Single-Nanoparticle Luminescence Nanothermometers with Enhanced Sensitivity in Physiological Temperature Range

Abstract

In this paper, we presented the unique optical properties of monodispersed and uniform luminescence nanothermometers (LNTs), composed of core (NaYF₄:Yb³⁺,Er³⁺) and core–shell (NaYF₄:Yb³⁺,Er³⁺@NaYF₄) upconverting nanoparticles (UCNPs). We observed a significant influence of the NaYF₄ shell on the reduction of the luminescence energy loss, which appeared as a steeper temperature vs. luminescence intensity ratio (I _G2/I _G1) curve. Moreover, the addition of nonionic IGEPAL CO-520 surfactant and a small amount of deionized water led to the additional improvement of the nanothermometer’s sensitivity, especially in the physiological temperature range between 35 and 40 °C. Such behavior was explained by the physical and chemical interactions of the surfactant molecules with the particle surface, which led to a significant reduction in luminescent energy loss. The developed method of lanthanide-based LNT synthesis and characterization is suitable for the preparation of in vitro nanothermometers and could be applied, for example, in microelectronics or environmental monitoring.

1. Introduction

The great demand for the invention of highly sensitive, precise, quick-responded, and noninvasive nanothermometers, with high spatial resolution and applicability in various environments and temperature ranges, has arisen over a decade ago due to the fast development of nanotechnology and biomedicine. So far, various luminescence and nonluminescence nanothermometers have been used in the form of, e.g., quantum dots, carbon nanotubes, organic dyes, fluorescence proteins, thermoresponsive polymers or lanthanide-doped nanoparticles. − Nevertheless, there are still some challenges to overcome, e.g., the localized heating problems and volume shrinking of a sensor, which often leads to its sensitivity reduction and makes the precise calibration impossible.

Recently, lanthanide-doped upconverting nanoparticles (UCNPs) have gained much attention in luminescence nanothermometry due to their ability for precise nanoscale temperature monitoring, biological compatibility, and unique optical properties, such as temperature-sensitive and long-lifetime luminescence, sharp emission lines, large anti-Stokes shifts, negligible autofluorescence, and high penetration depth into living cells. In addition, their remarkable thermal and chemical stability makes them very promising for the long-term biosensing and detection, multimodal cell imaging or phototherapy of critical diseases, − as well as for the real-time temperature sensing in microelectronics. UCNPs can be useful in lithium-ion batteries to prevent overheating and in microreactors for the precise thermal control, as well as for temperature mapping in catalytic reactions to evaluate temperature gradients and in nanoscale thermography in harsh environments due to their high chemical and thermal stability. −

One of the key applications of luminescence nanothermometers (LNTs) is mapping the temperature of living cells, which leads to the faster diagnosis and treatment of diseases. Accordingly, any activity at the cellular level, e.g., enzymatic reactions, cell division, increased metabolism or gene expression, produces a large amount of energy and causes an increase in cellular temperature, which is usually small and transient, leading to the difficulties in the precise temperature detection. , Therefore, there is still a challenge to invent a super sensitive and noninvasive sensing probe, able to detect temperature changes at the subcellular level. − Luminescence nanothermometry seems to be the accurate tool to meet these criteria, but still it is limited due to the photobleaching and photoblinking effects, which influence the intensity of luminescence. In order to improve the single-cell thermometry, various fluorescent nanomaterials have been developed, including quantum dots, nanodiamonds, or organic dyes. Nevertheless, most of them induce photodamage to living cells, due to the high-energy light wave excitation (UV–vis). −

Herein, we chose the lanthanide-based UCNPs, which are able to convert the low-energy light (from the NIR range 920–980 nm) to the high-energy photons (green and red light), which are then emitted, as the result of the sequential absorption of multiple low-energy photons and further two-step energy transfer between lanthanide ions. To be more precise, the Er³⁺ ions are being excited from their ground states (⁴I_15/2) to the intermediate states (⁴I_11/2) and further to the excited states (⁴F_7/2) (Figure ).

1.

Simplified energy level diagram of the Yb³⁺ and Er³⁺ ions with the two-step excitation mechanism.

The NIR light is less harmful to biological tissues during imaging than ultraviolet of visible light, as the excited anti-Stokes emission process reduces the cellular autofluorescence, cell photodamage and allows deep penetration of light into biological tissues, leading to the significant contrast improvement. This is crucial, especially in in vivo studies, where it is important to see deeper layers without cutting or disturbing the tissue. It was reported that the UCNP-based nanothermometers allow monitoring the temperature change of living cells and thermal dynamics of mitochondria. ,,

The UCNPs are generally composed of the inorganic host matrix (e.g., halides, oxides, phosphates, vanadates), sensitizer, and activator, which are two types of lanthanide ions (e.g., Yb³⁺-Er³⁺, Yb³⁺-Tm³⁺, Yb³⁺-Ho³⁺). Most often, fluorides are used as host materials (e.g., NaYF₄, CaF₂, BaGdF₅), as they exhibit high chemical stability and low phonon energy, leading to the maximum of the radiative emission. The most effective fluoride-based matrix, composed of harmless components and revealing negligible cytotoxicity, is β-NaYF₄ possessing lower symmetry and a hexagonal crystalline phase; it is broadly used in biological and biomedical applications. − Moreover, the intensity of multicolor upconversion luminescence can be controlled by the appropriate doping concentration of the sensitizer–activator ion pairs in the host matrix. Nevertheless, the first coordination sphere of lanthanides is sensitive to the presence of water and other specific compounds, due to the vibronic coupling with the O–H, N–H, or C–H groups and quenching of excited states (via nonradiative recombination). Also, in highly diluted water dispersions, the constituting ions can be released, causing cell death. In order to overcome these difficulties, the UCNPs are usually being covered by the protecting shell, which has often the same chemical structure as the host matrix (e.g., NaYF₄) and can be further surface-functionalized, ,− which may contribute to the formation of separated single nanoparticles. Nevertheless, the typical surface functionalization is often not enough to stop the aggregation process, and thus, surface-active compounds have to be admixed, e.g., surfactants.

It is commonly known that ionic, as well as nonionic, surfactants are able to form self-associates called micelles, which consist of polar and nonpolar interfaces and have sizes in the range of 10 to 100 nm. Introducing micelles into a dispersion system can help to improve separation between nanoparticles and other compounds. Accordingly, the nanoparticles subjected to separation can be localized in the micellar interior as well as on the micellar boundary. Recently, nonionic surfactants have been broadly studied, as they exhibit a high stability in aqueous solutions even though they do not undergo ionization. Widely used are especially ethoxylated alkyl phenol surfactants, which exhibit strongly temperature-dependent physicochemical properties. ,

Herein, we present the excellent properties of single core NaYF₄:Yb³⁺,Er³⁺ and core–shell NaYF₄:Yb³⁺,Er³⁺@NaYF₄ nanoparticles, stabilized with the nonionic surfactant IGEPAL CO-520 (polyoxyethyelene (5) nonylphenylether). These UCNPs exhibit an enhanced sensitivity in the physiological temperature range of 35–40 °C, superb photostability, and absence of photoblinking. All of these features, together with the monodispersity of UCNPs, make them excellent candidates for single-particle nanothermometers.

2. Materials and Methods

Anhydrous yttrium­(III), ytterbium­(III) and erbium­(III) chlorides (99%), octadec-1-ene (90%), sodium hydroxide, ammonium hydrogen difluoride, and IGEPAL CO-520 (polyoxyethylene (5) nonylphenylether) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Oleic acid (98%), hexane, methanol, ethanol, and acetone were obtained from Lach-Ner (Neratovice, Czech Republic). All of the chemicals were used as received.

2.1. Synthesis of the Core NaYF₄:Yb³⁺,Er³⁺ UCNPs

The core nanoparticles were synthesized according to Kostiv et al. Briefly, 0.78 mmol of YCl₃, 0.2 mmol of YbCl₃, 0.02 mmol of ErCl₃, 6 mL of oleic acid, and 15 mL of octadec-1-ene were placed in a 100 mL three-neck round-bottom flask and heated at 160 °C for 30 min under an argon atmosphere. Further, the mixture was cooled to room temperature, and in the next step, a solution of NaOH (4 mmol) and NH₄F·HF (2.6 mmol) in methanol (5 mL) was added. Then, the mixture was slowly heated at 120 °C under an argon atmosphere, in order to evaporate methanol. The reaction proceeded for 1.5 h at 300 °C under an argon atmosphere and finally, the mixture was cooled to room temperature. Such UCNPs were separated by centrifugation (6000 rpm; 30 min), redispersed in hexane, precipitated in ethanol and again separated by centrifugation (6000 rpm; 30 min). All of these purification steps were repeated two times, and the resulting nanoparticles were dispersed in hexane to a concentration of 10 mg/mL.

2.2. Synthesis of the Core–Shell NaYF₄:Yb³⁺,Er³⁺@NaYF₄ UCNPs

The core–shell nanoparticles were synthesized similarly to the core UCNPs, according to Podhorodecki et al. Namely, 0.2 mmol of YCl₃, 6 mL of oleic acid, and 15 mL of octadec-1-ene were placed in a 100 mL three-neck round-bottom flask and heated at 160 °C for 30 min under an argon atmosphere. Further, the mixture was cooled to room temperature and in the next step, 50 mg of core NaYF₄:Yb³⁺,Er³⁺ particles in hexane and the methanolic solution of NaOH (0.8 mmol) and NH₄F·HF (0.5 mmol) were added drop by drop. Then, the mixture was slowly heated at 120 °C under an argon atmosphere and the reaction continued for 1.5 h at 300 °C under an argon atmosphere. The obtained core–shell NaYF₄:Yb³⁺,Er³⁺@NaYF₄ nanoparticles were precipitated in acetone (10 mL), washed three times in ethanol and redispersed in hexane.

2.3. Preparation of Monodispersed UCNPs for Optical Measurements

In order to investigate the optical properties of the single core NaYF₄:Yb³⁺,Er³⁺ and core–shell NaYF₄:Yb³⁺,Er³⁺@NaYF₄ nanoparticles, which have the tendency to sediment over time in a hexane solution, 1 mg/mL of UCNPs was dispersed in 2 mL of hexane, and in the next step, 100 μL (0.22 mmol) of nonionic IGEPAL CO-520 surfactant and 10 μL (0.56 mmol) of deionized water were admixed. , The composition of dispersion was characterized by a water/surfactant molar ratio (ω₀), which was equal to ω₀ = 2.55, following eq :

$${\omega }_0=\frac{m({\mathrm{H}}_2\mathrm{O})·M(\mathrm{surf}.)}{m(\mathrm{surf}.)·M({\mathrm{H}}_2\mathrm{O})}$$ 1

where m (H₂O) - mass of water; m (surf.) - mass of surfactant; M (H₂O) - molecular weight of water; and M (surf.) -molecular weight of surfactant.

Thus, the number of water molecules in the UCNPs/hexane/IGEPAL CO-520/H₂O system was 2.55 times larger than the number of present surfactant chains, which caused the formation of rather small reverse micelles.

Before measurement, all UCNP-based dispersions were sonicated for 5 min using a UP400 St ultrasonic device (Hielscher Ultrasonics GmbH, Germany). In the next step, 10 μL of each stabilized sample was diluted with 90 μL of hexane and sonicated for 5 min. The procedure was repeated until appropriate spatial distribution of individual nanoparticles was observed in the microscopic system described below (point 2.4.).

2.4. X-ray Diffraction (XRD) Measurements

X-ray diffraction (XRD) structural analysis was performed by using a Malvern Panalytical Empyrean diffractometer. The instrument was equipped with a copper X-ray tube (λ = 1.540598 Å, CuKα₁ radiation). The sample suspension was loaded into a quartz capillary, which was subsequently mounted on a capillary spinner stage. Data were collected in the θ–2θ configuration over an angular range of 10 to 110°. A linear X-ray focus and a 1/4° antiscatter slit were employed to optimize beam collimation. Diffraction data were recorded using a Pixcel3D detector, with an acquisition time of 1500 s per step.

2.5. Transmission Electron Microscopy (TEM) Measurements

The morphology of synthesized UCNPs was analyzed using a Talos F200i transmission electron microscope (Thermo Fisher Scientific, Eindhoven, Netherlands) with an accelerating voltage of 200 kV and a fiber-coupled Ceta-M electron camera. The sample dispersion (4 μL) was deposited on a copper grid with the carbon support film, and after evaporation of hexane, the UCNPs were observed in the bright-field imaging mode. The size of the UCNPs was calculated using ImageJ software and averaged from at least 200 nanoparticles. The number-average diameter D _n was determined following eq :

$$D_n=\frac{\sum N_i·D_i}{\sum N_i}$$ 2

where N _i and D _i are the number and diameter of particles, respectively.

2.6. Dynamic Light Scattering (DLS) Measurements

The hydrodynamic diameter (D _h) and polydispersity index (PDI) of UCNPs were determined by the dynamic light scattering (DLS) method using a Zetasizer Nano ZS (Malvern Panalytical) apparatus with a detection angle of 173°, equipped with a 4 mW He–Ne laser (λ = 633 nm). Measurements were carried out at 25 °C using PCS1111 glass cuvettes. Each measurement was repeated three times for each sample, and the D _n and PDI results were given as the mean ± standard deviation (SD). The samples were filtered before each experiment, using a 0.22 μL PTFE filter. The concentration of each sample was ∼0.5 mg/mL.

2.7. Fluorescence Microscopy (FM) Measurements

In order to sequentially detect the luminescence of single UCNPs, we built the two-color wide-field microscopic system, which was described in details in our previous publication.

The tested sample containing UCNPs (0.1 mg/mL) was dropped onto the glass slide (5 μL), which, after solvent evaporation, was placed in the measuring chamber. In the next step, 2 mL of distilled water, used as a heat transmitter, was applied on the top of glass with UCNPs. Further, the spatial distribution of nanoparticles was observed using a CMOS camera and multiaxis motorized stage. After choosing the best place on the glass slide, covered with UCNPs, the sample heating started and the emission spectra (luminescence intensity vs wavelength) of individual nanoparticles were recorded. The microscopic system was equipped with three bandpass filters, limiting the signal reaching the detector to the range of the selected green band (filter 520/35 and filter 543/22) or both of these bands, cutting off the red emission, which also appeared in the observed luminescence spectrum. The obtained luminescence intensity maps were recorded with the same optical experimental setup and detector parameters (CMOS camera settings: excitation power, exposure time, binning, and detector gain).

2.8. Luminescence Spectra Interpretation

The emission spectrum of Er³⁺ ions constitutes of two intensity bands (Figure ), which are the result of energy transfer from the two excited states (²H_11/2 and ⁴S_3/2) to the ground state (⁴I_15/2) (Figure ). The first band is localized between 513 and 533 nm of the wavelength range, while the second one lies between 534 and 564 nm. In order to simplify the result interpretation, the above-mentioned bands were denoted as G1 and G2, respectively, where:

2.

Emission spectrum of Er³⁺ ions (in the range of green light) obtained at 24.4 °C for the core UCNPs, excited by laser light with a wavelength of 980 nm. The transmission bands of the bandpass filters representing channels G1 and G2 are framed on the graph.

G1: ²H_11/2 → ⁴I_15/2

G2: ⁴S_3/2 → ⁴I_15/2

3. Results and Discussion

The crystal structure of synthesized core NaYF₄:Yb³⁺,Er³⁺ and core–shell NaYF₄:Yb³⁺,Er³⁺@NaYF₄ UCNPs was determined using the XRD method and the normalized patterns are shown in Figure . We observed that the positions of recorded peaks were consistent with the reference pattern from JCPDS (card no. 00–064–0156), indicating the pure hexagonal (β-phase) structure of the core, as well as core–shell nanoparticles, with the space group of P6₃/m. Such a result was expected, as it is commonly known that the β-phase NaYF₄ crystals predominantly grow at high reaction temperatures and they exhibit higher luminescence intensity than the cubic (α-phase) of NaYF₄, forming at low reaction temperatures. ,

3.

Normalized X-ray diffraction pattern of synthesized β-NaYF₄:Yb³⁺,Er³⁺ core and β-NaYF₄:Yb³⁺,Er³⁺@NaYF₄ core–shell UCNPs, as well as the standard data of the β-NaYF₄ lattice according to JCPDS no. 00–064–0156.

The synthesized core (NaYF₄:Yb³⁺,Er³⁺) UCNPs were roughly spherical and monodispersed, revealing a narrow size distribution (Figure ). In contrast, the shape of core–shell (NaYF₄:Yb³⁺,Er³⁺@NaYF₄) UCNPs was rather ellipsoidal, due to the introduction of an anisotropic NaYF₄ shell, through epitaxial layer-by-layer growth, which may be nonhomogenous (Figure ). The number-average diameter (D _n ), determined using TEM microscopy and eq , was 23 and 26 nm (±1 nm) in the case of core and core–shell nanoparticles, respectively. Thus, the shell thickness was between 1 and 1.5 nm. It is necessary to mention that the number-average particle diameter and the shell thickness of the core–shell UCNPs were calculated based on the length of the shortest particle axis. Importantly, any signs of particle aggregation could be detected, which is highly desirable in nanothermometry.

4.

TEM micrographs of NaYF₄:Yb³⁺,Er³⁺ (a, b) and NaYF₄:Yb³⁺,Er³⁺@NaYF₄ (c, d) UCNPs, with corresponding particle size distributions.

The size of UCNPs expressed by the hydrodynamic diameter (D _h), calculated using the DLS method, was slightly larger (Figure a) and was equal to 24.6 and 27.7 nm for the core and core–shell UCNPs, respectively. Moreover, the polydispersity index (PDI) of the core and core–shell UCNPs in hexane was slightly lower than in the hexane/IGEPAL CO-520/H₂O system (Figure a), which might be caused by the presence of free surfactant molecules in the case of the latter. In order to obtain the size of IGEPAL CO-520/H₂O-based micelles, we allowed UCNPs to settle down for a week and we measured the DLS of the upper emulsion layer. The detected reverse micelles had D _h = 9.3 nm (Figure b).

5.

Number-weighted size distributions of (a) core and core–shell UCNPs, with/without the addition of IGEPAL CO-520/H₂O and (b) IGEPAL CO-520/H₂O reverse micelles present in the core/hexane/IGEPAL CO-520/H₂O system. The hydrodynamic diameters (D _h) and the polydispersity indexes (PDI) were marked on each graph.

Based on the work of van der Loop et al. we proposed the structure of IGEPAL CO-520/H₂O reverse micelles (Scheme ), in which the poly­(ethylene oxide) chains of IGEPAL CO-520 do not penetrate into the micellar center containing water. Such a conclusion was made based on the observations of water behavior (molecular reorientation), which was the same for water inside reverse micelles and water in the bulk.

1. Water-in-Oil UCNPs/Hexane/IGEPAL CO-520/H₂O Emulsion.

It was already published that the NaYF₄:Yb³⁺,Er³⁺-based UCNPs sediment fast in hexane. Thus, the addition of nonionic IGEPAL CO-520 surfactant and deionized water becomes justified as we observed an immediate change in the sample transparency and stability (Figure ). Accordingly, the UCNPs with surfactant did not sediment for several weeks. Also, the formed water-in-oil emulsion contributed to the improvement in the quality of optical measurements.

6.

Optical difference between the dispersion of core UCNPs (1 mg/mL) in hexane (a, b) and in a hexane/IGEPAL CO-520/H₂O emulsion (c, d).

In detail, the applied two-color wide-field microscopic method allowed us to obtain the luminescence intensity maps for the G1 and G2 bands (described in point 2.8.), which were analyzed using the ImageJ program (Figure ). We chose a representative group of 45 nanoparticles, the same on each map, which revealed similar size, shape and luminescence intensity. In the next step, the maximum of luminescence intensity was measured for each single nanoparticle and finally, the average value of the luminescence intensity ratio (I _G2/I _G1), as well as the standard deviation (type A), was calculated. The obtained temperature dependencies of the luminescence intensity ratios (I _G2/I _G1) of single nanoparticles are shown in Figure .

7.

Luminescence intensity maps for the bands: G1 (filter 520/35) (a) and G2 (filter 543/22) (b) recorded for the same field of view with the CMOS camera at 25 °C. Images were acquired sequentially (<1 s apart) with identical settings (excitation at 980 nm; exposure time and detector gain fixed). Chosen single NaYF₄:Yb³⁺,Er³⁺ UCNPs, which were taken into account for further calculations of the luminescence intensity, are marked in green circles.

8.

Temperature dependency of the normalized luminescence intensity ratio (I _G2/I _G1) obtained for the core NaYF₄:Yb³⁺,Er³⁺(a) and core–shell NaYF₄:Yb³⁺,Er³⁺@NaYF₄ (b) nanoparticles dispersed in hexane, as well as for the core (c) and core–shell (d) nanoparticles in the hexane/water emulsion, stabilized with IGEPAL CO-520 surfactant. Standard deviations of the I _G2/I _G1 values marked on the graphs were in the range of 1 to 5.3%. The intensity ratio I _G2/I _G1 values were normalized relative to the maximum value for the given particles.

At this stage, we have to mention that we used the IGEPAL CO-520 surfactant in order to obtain well-dispersed, single UCNPs in hexane, which could be easily detectable with an optical microscope. It is very important for measurements on single nanoparticles that they are well-separated so that the signal is collected from a single nanoparticle and not from their aggregates. Therefore, we used the previously described micelle strategy, which drastically changed the luminescence intensity of the UCNPs in the temperature region between 35 and 40 °C (Figure c,d). Such behavior seemed very interesting, especially in terms of UCNPs' practical use. Thus, we tried to explain this phenomenon and our clarification was presented in the following sections of the manuscript.

Analysis of the obtained temperature vs. luminescence intensity ratio (I _G2/I _G1) dependencies (Figure ) allowed us to observe a typical linear trend only in the case of core and core–shell nanoparticles dispersed in hexane (Figure a,b). Namely, an increase in temperature caused a linear decrease in the I _G2/I _G1 ratio, which is in accordance with the Boltzmann factor: ,

$$\frac{I_{\mathrm{G}1}}{I_{\mathrm{G}2}}=C·{\exp }^{-\mathrm{\Delta }E/k·T}$$ 3

where I _G1 and I _G2 are integrated intensities of the ²H_11/2 → ⁴I_15/2 and ⁴S_3/2 → ⁴I_15/2 transitions, C is a constant dependent on the parameters of the tested material, e.g., its degeneracy and spontaneous emission rate, ΔE is the energy gap between two excited states, k is the Boltzmann constant, and T is the absolute temperature.

An additional explanation for the decrease in the I _G2/I _G1 ratio with increasing temperature (Figure a,b) could be an observed significant decrease in the intensity of the G2 band (Figure a,b). It is visible (Figure ) that with increasing temperature of the core and core–shell UCNPs less energetic G2 transition is more significant, compared to the higher energetic G1, which could be explained by the higher susceptibility of the G2 transition to disturbances from the surrounding environment. Moreover, the slope of the I _G2/I _G1 vs temperature curve (Figure ) was steeper in the case of core–shell UCNPs (Figure b), which indicates higher sensitivity of NaYF₄:Yb³⁺,Er³⁺@NaYF₄ UCNPs to changes in temperature.

9.

Temperature dependency vs. normalized intensity of G1 and G2 transitions obtained for the core NaYF₄:Yb³⁺,Er³⁺ (a) and core–shell NaYF₄:Yb³⁺,Er³⁺@NaYF₄ (b) nanoparticles dispersed in hexane, as well as for the core (c) and core–shell (d) nanoparticles in the hexane/water emulsion stabilized with IGEPAL CO-520 surfactant. The intensity of G1 and G2 was normalized for each sample separately, relative to the maximum of the strongest band.

Such materials, especially core–shell UCNPs, might be used in nanothemometry, in which the intensity ratio of the two energy transfers, revealing different temperature dependencies, is monitored. Importantly, in the case of the core–shell UCNPs, the intensity of the G1 band remained nearly constant with increasing temperature (Figure b), which could be explained by the fact that the shell surrounding the core ions led to the decrease in the number of nonradiative transitions. Therefore, we concluded that the intensity of the higher energetic G1 band might be strongly dependent on the surface quenching, which is minimized in the case of the core–shell UCNPs.

In contrast, the addition of IGEPAL CO-520 and water into the UCNP/hexane system caused an unexpected jump in the luminescence intensity ratio (I _G2/I _G1), in the temperature region between 35 and 40 °C (Figure c,d). Moreover, the slope of the I _G2/I _G1 vs temperature curve was more steep for the UCNP-based systems containing the IGEPAL CO-520 surfactant, except the temperature region between 35 and 40 °C, which indicated more effective luminescence and thus better optical properties and higher sensitivity of the core and core–shell nanoparticles. This unusual behavior could be caused by the formation of thermodynamically stable water-in-oil emulsion with the reverse IGEPAL CO-520/H₂O micelles (Scheme ), which contributed to the better nanoparticle separation, reducing the amount of aggregates. We concluded that the presence of surfactant molecules caused stabilization of UCNPs and reduction of the quenching process, which led to insignificant changes in G1 transition. Thus, the observed drop in the luminescence intensity was mainly caused by the decrease in less energetic G2 transition (Figures c,d and c,d).

According to the literature, , when the temperature increases, water droplets, which are inside reverse micelles of IGEPAL CO-520 (Scheme ), also slightly increase in size due to the partial breaking of the hydrogen bonds between the water center of the micelle and the surrounding IGEPAL molecules, leading to the increased amount of free surfactant molecules in the solution. Therefore, we concluded that in our system the situation must be similar. Namely, when the particle dispersion in hexane is applied onto the microscopic slide, prior to microscopic measurement, hexane is being evaporated and in the next step the remaining UCNPs are being covered with 2 mL of distilled water. Therefore, the UCNPs and the hydrophobic chains of IGEPAL CO-520 are most probably directed toward hydrophobic glass (Scheme ) due to the repulsive interactions with surrounded water molecules.

2. Process of the Structural Changes in Reverse Micelles and Further Formation of Surfactant Self-Assemblies Occurring Most Probably under the Increase in Temperature of the System.

In the next step, when the temperature reaches ∼35 °C, most probably the water-based micellar centers slightly increase in size and their concentrations slightly decrease (Scheme ), as the hydrogen bonds between water droplets and surfactant molecules begin to break. Thus, the number of UCNPs stabilized with IGEPAL CO-520 molecules increases with an increasing temperature. Importantly, the surfactant chains can be oriented to the particle surface by the hydrophobic or hydrophilic part of the chain, bearing a hydroxyl group. We estimated that if there are free places on the particle surface, which are not occupied by oleic acid, the hydroxyl group of the IGEPAL CO-520 can be chemically attached. The rest of the surfactant molecules can stabilize UCNPs by the hydrophobic interactions between the hydrophobic part of the IGEPAL CO-520 chain and oleic acid, localized on the particle surface (Scheme ). Such behavior of the surfactant molecules most probably caused the detected jump in the I _G2/I _G1 ratio (Figure c,d), which is most probably related to the decrease in the number of nonradiative transfers leading to the lower loss of the energy emitted by the erbium ions. An explanation of this phenomenon may also be the fact that the polarity of the hydroxyl group of IGEPAL CO-520 is lower than the polarity of the carboxyl group of oleic acid surrounding UCNPs. Thus, the electronic configuration of UCNPs changes, as well as their mutual interactions with surface-coordinated molecules (i.e., oleic acid/IGEPAL CO-520). Our considerations could be also supported by the fact that an increasing amount of free IGEPAL CO-520 molecules (separated from the reverse micelles) in the system caused an increase in fluorescence intensity of the tested material.

Another fact is that the observed jump in the I _G2/I _G1 ratio (Figure c,d) was more pronounced in the case of the core–shell UCNPs (Figure d), which might be caused by the denser packing of the oleic acid and IGEPAL CO-520 chains around the protective shell of UCNPs and thus lower loss of energy in the form of nonradiative transfers. Such conclusion becomes from the fact that the core–shell UCNPs are larger than the core ones and thus have also larger specific surface area, allowing them to coordinate more oleic acid/surfactant chains.

It is important to mention that the properties of the interfacial nonionic surfactant layer, surrounding the UCNPs covered with the oleic acid, strongly depend on the temperature, the elastic properties of the surfactant molecules, i.e., the radius of their spontaneous curvature, and on their mutual interactions with water molecules. ,− Thus, in the temperature range of 35–40 °C, the poly­(ethylene glycol) chain configuration of IGEPAL CO-520 most probably is changing from more organized, densely packed zig-zag, present in the reverse micelles, to nonordered coil-like structures, present in the water solution, due to the breaking of hydrogen bonds between surfactant molecules and water droplet centers of micelles.

Above 40 °C, the IGEPAL CO-520 molecules released from the reverse micelles could start forming self-assemblies, e.g., surfactant-based micelles with hydrophobic centers, due to the increase in thermal motions of water molecules, leading to the movement of the surfactant molecules away from the particle surface. Thus, the UCNPs lost their additional protection against the nonradiative energy loss, which led to the further linear decrease in the luminescence intensity ratio (I _G2/I _G1) with increasing temperature (Scheme and Figure c,d). We estimated that in the temperature range between 40 and 50 °C some partially destroyed micelles are still present (Scheme ), as it was published that the water begins to be ejected from the reverse micelles of IGEPAL CO-520 only from 50 °C.

4. Conclusions

In summary, we synthesized the monodispersed and uniform UCNPs, due to which it was possible to reduce the luminescence measurement error to 1–5.3%. Accordingly, the number-average size of the core NaYF₄:Yb³⁺,Er³⁺ nanoparticles was 23 nm, whereas the core–shell NaYF₄:Yb³⁺,Er³⁺@NaYF₄ UCNPs were slightly larger (D _n = 26 nm), with the average shell thickness between 1 and 1.5 nm.

We proposed a mechanism explaining how the addition of nonionic IGEPAL CO-520 surfactant and a small amount of deionized water affects the luminescence intensity of the UCNPs. In detail, we presented unique optical properties of the core NaYF₄:Yb³⁺,Er³⁺ and core–shell NaYF₄:Yb³⁺,Er³⁺@NaYF₄ UCNPs stabilized with the IGEPAL CO-520/H₂O reverse micelles at 25 °C, as well as with free surfactant molecules at higher temperatures. We observed that the most promising UCNPs for the use in nanothermometry were the core–shell nanoparticles, as they revealed more steep temperature dependency of the luminescence intensity ratio (I _G2/I _G1) and thus were more sensitive to temperature changes. Addition of IGEPAL CO-520/H₂O into the hexane dispersion of core or core–shell UCNPs led to the unexpected increase in the I _G2/I _G1 ratio in the physiological temperature range (35–40 °C), which might be due to the breaking of hydrogen bonds between surfactant and water molecules, constituting centers of the reverse micelles, and subsequent coordination of the free surfactant molecules on the UCNP surface. Such a mechanism provides greater protection of the latter against the loss of the emitted luminescent energy and thus, the enhanced UCNPs are more easily detectable.

Importantly, we observed that the presence of the NaYF₄ shell significantly reduced the number of nonradiative transitions and, thus, the loss of luminescent energy, which might be related to the unchanged intensity of the G1 band with increasing temperature. Therefore, such uniform core–shell NaYF₄:Yb³⁺,Er³⁺@NaYF₄ nanoparticles dispersed in hexane exhibit promising potential for biomedical applications, particularly as in vitro single-nanoparticle luminescence nanothermometers capable of monitoring temperature changes at the cellular or subcellular level within cultured cells or biological tissues. In detail, the synthesized UCNPs could be immobilized onto a substrate (e.g., Petri dish) via solvent evaporation, followed by direct seeding of biological cells for in vitro temperature sensing experiments. Moreover, the properties of UCNPs could be improved by the addition of a small amount of IGEPAL CO-520 and deionized water, leading to higher accuracy of such LNTs in the physiological temperature range. Accordingly, the use of IGEPAL CO-520 prevents the aggregation of UCNPs, which allows to achieve an appropriate dilution series and a sparse distribution of nanoparticles in the prepared samples. Such a feature is highly desirable in nanothermometry, to precisely control the temperature at a given point, by single upconverting nanoparticles.

The obtained results constitute an important step toward developing a sensitive method for measuring temperature at the microscale using single UCNPs. Such a single luminescent nanoparticle placed in a biological system would not disturb its functioning. Thus, our next goal is to assess the compatibility of the monodispersed NaYF₄:Yb³⁺,Er³⁺@NaYF₄ UCNPs with living tissues, since we are able to transfer the particles into aqueous solution. Once this is achieved, we will be able to measure the single nanoparticle temperature by monitoring changes in its luminescence.

B.K.: Project administration, conceptualization, methodology, supervision, validation, and writingreview and editing. M.Ś.: Conceptualization, methodology, investigation, supervision, validation, visualization, writingoriginal draft, and writingreview and editing. K.H.-M.: Data curation, formal analysis (FM), investigation, methodology, validation, visualization, and software. D.H.: Resources and writingreview and editing. M.W.: Formal analysis (TEM). R.K.: Formal analysis (DLS). J.S.: Formal analysis (XRD). A.P.: Writingreview and editing.

The authors declare no competing financial interest.