each state or via energy transfer from the core state to the molecular state followed by radiative recombination. Progressive carbonization with increasing reaction time promoted growth of the sp² core network at the expense of the molecular state, enabling controlled decoupling of the two emissive centers and facilitating selective attribution of blue and red emission to their respective origins.

Fig. 19 (a) Diagram illustrating the relationship among different products within the one-pot hydrothermal system of CA and EDA. Reproduced from ref. 173 with permission from Royal Society of Chemistry, Copyright 2015. (b) Schematic illustration of dual-fluorescence in CNDs synthesized from glutathione and formamide via solvothermal reaction at varying reaction times. The blue and red emission signatures arise from carbon-core states and molecular states, respectively. Reproduced from ref. 196 with permission from Elsevier, Copyright 2022.

6.6 Defect-mediated emission

Defect-mediated emission is an important mechanism governing the PL of CNDs. During synthesis or post-synthetic treatments, processes such as oxidation, heteroatom doping, and structural reconstruction introduce defects within the carbon framework or on the surface. These defects generate localized electronic states within the band structure that act as trapping centers for photoexcited charge carriers. The radiative recombination of electrons and holes through these defect-induced energy levels results in fluorescence emission. The emission properties can be tuned depending on the type, density, and distribution of defects as well as the associated surface functional groups.

Bao et al. synthesized carbon nanodots by nitric acid oxidation of carbon fibers followed by ultrafiltration separation.⁷ The obtained CNDs exhibited tunable PL ranging from blue to red. Spectroscopic analysis suggested that emission originates from defect-mediated surface states formed due to oxidation-induced surface functional groups, where exciton recombination at these surface sites produces fluorescence. Zhang et al. demonstrated that oxidation of CNDs using IBX introduces oxygen-related defect sites in the carbon core by substituting graphitic nitrogen with oxygen atoms. These defects create localized electronic states and unpaired electrons, which act as electron-trapping centers and modify the electronic structure of the nanodots. As a result, enhanced NIR absorption and emission were observed. This indicates that the PL mainly originates from defect-mediated electronic states rather than intrinsic core emission. Shabbir et al. synthesized nitrogen-doped CQDs via a hydrothermal method using CA and urea as precursors.¹⁹⁷ The resulting CQDs contained abundant oxygen and nitrogen-containing surface functional groups, and the PL mainly originated from surface defect states associated with these functional groups.

6.7 Molecular state emission

Molecular state emission in CNDs originates from fluorescent molecular species formed during the precursor transformation and carbonization process. These MFs can act as independent emissive centers and significantly influence the PL properties of CNDs. In many cases, the optical behavior of CNDs is strongly related to the structure and chemical state of these fluorophore molecules, which may remain attached to or embedded within the graphitic carbon framework. Zhang et al. synthesized CNDs simultaneously exhibiting one-photon red FL emission (620 nm) and NIR 808 nm-induced two-photon red FL emission (630 and 680 nm), prepared by protonating CNDs derived from OPDA.¹⁹⁸ Protonation of the 2,3-diaminophenazine (2,3-DAPN) fluorophore significantly alters the chemical state of CNDs, lowering the photon transition bandgap and causing one-photon and NIR-induced two-photon high-color-purity red fluorescence emission. Furthermore, their discovery of 2,3-DAPN as the CND PL determinant implies that fluorophore products of precursor conversion may serve as a predictor for achieving targeted PL features, since 2,3-DAPN fluorophore is the OPDA oxidation product. As a result, this novel approach for controlling and predicting high-color-purity red emission on CNDs could serve as a roadmap for fabricating high-performance CNDs for a variety of diverse applications. Righetto et al. synthesized CNDs through a solvothermal treatment of phenylenediamine isomers (ortho-, meta-, and para-phenylenediamine) in ethanol at 180 °C for 12 h. Detailed spectroscopic investigations, including fluorescence correlation spectroscopy (FCS), HPLC-MS, and NMR, revealed that the observed PL originates predominantly from MFs formed during the synthesis rather than from the carbon core.¹⁹⁹ These emissive molecules are small diazocyclic compounds generated through self-condensation reactions of phenylenediamine, which remain freely dispersed in solution and dominate the fluorescence behavior of the system. Li et al. prepared a MF named 5,14-dihydroquinoxalino[2,3-b] phenazine (DHQP) and derived CNDs by a sequential reaction procedure (Fig. 20).²⁵ Initially, OPDA interacts with catechol (CAT) to produce intermediates such as 2,3-DAPN and phthalazine. The intermediates subsequently react with OPDA to yield the DHQP molecule, which exhibits PL similar to the final CNDs. As temperature and reaction duration increase, DHQP molecules undergo carbonization and merge into a monolayer graphene-like configuration, resulting in CNDs characterized by sp³ bonds. NMR findings confirmed the integration of DHQP's benzene rings into the conjugated CND core, resulting in zero-dimensional CNDs with DHQP units embedded or linked within their architecture.

Fig. 20 A schematic representation of the red-emissive CNDs building-up process utilizing a solvent-free method system including OPDA and CAT as precursors. Reproduced from ref. 25 with permission from Springer Nature, Copyright 2022.

CNDs demonstrate either excitation-dependent or excitation-independent emission, with instances of both behaviors occurring simultaneously. Significant differences are observed in the emission spectra of CNDs before and after purification. Prior to 2021, when column purification was not utilized in this domain, the excitation-dependent and independent emission characteristics of CNDs were predominantly ascribed to surface functional groups or the presence of various emissive sites. Functional groups such as –NH₂, –COOH, and –OH significantly impacted the degree and type of aggregation, thus influencing the emission properties. The literature indicates that the presence or absence of these functional groups, together with their interactions via the distorted sp² carbon framework, significantly influences the emission behavior. Furthermore, before purification, the emission spectra were frequently broad and poorly characterized owing to the presence of MFs, reaction byproducts, and unreacted precursors. Following the introduction of column purification, the emission characteristics were more pronounced and well defined. After purification, the CNDs exhibit more defined emission properties due to the separation of MFs (having molecular states only), quasi-CNDs (having fewer core states and more surface states on their surface), and CNDs (having more core states and fewer surface states).

Overall, the PL of CNDs originates from multiple emissive centers, including the carbon core, surface states, edge defects, and MFs formed during synthesis. Core states arising from sp²-conjugated domains contribute to intrinsic emission, while surface functional groups and defects introduce surface energy levels that strongly influence PL behavior. In addition, edge states and MFs can act as independent emissive centers, and sometimes dominate the overall fluorescence. Therefore, the PL mechanism of CNDs generally results from the combined contribution of core, surface, defect, and molecular states.

7. Emission behaviour in CNDs

In the literature, three categories of CNDs have been presented, which show (i) excitation-dependent emission,^200–202 (ii) excitation-independent emission,^63,91,203 and (iii) both excitation-dependent as well as independent emission.^2,204

7.1 Excitation-dependent and independent emission

The excitation dependency in CNDs occurs because of the presence of distinct surface-exposed functional groups or surface states, elemental doping, defects, different types of aggregates, or molecular heterogeneity, and environmental interactions (e.g., solvation), and originates from size heterogeneity, leading to multiple discrete electronic states.²⁰⁰ In contrast, excitation-independent emission in CNDs is attributed to the uniform emissive centre, dominant MF, independent fluorophore, homogeneity in their size distribution, and large number of homogeneous surface states/molecular states. In the reaction mixture, after hydrothermal treatment, fluorophores may exist as free-floating molecules in solution or adhered to the surface of the CNDs, or may be bonded with the carbon core. Nonetheless, these fluorophores alone can account for excitation-independent PL emission. In the literature, the majority of the reported CNDs exhibit either excitation-dependent or independent emission; nevertheless, there are a limited number of studies that describe dual emission behavior of CNDs. The emission of CNDs, both excitation-dependent and excitation-independent, is influenced by the contributions of molecular and core states. When molecular states predominate, excitation-independent emission is more pronounced with increased intensity, while a greater contribution from core states results in excitation-dependent behaviour.

Khan et al. further explained that the excitation-dependent emission in CNDs arises from inhomogeneous broadening and time-dependent spectral migration within individual particles (Fig. 21a).²⁰⁵ Rather than being caused by a mixture of sizes or multiple chromophores, it results from an ensemble of closely spaced energy substates created by surface functionalities, especially oxygenated groups. Different excitation wavelengths access different substates, leading to variable emission behavior further influenced by the red-edge effect. This internal energy relaxation process explains the redshift and complex PL often observed in CNDs. Cushing et al. reported that excitation-dependent emission occurs when solvent relaxation (solvation) is not faster than fluorescence (Fig. 21b).²⁰⁶ In polar solvents like water, the excited fluorophore and solvent interact during emission, causing the emission energy to shift based on the excitation wavelength. This results in a red-edge effect, where emission peaks red shift with longer excitation due to incomplete relaxation before photon emission. In contrast, in nonpolar solvents, rapid relaxation leads to excitation-independent emission. Fu et al. further demonstrated that the excitation-dependent emission in CNDs arises from the presence of multiple polycyclic aromatic hydrocarbon (PAH) components with slightly different energy gaps embedded in the CND structure (Fig. 21c).²⁰⁷ When CNDs are excited at different wavelengths, different PAH species are selectively excited, each emitting light at its own characteristic energy. Additionally, energy transfer between these PAHs further broadens and shifts the emission, contributing to the observed wavelength-dependent PL. Thus, the complex emission behavior of CNDs is due to heterogeneous molecular environments and exciton self-trapping, rather than a single uniform electronic structure.

Fig. 21 (a) Diagram illustrating the dynamical aspects of diverse fluorescence phenomena. Solvent relaxation with a timescale of fluorescence decay (τ_R ≈ τ_F) leads to emission from many substates with differing degree of relaxation, thus violating both Kasha–Vavilov rule and resulting in inhomogeneous broadening. Reproduced from ref. 205 with permission from American Chemical Society, Copyright 2015. (b) Fluorescence dependent on wavelength and affected by solvent environment. In nonpolar solvents, fluorescence is independent of excitation, whereas in polar solvents, solvent relaxation alters and broadens emission. When solvent interactions happen along with fluorescence, time-dependent red-edge emission manifests, resulting in excitation-dependent fluorescence, as evidenced for GO in water. Reproduced from ref. 206 with permission from American Chemical Society, Copyright 2014. (c) Schematics demonstrate that the emission of CNDs, which is dependent on excitation, results from the selective excitation of PAHs with marginally varying energy gaps and the energy transfer between them, while the Stokes shift is attributed to exciton self-trapping within the PAH network. Reproduced from ref. 207 with permission from American Chemical Society, Copyright 2015.

Li et al. modified the surface of CNDs through amine functionalization and investigated the PL dependence on surface chemistry (Fig. 22a).²⁰⁸ During the synthesis of CNDs via the hydrothermal technique, the density of amino groups on the surface of the CNDs was adjusted. It has been found that a reduced density of amine functional groups on the surface induces excitation-dependent emission where the emission peak shifts to longer wavelengths (red shifts) as the excitation wavelength increases. This behavior reflects the presence of multiple transition modes, primarily arising from diverse surface trap states. As a result, emission becomes highly dependent on excitation energy, showing broad, asymmetric, and red-shifted spectra. Consequently, a single emission state predominates in amine-rich CNDs, resulting in excitation-independent behavior. In instances of CNDs with reduced amine content, the surface exhibits several functional groups, including C–O, C–O–C, and O=C–OH, in addition to the amine group. Time-resolved spectroscopy was employed to provide a surface chemistry-dependent PL behavior. Enhancing the amine density on the surface of CNDs eliminates the non-radiative self-trapping and enhances PL efficiency. QY is also increased from 20.8% to 44.7% following complete surface alteration. In another report Wang et al. reported that if the CND surface is unpassivated, multiple surface states with a sequence of particular energies will play the key role in exhibiting excitation-dependent emission. And after passivation of CNDs, the uniform surface states facilitate excitation-independent behavior, but also a more uniform spectral distribution and an increased QY, as illustrated in Fig. 22b.²⁰⁹ Soni et al. showed that excitation-independent and excitation-dependent emission arises from molecular as well as core states, respectively (Fig. 22c).² They synthesized CNDs using OPDA as a precursor through hydrothermal conditions and separated three emissive components – blue, green, and red – using column chromatography. The red component shows only MF-like transitions with nicely arranged vibrational overtone bands observed both in absorption and emission spectra, identified as a MF. The green and blue components represent molecular state, surface state, and core state in both absorption and emission, while the surface and core states are stronger in blue emission. In summary, understanding and controlling these factors is essential for the rational design of CNDs for applications like bioimaging, sensing, and optoelectronics.

Fig. 22 (a) L-CNDs enriched with amino groups show excitation-independent behavior, L-CNDs with a variety of surface states or fewer surface amino groups show excitation-dependent emission. Reproduced from ref. 208 with permission from Springer Nature, Copyright 2014. (b) Passivated N-CNDs emit via a single radiative channel and exhibit excitation-independent behavior, whereas unpassivated N-CNDs exhibit excitation-dependent emission. Reproduced from ref. 209 with permission from Elsevier, Copyright 2017. (c) CNDs synthesized using OPDA and purified through column purification. The acquired green and blue components exhibit both excitation-dependent and independent emission. Reproduced from ref. 2 with permission from Royal Society of Chemistry, Copyright 2021.

7.2 Correlations between structure and emission

Significant progress has been made in understanding the structure–property relationships governing the PL behaviour of CNDs. Currently, key factors such as quantum size effects, surface chemistry, heteroatom doping, and molecular state emissions have been explored through theoretical calculations, as well as experimentally. Nevertheless, significant challenges remain, especially in precisely characterizing CND structures, due to the heterogeneous nature of synthesis products and reaction variables. A unified structural model that covers all types of synthesis and reaction variables is still lacking, which hinders the complete mechanistic understanding of PL in CNDs. Future progress relies heavily on combining experimental work with advanced theoretical simulations. Although DFT has evolved considerably to explain, predict, and guide experiments, it still faces limitations in accurately modeling the heterogeneous structures and dynamic reaction pathways involved in CND synthesis. Ultimately, interdisciplinary strategies integrating theory, computation, and experiment will be essential for achieving precise control over the structure and PL properties of CNDs for targeted applications.

8. Theoretical aspects of optical and structural relationships in CNDs

The extensive use of DFT and other computational techniques has greatly advanced our theoretical and optical understanding of CNDs. The photophysical characteristics of CNDs and their interactions with other molecules have shed light on various aspects of CND behavior. Establishing the basic connections between CND structure and its optical properties was the main goal of early research. The essence of this study lies in the carbon core, where the size of the sp² domains dictates the bandgap and influences the emission. As this core enlarges, the bandgap narrows, causing fluorescence emissions to redshift. Notably, suppression of π–π stacking through nonplanar structures remarkably enhances the fluorescence properties, especially when CNDs aggregate. Furthermore, molecular conjugation directly governs the emission spectrum; an increase in conjugation can lead to a significant redshift. Furthermore, the precise positioning of substituents or functional groups like amino and carbonyl on the CND surface can strategically change the absorption and emission spectra by modulating molecular orbital energy levels. Additionally, environmental factors transform sp² domains into sp³-hybridized structures and cause a noticeable shift in emission spectra. Finally, the introduction of doping elements can precisely guide electron and ion migration, indirectly influencing optical behavior by optimizing charge transfer pathways. Beyond the core, the surface states play a pivotal role in shaping the PL, determining broad red emissions, and even influencing the generation of reactive oxygen species (ROS). These theoretical frameworks are essential for predicting and engineering the tunable optical properties of CNDs for targeted applications.

Early investigations focused on establishing the fundamental links between CND structure and their optical responses. Liu et al. showed through theoretical calculations how the precise positioning of functional groups, such as chlorine atoms, on the aromatic core of CNDs affects their electronic structure and, consequently, their emission wavelengths, and alters molecular orbital energy levels.¹⁰⁶ This demonstrates the crucial role played by atomic-level structure in tuning CND luminescence. The optical characteristics of GQDs were examined in the study by Zhao et al. (2014) using TD-DFT, with an emphasis on the effects of edge functionalization on the absorption and fluorescence behavior of GQDs. The authors conducted a comprehensive analysis of the excited-state features of various GQD models using the 6-31G(d) basis set and a variety of exchange–correlation functionals, including B3LYP, PBE0, and CAM-B3LYP. Their computations showed that adding polar groups, such as –COOH and –OH to the GQD edges caused the HOMO–LUMO gaps to decrease, which led to notable red shifts in the optical spectra that were in good agreement with experimental findings. This study shows that TD-DFT is a reliable approach for predicting the photophysical behavior of carbon-based nanomaterials.

Xu et al. performed theoretical modeling of HOMO–LUMO energy levels as the number of aromatic rings in CNDs increases (Fig. 23a).²¹⁰ The figure demonstrates that as the sp²-conjugated carbon core grows (i.e., more aromatic rings), the energy gap between HOMO and LUMO decreases. This reduction in bandgap energy accounts for the redshift in fluorescence emission from yellow (570 nm) to NIR (721 nm), consistent with the experimental observations. The results support the mechanism that fluorescence tunability in these solid-state CNDs arises primarily from the quantum size effect and extended conjugation within the carbon core. In addition to indicating an energy transfer mechanism between these states, Zhang et al. employed DFT calculations in conjunction with experimental data to attribute the broad red emission of CNDs to O-containing surface groups and the green/yellow emission to conjugated sp²-domains (core states).²¹¹ The effects of external factors on the characteristics of CNDs have also been clarified using theoretical methods (Fig. 23b). Expanding on this, Umami et al. employed TD-DFT to rationally design CNDs with NIR absorption, finding that the carbonyl group is particularly effective for this purpose.²¹² Through theoretical calculations, Yan et al. confirmed through DFT calculations that a synergistic effect between the carbon core and surface states is responsible for the PL of white-light-emitting CNDs.²¹³ Their work demonstrated that red-emitting CNDs are primarily influenced by carbon core states, and the red shift in emission spectra is regulated by increasing amide content and conjugation (Fig. 23c). Using material simulation computations, Haotian Hao et al. illustrated how boron doping in CNDs affects aluminum ion migration in memory devices, resulting in more stable conductive filaments and improved memory behavior in memristors.²¹⁴ Through theoretical methods, doping strategies and their effects on CND performance have also been clarified (Fig. 23d). With emissions originating from both core and surface states, Sharma et al. employed DFT simulations to obtain structural insights into urea-based CNDs, connecting their tunable dual PL to particle size and band gap.²¹⁵ To provide a computational framework for anticipating luminescence, Sheardy et al. conducted comprehensive TD-DFT modeling on a variety of CND structures in order to clarify the nature of their optical characteristics.²¹⁶ Beyond intrinsic optical properties, theoretical studies have explored how CNDs react to external stimulus and their interactions with their environment. Ambrusi et al. developed a DFT and TD-DFT model to investigate the impact of various O functional groups on CND surface interactions with silver nanoparticles.²¹⁷ In the context of memory devices, Lu et al. used first-principles calculations to explain the piezochromic behavior of CNDs, supporting the conversion of sp² domains into sp²-hybridized domains under high pressure.²¹⁸ This study clarified how the CND electronic structure and optical response can be altered by mechanical forces (Fig. 23e). The nonplanar triphenylamine structure in CNDs can efficiently limit π–π stacking of aromatic skeletons, improving their fluorescence properties in the aggregated form, according to theoretical calculations conducted by Zhao et al. to further explore structural impacts.²¹⁹ This result provides a method overcoming quenching in solid-state CND applications induced by aggregation. Yang et al. used DFT simulations to examine the interaction between functionalized CNDs and curcumin for drug sensing, enhanced by symmetry-adapted perturbation theory (SAPT) energy decomposition and quantum theory of atoms in molecules (QTAIM).²²⁰ Their findings reveal that functionalization causes notable changes in energy gaps and increases adsorption capacity. In order to understand the configuration and distribution of N dopants within N-doped CNDs (NCNDs), Nguyen et al. conducted theoretical simulations. According to their findings, the HOMO–LUMO levels, bandgap, and ultimately the wavelengths of light absorption and emission are affected by the preferential localization of pyridinic and graphitic N near to the edges.²²¹ This offered a theoretical foundation for doping strategy optimization to regulate optoelectronic characteristics. Furthermore, theoretical calculations have been crucial in understanding the reactivity and functional processes of CNDs in photocatalysis, going beyond their optical features. Zhang et al. explained the controllable formation of ROS in red-emissive CNDs from the perspective of biomedical applications using theoretical calculations in conjunction with experimental characterization studies.²²² Their results showed that the core sizes and surface states of the CNDs, which determine their redox potentials and energy gaps between singlet and triplet states, are responsible for the distinct type I/II ROS formation. This gave a mechanistic understanding for designing CNDs with specific therapeutic functions. Furthermore, Wei et al. showed through theoretical calculations that the improved photocatalytic hydrogen generation in N-doped CNDs is primarily attributable to their electron and hole-trapping ability.²²³ Geng et al. used TD-DFT to design phosphorescent CNDs for sonodynamic treatment (SDT), utilizing theoretical underpinnings.²²⁴ To assess how surface chemistry affects electronic characteristics, they constructed simplified molecular models of CNDs functionalized with electron-donating (pyrrolic and graphitic N) and electron-withdrawing (sulfonic acid) groups. The HOMO–LUMO gap is greatly reduced by the donor and acceptor groups, from 4.07 eV in sulfonic-only systems to 1.92 eV in fully modified dots, according to their TD-DFT calculations. This reduction enhanced NIR absorbance and promoted long-lived triplet excited states. The predictive ability of DFT in nanomaterial design was confirmed by the direct correlation between these electronic properties and the superior ROS production and experimentally observed NIR phosphorescence under ultrasonic irradiation. Nevertheless, the collective theoretical and experimental studies underscore the indispensable role played by theoretical calculations in elucidating the complex photophysical mechanisms of CNDs.

Fig. 23 (a) HOMO and LUMO states of the established model by increasing the aromatic rings. Reproduced from ref. 210 with permission from Wiley-VCH, Copyright 2023. (b) Structural models of graphene sheets functionalized with varying quantities of C=O or OH groups, as indicated and molecular orbital diagrams illustrating the predicted HOMO/LUMO energy levels and corresponding bandgap of these structures. The red and green colors indicate distributions of negative and positive charges, respectively. Reproduced from ref. 211 with permission from Wiley-VCH, Copyright 2023. (c) Theoretical calculations based on the pyrene model of BCNDs, GCNDs, YCNDs, and RCNDs. Reproduced from ref. 213 with permission from Wiley-VCH, Copyright 2023. (d) DFT molecular simulation results (i.e., HOMO and LUMO). Reproduced from ref. 214 with permission from Wiley-VCH, Copyright 2024. (e) Proposed fluorescence transition mechanism from sp2 to sp3 hybridization determined by first-principles calculations. The top and side view of the structures at ambient and high pressures. The HOMO and LUMO orbitals of the ambient and high-pressure structures. Reproduced from ref. 218 with permission from Wiley-VCH, Copyright 2017.

Tetsuka et al. demonstrated how N functional groups modulate the electronic structure of GQDs.²²⁵ Theoretical calculations revealed that different N moieties shift the HOMO and LUMO levels via orbital interactions (Fig. 24a), a finding confirmed experimentally through photoelectron yield spectroscopy and PL measurements, showing clear changes in energy levels and emission colors. The figure highlights that surface functionalization enables precise control over GQD optical properties, critical for optoelectronic applications. Paloncyová et al. demonstrated that pure CNDs (∼2.1 nm) remain structurally stable in water, forming spherical multilayer stacks with water surrounding but not entering the structure.²²⁶ In contrast, they also demonstrated the destabilizing effects of surface functionalization with negatively charged carboxyl groups (Fig. 24b). Several simulation outcomes are shown, ranging from mild deformation of the CND structure to complete exfoliation of individual layers into solution. As the density of deprotonated carboxyl groups increases, especially in smaller dots, electrostatic repulsion between layers overcomes stabilizing interactions, despite the presence of sodium counterions. This leads to interlayer slippage, partial fragmentation, or full dissolution of layers. These results emphasize that while functional groups play a key role in solubility and reactivity, excessive negative surface charge can compromise the structural integrity of CNDs.

Fig. 24 (a) Energy levels and PL for N-functionalized graphene quantum dots (NGQDs). (i) Predicted energy level diagrams for graphene with various functional groups. The isosurface displays HOMO and LUMO. (ii) Measured energy level diagram for NGQDs. The discrepancy between the anticipated energy levels and observed ones arises from the difference in the size of NGQDs and the quantity of nitrogenous functional groups. (iii) PL image of NGQDs in aqueous solution excited using a UV light (365 nm): (i) Azo-GQDs; (ii) NH2-GQDs; (iii) OPD-GQDs; and (iv) DAN-GQDs. (iv) Normalized PL spectra (excited at 380 nm) from aqueous dispersions of NGQDs. Reproduced from ref. 225 with permission from Wiley-VCH, Copyright 2016. (b) (i) Density plot (left) and side view (right) of pure CNDs with diameter of gyration of 2.1 nm (black) in water (red). The layer undulations are seen in the widths of the density peaks of each layer. H₂O molecules surround the CNDs with a hydrophobic gap above and below the graphene-like planes; however, they do not penetrate inside the CND structure. The CND structure is a representation from MD simulation, depicted as sticks with cyan carbons, white hydrogens, and red oxygens (only water molecules within 0.5 nm are shown for clarity). CNDs with deprotonated carboxyl groups (sodium counterions represented as dots): (ii) large CNDs can stay as a deformed dot, (iii) layers may move further from each other with ions connecting the edge carboxyl (−1) groups between the layers, (iv) dots may fragment into components that are either completely detached or rejoin, and (v) individual layers (in the case of small layers) may leave the CNDs fully and remain solvated in solution. Reproduced from ref. 226 with permission from American Chemical Society, Copyright 2018.

9. Pro and cons

Since the accidental discovery of fluorescent CNDs in 2004, two decades of systematic research have progressively established a comprehensive multi-state mechanistic framework for understanding the origin of their PL. Four distinct emission contributors have been identified: the graphitic sp²-hybridised carbon core, surface functional groups, edge defects, and MFs, each playing a unique and often competing role in governing the overall optical behavior of CNDs. The graphitic core contributes intrinsic bandgap emission through π–π* electronic transitions, where the size of sp²-conjugated domains directly determines the emission wavelength via quantum confinement effects. As these domains enlarge, the HOMO–LUMO gap narrows, enabling progressively red-shifted and NIR emission. Surface states also introduce intermediate energy levels between the HOMO and LUMO of the carbon core, acting as trap states that redirect photoexcited electrons toward red-shifted radiative recombination. Critically, the controlled oxidation studies demonstrated that surface oxidation degree can independently tune emission wavelength without altering particle size, establishing surface states as a genuinely controllable emission handle. MFs formed during synthesis play a significant role in contributing to excitation-independent emission with high QY. Their identification has been crucial in resolving longstanding mechanistic ambiguities in this field. Early reports on the PL of CNDs were often attributed to emissions originating from the carbonized core. However, subsequent studies revealed that, in many cases, the observed fluorescence predominantly arose from small MFs rather than from the intrinsic carbon core itself. This realization clarified a major confusion in the early literature and highlighted the critical importance of purification and structural characterization in accurately interpreting the PL behavior of CNDs. Edge states originating from dangling bonds and structural defects at the periphery of sp²-hybrised domains further add diversity to the emission landscape. Heteroatom doping with nitrogen, sulfur, fluorine, silicon, boron, and phosphorus has emerged as a powerful strategy to modify electronic structure, introduce new emissive energy levels, and enhance QY, in some optimized systems exceeding up to 90–99%. Most recently, single-particle investigations of heteroatom-doped CNDs have revealed QDs with multilevel PL blinking driven by charged exciton formation, providing the most direct experimental evidence yet that the surface chemical environment actively governs exciton dynamics in CNDs. Along with this, advanced computational approaches including DFT and TD-DFT have increasingly complemented experimental work by predicting HOMO–LUMO gaps, rationalizing doping, functionalization effects, and supporting the rational design of CNDs with targeted optical properties.

Despite these substantial advances, the origin of PL in CNDs remains one of the most debated and unresolved problems in nanomaterials science. The central difficulty is the intrinsic structural heterogeneity of CNDs, which simultaneously contain graphitic cores, amorphous carbon regions, surface functional groups, edge defects, and MFs in varying proportions that are exquisitely sensitive to synthesis conditions. This coexistence of multiple emissive centers makes it extremely difficult to isolate, characterize, and independently assign individual contributions to the observed emission. The systematic misidentification of MF contaminants as genuine carbon-core emission has been among the most persistent problems in this field, affecting a significant portion of the published literature, particularly in early bottom-up synthesis studies conducted without rigorous purification. Conventional characterization techniques including TEM, AFM, and ensemble spectroscopy are insufficient to distinguish genuine CNDs from aggregated molecular species or incompletely carbonized structures. Hydrogen-bonded nanoassemblies of molecular precursors can appear as spherical nanoparticles with graphitic lattice spacings under TEM, leading to systematic misassignment. The excitation-dependent emission behavior of CNDs remains mechanistically ambiguous. It may arise independently or from multiple factors, including quantum confinement, distribution of surface-state energy levels, molecular heterogeneity, solvent relaxation effects, and inhomogeneous broadening from oxygenated surface substates. In most cases, these mechanisms likely coexist and collectively contribute to the observed PL. Reproducibility in the synthesis is also a major challenge, as minor variations in precursor selection, reaction conditions, solvent, pH, or purification can drastically shift the relative contributions of core, surface, molecular, and edge states, resulting in conflicting mechanistic interpretations. At the theoretical level, computational models are not yet capable of reliably predicting PL properties solely from structural parameters. Moreover, the quantitative relationships between the sp²/sp³ carbon ratio, heteroatom doping configurations, surface functional group density, and QY remain poorly defined.

10. Discussion and outlook

The PL of CNDs represents one of the most complex and contested optical phenomena in contemporary nanomaterials science. Unlike conventional semiconductor QDs, whose emission is governed predominantly by quantum confinement effects, the PL of CNDs does not conform to a single unified mechanism. Instead, it emerges from the intricate interplay of multiple structural and chemical factors, each of which can dominate or recede depending on the synthesis route, precursor composition, and post-synthetic treatment applied.

The earliest mechanistic interpretations of CNDs PL, largely based on quantum confinement analogies, provided a useful but ultimately incomplete framework. As synthetic methodologies diversified and characterization techniques advanced, it became evident that the optical behavior of CNDs arises from a complex interplay of multiple contributions, including carbon core states, localized sp² domains, surface functional groups, heteroatom dopants, and MFs formed as byproducts during synthesis. Importantly, the relative contribution of each of these factors is not fixed but varies with the structural heterogeneity introduced by different synthesis conditions, thereby making direct comparisons across systems inherently challenging. A key turning point in CND research was the realization that MFs can dominate the observed fluorescence, shifting the focus from intrinsic nanodot emission to contributions from co-synthesized species. This fundamentally changed how PL is interpreted. Subsequently, surface chemistry emerged as a central factor, with different functional groups creating environmentally sensitive emissive states that complicate mechanistic understanding. The role played by surface chemistry in governing PL is now well established, yet mechanistically underresolved. At the same time, heteroatom doping has been widely used to tune emission and enhance QY, although its exact role remains unclear as it can simultaneously influence core, surface, and molecular states. Together, these factors, combined with the lack of standardized synthesis and characterization protocols, have made it challenging to establish clear and universal structure–property relationships in CNDs. As a result, a universal structural framework capable of explaining PL behavior across different CND systems has yet to be established. Increasing experimental evidence, particularly from advanced purification methods such as column chromatography and high-resolution spectroscopic analyses, has demonstrated that MFs can significantly contribute to, or even dominate, the fluorescence observed in some CND systems. In addition, environmental factors such as pH, solvent polarity, and temperature influence the PL origin by altering surface states, defect energy levels, and molecular emissive centers, thereby modulating electronic transitions and radiative relaxation pathways. These findings highlight the importance of carefully controlling synthesis parameters and reaction conditions during CND preparation, as summarized in Fig. 25. A clear understanding of the structural features of CNDs is crucial for determining the fundamental origin of their PL. The synthesis strategy plays a key role in defining the structural heterogeneity and optical properties of CNDs. Therefore, establishing a clear relationship between synthesis conditions, structural characteristics, and emission behavior is essential for understanding the PL mechanisms of CNDs.

Fig. 25 Schematic illustration of the key factors governing PL origin in CNDs, including precursor selection, elemental doping, reaction variables, synthesis methods, purification, storage conditions, and surface structure regulation.

Looking forward, several fundamental questions must be addressed to advance this field. What is the precise energy level structure of oxygen-related surface states, and how does it vary with surface chemistry? How do core and surface states interact through energy transfer, and what governs their relative emission contributions? Can molecular state emission be completely suppressed through synthesis design alone? What quantitatively determines QY in heteroatom-doped systems? And, most fundamentally, can a single unified mechanistic model explain PL behavior across diverse CND synthesis routes and structural variations? Despite substantial advancements, a unified structural model that consistently explains the PL behavior of CNDs across different systems has yet to be established. Addressing this challenge will require systematic classification of precursors such as aromatic molecules, aliphatic carbon sources, heteroatom-containing compounds, and fused ring systems and their corresponding CND structures. In addition, rigorous purification protocols are also necessary to effectively separate MFs and other emissive products, enabling accurate identification of the true origin of PL. Answering these questions will require the development of advanced single-particle characterization tools, standardized synthesis as well as purification protocols, and deeper integration of experimental investigations with theoretical simulations. Together, these efforts will enable the transition from empirical observation to truly rational design of CNDs with precisely controlled PL properties.

Conflicts of interest

The authors declare no conflict of interest.

Data availability

No primary research results, software or code have been included and no new data were generated or analysed as part of this review.

Acknowledgements

C. K. N. thanks the Indian Institute of Technology Mandi, India, for providing facilities for academic research work. R. G. thanks the Prime Minister's Research Fellowship (PMRF) for the fellowship. A. S. acknowledges the Ministry of Education (MoE), India, for the scholarship.

References

1. X. Xu, et al., Electrophoretic analysis and purification of fluorescent single-walled carbon nanotube fragments, J. Am. Chem. Soc., 2004, 126, 12736–12737.
2. Y. P. Sun, Quantum-sized carbon dots for bright and colorful photoluminescence, J. Am. Chem. Soc., 2006, 128, 7756–7757.
3. H. Peng and J. Travas-Sejdic, Simple aqueous solution route to luminescent carbogenic dots from carbohydrates, Chem. Mater., 2009, 21, 5563–5565.
4. F. Ehrat, et al., Tracking the Source of Carbon Dot Photoluminescence: Aromatic Domains versus Molecular Fluorophores, Nano Lett., 2017, 17, 7710–7716.
5. H. Guo, et al., Machine learning-guided realization of full-color high-quantum-yield carbon quantum dots, Nat. Commun., 2024, 15, 1–10.
6. W. Wang, et al., Shedding light on the effective fluorophore structure of high fluorescence quantum yield carbon nanodots, RSC Adv., 2017, 7, 24771–24780.
7. L. Bao, C. Liu, Z.-L. Zhang and D.-W. Pang, Photoluminescence–Tunable Carbon Nanodots: Surface–State Energy–Gap Tuning, Adv. Mater., 2015, 27, 1663–1667.
8. H. Liu, Z. He, L. P. Jiang and J. J. Zhu, Microwave-assisted synthesis of wavelength-tunable photoluminescent carbon nanodots and their potential applications, ACS Appl. Mater. Interfaces, 2015, 7, 4913–4920.
9. U. Sahrawat, et al., One–Pot Synthesis of Orange Emissive Carbon Dots Specific for Staining of Mitochondria in both Cancer and Non–Cancer Cells, ChemNanoMat, 2024, 10, e202300628.
10. Y. Choi, et al., Highly Biocompatible Carbon Nanodots for Simultaneous Bioimaging and Targeted Photodynamic Therapy In Vitro and In Vivo, Adv. Funct. Mater., 2014, 24, 5781–5789.
11. Y. Zhang, et al., One-step microwave synthesis of N-doped hydroxyl-functionalized carbon dots with ultra-high fluorescence quantum yields, Nanoscale, 2016, 8, 15281–15287.
12. R. Patra, et al., Facile synthesis of yellow emissive heteroatom doped carbon quantum dots fluorescent probe with superior quantum yield for fast-track ionic detection in aqueous media, Mater. Res. Bull., 2024, 180, 113040.
13. K. M. Chan, W. Xu, H. Kwon, A. M. Kietrys and E. T. Kool, Luminescent Carbon Dot Mimics Assembled on DNA, J. Am. Chem. Soc., 2017, 139, 13147–13155.
14. Y. Liu, Y. Tian, Y. Tian, Y. Wang and W. Yang, Carbon-Dot-Based Nanosensors for the Detection of Intracellular Redox State, Adv. Mater., 2015, 27, 7156–7160.
15. S. Mondal, et al., Efficient Photosensitizing Capabilities and Ultrafast Carrier Dynamics of Doped Carbon Dots, J. Am. Chem. Soc., 2019, 141, 15413–15422.
16. T. H. Kim, et al., Yellow-Emitting Carbon Nanodots and Their Flexible and Transparent Films for White LEDs, ACS Appl. Mater. Interfaces, 2016, 8, 33102–33111.
17. S. Adrita, K. Tasnim, J. Ryu and S. Sharker, Nanotheranostic Carbon Dots as an Emerging Platform for Cancer Therapy, J. Nanotheranostics, 2020, 1, 58–77.
18. B. Zhi, et al., Malic Acid Carbon Dots: From Super-resolution Live-Cell Imaging to Highly Efficient Separation, ACS Nano, 2018, 12, 5741–5752.
19. L. Wu, X. Li, Y. Ling, C. Huang and N. Jia, Morpholine Derivative-Functionalized Carbon Dots-Based Fluorescent Probe for Highly Selective Lysosomal Imaging in Living Cells, ACS Appl. Mater. Interfaces, 2017, 9, 28222–28232.
20. R. Garg, et al., Carbon Nanodots as a Red Emissive Fluorescent Probe for the Super-Resolution Microscopy of DNA Dynamics during Paclitaxel Treatment, Small Struct., 2025, 6, 1–9.
21. N. Suzuki, et al., Chiral Graphene Quantum Dots, ACS Nano, 2016, 10, 1744–1755.
22. H. Pei, et al., Chiral carbonized polymer dots:A comprehensive review, Sustainable Mater. Technol., 2025, 43, e01271.
23. M. Alafeef, I. Srivastava, T. Aditya and D. Pan, Carbon Dots: From Synthesis to Unraveling the Fluorescence Mechanism, Small, 2024, 20, 1–13.
24. M. L. Liu, et al., The light of carbon dots: From mechanism to applications, Light: Sci. Appl., 2022, 11, 1–13.
25. P. Li, et al., Formation and fluorescent mechanism of red emissive carbon dots from o-phenylenediamine and catechol system, Light: Sci. Appl., 2022, 11, 1–11.
26. M. L. Liu, B. B. Chen, C. M. Li and C. Z. Huang, Carbon dots: Synthesis, formation mechanism, fluorescence origin and sensing applications, Green Chem., 2019, 21, 449–471.
27. Y. P. Sun, et al., Quantum-sized carbon dots for bright and colorful photoluminescence, J. Am. Chem. Soc., 2006, 128, 7756–7757.
28. L. Bao, et al., Electrochemical tuning of luminescent carbon nanodots: From preparation to luminescence mechanism, Adv. Mater., 2011, 23, 5801–5806.
29. M. J. Krysmann, A. Kelarakis, P. Dallas and E. P. Giannelis, Formation mechanism of carbogenic nanoparticles with dual photoluminescence emission, J. Am. Chem. Soc., 2012, 134, 747–750.
30. T. Zhang, et al., Constructing Oxygen–Related Defects in Carbon Nanodots with Janus Optical Properties: Noninvasive NIR Fluorescent Imaging and Effective Photocatalytic Therapy, Adv. Mater., 2023, 35, 2302705.
31. Z. Wei, et al., Heteroatom Doping Effects on the Exciton Behavior in Carbonized Polymer Dots, Nano Lett., 2025, 25, 4013–4021.
32. H. Liu, et al., A review of carbon dots in synthesis strategy, Coord. Chem. Rev., 2024, 498, 215468.
33. Y. Wang, Q. Wang, W. Liu, X. Xin and B. Zhao, A Review on the Synthesis of Carbon Dots and Their Applications in Environmental Analysis, Crystals, 2025, 15, 1–22.
34. M. Pykal, et al., Thermodynamics and kinetics of early stages of carbon dot formation: a case of citric acid and ethylenediamine reaction, Nanoscale, 2025, 17, 7780–7789.
35. S. Qu, et al., Toward Efficient Orange Emissive Carbon Nanodots through Conjugated sp2-Domain Controlling and Surface Charges Engineering, Adv. Mater., 2016, 28, 3516–3521.
36. S. Cho, et al., Insights into glucose-derived carbon dot synthesis via Maillard reaction: from reaction mechanism to biomedical applications, Sci. Rep., 2024, 14, 1–13.
37. A. B. Siddique, A. K. Pramanick, S. Chatterjee and M. Ray, Amorphous Carbon Dots and their Remarkable Ability to Detect 2,4,6-Trinitrophenol, Sci. Rep., 2018, 8, 1–10.
38. A. Pal, M. P. Sk and A. Chattopadhyay, Recent advances in crystalline carbon dots for superior application potential, Mater. Adv., 2020, 1, 525–553.
39. R. Wang, K. Q. Lu, Z. R. Tang and Y. J. Xu, Recent progress in carbon quantum dots: synthesis, properties and applications in photocatalysis, J. Mater. Chem. A, 2017, 5, 3717–3734.
40. X. T. Zheng, A. Ananthanarayanan, K. Q. Luo and P. Chen, Glowing graphene quantum dots and carbon dots: Properties, syntheses, and biological applications, Small, 2015, 11, 1620–1636.
41. Y. Shen, et al., High-Temperature Resistance Photoluminescence Carbonized Polymer Dots Through Equilibrium Bi-Confinement Effects, Adv. Mater., 2025, 37, 1–10.
42. R. Garg, et al., Tracking the super resolved structure of mitochondria using red emissive carbon nanodots as a fluorescent biomarker, Chem. Commun., 2023, 59, 13454–13457.
43. Q. Wang, et al., Pressure-triggered aggregation-induced emission enhancement in red emissive amorphous carbon dots, Nanoscale Horiz., 2019, 4, 1227–1231.
44. Y. Zhang, et al., Solid-State Red Laser with a Single Longitudinal Mode from Carbon Dots, Angew. Chem., Int. Ed., 2021, 60, 25514–25521.
45. X. Mao, et al., Hydrophobic Carbon Dots with Aggregation-Induced Emission for Multistage Anticounterfeiting, ACS Appl. Nano Mater., 2023, 6, 20346–20354.
46. R. Gopal, et al., Aggregation-Induced White Light Emission and Solvent-Induced Dual Mode Multi-Analyte Sensing by Highly Crystalline Single Component Derived N Co-Doped Carbon Nanodots, ACS Appl. Opt. Mater., 2023, 1, 701–714.
47. H. Ding, J. S. Wei, N. Zhong, Q. Y. Gao and H. M. Xiong, Highly Efficient Red-Emitting Carbon Dots with Gram-Scale Yield for Bioimaging, Langmuir, 2017, 33, 12635–12642.
48. X. Chen, et al., Recent Progress on Chiral Carbon Dots: Synthetic Strategies and Biomedical Applications, ACS Biomater. Sci. Eng., 2023, 9, 5548–5566.
49. Y. Ru, et al., Recent advances in chiral carbonized polymer dots: From synthesis and properties to applications, Nano Today, 2020, 34, 100953.
50. A. Döring, E. Ushakova and A. L. Rogach, Chiral carbon dots: synthesis, optical properties, and emerging applications, Light: Sci. Appl., 2022, 11, 75.
51. N. C. Verma, A. Yadav and C. K. Nandi, Paving the path to the future of carbogenic nanodots, Nat. Commun., 2019, 10, 2391.
52. X. Chen, et al., One-Step Synthesis of Ultrasmall and Ultrabright Organosilica Nanodots with 100% Photoluminescence Quantum Yield: Long-Term Lysosome Imaging in Living, Fixed, and Permeabilized Cells, Nano Lett., 2018, 18, 1159–1167.
53. J. Guo, et al., Yellow-Emissive Carbon Dots with High Solid-State Photoluminescence, Adv. Funct. Mater., 2022, 32, 1–11.
54. Z. Sun, et al., High-efficient and pH-sensitive orange luminescence from silicon-doped carbon dots for information encryption and bio-imaging, J. Colloid Interface Sci., 2022, 607, 16–23.
55. H. Jia, et al., Electroluminescent Warm White Light–Emitting Diodes Based on Passivation Enabled Bright Red Bandgap Emission Carbon Quantum Dots, Adv. Sci., 2019, 6, 1900397.
56. Y. Zhang, et al., Carbon Dots with Blue–to–Near–Infrared Lasing for Colorful Speckle–Free Laser Imaging and Dynamical Holographic Display, Adv. Mater., 2023, 35, 1–10.
57. Y. Liu, et al., Toward Strong Near–Infrared Absorption/Emission from Carbon Dots in Aqueous Media through Solvothermal Fusion of Large Conjugated Perylene Derivatives with Post–Surface Engineering, Adv. Sci., 2022, 9, 2202283.
58. H. Bai, et al., Efficient preparation strategy of high quantum yield multicolor CDs for warm white LED, Chem. Eng. J., 2024, 488, 150980.
59. L. Liao, X. Lin, K. Wang, Z. Hu and F. Wu, Highly luminescent nitrogen and sulfur co-doped carbon dots for Sn2+ and S2− sensing and dual-mode anti-counterfeiting, Microchim. Acta, 2023, 190, 335.
60. J. Li, H. Zhao, X. Zhao and X. Gong, Boosting efficiency of luminescent solar concentrators using ultra-bright carbon dots with large Stokes shift, Nanoscale Horiz., 2022, 8, 83–94.
61. Q. Lou, et al., Carbon Nanodots with Nearly Unity Fluorescent Efficiency Realized via Localized Excitons, Adv. Sci., 2022, 9, 1–10.
62. Y. Tao, et al., Acid-mediated synthesis coupled with liquid–liquid extraction separation for obtaining red and orange double-color carbon dots: Application for pH, water sensing and cell-imaging, Microchem. J., 2022, 182, 107901.
63. Y. Guo, et al., Ultrabright Green-Emissive Nanodots for Precise Biological Visualization, Nano Lett., 2024, 24, 2264–2272.
64. Z. F. Wu, et al., Green Fluorescent Carbon Dots with Critically Controlled Surface States: Make Silk Shine via Feeding Silkworms, Nano Lett., 2024, 24, 9675–9682.
65. W. Li, et al., Controllable and large-scale synthesis of carbon quantum dots for efficient solid-state optical devices, Nano Energy, 2024, 122, 109289.
66. J. Lin, L. Wang, Q. Jing and H. Zhao, Highly efficient and high color rendering index multilayer luminescent solar concentrators based on colloidal carbon quantum dots, Chem. Eng. J., 2024, 481, 148441.
67. K. Hosseinpanahi, M. H. Abbaspour-Fard, M. R. Golzarian, E. K. Goharshadi and A. Vomiero, Luminescent Solar Concentrators for Greenhouse Applications Based on Highly Luminescent Carbon Quantum Dots, Sol. RRL, 2024, 8, 2400442.
68. Y. Liu, et al., Perylenedioic Acid–Derived Carbon Dots with Near 100% Quantum Yield in Aqueous Solution for Lasing and Lighting, Adv. Funct. Mater., 2024, 34, 2401353.
69. J. Guo, et al., Easy-to-perform organic-solvent-free synthesis of carbon dots with strong green photoluminescence, Chin. Chem. Lett., 2024, 35, 6–11.
70. R. Cheng, T. Zhang, X. Huang and J. Yu, Facile synthesis of high-brightness green-emitting carbon dots with narrow bandwidth towards backlight display, Chin. Chem. Lett., 2024, 35, 108763.
71. T. Yuan, et al., Carbon Quantum Dots with Near–Unity Quantum Yield Bandgap Emission for Electroluminescent Light–Emitting Diodes, Angew. Chem., 2023, 135, e202218568.
72. X. Gong, S. Zheng, X. Zhao and A. Vomiero, Engineering high-emissive silicon-doped carbon nanodots towards efficient large-area luminescent solar concentrators, Nano Energy, 2022, 101, 107617.
73. D. Zhang, L. Liu and C. Li, Carbon dots with high quantum yields used for Fe3+ detection, information encryption and anti-counterfeiting, New J. Chem., 2023, 47, 20061–20069.
74. X. Zhao, M. Zhang and Z. Zhang, Ultra-fast microwave-assisted synthesis of photoluminescent carbon dots with an ultra-high quantum yield for H₂O₂ detection, J. Environ. Chem. Eng., 2022, 10, 109008.
75. A. Kolanowska, et al., Carbon Quantum Dots from Amino Acids Revisited: Survey of Renewable Precursors toward High Quantum-Yield Blue and Green Fluorescence, ACS Omega, 2022, 7, 41165–41176.
76. Y. Zhang, et al., Heteroatom-Doping and Conjugation-Extension Enable Ultrahigh-Bright Full-Color Carbon Quantum Dots for Tunable Liquid-State Lasers and 3D Printing, Adv. Mater., 2026, 38, 1–13.
77. S. Lu, et al., Near-Infrared Photoluminescent Polymer–Carbon Nanodots with Two-Photon Fluorescence, Adv. Mater., 2017, 29, 1–6.
78. H. Ding, et al., Solvent–Controlled Synthesis of Highly Luminescent Carbon Dots with a Wide Color Gamut and Narrowed Emission Peak Widths, Small, 2018, 14, 1800612.
79. D. Li, et al., Thermally Activated Upconversion Near-Infrared Photoluminescence from Carbon Dots Synthesized via Microwave Assisted Exfoliation, Small, 2019, 15, 1–7.
80. F. Yuan, et al., Highly efficient and stable white LEDs based on pure red narrow bandwidth emission triangular carbon quantum dots for wide-color gamut backlight displays, Nano Res., 2019, 12, 1669–1674.
81. F. Yuan, et al., Ultrastable and Low–Threshold Random Lasing from Narrow–Bandwidth–Emission Triangular Carbon Quantum Dots, Adv. Opt. Mater., 2019, 7, 1801202.
82. Y. Liu, C. Y. Liu and Z. Y. Zhang, Graphitized carbon dots emitting strong green photoluminescence, J. Mater. Chem. C, 2013, 1, 4902–4907.
83. J. Wang, F. Zhang, Y. Wang, Y. Yang and X. Liu, Efficient resistance against solid-state quenching of carbon dots towards white light emitting diodes by physical embedding into silica, Carbon, 2018, 126, 426–436.
84. M. Zheng, et al., On-off-on fluorescent carbon dot nanosensor for recognition of chromium(VI) and ascorbic acid based on the inner filter effect, ACS Appl. Mater. Interfaces, 2013, 5, 13242–13247.
85. C. Zheng, X. An and J. Gong, Novel pH sensitive N-doped carbon dots with both long fluorescence lifetime and high quantum yield, RSC Adv., 2015, 5, 32319–32322.
86. H. Liu, et al., Synthesis of Luminescent Carbon Dots with Ultrahigh Quantum Yield and Inherent Folate Receptor-Positive Cancer Cell Targetability, Sci. Rep., 2018, 8, 1086.
87. S. Wang, S. Liu, J. Zhang and Y. Cao, Highly fluorescent nitrogen-doped carbon dots for the determination and the differentiation of the rare earth element ions, Talanta, 2019, 198, 501–509.
88. H. J. Wang, et al., Novel yellow solid-state fluorescent-emitting carbon dots with high quantum yield for white light-emitting diodes, Spectrochim. Acta, Part A, 2021, 250, 119340.
89. D. Li, et al., Near–Infrared Excitation/Emission and Multiphoton–Induced Fluorescence of Carbon Dots, Adv. Mater., 2018, 30, 1–8.
90. L. Jiang, et al., UV–Vis–NIR Full–Range Responsive Carbon Dots with Large Multiphoton Absorption Cross Sections and Deep–Red Fluorescence at Nucleoli and In Vivo, Small, 2020, 16, 2000680.
91. Z. Wang, et al., 53% Efficient Red Emissive Carbon Quantum Dots for High Color Rendering and Stable Warm White–Light–Emitting Diodes, Adv. Mater., 2017, 29, 1702910.
92. D. Zhang, et al., One-Step Green Solvothermal Synthesis of Full-Color Carbon Quantum Dots Based on a Doping Strategy, J. Phys. Chem. Lett., 2021, 12, 8939–8946.
93. Y. Han, et al., Carbon quantum dots capped with metal ions for efficient optoelectronic applications, J. Mater. Chem. C, 2024, 12, 5818–5825.
94. S. Zhang, L. Yuan, G. Liang and A. Gu, Preparation of multicolor-emissive carbon dots with high quantum yields and their epoxy composites for fluorescence anti-counterfeiting and light-emitting devices, J. Mater. Chem. C, 2022, 10, 8441–8458.
95. Y. Meng, et al., Facile synthesis of multifunctional carbon dots with 54.4% orange emission for label-free detection of morin and endogenous/exogenous hypochlorite, J. Hazard. Mater., 2022, 424, 127289.
96. Y. Wu, et al., High Quantum Yield Boron and Nitrogen Codoped Carbon Quantum Dots with Red/Purple Emissions for Ratiometric Fluorescent IO4-Sensing and Cell Imaging, ACS Sustainable Chem. Eng., 2022, 10, 5195–5202.
97. G. Liu, M. Zavelani-Rossi, G. Han, H. Zhao and A. Vomiero, Red-emissive carbon quantum dots enable high efficiency luminescent solar concentrators, J. Mater. Chem. A, 2023, 11, 8950–8960.
98. S. Waghmare, U. S. Sayyad, A. Chatterjee, P. Purkayastha and S. Mondal, Stereoselective Interaction between Chiral Carbon Dots and Chiral Ligands for Developing Chiral Sensing Devices, J. Phys. Chem. C, 2024, 128, 9345–9352.
99. Y. Jiang, Z. Wang and Z. Dai, Preparation of Silicon-Carbon-Based Dots@Dopamine and Its Application in Intracellular Ag+ Detection and Cell Imaging, ACS Appl. Mater. Interfaces, 2016, 8, 3644–3650.
100. H. Chen, et al., A ratiometric fluorescent nanoprobe based on terbium functionalized carbon dots for highly sensitive detection of an anthrax biomarker, Chem. Commun., 2015, 51, 5036–5039.
101. S. Sarkar, K. Das and P. K. Das, Estradiol Hemisuccinate-Modified Surface-Engineered Carbon Dots: Target-Specific Theranostic Agent, ACS Sustainable Chem. Eng., 2017, 5, 8356–8369.
102. H. Li, Y. Zhang, L. Wang, J. Tian and X. Sun, Nucleic acid detection using carbon nanoparticles as a fluorescent sensing platform, Chem. Commun., 2011, 47, 961–963.
103. M. Lan, et al., Carbon nanoparticle-based ratiometric fluorescent sensor for detecting mercury ions in aqueous media and living cells, ACS Appl. Mater. Interfaces, 2014, 6, 21270–21278.
104. X. Zhu, H. Jin, C. Gao, R. Gui and Z. Wang, Ratiometric, visual, dual-signal fluorescent sensing and imaging of pH/copper ions in real samples based on carbon dots-fluorescein isothiocyanate composites, Talanta, 2017, 162, 65–71.
105. S. Mohammadi, A. Salimi, S. Hamd-Ghadareh, F. Fathi and F. Soleimani, A FRET immunosensor for sensitive detection of CA 15–3 tumor marker in human serum sample and breast cancer cells using antibody functionalized luminescent carbon-dots and AuNPs-dendrimer aptamer as donor-acceptor pair, Anal. Biochem., 2018, 557, 18–26.
106. T. Liu, et al., Solid-State Luminescence in Self-Assembled Chlorosalicylaldehyde-Modified Carbon Dots, ACS Mater. Lett., 2023, 5, 846–853.
107. J. Li, J. Liu, L. Xu and J. Chen, Preparation of thermoresponsive fluorescent carbon dots for cellular imaging, Polym. Int., 2017, 66, 92–97.
108. C. Wang, K. Jiang, Z. Xu, H. Lin and C. Zhang, Glutathione modified carbon-dots: From aggregation-induced emission enhancement properties to a ‘turn-on’ sensing of temperature/Fe3+ ions in cells, Inorg. Chem. Front., 2016, 3, 514–522.
109. K. Gao, et al., Quaternary ammonium-functionalized carbon dots for sensitive and selective detection of 2,4,6-trinitrophenol in aqueous medium, Sens. Actuators, B, 2018, 262, 298–305.
110. I. P. J. Lai, et al., Solid-state synthesis of self-functional carbon quantum dots for detection of bacteria and tumor cells, Sens. Actuators, B, 2016, 228, 465–470.
111. S. Li, et al., Hyperbranched polyglycerol conjugated fluorescent carbon dots with improved: In vitro toxicity and red blood cell compatibility for bioimaging, RSC Adv., 2017, 7, 4975–4982.
112. C. Yuan, B. Liu, F. Liu, M. Y. Han and Z. Zhang, Fluorescence ‘turn On’ Detection of Mercuric Ion Based on Bis(dithiocarbamato)copper(II) Complex Functionalized Carbon Nanodots, Anal. Chem., 2014, 86, 1123–1130.
113. S. Khan, et al., Small molecular organic nanocrystals resemble carbon nanodots in terms of their properties, Chem. Sci., 2017, 9, 175–180.
114. H. Liu, T. Ye and C. Mao, Fluorescent carbon nanoparticles derived from candle soot, Angew. Chem., Int. Ed., 2007, 46, 6473–6475.
115. Q. Zhang, et al., Synthesis of Novel Fluorescent Carbon Quantum Dots From Rosa roxburghii for Rapid and Highly Selective Detection of o-nitrophenol and Cellular Imaging, Front. Chem., 2020, 8, 1–9.
116. C. Fan, et al., Fluorescent Nitrogen-Doped Carbon Dots via Single-Step Synthesis Applied as Fluorescent Probe for the Detection of Fe 3+ Ions and Anti-Counterfeiting Inks, Nano, 2018, 13, 1850097.
117. P. Uthirakumar, M. Devendiran, T. H. Kim and I. H. Lee, A convenient method for isolating carbon quantum dots in high yield as an alternative to the dialysis process and the fabrication of a full-band UV blocking polymer film, New J. Chem., 2018, 42, 18312–18317.
118. H. Ding, S. B. Yu, J. S. Wei and H. M. Xiong, Full-color light-emitting carbon dots with a surface-state-controlled luminescence mechanism, ACS Nano, 2016, 10, 484–491.
119. F. Yuan, et al., Engineering triangular carbon quantum dots with unprecedented narrow bandwidth emission for multicolored LEDs, Nat. Commun., 2018, 9, 1–11.
120. B. Vercelli, et al., Hydrothermal Approach for the Preparation of Blue–Emitting Carbon Quantum Dots: An Insight into the Influence of the Reaction Parameters, Small Struct., 2025, 6, 2400481.
121. Y. Han, L. Liccardo, E. Moretti, H. Zhao and A. Vomiero, Synthesis, optical properties and applications of red/near-infrared carbon dots, J. Mater. Chem. C, 2022, 10, 11827–11847.
122. B. De and N. Karak, A green and facile approach for the synthesis of water soluble fluorescent carbon dots from banana juice, RSC Adv., 2013, 3, 8286–8290.
123. S. A. A. Vandarkuzhali, et al., Highly fluorescent carbon dots from Pseudo-stem of banana plant: Applications as nanosensor and bio-imaging agents, Sens. Actuators, B, 2017, 252, 894–900.
124. N. Chaudhary, P. K. Gupta, S. Eremin and P. R. Solanki, One-step green approach to synthesize highly fluorescent carbon quantum dots from banana juice for selective detection of copper ions, J. Environ. Chem. Eng., 2020, 8, 103720.
125. M. He, et al., Material and Optical Properties of Fluorescent Carbon Quantum Dots Fabricated from Lemon Juice via Hydrothermal Reaction, Nanoscale Res. Lett., 2018, 13, 175.
126. A. Gholipour and S. Rahmani, The Green Synthesis of Carbon Quantum Dots through One-step Hydrothermal Approach by Orange Juice for Rapid, and Accurate Detection of Dopamine, J. Fluoresc., 2024, 34, 2665–2677.
127. S. Sahu, B. Behera, T. K. Maiti and S. Mohapatra, Simple one-step synthesis of highly luminescent carbon dots from orange juice: Application as excellent bio-imaging agents, Chem. Commun., 2012, 48, 8835–8837.
128. M. Preethi, R. Murugan, C. Viswanathan and N. Ponpandian, Potato starch derived N-doped carbon quantum dots as a fluorescent sensing tool for ascorbic acid, J. Photochem. Photobiol., A, 2022, 431, 114009.
129. S. Min, P. Ezati and J. W. Rhim, Gelatin-based packaging material incorporated with potato skins carbon dots as functional filler, Ind. Crops Prod., 2022, 181, 114820.
130. Y. Zhao, et al., Novel carbon quantum dots from egg yolk oil and their haemostatic effects, Sci. Rep., 2017, 7, 1–8.
131. J. Wang, C. F. Wang and S. Chen, Amphiphilic egg-derived carbon dots: Rapid plasma fabrication, pyrolysis process, and multicolor printing patterns, Angew. Chem., Int. Ed., 2012, 51, 9297–9301.
132. X. Liu, et al., High yield conversion of biowaste coffee grounds into hierarchical porous carbon for superior capacitive energy storage, Sci. Rep., 2020, 10, 1–12.
133. D. J. Kim, et al., Graphene Quantum Dots from Carbonized Coffee Bean Wastes for Biomedical Applications, Nanomaterials, 2021, 11, 1423.
134. Y. Zhou, et al., Synthesis of carbon dots from spent coffee grounds: transforming waste into potential biomedical tools, Nanoscale, 2025, 17, 9947–9962.
135. A. I. Costa, P. D. Barata, B. Moraes and J. V. Prata, Carbon Dots from Coffee Grounds: Synthesis, Characterization and Detection of Noxious Nitroanilines, Chemosensors, 2022, 10, 1–17.
136. N. Papaioannou, et al., Structure and solvents effects on the optical properties of sugar-derived carbon nanodots, Sci. Rep., 2018, 8, 1–10.
137. B. Mukherjee, et al., Sugar-Derived Carbon Dots with Variable Fluorescence Blinking Property as a Super-Resolution Imaging Probe, J. Phys. Chem. C, 2024, 128, 19260–19269.
138. S. G. Eswaran, D. Thiruppathi and N. Vasimalai, Synthesis of highly fluorescent carbon dots from bread waste and their nanomolar lead ions sensor application, Environ. Nanotechnol., Monit. Manage., 2022, 18, 100748.
139. S. Supajaruwong, et al., Scaling-up of carbon dots hydrothermal synthesis from sugars in a continuous flow microreactor system for biomedical application as in vitro antimicrobial drug nanocarrier, Sci. Technol. Adv. Mater., 2023, 24, 2260298.
140. M. P. Sk, A. Jaiswal, A. Paul, S. S. Ghosh and A. Chattopadhyay, Presence of Amorphous Carbon Nanoparticles in Food Caramels, Sci. Rep., 2012, 2, 383.
141. A. G. Kolekar, et al., Prompt in situ synthesis of sulphur doped carbon dots from jaggery for parallel determination of iron, chromium and manganese in environmental samples, J. Photochem. Photobiol., A, 2024, 454, 115672.
142. M. Picard, S. Thakur, M. Misra and A. K. Mohanty, Miscanthus grass-derived carbon dots to selectively detect Fe3+ ions, RSC Adv., 2019, 9, 8628–8637.
143. M. Sabet and K. Mahdavi, Green synthesis of high photoluminescence nitrogen-doped carbon quantum dots from grass via a simple hydrothermal method for removing organic and inorganic water pollutions, Appl. Surf. Sci., 2019, 463, 283–291.
144. Y. An, et al., Red, green, and blue light-emitting carbon dots prepared fromo-phenylenediamine, RSC Adv., 2021, 11, 26915–26919.
145. J. D. Stachowska, et al., A rich gallery of carbon dots based photoluminescent suspensions and powders derived by citric acid/urea, Sci. Rep., 2021, 11, 1–14.
146. J. Gogoi and D. Chowdhury, Calcium-modified carbon dots derived from polyethylene glycol: fluorescence-based detection of Trifluralin herbicide, J. Mater. Sci., 2020, 55, 11597–11608.
147. N. Sahiner, S. S. Suner, M. Sahiner and C. Silan, Nitrogen and Sulfur Doped Carbon Dots from Amino Acids for Potential Biomedical Applications, J. Fluoresc., 2019, 29, 1191–1200.
148. X. Lin and X. Chen, Gold nanoparticles stabilized with four kinds of amino acid-derived carbon dots for colorimetric and visual discrimination of proteins and microorganisms, Microchim. Acta, 2019, 186, 513.
149. V. B. Kumar, S. K. Mirsky, N. T. Shaked and E. Gazit, High Quantum Yield Amino Acid Carbon Quantum Dots with Unparalleled Refractive Index, ACS Nano, 2024, 18, 2421–2433.
150. T. Prathumsuwan, S. Jamnongsong, S. Sampattavanich and P. Paoprasert, Preparation of carbon dots from succinic acid and glycerol as ferrous ion and hydrogen peroxide dual-mode sensors and for cell imaging, Opt. Mater., 2018, 86, 517–529.
151. D. Uriarte, C. Domini and M. Garrido, New carbon dots based on glycerol and urea and its application in the determination of tetracycline in urine samples, Talanta, 2019, 201, 143–148.
152. L. Lin, et al., Hydrothermal synthesis of nitrogen and copper co-doped carbon dots with intrinsic peroxidase-like activity for colorimetric discrimination of phenylenediamine isomers, Microchim. Acta, 2019, 186, 288.
153. H. Barhum, et al., Multicolor Phenylenediamine Carbon Dots for Metal-Ion Detection with Picomolar Sensitivity, ACS Appl. Nano Mater., 2021, 4, 9919–9931.
154. C. Ji, et al., Phenylenediamine-derived near infrared carbon dots: The kilogram-scale preparation, formation process, photoluminescence tuning mechanism and application as red phosphors, Carbon, 2022, 192, 198–208.
155. S. Wu, et al., N-Acetylcysteine-Derived Carbon Dots for Free Radical Scavenging in Intervertebral Disc Degeneration, Adv. Healthcare Mater., 2023, 12, 1–15.
156. A. Döring, Y. Qiu and A. L. Rogach, Improving the Accuracy of Carbon Dot Temperature Sensing Using Multi-Dimensional Machine Learning, ACS Appl. Nano Mater., 2024, 7, 2258–2269.
157. J. Guo, et al., Solvent-Engineered Synthesis of Multicolor Emissive Carbon Dots from Citric Acid and Cyanamide, Ind. Eng. Chem. Res., 2024, 63, 9050–9057.
158. Y. Ru, et al., Rational Design of Multicolor-Emitting Chiral Carbonized Polymer Dots for Full-Color and White Circularly Polarized Luminescence, Angew. Chem., Int. Ed., 2021, 60, 14091–14099.
159. R. Kumari and S. K. Sahu, Effect of Solvent-Derived Highly Luminescent Multicolor Carbon Dots for White-Light-Emitting Diodes and Water Detection, Langmuir, 2020, 36, 5287–5295.
160. Y. Yan, L. Xia and L. Ma, Solvent-controlled synthesis of multicolor photoluminescent carbon dots for bioimaging, RSC Adv., 2019, 9, 24057–24065.
161. J. Zhan, et al., A solvent-engineered molecule fusion strategy for rational synthesis of carbon quantum dots with multicolor bandgap fluorescence, Carbon, 2018, 130, 153–163.
162. L. Wang, et al., Full-color fluorescent carbon quantum dots, Sci. Adv., 2020, 6, 1–8.
163. B. Zheng, et al., One pot selective synthesis of water and organic soluble carbon dots with green fluorescence emission, RSC Adv., 2015, 5, 11667–11675.
164. N. Papaioannou, M. M. Titirici and A. Sapelkin, Investigating the Effect of Reaction Time on Carbon Dot Formation, Structure, and Optical Properties, ACS Omega, 2019, 4, 21658–21665.
165. H. Kalita, et al., Efficient synthesis of rice based graphene quantum dots and their fluorescent properties, RSC Adv., 2016, 6, 23518–23524.
166. K. Zhao, et al., Multi-color fluorescent carbon dots with single wavelength excitation for white light-emitting diodes, J. Alloys Compd., 2019, 793, 613–619.
167. Y. Zhang, et al., Effect of reaction temperature on structure and fluorescence properties of nitrogen-doped carbon dots, Appl. Surf. Sci., 2016, 387, 1236–1246.
168. A. F. Sonsin, et al., Temperature-dependence on the optical properties of chitosan carbon dots in the solid state, RSC Adv., 2021, 11, 2767–2773.
169. J. Ahn, C. Pealer and S. Kim, Analysis of Carbon Dots Synthesized at Different Hydrothermal Carbonization Temperatures and Their Ferric Ion Detection Performances, Phys. Status Solidi B, 2024, 261, 1–7.
170. P. Yu, X. Wen, Y. R. Toh and J. Tang, Temperature-dependent fluorescence in carbon dots, J. Phys. Chem. C, 2012, 116, 25552–25557.
171. Q. Wu, et al., Effect of reaction temperature on properties of carbon nanodots and their visible-light photocatalytic degradation of tetracyline, RSC Adv., 2015, 5, 75711–75721.
172. N. Santos, et al., Effect of temperature on the carbonization process of cationic carbon dots: a physicochemical and in vitro study, RSC Adv., 2025, 15, 12814–12824.
173. Y. Song, et al., Investigation from chemical structure to photoluminescent mechanism: A type of carbon dots from the pyrolysis of citric acid and an amine, J. Mater. Chem. C, 2015, 3, 5976–5984.
174. M. Shamsipur, A. Barati, A. A. Taherpour and M. Jamshidi, Resolving the Multiple Emission Centers in Carbon Dots: From Fluorophore Molecular States to Aromatic Domain States and Carbon-Core States, J. Phys. Chem. Lett., 2018, 9, 4189–4198.
175. J. Wang, et al., Acid-Controlled Fabrication of Multicolor Carbon Dots with Switchable Organelle-Targeting Capability for Visualizing Organelle Interactions, Anal. Chem., 2025, 97, 5668–5677.
176. R. Dai, X. Chen, N. Ouyang and Y. Hu, A pH-controlled synthetic route to violet, green, and orange fluorescent carbon dots for multicolor light-emitting diodes, Chem. Eng. J., 2022, 431, 134172.
177. S. Lu, et al., PH-Dependent Synthesis of Novel Structure-Controllable Polymer-Carbon NanoDots with High Acidophilic Luminescence and Super Carbon Dots Assembly for White Light-Emitting Diodes, ACS Appl. Mater. Interfaces, 2016, 8, 4062–4068.
178. Z. Sun, et al., High-efficient and pH-sensitive orange luminescence from silicon-doped carbon dots for information encryption and bio-imaging, J. Colloid Interface Sci., 2022, 607, 16–23.
179. M. Wu, J. Li, Y. Wu, X. Gong and M. Wu, Design of a Synthetic Strategy to Achieve Enhanced Fluorescent Carbon Dots with Sulfur and Nitrogen Codoping and Its Multifunctional Applications, Small, 2023, 19, 1–11.
180. N. Li, et al., One-step synthesis of N, P Co-doped orange carbon quantum dots with novel optical properties for bio-imaging, Opt. Mater., 2021, 111, 110618.
181. F. Li, et al., Mg/N double doping strategy to fabricate extremely high luminescent carbon dots for bioimaging, RSC Adv., 2014, 4, 3201–3205.
182. Z. Deng, et al., High quantum yield blue- and orange-emitting carbon dots: One-step microwave synthesis and applications as fluorescent films and in fingerprint and cellular imaging, Analyst, 2019, 144, 4569–4574.
183. S. Y. Song, et al., Deep-Ultraviolet Emissive Carbon Nanodots, Nano Lett., 2019, 19, 5553–5561.
184. Z. Lin, W. Xue, H. Chen and J. M. Lin, Classical oxidant induced chemiluminescence of fluorescent carbon dots, Chem. Commun., 2012, 48, 1051–1053.
185. K. Holá, et al., Graphitic Nitrogen Triggers Red Fluorescence in Carbon Dots, ACS Nano, 2017, 11, 12402–12410.
186. G. Hu, et al., Rational synthesis of silane-functionalized carbon dots with high-efficiency full-color solid-state fluorescence for light emitting diodes, Carbon, 2023, 203, 1–10.
187. M. Lei, et al., Surface state modulation of blue-emitting carbon dots with high quantum yield and high product yield, RSC Adv., 2022, 12, 27431–27441.
188. Q. Zhao, et al., Surface amino group modulation of carbon dots with blue, green and red emission as Cu2+ ion reversible detector, Appl. Surf. Sci., 2022, 598, 153892.
189. H. Zhang, et al., Surface State-Based panchromatic luminescent carbon dots, J. Colloid Interface Sci., 2025, 678, 77–87.
190. X. Wen, P. Yu, Y. R. Toh, X. Hao and J. Tang, Intrinsic and Extrinsic Fluorescence in Carbon Nanodots: Ultrafast Time-Resolved Fluorescence and Carrier Dynamics, Adv. Opt. Mater., 2013, 1, 173–178.
191. K. Jiang, Y. Wang, C. Cai and H. Lin, Conversion of Carbon Dots from Fluorescence to Ultralong Room-Temperature Phosphorescence by Heating for Security Applications, Adv. Mater., 2018, 30, 1–8.
192. Q. Xu, et al., Single-particle spectroscopic measurements of fluorescent graphene quantum dots, ACS Nano, 2013, 7, 10654–10661.
193. V. Nidhisha, et al., p-Phenylenediamine-derived carbon nanodots for probing solvent interactions, Nanoscale Adv., 2024, 6, 1535–1547.
194. N. Basu and D. Mandal, Solvatochromic Response of Carbon Dots: Evidence of Solvent Interaction with Different Types of Emission Centers, J. Phys. Chem. C, 2018, 122, 18732–18741.
195. C. M. Singaravelu, X. Deschanels, C. Rey and J. Causse, Investigation on Fluorescence Origin and Spectral Heterogeneity in Carbon Dots: A Dynamic Perspective, ChemPhotoChem, 2024, 8, e202400044.
196. J. R. Macairan, et al., Elucidating the mechanism of dual-fluorescence in carbon dots, J. Colloid Interface Sci., 2022, 606, 67–76.
197. H. Shabbir, E. Csapó and M. Wojnicki, Carbon Quantum Dots: The Role of Surface Functional Groups and Proposed Mechanisms for Metal Ion Sensing, Inorganics, 2023, 11, 262.
198. Q. Zhang, R. Wang, B. Feng, X. Zhong and K. K. Ostrikov, Photoluminescence mechanism of carbon dots: triggering high-color-purity red fluorescence emission through edge amino protonation, Nat. Commun., 2021, 12, 1–13.
199. M. Righetto, et al., The Elusive Nature of Carbon Nanodot Fluorescence: An Unconventional Perspective, J. Phys. Chem. C, 2020, 124, 22314–22320.
200. B. van Dam, et al., Excitation–Dependent Photoluminescence from Single–Carbon Dots, Small, 2017, 13, 1–5.
201. S. Elkun, M. Ghali, T. Sharshar and M. M. Mosaad, Green synthesis of fluorescent N-doped carbon quantum dots from castor seeds and their applications in cell imaging and pH sensing, Sci. Rep., 2024, 14, 27927.
202. B. Chen, et al., Large scale synthesis of photoluminescent carbon nanodots and their application for bioimaging, Nanoscale, 2013, 5, 1967–1971.
203. Y. Zhang, et al., Carbon Dots with Blue–to–Near–Infrared Lasing for Colorful Speckle–Free Laser Imaging and Dynamical Holographic Display, Adv. Mater., 2023, 35, 2302536.
204. S. Mohandoss, et al., Excitation-dependent multiple luminescence emission of nitrogen and sulfur co-doped carbon dots for cysteine sensing, bioimaging, and photoluminescent ink applications, Microchem. J., 2021, 167, 106280.
205. S. Khan, A. Gupta, N. C. Verma and C. K. Nandi, Time-Resolved Emission Reveals Ensemble of Emissive States as the Origin of Multicolor Fluorescence in Carbon Dots, Nano Lett., 2015, 15, 8300–8305.
206. S. K. Cushing, M. Li, F. Huang and N. Wu, Origin of strong excitation wavelength dependent fluorescence of graphene oxide, ACS Nano, 2014, 8, 1002–1013.
207. M. Fu, et al., Carbon Dots: A Unique Fluorescent Cocktail of Polycyclic Aromatic Hydrocarbons, Nano Lett., 2015, 15, 6030–6035.
208. X. Li, S. Zhang, S. A. Kulinich, Y. Liu and H. Zeng, Engineering surface states of carbon dots to achieve controllable luminescence for solid-luminescent composites and sensitive Be2+ detection, Sci. Rep., 2014, 4, 1–8.
209. Y. Wang, J. Zheng, J. Wang, Y. Yang and X. Liu, Rapid microwave-assisted synthesis of highly luminescent nitrogen-doped carbon dots for white light-emitting diodes, Opt. Mater., 2017, 73, 319–329.
210. B. Xu, et al., Solid-State Fluorescent Carbon Dots with Unprecedented Efficiency from Visible to Near-Infrared Region, Adv. Sci., 2023, 10, 1–8.
211. B. Zhang, et al., Assignment of Core and Surface States in Multicolor-Emissive Carbon Dots, Small, 2023, 19, 1–11.
212. R. Umami, F. A. Permatasari, C. D. D. Sundari, A. S. Santika and F. Iskandar, Positioning of functional group for tailoring absorption spectrum of carbon dots: Insights from density functional theory, Mater. Chem. Phys., 2024, 318, 129243.
213. Z. Yan, et al., One-Step Synthesis of White-Light-Emitting Carbon Dots for White LEDs with a High Color Rendering Index of 97, Adv. Sci., 2023, 10, 1–10.
214. H. Hao, et al., Boron-Doped Engineering for Carbon Quantum Dots-Based Memristors with Controllable Memristance Stability, Small Methods, 2024, 8, 1–13.
215. S. Sharma and P. Chowdhury, Tunable dual photoluminescence from synthesized urea-based carbon quantum dots: A DFT based simulation on structural insights, Opt. Mater., 2024, 153, 115617.
216. A. T. Sheardy, D. M. Arvapalli and J. Wei, Experimental and Time-Dependent Density Functional Theory Modeling Studies on the Optical Properties of Carbon Nanodots, J. Phys. Chem. C, 2020, 124, 4684–4692.
217. R. E. Ambrusi, et al., Density functional theory model for carbon dot surfaces and their interaction with silver nanoparticles, Phys. E, 2019, 114, 113640.
218. S. Lu, et al., Piezochromic Carbon Dots with Two-photon Fluorescence, Angew. Chem., Int. Ed., 2017, 56, 6187–6191.
219. B. Zhao, et al., Triphenylamine–Derived Solid–State Emissive Carbon Dots for Multicolor High–Efficiency Electroluminescent Light–Emitting Diodes, Angew. Chem., 2023, 135, e202301651.
220. L. Yang, X. Li, Y. Zhang, P. Yin and X. Sun, DFT-D3 and TD-DFT studies on the sensor performance of the functionalized carbon quantum dots for curcumin, Surf. Interfaces, 2024, 45, 103870.
221. K. G. Nguyen, et al., Controlling the optoelectronic properties of nitrogen-doped carbon quantum dots using biomass-derived precursors in a continuous flow system, Carbon, 2024, 230, 119623.
222. Y. Zhang, et al., Carbon dots nanophotosensitizers with tunable reactive oxygen species generation for mitochondrion-targeted type I/II photodynamic therapy, Biomaterials, 2023, 293, 121953.
223. P. Wei, et al., Manipulation of Charge-Transfer Kinetics via Ti3C2Tx (T = −O) Quantum Dot and N-Doped Carbon Dot Coloading on CdS for Photocatalytic Hydrogen Production, ACS Catal., 2023, 13, 587–600.
224. B. Geng, et al., Near-infrared phosphorescent carbon dots for sonodynamic precision tumor therapy, Nat. Commun., 2022, 13, 5735.
225. H. Tetsuka, A. Nagoya, T. Fukusumi and T. Matsui, Molecularly Designed, Nitrogen-Functionalized Graphene Quantum Dots for Optoelectronic Devices, Adv. Mater., 2016, 28, 4632–4638.
226. M. Paloncýová, M. Langer and M. Otyepka, Structural Dynamics of Carbon Dots in Water and N, N -Dimethylformamide Probed by All-Atom Molecular Dynamics Simulations, J. Chem. Theory Comput., 2018, 14, 2076–2083.

---