Efficient bioimaging using rare-earth-doped carbon quantum dots: A doping strategy

Summary

This study addressed the limitations of carbon quantum dots (CQDs) in bioimaging, specifically low photoluminescence quantum yield (PLQY) and complex purification processes, by developing lanthanide-doped CQDs (L-CQDs). Using a microwave-assisted solvothermal method in a glycerol-water system with precursor mass ratios of citric acid monohydrate: urea: NdCl₃·6H₂O:CeCl₃ at 30:3:10:2 and a glycerol/water volume ratio of 2:1, we achieved one-pot synthesis with cerium and neodymium doping. This approach significantly enhanced the PLQY from 0.43% to 69%. The resulting L-CQDs exhibited a graphene-like structure (≈8 nm) and demonstrated anti-Stokes blue emission under 740 and 800 nm excitation. Besides, ethanol precipitation facilitated purification and indicated surface hydrophobicity post-doping. The materials showed excellent biosafety in HUVEC cells, RAW264.7 cells, and C57 mice, with enhanced fluorescence intensity and specific accumulation in highly perfused organs, including the brain, lungs, spleen, and kidneys, while demonstrating blood-brain barrier penetration capability. These findings highlighted the potential of L-CQDs for cellular labeling and in vivo fluorescence imaging applications.

Graphical abstract

Highlights

• •High FLQY carbon quantum dots with blue light were prepared by a microwave one-pot method
• •L-CQDs showed an Excellent biocompatibility both in vivo and in vitro
• •L-CQDs exhibited enhanced bioimaging with multi-organ, compared to the normal CQDs

Optical imaging; Applied sciences

Introduction

Carbon quantum dots (CQDs), an emerging class of nanomaterials in recent years, have attracted attention in the field of biomaterials.¹ Numerous studies have applied C-dots in biosensing,^2,3 drug delivery systems,^4,5 antibacterial and antioxidant agents,^6,7 fluorescent inks,⁸ and bioimaging.⁹ However, CQDs still faced challenges such as insufficient fluorescence intensity, difficulty in controlling upconversion luminescence¹⁰ and tissue specific regulation in organ imaging.

Rare-earth ion-doped fluorescent nanomaterials typically exhibit high quantum yields and brightness, providing strong fluorescent signals in complex biological environments,¹¹ achieve upconversion luminescence.¹² Nanocomposites improved light, heat, and chemical stability.¹³ However, current rare-earth-doped CQDs suffer from various issues such as rare-earth segregation phenomena,¹⁴ low reaction kinetic,¹⁵ difficulty in separation¹⁶ and non standardized.^17,18 From a fluorescent perspective, energy back to the low-energy states causes unnecessary low photoluminescence quantum yield (LQY).¹⁹ From a biological application perspective, the introduction of rare-earth ions (e.g., Gd³⁺ and Dy³⁺) may induce mitochondrial damage²⁰ and oxidative stress,^21,22 leading to unintended toxicity. However, neodymium ions (Nd³⁺) have infrared absorption at 740 nm and 800 nm, enabling efficient infrared or upconversion emission.²³ Cerium ions (Ce³⁺) have excellent biocompatibility and UV absorption, can enhance the PLQY.²⁴ Furthermore, cerium and neodymium showed good biocompatibility²⁵ and excellent near-infrared fluorescence performance, fluorescence sensitization ability, and reaction catalytic capability.²⁶

In this study, we plan to design a type of upconversion carbon quantum dots with good biocompatibility and high fluorescence quantum efficiency, and to equip them with a standardized and readily available purification method to achieve in vitro and in vivo cell imaging. They are incorporated into CQDs via a microwave-assisted solvothermal method under acidic conditions. Utilizing the catalytic performance of cerium ions to achieve high crystallinity and surface passivation, the resulting CQDs can be rapidly purified using ethanol as a precipitant. Address the issues of existing rare-earth doping, such as non-radiative losses, complex preparation and purification processes, and potential biosafety risks. This yields CQDs with uniform doping, good biocompatibility, and high fluorescence intensity, aiming to develop this material into an ideal candidate for bioimaging (Scheme 1).

Scheme 1.

Preparation process and application of L-CQDs

Results and discussion

Preparation and optimization of lanthanide-doped carbon quantum dots

The screening of reaction raw material types is shown in Figure S1. Matrix screening was performed using an orthogonal method (Figure S1A), and doping screening was performed through the orthogonal analysis of single-doped and double-doped groups (Figure S1B). Based on statistically significant fluorescence differences, glycerol, citric acid monohydrate, neodymium chloride hexahydrate, and cerium chloride were selected as reaction raw materials.

The preparation of L-CQDs was optimized by separately investigating the effects of reactant types, reactant mass ratios, reaction time, and volume percentage of organic solvent on the fluorescence properties of L-CQDs, aiming to obtain the optimal preparation parameters.

The effect of N-element doping on the fluorescence properties of L-CQDs was examined (Figure S5A), which showed photographs of the prepared L-CQDs, and (Figure S5B) showed fluorescence photographs of the prepared L-CQDs. Comparison of mass ratios 10:0 and 10:1 indicated that doping nitrogen elements significantly enhanced the fluorescence intensity of L-CQDs. However, comparison of 10:1 and 10:2 showed that excessive nitrogen doping caused the L-CQDs solution to turn significantly yellow, and the fluorescence intensity decreased. This may be due to excessive nitrogen element doping on the surface, causing the surface state luminescence centers to radiate toward defect states, leading to significant fluorescence concentration quenching effects.²⁷ Therefore, L-CQDs obtained with urea providing nitrogen at a mass ratio of 10:1 were selected for further optimization.

The effect of the rare-earth element doping ratio on the fluorescence properties of L-CQDs was examined (Figure S6) shows fluorescence photographs of the prepared L-CQDs. As the mass ratio increased from 30:3:2:2 to 30:3:10:2, the fluorescence intensity gradually increased. For mass ratios from 30:3:10:1 to 30:3:10:3, the fluorescence intensity first increased and then decreased. During the experiment, we also observed the reaction solution color gradually turning dark brown. This might be due to excessive cerium ions over-catalyzing the reaction,²⁸ leading to over-carbonization,²⁹ which disrupts the original surface structure of the carbon quantum dots and reduces luminescence intensity. Therefore, L-CQDs obtained at a mass ratio of 30:3:10:2 were selected for further optimization.

The effect of reaction time on the fluorescence properties of L-CQDs was examined (Figure S7) shows fluorescence photographs of the prepared L-CQDs. As the reaction time increased from 40 s to 60 s, the fluorescence intensity gradually increased. L-CQDs with a reaction time of 40 s showed almost no fluorescence, indicating that L-CQDs were barely formed at 40 s. As the reaction time increased from 60 s to 70 s, the fluorescence intensity decreased. This might be because, with longer reaction times, the carbonization degree of the carbon quantum dots deepens, surface passivation improves, and fluorescence intensity increases. However, with further prolongation of the reaction time, the carbon quantum dots become over-carbonized, defect states increase,³⁰ and aggregation-induced quenching occurs, leading to decreased fluorescence intensity.^31,32 Therefore, L-CQDs obtained with a reaction time of 60 s were selected for further optimization.

The effect of the volume ratio of water to glycerol in the reaction solvent on the fluorescence properties of L-CQDs was examined (Figure S8A), showing fluorescence photographs of the prepared L-CQDs. As the volume ratio increased from 25% to 50%, the fluorescence intensity gradually increased; as the volume ratio increased from 50% to 100%, the fluorescence intensity decreased (Figure S8B) shows the curve of the reaction system temperature versus time for a volume ratio of 50%. When water and glycerol coexist, the reaction undergoes Phase I: rapid heating from 0°C to 126°C; Phase II: slow rise to 257°C; Phase III: natural standing after reaction ends; and Phase IV: rapid cooling to 93°C after adding water to quench the reaction. Phase I corresponds to the evaporation of water in the microwave field. During this process, rapid boiling of water due to evaporation increases local pressure in the reaction solution, pushing the reaction precursors into a rapid polymerization phase.³³ Phase II corresponds to the heating of glycerol in the microwave field. During this process, the reaction precursors continue to polymerize and carbonize to form carbon quantum dots upon reaching the peak temperature.³⁴ After stopping the reaction, if quenching is not performed, the temperature decreases slowly due to polymerization exotherm and glycerol’s heat retention capacity, and the carbonization process of the carbon quantum dots continues. Quenching rapidly lowers the temperature, terminating the reaction and ensuring uniform particle size and good crystallinity of the carbon quantum dots. When the volume ratio is 50%, the durations of Phases I and II are appropriate, promoting efficient synthesis of carbon quantum dots, hence the better fluorescence intensity.^33,35 Therefore, a volume ratio of water to glycerol of 50% in the reaction solvent was selected as the reaction condition for L-CQDs.

In summary, the optimal synthesis conditions for L-CQDs are as follows: reactant mass ratio set to 30:3:10:2, volume ratio of water to glycerol in the reaction solvent set to 50%, and a reaction time of 60 s. L-CQDs obtained under these conditions exhibit the best fluorescence intensity. The preparation flowchart of L-CQDs is shown schematically in (Figure 1A).

Figure 1.

The preparation flowchart of L-CQDs, TEM particle size and morphology image, SAED pattern image, and EDS image

(A) Schematic diagram of the preparation process of L-CQDs.

(B) TEM particle size distribution of L-CQDs.

(C) TEM image of L-CQDs, inset: HRTEM image of L-CQDs. (Scale bars, 100 nm).

(D) SAED pattern of L-CQDs. (Scale bars, 10 1/nm).

(E) EDS energy spectrum of L-CQDs under SEM scanning.

Structural characterization of lanthanide-doped carbon quantum dots

To elucidate the morphology and carbon core structure of L-CQDs, transmission electron microscopy (TEM) images were acquired, and the particle size of the carbon quantum dots was calculated. Selected area electron diffraction (SAED) was performed to observe the diffraction patterns of the carbon quantum dots. Energy-dispersive X-ray spectroscopy (EDS) was measured under SEM to determine the elemental composition. TEM images (Figures 1B and 1C) showed that L-CQDs had a quasi-spherical morphology and were uniformly dispersed. Specifically, the average particle size of L-CQDs was 8.44 ± 2.03 nm. High-resolution transmission electron microscopy (HRTEM, as shown in the inset of Figure 1B) analysis revealed clear lattice fringes in L-CQDs, indicating a graphite-like carbon core structure. The lattice spacing of 0.210 nm for L-CQDs corresponds to the (100) plane of graphitic carbon,³⁶ the 0.318 nm spacing corresponds to the (002) plane,³⁷ further confirming their graphitic nature, and the 0.235 nm spacing corresponds to the (100) plane of amorphous carbon-doped graphitic carbon.³⁸ SAED (Figure 1D) showed distinct graphite (002) and (100) plane diffraction rings for L-CQDs.³⁷ Since the selected area contains numerous carbon quantum dot monomers, it appears as ring-shaped diffraction bands, further confirming the existence of a graphitic structure within the carbon cores. EDS spectra obtained under SEM (Figure 1E) confirmed that the carbon quantum dots consisted mainly of carbon and oxygen, doped with nitrogen, neodymium, and cerium elements.

Chemical environment characterization of lanthanide-doped carbon quantum dots

To further confirm the surface elemental composition and structure of L-CQDs, Fourier transform infrared spectroscopy (FT-IR), ultraviolet-visible (UV-Vis) spectroscopy, X-ray diffraction (XRD), Raman spectroscopy, and X-ray photoelectron spectroscopy (XPS) characterizations were performed. The FT-IR spectrum of L-CQDs showed absorption peaks at 3,402 cm⁻¹ attributed to the stretching vibrations of N-H and O-H bonds,³⁹ a peak at 2,965 cm⁻¹ corresponding to sp³ C-H stretching vibration, and peaks at 1,721, 1,580, and 1,416 cm⁻¹ corresponding to C=O stretching vibration,³⁸ and C-N, C=C,³⁴ and C-O stretching vibrations,³⁶ respectively. Notably, R-CQDs lack the C=C stretching vibration peak around 1,580 cm⁻¹, and the intensity of the C=O stretching vibration peak at 1,724 cm⁻¹ is lower. This reduction in C=O stretching vibration intensity is also observed in N-CQDs (Figure 2A). These observations suggest the formation of N, O-conjugated aromatic rings and sp² graphene structures within L-CQDs, confirming the successful doping of N atoms and the formation of a large conjugated system. Rare-earth doping effectively catalyzes the formation of sp² conjugated structures in carbon quantum dots, and cerium doping enhances surface passivation efficiency.

Figure 2.

Characterization of L-CQDs by FT-IR, UV-Vis, XRD, Raman spectroscopy, and XPS

(A) FT-IR spectrum of L-CQDs.

(B) UV-Vis absorption spectrum of L-CQDs.

(C) XRD spectrum of L-CQDs.

(D) Raman spectrum of L-CQDs.

(E) High-resolution C 1s XPS spectrum of L-CQDs.

(F) High-resolution N 1s XPS spectrum of L-CQDs.

(G) High-resolution O 1s XPS spectrum of L-CQDs.

UV-Vis spectra (Figures 2B and S2) show a significant tailing absorption below about 240 nm, attributed to sp² π-π∗ transitions,⁴⁰ indicating the presence of a graphene carbon core structure in L-CQDs. Due to the special position of Ce among the lanthanides, L-CQDs exhibit band-like Ce 4f-5d transition peaks at 247 and 263 nm compared to R-CQDs, indicating successful Ce doping. Compared to R-CQDs, L-CQDs lack the n-π∗ transition of C=O/C=N around 330 nm, which we speculate is due to ligand-to-metal charge transfer (LMCT) coordination interactions between rare-earth ions and surface amino and carboxyl groups.^37,38 Multiple linear absorption peaks at 580, 730, 740, and 800 nm are attributed to the spectral term transitions of Nd based on ⁴I₉/₂, indicating successful Nd doping.

XRD patterns (Figures 2C and S3) show that both R-CQDs and L-CQDs have two significant diffraction peaks at approximately 21.7°/37.2° and 29.8°/44.2°, attributed to the (002)/(100) crystal planes of graphitic carbon.³⁷ The diffraction peaks of N-CQDs and C-CQDs are similar to those of L-CQDs. The expanded interplanar spacing in R-CQDs is due to the interlayer doping of amorphous carbon.³⁸ This further confirms the existence of a graphitic structure within the carbon cores, and the graphitization degree is higher in rare-earth-doped carbon cores.

Raman spectra (Figure 2D) show that both R-CQDs and L-CQDs have characteristic peaks near 1,380 cm⁻¹ (D band) and 1,600 cm⁻¹ (G band), corresponding to carbon lattice defects and the in-plane stretching vibrations of sp² hybridized carbon atoms, respectively. The D band of L-CQDs is remarkably weak, coupled with neighboring nitrogen-containing vibration peaks and thus undetectable. The intensity ratio of the G band to the D band (I_G/I_D) is an indicator of the graphitization degree, with a higher ratio indicating a higher degree of graphitization. The I_G/I_D value for R-CQDs is about 1.2, indicating a relatively high degree of graphitization for both, while the graphitization degree of L-CQDs is significantly higher than that of R-CQDs. This proves that using rare-earth doping indeed increases the graphitization degree of CQDs, which may be related to the direct action of rare earths or their catalytic effect on the reaction. In result, this is beneficial for enhancing the fluorescence emission intensity of CQDs.⁴¹

XPS spectra (Figures 2E, 2F, 2G, and S8) show that L-CQDs are composed of C, N, O, Nd, and Ce elements. The surface structure of L-CQDs was further analyzed from high-resolution XPS spectra. In L-CQDs, the C element exists in the forms of C=C/C-C, C-O, and O-C=O. The O element exists as C=O and C-O.⁴¹ Furthermore, the N element in L-CQDs exists primarily as graphitic nitrogen.^42,43 The presence of Nd and Ce elements in the XPS spectra confirms the successful doping of both Nd and Ce atoms into L-CQDs.

Optical properties of lanthanide-doped carbon quantum dots

The synthesized CQDs exhibited excellent blue fluorescence emission properties (Figure 3E). Among the synthesized CQDs, the optimal emission wavelength of L-CQDs was 430 nm (Figure 4A), with a corresponding quantum yield (QY) of approximately 69%. The optimal emission wavelengths of R-CQDs, C-CQDs, and N-CQDs are 440 nm, 430 nm, and 430 nm (Figures 4B–4D), with QYs of 0.43%, 3.55%, and 52.00%, respectively. Fluorescence emission spectra (Figures 3A and 3B) show that N-CQDs, C-CQDs, and L-CQDs all exhibit excitation-independent properties, meaning the emission peak positions remain unchanged despite variations in the excitation wavelength, indicating similar surface structures, which is beneficial for improving reproducibility. Additionally, as the excitation wavelength increases from 310 nm to 450 nm, the emission peak of R-CQDs shows a significant redshift, and the Stokes shift increases, indicating that R-CQDs have more than two emission centers. Upconversion emission spectra demonstrated that rare-earth-coated carbon quantum dots exhibited enhanced blue upconversion emission under 740 nm and 800 nm excitation compared to R-CQDs (Figure 3C). Due to the large differences in fluorescence intensity between groups, the images are normalized. In the non-normalized images (Figure S9), the emission under 800 nm excitation is primarily dominated by Ce, showing a single emission peak centered at 450 nm. The emission under 740 nm excitation is significantly stronger than under 800 nm excitation, dominated by Nd, showing a single emission peak centered at 440 nm (Figure 3C). This results in L-CQDs having a constant emission wavelength under various excitation wavelengths, which is unique. The main reason is the uniform surface structure of L-CQDs and the unique emission center after rare-earth doping, leading to a constant emission wavelength. We also studied the photostability of L-CQDs under 365 nm excitation (Figure 3D). Under continuous excitation for 120 min, L-CQDs demonstrated good resistance to photobleaching. The fluorescence decay rate (RFD) was approximately 1.102 × 10⁻³%/min.

Figure 3.

Optical properties of L-CQDs

(A) Fluorescence emission spectrum (normalized) of synthesized CQDs.

(B) Emission spectrum of L-CQDs after altering the excitation wavelength.

(C) Upconversion fluorescence emission spectrum (normalized) of L-CQDs. (D) Fluorescence decay curve of L-CQDs.

(E) Fluorescence emission of synthesized CQDs (captured by camera).

Figure 4.

Optical mechanism of L-CQDs

(A) EEM spectrum of R-CQDs.

(B) EEM spectrum of C-CQDs.

(C) EEM spectrum of N-CQDs.

(D) EEM spectrum of L-CQDs.

(E) Fluorescence mechanism of R-CQDs.

(F) Fluorescence mechanism of L-CQDs.

Fluorescence emission mechanism of lanthanide-doped carbon quantum dots

The mechanism behind the rare-earth enhancement of CQD fluorescence and upconversion emission was further investigated. First, the fluorescence emission EEM spectra of the four types of CQDs were measured. The EEM spectra of the CQDs (Figures 4A–4D) show that R-CQDs have an emission peak at 420 nm under 320 nm excitation and a peak at 550 nm under 460 nm excitation, indicating that R-CQDs possess more than two emission centers. In contrast, the emission peak under 460 nm excitation disappears in rare-earth-doped carbon quantum dots, suggesting that rare-earth doping can alter the chemical environment of the emission center, primarily absorbing at this wavelength, thereby forbidding this transition process. Meanwhile, the emission peaks of rare-earth-doped carbon quantum dots are redshifted, and the Stokes shift decreases. We also note that C-CQDs have weak emission under 230–280 nm excitation wavelengths, L-CQDs have strong emission in this range, while N-CQDs have no emission in this interval. This indicates a synergistic relationship between the two rare-earth elements: cerium ions can undergo Ce³⁺/Ce⁴⁺ conversion upon excitation,⁴⁴ acting as a sensitizer to absorb energy. Neodymium ions are closely bound to the carbon quantum dots, promoting energy transfer to and the number of radiative transitions with cerium ions, thereby enhancing luminescence.⁴⁵

The XRD patterns of the carbon quantum dots (Figures 2C and S7) show the X-ray diffraction patterns of the four types of carbon quantum dots. Crystallinity calculated based on peak area integration indicates that the crystallinity of cerium-doped carbon quantum dots is significantly higher than that of undoped ones. Since cerium ions can significantly catalyze organic reactions, especially nitrogen-containing oxidation reactions,⁴⁶ we observed that the synthesis process involving rare-earth elements proceeded at a significantly higher rate than the process without rare-earth elements, specifically manifested by an earlier transition point from transparent solution to turbid solution (about 10–15 s earlier under the optimal preparation conditions mentioned in the methods). The contribution of cerium ions to improving the conjugation of surface-isolated sp² clusters during the preparation of carbon quantum dots can also enhance luminescence.

The upconversion fluorescence emission spectra of the CQDs (Figure 3C) show that Ce has a weak and broad emission centered at 450 nm under 800 nm excitation, while Nd has a stronger and narrower emission at 440 nm under 740 nm excitation. Since Ce itself does not absorb in the near-infrared region, it is speculated that it primarily forms hybrid bands with narrow energy levels through oxygen-metal coordination doping on the graphene surface of the carbon quantum dots⁴⁷ and conjugation with nearby nitrogen-doped isolated sp² clusters.⁴⁸ Nd has sharp absorption peaks at 730/740 nm, attributed to ⁴I₉/₂ - ⁴D₃/₂/⁴I₉/₂ - ⁴G₅/₂ transitions,⁴⁹ and can itself act as an absorption center providing energy for the fluorescence emission of the carbon quantum dots. The emission at 440, 450 nm indicates that the upconversion luminescence of L-CQDs is still based on the luminescence mode of the carbon quantum dots.

We assumed that doping carbon nanomaterials with inorganic metal particles can transform the excitation-dependent fluorescence emission behavior of carbon quantum dots into excitation-independent behavior through surface passivation³⁴ (Figures 4E and 4F). Moreover, the chemical adsorption or covalent bonding between inorganic compounds and organic compounds can effectively increase the fluorescence quantum efficiency of carbon quantum dots by modulating the excitation energy gap.⁵⁰ To further illustrate that the effect of doped rare-earth elements on the surface state of carbon quantum dots is an important reason for enhancing their fluorescence activity, literature suggests⁵¹ that nitrogen-containing carbon quantum dots prepared from citric acid usually contain a large number of PL conjugating units (connected nitrogen-containing derivatives and some isolated sp² clusters), graphitic cores, and abundant defect states (C-related dangling bonds). Under UV light irradiation in the range of 280–380 nm, electrons in the π orbitals of the luminescent conjugating units can be excited to the π∗ orbitals (Process a). Then, some of the excited electrons in the π∗ orbitals may recombine directly with holes in the π orbitals, emitting PL signals centered at 420–440 nm, which can be called intrinsic emission, similar to the band-edge emission observed in QDs (Process b).⁵² However, other excited electrons in the π∗ orbitals may be trapped by defect states with energies lower than the π∗ orbitals before they finally recombine with holes in the π orbitals (Process c). Meanwhile, some electrons may be directly excited and trapped by defect states (Process d). Then, the electrons trapped by the defect states relax through either radiative (Process e) or non-radiative (Process f) pathways. The PL signal from radiative relaxation is defect emission, which is also observed in QDs.⁵³ Therefore, when the excitation wavelength is in the range of 280–380 nm, the carbon quantum dots produce both intrinsic PL and defect PL signals. When the excitation wavelength is greater than 380 nm, its energy is less than that required for the π-π∗ transition, and electrons can only be excited to the defect states, so only defect emission can be observed.

Therefore, the core luminescence pathway of citric acid-derived nitrogen-containing carbon quantum dots involves nitrogen-containing derivatives covalently linked to the CQDs, acting as luminescent centers to generate intrinsic PL signals. The main PL properties of the CQDs are similar to the corresponding nitrogen-containing derivatives and are less affected by doping elements. Furthermore, the QY of the CQDs mainly depends on the amount and PL properties of the contained nitrogen-containing derivatives and the number of defects. In other words, the different QYs of the four CQD samples should be the result of the synergistic effect of these factors. For example, R-CQDs have obvious defect state emission, indicating that R-CQDs are rich in surface defects, and their QY is relatively low. Rare-earth doping can alter the surface defect environment, thereby changing the defect state emission behavior and improving the QY of CQDs,⁵⁴ which is indeed the case. We also note that the solubility of rare-earth-doped carbon quantum dots in ethanol is significantly reduced compared to undoped rare-earth carbon quantum dots, allowing us to easily purify rare-earth-doped carbon quantum dots by precipitation with ethanol without relying on complex dialysis or freeze-drying processes. This also proves that rare-earth doping can alter the surface chemical environment of carbon quantum dots.

Based on the above intrinsic emission, we believe the core pathway for upconversion emission is that Ce, through conjugation with surface nitrogen-doped isolated sp² clusters, generates new broad absorption energy levels, transferring near-infrared energy into the intrinsic emission of the carbon quantum dots. Nd, based on its own spectral term transitions, generates absorption at 730/740 nm and 800 nm, and through the long decay time characteristic of rare-earth elements themselves, produces a multi-photon energy accumulation process and FRET process to transfer energy to the intrinsic emission of the carbon quantum dots. When Ce and Nd are co-doped, Ce can further sensitize Nd, enhancing the emission efficiency and producing upconversion luminescence.

Therefore, it can be concluded that the main emission spectra of L-CQDs depend primarily on the nature of the connected nitrogen-containing derivative conjugating units, while the QY is influenced by the nature of the conjugating units and the defect states. Thus, adding Ce and Nd can interact with the conjugating units and defect states, undergo FRET processes to enhance the intrinsic emission intensity of L-CQDs, increase the fluorescence intensity of L-CQDs, and generate upconversion luminescence.

Application of lanthanide-doped carbon quantum dots in bioimaging

Excellent biocompatibility and good intracellular fluorescence capability are important foundations for CQDs to achieve bioimaging (Figure 5). Therefore, HUVEC cell lines and RAW264.7 cell lines were used to investigate the cell imaging capability and biocompatibility of R-CQDs and L-CQDs. HUVEC and RAW264.7 cells treated with a load capacity of 1,000 μg/mL for 24 h all exhibited blue fluorescence under UV excitation (Figure 6A). The fluorescence intensity of HUVEC cells treated with R-CQDs was weaker, while that of cells treated with L-CQDs was stronger (Figure 6B), indicating that the intracellular imaging capability of L-CQDs is significantly better than that of R-CQDs without rare-earth doping. Cytotoxicity tests on HUVECs and RAW264.7 showed that, during 24, 72, and 144 h of co-culture with HUVEC and RAW264.7 cell lines, L-CQDs exhibited no toxicity at concentrations below 1,000 μg/mL, while demonstrating significant cytotoxicity at concentrations above 10,000 μg/mL, indicating that L-CQDs possess high safety within the concentration of 1,000 μg/mL (Figures 6C and 6D). Therefore, a dose of 1,000 μg/mL was selected for cell imaging.

Figure 5.

Schematic diagram of evaluation dimensions for the bioimaging application of L-CQDs

Figure 6.

Cell imaging of R-CQDs and L-CQDs in HUVECs and cytotoxicity of R-CQDs and L-CQDs in HUVECs and RAW264.7

(A) Confocal fluorescence images of HUVEC cells treated with R-CQDs and L-CQDs.

(B) Fluorescence quantitative imaging of HUVEC cells.

(C) Cytotoxicity curves of different concentrations of CQDS on HUVEC cells under different exposure times.

(D) Cytotoxicity curves of different concentrations of CQDS on RAW264.7 cells under different exposure times.

The length of the scale bar in A is 50 μm.

Data are represented as mean ± SEM.

^∗∗∗p < 0.001 vs. Blank group.

After confirming the good safety and effective imaging capability of L-CQDs for HUVEC cells and RAW264.7, C57BL/6J mice were selected as the target for bioimaging research. First, a toxicity experiment was conducted. In the 10-day toxicity experiment with a maximum dose of 2,000 mg/kg administered intraperitoneally (n = 6), no mice died. H&E staining of their organs (Figure 7A) showed no significant pathological changes. The evaluation of routine blood parameters (five classifications) and serum liver and kidney function indicators from the blood of mice exposed to the highest concentration in the toxicity experiment (Figures 7E and 7D) confirmed that L-CQDs at 2,000 mg/kg did not affect the hematopoietic system, liver, or kidney functions, demonstrating good biosafety. R-CQDs and L-CQDs were administered intraperitoneally at 2,000 mg/kg, and organs were collected 24 h later for frozen sectioning and fluorescence analysis (Figure 7B). Among the major organs, the brain showed a fluorescent response, indicating that the synthesized L-CQDs could effectively cross the blood-brain barrier (BBB). The distribution of L-CQDs in the brain exhibited zoning characteristics, suggesting tissue-specific uptake of L-CQDs in the brain. This is beneficial for the imaging application of carbon quantum dots in the nervous system. Furthermore, fluorescent responses were observed in the lungs, spleen, and kidneys, indicating that L-CQDs can enter organs rich in capillaries and frequent blood exchange, such as the lungs, spleen, and kidneys. Some studies indicate that carbon quantum dots are excreted through the kidneys.⁵⁵ Therefore, the fluorescence intensity in the kidneys can also prove that carbon quantum dots are excreted from the kidneys as early as 1 h after intake. No significant meaningful fluorescence was observed in the heart among the major organs, possibly because myocardial tissue has a fast metabolism and rapid blood circulation, resulting in a short residence time for L-CQDs there, preventing aggregation. The fluorescence intensity of L-CQDs was significantly stronger than that of R-CQDs, indicating that L-CQDs also have better fluorescence imaging capability in organ distribution within living organisms, possessing stronger application value (Figure 7C).

Figure 7.

Animal safety and imaging experiments of L-CQDs

(A) HE staining of organs in mice treated with L-CQDs.

(B) Organ fluorescence imaging of mice using L-CQDs.

(C) Fluorescence quantification of L-CQDs on mouse organs.

(D) Blood routine examination evaluation indicators of L-CQDs on mice.

(E) L-CQDs on renal and liver function evaluation indicators in mice. The length of the scale bar in A is 200 μm, in B is 1 mm

Data are represented as mean ± SEM. ^∗p < 0.05 vs. Sham group. ^∗∗p < 0.01 vs. Sham group. ^∗∗∗p < 0.001 vs. Sham group. ^#p < 0.05 vs. R-CQDs group. ^##p < 0.01 vs. R-CQDs group. ^###p < 0.001 vs. R-CQDs group.

Our systematic investigation established the optimal ratio of the CQDs, confirming their excellent biocompatibility and remarkable cortical tissue selectivity for bioimaging. These findings position the CQDs as a highly promising probe for cellular and tissue imaging.⁵⁶ Future work will focus on exploiting their specific targeting preferences to design self-selective labeling materials, thereby providing novel tools for applications ranging from cancer cell tracing,⁵⁷ metabolism studies,⁵⁸ to neurological function research.⁵⁹

In this work, highly efficient blue-light-emitting CQDs were synthesized using citric acid monohydrate, urea, glycerol, neodymium chloride hexahydrate, and cerium chloride as reaction precursors, and a mixed solution of glycerol and water as the reaction solvent via a one-step microwave-assisted solvothermal method. The results demonstrated that the prepared L-CQDs achieve a fluorescence quantum yield of 69%, exhibit upconversion luminescence, higher fluorescence activity in cell imaging, and good biosafety. L-CQDs have good enrichment ability in organs with rich blood flow, such as the brain, lungs, spleen, and kidneys, can cross the blood-brain barrier, and possess tissue-selective labeling capability in the brain. This research holds great application potential in the optimization of carbon quantum dot preparation processes and the field of bioimaging.

Limitations of the study

Although this study proposes a detailed hypothesis regarding the fluorescence and upconversion luminescence mechanisms of L-CQDs, the explanations provided still lack more direct experimental evidence to fully elucidate the underlying physicochemical processes. Furthermore, the study reveals that L-CQDs exhibit a compartmentalized distribution pattern in brain tissue, indicating tissue selectivity—a significant finding. However, the research does not delve into the molecular or cellular mechanisms responsible for this selective uptake. The specific surface properties or interactions that determine their enrichment in particular brain regions remain unclear.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Liang Jin (liangjin881219@foxmail.com).

Materials availability

CQDs generated in this study will be made available on request, but we may require a payment and/or a completed materials transfer agreement if there is potential for commercial application

Data and code availability

• •Data: All data are available in the article and in supplemental information and/or from the corresponding authors upon reasonable request.
• •Code: This article does not report original code.
• •Additional information: Any additional information required to reanalyze the data reported in this article is available from the lead contact upon request.

Acknowledgments

This research was supported by grants from the National Natural Science Foundation of China (82304761, 82374110), the Chinese funding program for postdoctoral researcher (GZC20231701), the Natural Science Foundation of Zhejiang Province (LQN25H280006, LY23H280005), the Zhejiang Province Traditional Chinese Medicine Science and Technology Project (2025ZR017), the Science and Technology Department of the State Administration of Traditional Chinese Medicine and the Zhejiang Provincial Administration of Traditional Chinese Medicine jointly build technology plan projects (GZY-ZJ-KJ-24065), the Zhejiang Xinmiao Talent Program (2024R410A038), the China Postdoctoral Science Foundation (2025M772300), and the National Natural Science Foundation of China Youth Science Fund Project (Category C, 52501313).

Author contributions

L.Z., L.J., H.W., and JY.S. designed and supervised this project; JY.S., Y.Z., CY.H., and J.R.L. performed most of the experiments; L.Z., L.J., H.W., J.Y.S., and S.W. analyzed the results; L.Z., L.J., H.W., J.Y.S., and Z.Y.B. contributed to the conception of the schematic representation of figures; L.J., H.W., J.Y.S., and Y.T.W. wrote the article with comments from all authors. All authors have read and approved the final article.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Chemicals, peptides, and recombinant proteins
Citric acid monohydrate	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#77-92-9;AR
L-ascorbic acid	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#50-81-7;99.99%
Urea	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#57-13-6;AR
Triethylamine	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#121-44-8;AR
Glycine	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#56-40-6;AR
Glycerol	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#56-81-5;AR
polyethylene glycol	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#25322-68-3;Mn=200
neodymium chloride hexahydrate	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#13477-89-9;4N
cerium chloride	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#7790-86-5;4N
europium chloride	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#13759-92-7;4N
ytterbium chloride	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#10035-01-5;4N
yttrium chloride	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#10025-94-2;4N
terbium chloride	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#13798-24-8;4N
potassium bromide	Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China)	CAS#7758-02-3;SP
Ethanol	Shanghai Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China)	CAS#64-17-5;99.8%
OCT embedding compound	SAKURA, USA	REF#4583
Glycerol	Macklin, Shanghai, China	REF#G810582
CCK-8 kit	Beyotime	REF#C0038
Dulbecco's Modified Eagle's Medium (DMEM)	C11885500BT	Gibco, Germany
10% fetal bovine serum	C04001	VivaCell, China
1% antibiotic-antimycotic	C100C8	NCM Biotech, Suzhou, China
Deionized water	Produced by the laboratory's ultrapure water system	The purity greater than 18.2 MΩ·cm-1
Experimental models: Cell lines
HUVEC	Zhejiang Badi Biotechnology Co., Ltd.	wt/wt
RAW264.7	Zhejiang Badi Biotechnology Co., Ltd.	wt/wt
Experimental models: Organisms/strains
Mouse	C57BL/6Jwt/wt	wt/wt

Experimental model and study participant details

Cell culture

HUVECs cells and RAW264.7 cells were purchased from Zhejiang Badi Biotechnology Co., Ltd. They were cultured in high-glucose Dulbecco's Modified Eagle's Medium (DMEM) (C11885500BT, Gibco, Germany). The medium was supplemented with 10% fetal bovine serum (FBS) (C04001, VivaCell, China) and 1% antibiotic-antimycotic (C100C8, NCM Biotech, Suzhou, China). HUVECs cells and RAW264.7 cells were cultured in an incubator (Thermo, USA) at 37°C with 5% CO₂.

Animal experiments

Forty 6-week-old female C57BL/6J mice (18-20 g) were purchased from Hangzhou Qizhen Experimental Animal Co., Ltd. All animal procedures and experimental protocols were approved by the Animal Ethics and Welfare Committee of Zhejiang Chinese Medical University (Approval No.: IACUC-20231218-09, Ethics Application No. 202312-1037) and followed the guidelines established by the NIH Guide for the Care and Use of Laboratory Animals. After a one-week acclimatization period under SPF laboratory conditions (23±2 °C, 50% relative humidity, 12 h light/dark cycle) with free access to food and water. All animal experiments were approved by the Zhejiang Chinese Medical University Animal Ethical and Welfare Committee (Approval No.: IACUC-20231218-09, Ethics Application No. 202312-1037).

Method details

Optimization of CQDs preparation process

To obtain rare-earth-doped carbon quantum dots with superior optical properties, reaction parameters were optimized based on the principles of the single-factor variable method. The fluorescence intensity at the optimal emission wavelength was taken as the evaluation parameter. The types of raw materials were selected from groups including glycerol and PEG-200 as liquid carbon sources/solvents; citric acid and L-ascorbic acid as solid carbon sources; urea, triethylamine, and glycine as nitrogen sources; and neodymium chloride, cerium chloride, europium chloride, ytterbium chloride, yttrium chloride, and terbium chloride as rare-earth doping elements. For the selected groups, the effects of reactant types, reactant molar ratios, reaction time, and volume percentage of organic solvent on the fluorescence properties were investigated. Matrix raw material screening was performed using 1,500 mg solid carbon source, 150 mg nitrogen source dissolved in 4 mL water, adding 10 mL liquid carbon source, thoroughly mixed, and reacted in a microwave oven at 700 W for 90 s.

Doping raw material screening was performed using 1,500 mg solid carbon source, 150 mg nitrogen source, 200 mg doped rare-earth element dissolved in 4 mL water, adding 10 mL liquid carbon source, thoroughly mixed, and reacted in a microwave oven at 700 W for 90 s.

With the reaction time fixed at 1 min and the volume ratio of water to glycerol in the reaction solvent set to 50%, the effect of N-element doping on the fluorescence properties of L-CQDs was investigated. When the mass of citric acid monohydrate was 1,500 mg, the mass ratios of citric acid monohydrate to urea were 10:0, 10:1, and 10:2. The appearance of the solution after reaction was observed and the fluorescence intensity was measured. The optimal combination was selected for subsequent optimization.

With the reaction time fixed at 1 min and the volume ratio of water to glycerol in the reaction solvent set to 50%, the effect of the rare-earth element doping ratio on the fluorescence properties of L-CQDs was investigated. When the mass of citric acid monohydrate was 1,500 mg, the reactant mass ratios were 30:3:10:3, 30:3:10:2, 30:3:10:1, 30:3:6:2, 30:3:2:2, and 30:3:2:1. The appearance of the solution after reaction was observed and the fluorescence intensity was measured. The optimal combination was selected for subsequent optimization.

The microwave reaction time with the volume was selected based on existing studies on microwave-assisted synthesis of carbon quantum dots, with a strategically designed solvent system to enhance reaction efficiency.ratio of water to glycerol in the reaction solvent set to 50% and the reactant mass ratio set to 30:3:10:2, the effect of reaction time on the fluorescence properties of L-CQDs was investigated. Reaction times were 40 s, 50 s, 60 s, and 70 s. The appearance of the solution after reaction was observed and the fluorescence intensity was measured. The optimal combination was selected for subsequent optimization.

With the reaction time fixed at 1 min and the reactant mass ratio set to 30:3:10:2, the effect of the volume ratio of water to glycerol in the reaction solvent on the fluorescence properties of L-CQDs was investigated. Since the raw materials could not dissolve at 0% volume ratio, this was not considered. Volume ratios were 25%, 50%, 75%, and 100%. The appearance of the solution after reaction was observed and the fluorescence intensity was measured. The optimal combination was selected for subsequent optimization.

Under the optimal reaction conditions, a thermocouple was fixed at the center of the solution to measure the temperature change during the reaction process, plot the graph, and determine the system temperature change to assist in inferring the reaction progress.

Preparation of L-CQDs

Using citric acid monohydrate, urea, glycerol, neodymium chloride hexahydrate, and cerium chloride as reaction precursors, and a mixed solution of glycerol and water as the reaction solvent, highly efficient blue-light-emitting CQDs were synthesized via a one-step microwave-assisted solvothermal method. The preparation process is as follows: First, accurately weigh an appropriate amount of citric acid monohydrate, urea, neodymium chloride hexahydrate, and anhydrous cerium chloride into a 50 mL beaker, add 4.00 mL distilled water, and stir until completely dissolved. Then add an appropriate amount of glycerol and stir until fully mixed. Place the beaker in the center of the microwave turntable and react at 700 W power. After removal, immediately add 10 mL distilled water to obtain the solution of neodymium and cerium co-doped carbon quantum dots (L-CQDs). For comparison of CQD structure, solid-state blue-light-emitting CDs without rare-earth doping elements, with only neodymium chloride hexahydrate, and with only cerium chloride were also prepared and named R-CQDs, N-CQDs, and C-CQDs, respectively. R-CQDs without rare-earth doping, N-CQDs with only Nd, and C-CQDs with only Ce were prepared by simply omitting the corresponding rare-earth raw materials from the reaction precursors, with other steps remaining unchanged. For long-term storage, the CQD solution can be lyophilized or precipitated with anhydrous ethanol and then dried. The obtained powder is ground in an agate mortar and stored at low temperature under sealed conditions for subsequent characterization and performance testing.

Characterization

High-resolution transmission electron microscopy (HRTEM, JEM-2010F) was used to observe the nanoscale morphology, crystal plane characteristics, and diffraction patterns of the CQDs. L-CQDs were dispersed in anhydrous ethanol at a concentration of 20 mg/mL, treated with a 130 W ultrasonic cell disruptor for 2 min, immediately dropped onto an ultrathin carbon film, dried, and then characterized for morphology and crystal plane features using TEM at an accelerating voltage of 200 kV. Diffraction patterns were captured in areas where CQDs aggregated.

Fourier transform infrared spectroscopy (FT-IR, Thermo IS50) was used to characterize the surface groups of the CQDs. Using potassium bromide (KBr) as the background, L-CQDs were mixed with KBr at a ratio of 1:100, and the FT-IR spectrum was measured using the pellet method.

Steady-state fluorescence spectrometer (Edinburgh FLS920) and upconversion fluorescence spectrometer (HORIBA FluoroMax-4) were used to characterize the fluorescence properties of the CQDs. Fluorescence spectra were measured using a quadrangle quartz cuvette. Quantum yield (QY) was measured using the integrating sphere method, ensuring the absorbance A of the sample at the detection wavelength was <0.2.

X-ray powder diffractometer (Shimadzu XRD6100) was used to characterize the crystallinity and crystal plane types of the CQDs. The X-ray diffractometer used a copper target with an emission wavelength of 1.5406 Å and a scanning rate of 5°/min.

UV-Vis-IR absorption spectrometer (Shimadzu UV2550) was used to characterize the absorption spectra of the CQDs.

Laser Raman spectrometer (Renishaw inVia) was used to analyze the vibrational modes of the CQDs. Spectra were acquired using powder mode with 532 nm excitation.

Cell counting kit-8 (CCK-8) assay

Cell viability and safe drug concentration for HUVECs and RAW264.7s were assessed using a CCK-8 kit (C0038, Beyotime) according to the manufacturer's instructions. HUVECs cells and RAW264.7s cells were seeded in a 96-well plate at a density of 1×10⁵ cells/mL and cultured in an incubator at 37°C and 5% CO₂. When cell confluence reached 90%, carbon quantum point solutions in serum-free high-glucose DMEM at concentrations ranging from 1 to 10,000 mcg/mL were added. High-glucose DMEM medium without carbon quantum dots was used as the control. After 24, 72, 144 hours of incubation, the medium was discarded, and the cells were washed three times with PBS. Then, 20 μL of CCK-8 reagent was added to each well containing 200 μL medium and co-incubated for 2 hours. The absorbance at 450 nm wavelength was measured using an EnSpire multimode plate reader (PerkinElmer, USA). The detected absorbance values were normalized.

In vitro fluorescence imaging

Cells were cultured in confocal dishes as described above. When cell density reached approximately 75%, the medium was discarded, and the cells were washed three times with PBS. Then, 1,000 μg/mL carbon quantum dot solution in serum-free high-glucose DMEM medium was added and co-incubated for 24 hours. The cells were washed three times with PBS, and 1 mL PBS solution was added for the last wash. Fluorescence images were then observed and recorded under a laser scanning confocal microscope (ZEISS, Germany).

Animal experiments

The mice were randomly divided into 4 groups (n=7 per group) for the toxicity test: control group (0.9% sodium chloride solution), low-dose group (L-CQDs, 500 mg/kg), medium-dose group (L-CQDs, 1,000 mg/kg), and high-dose group (L-CQDs, 2,000 mg/kg), administered via a single intraperitoneal injection. The selected L-CQDs doses were based on toxicological classification, with doses above 2,000 mg/kg corresponding to slightly toxic or non-toxic. Higher doses of carbon quantum dots could not dissolve and thus could not be administered. The acute toxicity test period was 10 days. Mortality was observed, and at the end, all mice were euthanized using tribromoethanol, and whole blood, serum, and organs were collected.

The highest concentration group with no mortality was selected, and mice were randomly divided into 3 groups (n=4 per group) for the imaging experiment: control group, positive drug group (R-CQDs, 2,000 mg/kg), and experimental group (L-CQDs, 2,000 mg/kg), administered via a single intraperitoneal injection. After waiting for 1 hour, all mice were euthanized using tribromoethanol, and organs were collected.

Hematoxylin and eosin (H&E) staining

Representative major organs (brain, heart, lungs, liver, spleen, kidneys) were fixed in 4% paraformaldehyde (PFA) at room temperature for 24 hours, subsequently dehydrated through a graded ethanol series, cleared in xylene, and embedded in paraffin blocks. The paraffin blocks were sectioned into 4 μm thick slices using a rotary microtome, mounted on glass slides, and dried overnight. After deparaffinization, the sections were stained with hematoxylin and eosin staining solutions. Histopathological evaluation was performed using an optical microscope, and digital images of all samples were systematically recorded.

In vivo fluorescence imaging

Organs, as described above, were dehydrated in 10% PBS sucrose solution in a 4°C refrigerator until they sank, embedded in OCT embedding compound (SAKURA, USA) using liquid nitrogen for rapid freezing, and sectioned into 4 μm thick slices using a cryostat (PEREDIA NX50, USA). The slices were mounted on glass slides and dried in a 40°C oven for 1 hour. Slides were mounted using glycerol (G810582, Macklin, Shanghai, China). Images of each slice were captured using a fluorescence inverted microscope (ZEISS, Germany) and subjected to statistical analysis.

Quantification and statistical analysis

All experiments were repeated at least three times independently. Data are expressed as mean ± standard error of the mean (Mean ± SEM) and analyzed using SPSS 27.0. T-test was used for comparisons between two groups, and one-way analysis of variance (One-way ANOVA, Tukey method as post test) was used for comparisons among multiple groups. All results are presented as mean ± SEM, and statistical significance was determined when ∗P < 0.05 vs. Sham group. ^∗∗P < 0.01 vs. Sham group. ^∗∗∗P < 0.001 vs. Blank/Sham group. ^#P < 0.05 vs. R-CQDs group. ^##P < 0.01 vs. R-CQDs group. ^###P < 0.001 vs. R-CQDs group.

Published: February 2, 2026