Broad-range flow velocimetry enabled by pulse-width-dependent luminescence of core–multishell upconversion nanoprobes

Abstract

Lanthanide-doped upconversion nanoparticles (UCNPs) have garnered extensive attention in fundamental research and cutting-edge applications due to their unique optical properties. Particularly in sensing, UCNP-based fluorescent probes provide a versatile platform for microfluidic flow velocity calibration. However, designing nanoprobes with efficient luminescence modulation for broad-range flow velocimetry remains challenging. Herein, we engineered a core–multishell UCNP probe: NaGdF₄:Tm³⁺/Yb³⁺@NaGdF₄@NaGdF₄:Eu³⁺@NaYF₄, in which spatially isolated Tm³⁺ (blue) and Eu³⁺ (red) activators enable dual emissions. The intensity ratio between these channels exhibits a laser pulse-width-dependent behavior, enabling real-time dynamic optical modulation. Leveraging this mechanism, we showed fluid velocity assessment by dispersing nanoprobes in a fluid stream under fixed laser excitation. The flowing medium underwent flow-velocity-dependent effective excitation pulse width variations, establishing a quantitative emission ratio–velocity mapping for precise calibration. This paradigm advances flow velocimetry technology while significantly broadening the measurable velocity range via energy migration-mediated kinetics. This sensing paradigm not only advances fluid velocimetry techniques but also expands the multifunctional utility of core–multishell UCNPs in emerging photonic technologies.

---

1. Introduction

Over the past three decades, nanoscience has witnessed remarkable advancements, with lanthanide-doped upconversion nanoparticles (UCNPs)—capable of efficiently converting infrared radiation to visible light—garnering extensive attention.^1–4 Their unique merits, including exceptional photostability and chemical robustness,⁵ prolonged luminescence lifetimes,⁶ narrow-band emissions,⁷ and low cytotoxicity,⁸ have enabled broad applications in displays,⁷ anti-counterfeiting,⁹ solid-state lasers,¹⁰ and chemosensors.^11–16 Notably, an emerging velocimetry technique based on upconversion luminescence has recently been developed,^17–19 which assesses flow velocity by leveraging the emission profiles of fluoride nanoparticle probes dispersed in the fluid, thereby eliminating the need for expensive high-speed cameras or rapid dual-pulse lasers required by conventional methods such as particle image velocimetry (PIV) and particle tracking velocimetry (PTV).^20,21 For example, NaGdF₄:Yb³⁺/Tm³⁺@NaGdF₄:Tb³⁺ (ref. 17) and NaYF₄:Yb³⁺/Ho³⁺/Ce³⁺@NaGdF₄ ¹⁹ utilize the emissions of blue and green light, and red and green light, respectively, as output signals for fluid velocity measurements. This technique offers significant advantages such as visualization, short responsive time, and high anti-interference ability, showing potential for frontier applications in automotive monitoring, environmental surveillance, and medical diagnostics (e.g., blood flow velocimetry). However, existing probes exhibit a limited measurable velocity range (<50 cm s⁻¹).^17–19 Moreover, the lifetime-based strategy employed by Capobianco's group requires repeated measurements of distance variations between different emission maxima using bandpass filters and high-speed cameras for different flow rates, making it unsuitable for multi-velocity calibration.¹⁷ The Ce/Ho cross-relaxation approach may cause reduced luminescence intensity, which is unfavorable for detection.^18,19 Additionally, Ho³⁺ emission arises from a two-photon process with a fast rise edge, resulting in a narrow non-steady-state time window available for measurement. Achieving efficient dynamic luminescence modulation to broaden this range poses a critical challenge in nanostructural design.

Lanthanide ions possess rich step-like energy level structures, enabling emissions across the entire visible spectrum. However, luminescence tuning through modulation of intrinsic factors (e.g., the type of doped rare-earth ion or crystal phase)^22,23 is often time-consuming and resource-intensive, requiring material re-synthesis. Furthermore, co-doping multiple lanthanides risks concentration quenching. Consequently, combining intrinsic control with extrinsic stimuli (e.g., temperature,²⁴ pressure²⁵ and excitation conditions^26–28) is essential for real-time dynamic modulation. Recent advances in core–multishell nanostructures offer a viable solution: multishell coating prevents energy transfer from core activators to surface defects, enhancing emission intensity and extending luminescence lifetimes; spatial confinement of lanthanides ions within distinct shells mitigates concentration quenching and cross-relaxation processes between different dopants.^29–33 Notably, studies have shown that upconversion emissive-state dynamics can be controlled through the modulation of excitation parameters,^34–36 thereby enabling the design of specific energy-transfer pathways in core–shell architectures to achieve dynamic switching of emission colors through pulse-width modulation. Furthermore, the ability of multishell structures to prolong non-steady-state upconversion processes shows promise for expanding the measurement range in velocity-sensing applications.

Here, we fabricated a core–multishell nanoparticle, NaGdF₄:Yb³⁺/Tm³⁺@NaGdF₄@NaGdF₄:Eu³⁺@NaYF₄, where two distinct activators—Tm³⁺ and Eu³⁺—are spatially segregated as dual luminescent centers to achieve blue and red emission bands. NaGdF₄ and NaYF₄ act as inert shells to enhance the luminescence intensity. Upon absorbing pump photons, the sensitizer Yb³⁺ excites adjacent Tm³⁺ ions to high-lying excited states. Subsequently, Gd³⁺ ions harvest energy from the excited Tm³⁺ and mediate its migration through the sublattice, ultimately transferring it to Eu³⁺ activators to generate their characteristic upconversion luminescence.³⁷ Furthermore, by optimizing the Tm³⁺ doping concentration, the red emission of Eu³⁺ (⁵D₀ → ⁷F₂) at 615 nm can be enhanced through the Tm–Gd–Eu energy transfer pathway.^33,38 Crucially, due to the distinct time periods for Tm³⁺ and Eu³⁺ emissions to reach their steady state, red emission from Eu³⁺ (⁵D₀ → ⁷F₂) dominates under 980 nm continuous-wave (C.W.) laser excitation. As the laser pulse width decreases, the emission color shifts from red to blue, enabling dynamic luminescence regulation. Leveraging this mechanism, we showed fluid velocity calibration by dispersing these probes in fluid and correlating the emission intensity ratio (I_Red/I_Blue) with flow velocity. The multishell's long-range energy migration achieves a velocity range of 0–118 cm s⁻¹—a 2.4-fold improvement over prior probes^17–19 (<50 cm s⁻¹). This work not only validates core–multishell nanoparticles for flow sensing but achieves exceptional range extension in velocimetry, advancing practical applications of this technology.

2. Results and discussion

We synthesized NaGdF₄:Yb³⁺/Tm³⁺, NaGdF₄:Yb³⁺/Tm³⁺@NaGdF₄, NaGdF₄: Yb³⁺/Tm³⁺@NaGdF₄@NaGdF₄:Eu³⁺, and NaGdF₄:Yb³⁺/Tm³⁺@NaGdF₄@NaGdF₄:Eu³⁺@NaYF₄ nanoparticles by the co-precipitation method,²⁹ which are denoted as C, CS, CSS, and CSSS, respectively. In these structures, the doping concentrations were fixed at 1.5 mol% for Tm³⁺, 30 mol% for Yb³⁺, and 15 mol% for Eu³⁺. As shown in Fig. 1(a), all the products were in the pure hexagonal phase β-NaGdF₄ (JCPDS no. 27-0699), and the diffraction peaks tended to become narrower as the particle size increased. The transmission electron microscopy (TEM) image in Fig. 1(b) shows that the CSSS nanocrystals were well-dispersed without agglomeration and had a relatively uniform size. Scanning electron microscopy (SEM) micrographs (Fig. S1) provided evidence that the core nanoparticles were monodispersed with the mean sizes of ∼17 nm and the particles sizes increased to 25 nm, 38 nm, and 47 nm (length, along the c axis) for the CS, CSS and CSSS nanoparticles, respectively. The high-resolution TEM image and the internal Fourier transform pattern in Fig. 1(c) confirm its single-crystal nature and high crystallinity. The lattice fringes were clearly resolved with a lattice spacing of 0.52 nm, which could be assigned to the (100) plane of the hexagonal NaYF₄ phase. The high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) image in Fig. 1(d) shows the structure of the sample. Due to the significant difference in atomic numbers between Gd (Z = 64) and Y (Z = 39), their nuclei and outer shells form a distinct bright-dark contrast in imaging. The energy-dispersive X-ray elemental mapping in Fig. 1(e–k) clearly shows the elemental distribution of Na, F, Yb, Tm, Gd, Eu, and Y in the CSSS particles.

Fig. 1 (a) XRD patterns of the C, CS, CSS and CSSS samples. Bars represent the standard diffraction data of the hexagonal NaGdF₄ crystal (JPCDS no. 27-0699), (b) TEM micrograph of the NaGdF₄:Yb³⁺/Tm³⁺@NaGdF₄@NaGdF₄:Eu³⁺@NaYF₄ CSSS NCs, and (c) side view of an individual CSSS NC. Insets show the corresponding fast Fourier transform (FFT) patterns, and (d) STEM image and EDX mappings of (e) Na, (f) F, (g) Yb, (h) Tm, (i) Gd, (j) Eu, and (k) Y elements.

Subsequently, the up-conversion emission characteristics after the layer-by-layer growth of different shells were investigated. As shown in Fig. 2(a), for the NaGdF₄:Yb³⁺/Tm³⁺ (C) core, under the steady-state excitation of a 980 nm laser, the typical emissions of Tm³⁺: ¹D₂ → ³F₄ (450 nm), ¹G₄ → ³H₆ (475 nm), ¹G₄ → ³F₄ (647 nm), and ³F_2,3 → ³H₆ (695 nm) were clearly observed. After the growth of the NaGdF₄ shell, NaGdF₄:Yb³⁺/Tm³⁺@NaGdF₄ (CS) was obtained. The positions of its emission peaks were basically the same as those of C, but the luminescence intensity was significantly enhanced. After the growth of the second NaGdF₄:Eu³⁺ shell, NaGdF₄:Yb³⁺/Tm³⁺@NaGdF₄@NaGdF₄:Eu³⁺ (CSS) nanocrystals were obtained, and up-conversion emission bands belonging to the Eu³⁺ transitions of ⁵D₀ → ⁷F₁ (591 nm), ⁵D₀ → ⁷F₂ (615 nm), and ⁵D₀ → ⁷F₃ (690 nm) were detected. After further coating with the third NaYF₄ layer, the luminescence intensity was also significantly enhanced. Comparing from C to CSSS, the overall up-conversion emission intensities of Tm³⁺ and Eu³⁺ after the growth of the corresponding shells were significantly enhanced. Specifically, the up-conversion luminescence of Tm³⁺ in CS was approximately 100 times that in C; the up-conversion luminescence of Eu³⁺ in CSSS was approximately 2.5 times that in CSS, which was attributed to the surface passivation effect of the shells.²⁸ For upconversion materials, a longer decay lifetime generally indicates a higher upconversion emission efficiency.^29,30 As shown in Fig. 2(b and c), the variation trend of the lifetime was consistent with that of the upconversion emission: the upconversion decay lifetimes of the emission states of Tm³⁺:¹D₂ and Eu³⁺:⁵D₀ gradually increase with the growth of the shell layer. Specifically, the transition lifetime of Tm³⁺ at its ¹D₂ state (450 nm) increases from 56 μs to 292 μs, and the transition lifetime of Eu³⁺ at its ⁵D₀ state (615 nm) increases from 4895 μs to 5780 μs. The microsecond-scale lifetime of the Tm^{3+ 1}D₂ level and the millisecond-scale lifetime of the Eu^{3+ 5}D₀ level were both consistent with the values reported in the literature.^38,39 Moreover, a noticeable increase in the population of the Tm^{3+ 1}D₂ energy level could be observed (Fig. 2b); this was likely because the shell layer could passivate the surface defects of the nanocrystals, protecting the luminescent lanthanide ions (especially those near the surface) from the non-radiative processes caused by surface defects.^37,40,41Fig. 2(d) shows the luminescence photographs of the samples from C to CSSS under the irradiation of a continuous-wave (C.W.) 980 nm laser. It could be clearly seen that the blue light emitted by the sample coated with an inert shell layer was significantly brighter than that of the sample without the shell layer. The emission of red light showed a similar situation. Considering that the concentration of the probes dispersed in the fluid may be diluted, the improvement of the luminescence intensity was beneficial for the successful capture of the luminescence spectra of the probes.

Fig. 2 (a) Upconversion emission spectra of the core@shell samples under 980 nm laser irradiation (from bottom to top: C, CS, CSS and CSSS), (b) UC decay curves by monitoring (b) Tm³⁺:¹D₂ → ³F₄ transition (450 nm emission) and (c) Eu³⁺:⁵D₀ → ⁷F₂ transition (615 nm emission) for the C, CS, CSS, and CSSS samples, (d) upconversion luminescent photographs for the samples dispersed in hexane solutions (from left to right: C, CS, CSS and CSSS), and (e) schematic core–multishell nanoarchitectures of CSSS, in the structures; Yb³⁺ was doped in the NaGdF₄ core to enable the excitation of a 980 nm laser, Tm³⁺ and Eu³⁺ were doped in different locations (the NaGdF₄ core and the second NaGdF₄ shell, respectively) to spatially separate them to induce intense Tm³⁺ and Eu³⁺ upconversion emissions, and the third NaYF₄ shell was used to reduce the surface quenching effect. Notably, the NaGdF₄ host was adopted in the second and third shells to enable energy migration upconversion from Tm³⁺ to Eu³⁺ with the assistance of Gd ions; ET represents the energy transfer in the individual core or shell and IET represents the interface energy transfer from one shell to the next one.

Fig. 2(e) shows the schematic diagram of the specific energy transfer mechanisms, where ET (energy transfer) denotes energy transfer within a single core or shell, and IET (interface energy transfer) represents interfacial energy transfer from one shell to the next. The sensitizing ions Yb³⁺ absorb excitation energy and subsequently transfer it to Tm³⁺. Notably, the NaGdF₄ matrix was employed in both the core and the first shell to facilitate energy migration from Tm³⁺ to Eu³⁺ mediated by Gd³⁺. To induce intense upconversion emission from Eu³⁺, spatial segregation of Tm³⁺ and Eu³⁺ was essential to prevent luminescence co-quenching; consequently, Tm³⁺ and Eu³⁺ were selectively doped into the core and the second shell, respectively. Detailed energy-level transition diagrams are provided in Fig. S2. Furthermore, to ensure significant blue–red emission variation in the samples, we synthesized a series of core–multishell structures – NaGdF₄:Yb³⁺/Tm³⁺ (30/x mol%)@NaGdF₄@NaGdF₄:Eu³⁺ (15 mol%)@NaYF₄ with different Tm³⁺ doping ratios (x = 0.5, 1, 1.5). As shown in Fig. S3 and S4, as the Tm³⁺ doping concentration increased, the red-to-blue emission intensity ratio increased correspondingly, exhibiting the most pronounced emission contrast at 1.5 mol%. Consequently, the NaGdF₄:Yb³⁺/Tm³⁺ (30/1.5 mol%)@NaGdF₄@NaGdF₄:Eu³⁺ (15 mol%)@NaYF₄ architecture was selected as the optimal system for detailed investigation. The enhanced luminescence intensity and distinct color difference collectively ensured reliable spectral acquisition and accurate differentiation during dynamic monitoring, even considering the risk of probe concentration dilution in fluids.

Subsequently, the non-steady-state upconversion spectra were measured and compared with the steady-state spectra. Fig. 3(a) shows a simple schematic diagram of short and long laser pulses. As can be seen from Fig. 3(b) and Table S1, under steady-state excitation, the red emission of Eu³⁺ at 615 nm was stronger than the blue emission of Tm³⁺ at 475 nm. As the pulse width decreased, the intensities of both the red and blue emissions weakened. However, the change amplitude of the red emission was greater than that of the blue emission. As a result, when the pulse width was 0.5 ms, the emission intensities of the two were basically equal. When the pulse width continued to decrease, the intensity of the blue emission was already higher than that of the red emission. A more pronounced contrast in the emission intensity could be observed in the normalized upconversion luminescence spectra in Fig. S5. To more intuitively understand the changes in the intensities of red and blue emissions, the ratio of red-to-blue emission intensities was calculated. As shown in Fig. 3(c), the ratio of red-to-blue emission intensities gradually decreased as the pulse width decreased. To explain this phenomenon, the time-resolved curves of the emission intensities were measured. As shown in Fig. 3(d), the energy levels ¹G₄ of Tm³⁺ and ⁵D₀ of Eu³⁺ did not reach the steady state simultaneously. Although the emission of Tm³⁺ was a multi-photon process and required a longer excitation time, the emission of Eu³⁺ needed an energy transfer process; the energy was transferred from Gd³⁺ in the core to the ⁶P_7/2 state of Gd³⁺ in the first-layer shell, then to Gd³⁺ in the second – layer shell, and finally to Eu³⁺, as shown in Fig. 3(e). This results in a longer time required for it to reach the steady state. When the pulse becomes shorter, the excitation energy decreases, and the population numbers reaching the excited states of Tm³⁺ and Eu³⁺ both decrease. However, the decrease in Eu³⁺ was more significant. Therefore, the blue emission gradually dominates. This non-steady-state characteristic indicates that the upconversion luminescence of the particles can be dynamically regulated by adjusting the pulse width.

Fig. 3 (a) Schematic diagram of long and short 980 nm laser pulses, (b) the upconversion spectra taken upon pulse excitation (pulse width varied from 0.1 to 5 ms), (c) dependence of the red-to-blue fluorescence intensity ratio (FIR) on laser pulse width, (d) time-dependent Eu³⁺ emission at 615 nm and Tm³⁺ emission at 475 nm profiles upon pulsed 980 nm laser excitation (20 ms pulse width), and (e) schematic representation detailing the non-steady-state upconversion mechanisms, under short-pulsed excitation and the blue emission at 475 nm from Tm³⁺; by comparison, it was seen that long-pulsed excitation promotes the population of the ⁶P_7/2 state of Gd³⁺ through a four-step energy transfer process (Yb → Tm → Gd → Gd → Eu).

To apply this non-steady-state luminescence characteristic to fluid velocity calibration, it is hypothesized that when a fluid containing nanoparticle probes flows through a 980 nm steady-state laser spot, the nanocrystals will undergo a pulsed excitation. As schematically illustrated in Fig. S6, under the condition where the excitation light (with a spot diameter d) remains stationary while the fluid is in motion, the sample region illuminated by the excitation beam can be regarded as an emission unit. Each unit is excited only when it passes through the excitation spot area, experiencing a pulse width τ = d/v, where d is the laser spot diameter and v is the velocity of the upconverting sample. As the flow velocity increases, the pulse width experienced by each unit decreases accordingly, leading to regular changes in the luminescence characteristics of the fluid with varying flow rates. Based on this principle, we constructed an upconversion luminescence velocimetry platform. A specific schematic diagram is shown in Fig. 4(a). The synthesized nanoprobes were dispersed in octadecene to form a colloidal solution. A peristaltic pump was used to make the liquid flow in a thin tube. The thin tube was fixed in the sample chamber of a spectrometer to ensure the smooth flow of the fluid and that the 980 nm steady-state laser spot falls on the thin tube for convenient spectral collection. Finally, the relationship between the flow velocity and the luminescence intensity ratio was calculated for fluid velocity calibration. Fig. 4(b) shows a physical image for observing the color change of the fluid in the tube. As shown in Fig. 4(c), the up-conversion emission color of the fluid in the thin tube intuitively changes from red to blue as the flow velocity increases. From the up-conversion luminescence spectrum in Fig. 4(d), Fig. S7, and Table S2, it can also be seen that the FIR (I_R/I_B) value decreases as the velocity increases. Fig. 4(e) shows the corresponding chromaticity coordinate change diagram. As the flow velocity increases, the up-conversion luminescence color of the fluid moves from the red region to the blue region. It should be noted that this material belongs to high-photon up-conversion samples, and the excitation distance has a significant impact on the luminescence phenomenon. In the experiment, the thin tube was fixed to ensure that the distance between the laser and the thin tube remains constant.

Fig. 4 (a) Schematic diagram of the experimental setup for flow velocity sensing, (b) the photo of this optical detection setup, (c) photos of the fluid in the tube with various flow rates, excited with a 980 nm laser, (d) the captured upconversion spectra on the 1.5% Tm³⁺ doped sample flowing at different rates, (e) the corresponding emission colors were indexed on the chromaticity diagram (CIE 1931), (f) the plots of FIR (I_R/I_B) versus the flow rate, accompanied by the polynomial fitting lines as the calibration curves, and (g) the calculated values of S_a and S_r for this sample.

By fitting the curve of FIR versus flow velocity with an exponential expression, we obtained the velocity calibration curve. As shown in Fig. 4(f), the measured data matched well with the exponential model, a conclusion that was further supported by the corresponding error analysis in Table S3. Therefore, the fitted line could be used as the calibration curve, and the following formulas could be further used to calculate the absolute sensitivity S_a and the relative sensitivity S_r:

Sensitivity increases with the flow rate. As shown in Fig. 4(g), the maximum S_a value was approximately 0.023 s cm⁻¹, and the maximum S_r value reached about 1.1%. It can be seen that it still had good sensitivity for high-speed flowing samples. Furthermore, 10 cycles of accelerating–decelerating tests conducted within the velocity range of 0–118 cm s⁻¹ showed small variations in the FIR values, confirming the stability of the velocimetry method (Fig. S8). This study primarily validated the relevant physical mechanism in an octadecene model system. Furthermore, our nanoprobes could be dispersed in various solvent environments through surface engineering according to application requirements, such as in aqueous or biologically relevant media (e.g., blood-mimicking solutions). Existing research has shown that such probes can achieve aqueous dispersion through surface modification.¹⁸ Therefore, our core–multishell nanoparticles provide a new option for probe materials in fluid velocity calibration.

3. Conclusions

In conclusion, in this work, a core–multishell nanoparticle, NaGdF₄:Yb³⁺/Tm³⁺@NaGdF₄@NaGdF₄:Eu³⁺@NaYF₄, which can dynamically regulate upconversion luminescence using laser pulse width, was prepared by the co-precipitation method. As a novel nanoprobe for fluid velocity calibration, the system achieves an extended measurable velocity range of 0–118 cm s⁻¹—representing a near 136% improvement over state-of-the-art probes (<50 cm s⁻¹)—while maintaining a rapid response, high spatial resolution, and remote detection capability, showing application potential in scenarios such as vehicle monitoring, mechanical control, and blood flow detection. The findings of this work establish core–multishell nanostructures as a viable choice for advanced flow sensing and highlight the potential of rare-earth-doped fluorides for future velocimetry systems.

4. Experimental section

All nanoparticles were synthesized via the co-precipitation method, and the specific procedures are detailed in the SI.

4.1. Characterization

The X-ray diffraction (XRD) patterns of upconversion nanoparticles were recorded using a powder diffractometer (Smart Lab 3KW) equipped with Cu-Kα radiation (λ = 0.154 nm). The micromorphology of all samples was investigated by scanning electron microscopy (SEM, J ZEISS Sigma 500). Furthermore, the size and morphological characteristics of the CSSS samples were observed through transmission electron microscopy (TEM, FEI Talos F200s). High-angle annual dark-field scanning transmission electron microscopy (HAADF-STEM) was carried out on an FEI aberration-corrected Titan Cubed S-Twin transmission electron microscope equipped with an energy dispersive X-ray spectroscope (EDS) operated at 200 kV. The TEM sample preparation was accomplished by dispersing the CSSS sample in cyclohexane, followed by subsequent dilution and drop-casting onto a carbon-coated copper grid. The steady state and non-steady-state upconversion spectra as well as the temporal response curves were obtained from a FLS1000 spectrofluorometer equipped with a 980 nm laser pump modulated with an impulse generator (square wave was generated with the pulse width ranging from 0.1–5 ms). The experiments employed a 980 nm laser (2 mm diameter circular spot) to excite the upconversion nanoparticles, which were dispersed in octadecene at a concentration of 0.25 mmol mL⁻¹. With the excitation power set to 0.6 W, the corresponding power density was approximately 19.10 W cm⁻². A micro-peristaltic pump (Runze LM60B) was employed to induce dynamic variations in fluid flow velocity through precise flow rate modulation to obtain dynamic spectral detection.

Conflicts of interest

There are no conflicts to declare.

Data availability

All relevant data are within the paper.

Supplementary information (SI) is available. Synthesis and SEM of core-multishell; The energy level transition diagram; Red-to-blue emission intensity ratio; Variation in emission and chromaticity coordinates; Upconversion emission spectrum; Schematic showing the mechanism of the velocity sensitive upconversion luminescence; The captured upconversion emission spectrum; The FIR during 10 cycles; Color coordinates; Calibration of flow velocity and measurement error. See DOI: https://doi.org/10.1039/d5nr03368c.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (12304456), the Youth Project of Natural Science Foundation of Xiamen City, the Fujian Province (3502Z202372042), the National Natural Science Foundation of China (52272141), the Natural Science Foundation of Fujian Province (2024J02014), the Industry-University-Research Collaboration Project of Fujian Provincial Universities (2024H6021), and the Major Project of Science and Technology of Xiamen City (3502Z20241024).

References

1. F. Auzel, Chem. Rev., 2004, 104, 139–174.
2. A. P. Alivisatos, Science, 1996, 127, 933–937.
3. A. R. Armstrong and P. G. Bruce, Nature, 1996, 381, 466–500.
4. W. C. W. Chan and S. Nie, Science, 1998, 281, 2016–2018.
5. B. Zhou, B. Shi, D. Jin and X. Liu, Nat. Nanotechnol., 2015, 10, 924–936.
6. T. Sun, R. Ma, X. Qiao, X. Fan and F. Wang, ChemPhysChem, 2016, 17, 766–770.
7. R. Deng, F. Qin, R. Chen, W. Huang, M. Hong and X. Liu, Nat. Nanotechnol., 2015, 10, 237–242.
8. F. Liu, X. He, Z. Lei, L. Liu, J. Zhang, H. You, H. Zhang and Z. Wang, Adv. Healthcare Mater., 2015, 4, 559–568.
9. A. Baride, J. M. Meruga, C. Douma, D. Langerman, G. Crawford, J. J. Kellar, W. M. Cross and P. S. May, RSC Adv., 2015, 5, 101338–101346.
10. X. Chen, L. Jin, W. Kong, T. Sun, W. Zhang, X. Liu, J. Fan, S. F. Yu and F. Wang, Nat. Commun., 2016, 7, 10304.
11. K. C. Liu, Z. Y. Zhang, C. X. Shan, Z. Q. Feng, J. S. Li, C. L. Song, Y. N. Bao, X. H. Qi and B. Dong, Light: Sci. Appl., 2016, 5, e16136.
12. L. N. Sun, H. Peng, M. I. Stich, D. Achatz and O. S. Wolfbeis, Chem. Commun., 2009, 5000–5002.
13. S. Liu, Y. Li, C. Zhang, L. Yang, T. Zhao, R. Zhang and C. Jiang, J. Colloid Interface Sci., 2017, 493, 10–16.
14. N. Wang, X. Yu, K. Zhang, C. A. Mirkin and J. Li, J. Am. Chem. Soc., 2017, 139, 12354–12357.
15. R. Liu, Y. Cheng, H. Lin, J. Xu and Y. S. Wang, J. Colloid Interface Sci., 2025, 689, 137243.
16. Q. Liu, J. Peng, L. Sun and F. Li, ACS Nano, 2011, 5, 8040–8048.
17. G. Tessitore, S. L. Maurizio, T. Sabri, C. D. Skinner and J. A. Capobianco, Adv. Mater., 2020, 32, e2002266.
18. L. Zhang, S. Hu, X. Fu, N. Zhang, F. Huang, T. Pang, H. Huang, A. Xie and D. Chen, J. Mater. Chem. C, 2025, 13, 10602–10610.
19. H. Huang, S. Wang, R. Chen, N. Zhang, H.-R. Yao, Y. Zheng, F. Huang and D. Chen, Nano Res., 2022, 16, 1212–1219.
20. K. D. Hinsch, Meas. Sci. Technol., 1995, 6, 742–753.
21. N. Chen, C. Wang and W. Heidrich, Laser Photonics Rev., 2021, 15, 2100008.
22. G. Chen, H. Qiu, P. N. Prasad and X. Chen, Chem. Rev., 2014, 114, 5161–5214.
23. Y. Wang, K. Zheng, S. Song, D. Fan, H. Zhang and X. Liu, Chem. Soc. Rev., 2018, 47, 6473–6485.
24. A. Sedlmeier, D. E. Achatz, L. H. Fischer, H. H. Gorris and O. S. Wolfbeis, Nanoscale, 2012, 4, 7090–7096.
25. C. Renero-Lecuna, R. Martín-Rodríguez, R. Valiente, J. González, F. Rodríguez, K. W. Krämer and H. U. Güdel, Chem. Mater., 2011, 23, 3442–3448.
26. L. Yan, J. Huang, Z. An, Q. Zhang and B. Zhou, Nat. Commun., 2024, 15, 1923.
27. J. Huang, L. Yan, Z. An, H. Wei, C. Wang, Q. Zhang and B. Zhou, Adv. Mater., 2024, 36, e2310524.
28. J. Huang, L. Tao, H. Wei, H. Huang, Q. Zhang and B. Zhou, Nat. Commun., 2025, 16, 2378.
29. Z. An, Q. Li, J. Huang, L. Tao and B. Zhou, Nano Lett., 2023, 23, 6241–6248.
30. S. Liu, L. Yan, J. Huang, Q. Zhang and B. Zhou, Chem. Soc. Rev., 2022, 51, 1729–1765.
31. S. Zhou, Y. Wang, P. Hu, W. Zhong, H. Jia, J. Qiu and J. Fu, Laser Photonics Rev., 2022, 17, 2200531.
32. L. Li, Z. Su, W. Zhang, Y. Li, D. Li, G. Bai and L. Chen, Laser Photonics Rev., 2025, 2500346.
33. H. Huang, J. Chen, Y. Liu, J. Lin, S. Wang, F. Huang and D. Chen, Small, 2020, 16, 2000708.
34. Y. Han, C. Gao, T. Wei, K. Zhang, Z. Jiang, J. Zhou, M. Xu, L. Yin, F. Song and L. Huang, Angew. Chem., Int. Ed., 2022, 61, e202212089.
35. A. Teitelboim, B. Tian, D. J. Garfield, A. Fernandez-Bravo, A. C. Gotlin, P. J. Schuck, B. E. Cohen and E. M. Chan, J. Phys. Chem. C, 2019, 123, 2678–2689.
36. A. Casillas-Rubio, K. Hamraoui, D. Mendez-Gonzalez, M. Laurenti, J. Rubio-Retama, O. G. Calderon and S. Melle, Nanoscale, 2025, 17, 18173–18184.
37. F. Wang, J. Wang and X. Liu, Angew. Chem., Int. Ed., 2010, 49, 7456–7460.
38. F. Wang, R. Deng, J. Wang, Q. Wang, Y. Han, H. Zhu, X. Chen and X. Liu, Nat. Mater., 2011, 10, 968–973.
39. S. Guo, M. K. Tsang, W. S. Lo, J. Hao and W. T. Wong, Nanoscale, 2018, 10, 2790–2803.
40. F. Shi, J. Wang, X. Zhai, D. Zhao and W. Qin, CrystEngComm, 2011, 13, 3782–3787.
41. Q. Cheng, J. Sui and W. Cai, Nanoscale, 2012, 4, 779–784.

---