Hydrothermal Synthesis and Upconversion Properties of About 19 nm Sc₂O₃: Er³⁺, Yb³⁺ Nanoparticles with Detailed Investigation of the Energy Transfer Mechanism

Abstract

The Sc₂O₃: Er³⁺, Yb³⁺ nanoparticles (NPs) with the size of about 19 nm were synthesized by a simple oleic acid-mediated hydrothermal (HT) process. X-ray diffraction (XRD), transmission electron microscopy (TEM), upconversion luminescence (UCL) spectra, and decay curves were used to characterize the resulting samples. The Sc₂O₃: Er³⁺, Yb³⁺ NPs made by HT method exhibit the stronger UCL, of which the red UCL are enhanced by a factor of 4, in comparison with those samples prepared by solvothermal (ST) method at the same optimized lanthanide ion concentrations. The UCL enhancement can be attributed to the reduced surface groups and longer lifetimes. Under 980 nm wavelength excitation, the decay curves of Er³⁺: (²H_11/2, ⁴S_3/2) → ⁴I_15/2 and ⁴F_9/2 → ⁴I_15/2 emissions for Sc₂O₃: Er³⁺, Yb³⁺ NPs samples are both close to each other, resulting from the cross relaxation energy transfer from Er³⁺ to Yb³⁺, followed by an energy back transfer within the same Er³⁺-Yb³⁺ pair. Also, under the relatively low-power density, the slopes of the linear plots of log(I) vs. log(P) for red and green emissions are 2.5 and 2.1, implying the existence of three-photon processes. Our results indicate that Sc₂O₃: Er³⁺, Yb³⁺ NPs is an excellent material for achieving intense UCL with small size in the biological fields.

Introduction

Infrared to visible upconversion luminescence (UCL) has been extensively studied for its fundamental value [1,2,3] and its various potential applications in upconversion lasers, bioimaging, infrared imaging, solar cells, etc. [4,5,6,7,8]. The co-doping of Er³⁺ and a high concentration of sensitizer Yb³⁺ forms the most attractive energy transfer (ET) upconversion system [1]. Under 980 nm infrared excitation of the sensitizer Yb³⁺, this system can generate green and red emission originating from the (²H_11/2, ⁴S_3/2) → ⁴I_15/2 and ⁴F_9/2 → ⁴I_15/2 transitions of Er³⁺, respectively [9]. Selection of appropriate host material is essential in the synthesis of lanthanide-doped nanocrystals (NCs) with favorable optical properties such as high UC efficiency and controllable emission profile. The practical applications require the development of more efficient, high stability UC materials with low excitation density [10, 11]. Oxide materials are usually very stable chemically, mechanically, and thermally, and could therefore be promising hosts for UC applications [3, 12,13,14,15,16]. The cubic sesquioxide materials (such as Y₂O₃, Lu₂O₃, Sc₂O₃, etc.) display particular structural characteristics and physical properties. For example, Y₂O₃ shows up the outstanding UCL as the typical oxide host [3, 17]. The Sc₂O₃ has the smallest lattice parameter. The short Sc–Sc bond length in Sc₂O₃ can produce the short distance within an Yb³⁺-Er³⁺ pair, speeding up the Yb³⁺ → Er³⁺ energy transfer. In our previous work, Sc₂O₃: Er³⁺, Yb³⁺ nanostructures were obtained using a biphasic solvothermal (ST) method [17]. The red UCL in this samples are enhanced, compared with the bulk sample synthesized using a solid-state (SS) reaction. The average crystal size of nanostructures has reduced to about 200 nm, which favors the application in fluorescence imaging.

A variety of chemical techniques, including coprecipitation, solvothermal synthesis (ST), hydrothermal method (HT), sol–gel processing, thermal decomposition, etc., have been demonstrated to synthesize lanthanide-doped UC NCs [14, 18,19,20,21,22]. Optimization of synthesis procedure is critical to obtain NCs with tailored crystal size, morphology, surface functionalization, and optical properties. The HT approach is a good choice due to its convenience, exemption from pollution, and the possibility of achieving satisfying crystallinity at a relatively low temperature [23]. Zhao et al. utilized an oleic acid-mediated HT method for the synthesis of UC NaYF₄ nanorods, nanotubes, and flower-patterned nano-disks [20]. Chen et al. prepared Fe³⁺ co-doped NaYF₄: Er, Yb UC NCs by a HT method using oleic acid as a capping ligand and a surface modifier [24]. In this work, Sc₂O₃: Er³⁺, Yb³⁺ nanoparticles (NPs) of 19 nm in average diameters have been first synthesized through a simple oleic acid-mediated HT method. We found the stronger UCL in this Sc₂O₃: Er³⁺, Yb³⁺NPs samples, of which the red UCL are enhanced by a factor of 4, in comparison with that in the same optimized concentration Sc₂O₃ samples by ST method. The UCL enhancement can be attributed to the reduced surface groups and longer lifetimes. Additionally, the UCL property and mechanism of HT-Sc₂O₃: Er³⁺, Yb³⁺ NPs were investigated by the spectra distribution, power dependence, and lifetime measurement.

Experimental

Sample Preparation

The Sc₂O₃: Er³⁺, Yb³⁺ samples were prepared by the HT method via the hydrolysis of relevant mineral salts in an ethanol scheme. The high purity raw materials of Sc₂O₃, Er₂O₃, and Yb₂O₃ powers were dissolved in dilute HNO₃ and deionized water to obtain cationic nitrates solutions, respectively. The Sc(NO₃)₃, Er(NO₃)₃, and Yb(NO₃)₃ solutions with corresponding mole ratios were dissolved in absolute ethanol (20 ml), stirring to form a homogeneous solution. Then an aqueous sodium hydroxide solution (2 ml) was added dropwise to the above mixture with stirring for 30 min, followed by adding oleic acid (1 ml), then vigorous stirring for 1 to 2 h. The resulting suspension was placed in a close Teflon-lined stainless steel autoclave with 50 ml capacity and heated at 180 °C for 24 h. After the autoclave was cooled to room temperature, naturally the precipitate was then centrifuged and washed several times with deionized water and absolute ethanol, respectively. The powder was obtained after being dried in a vacuum oven at 80 °C for 15 h and annealed 700 °C for 2 h. For comparison, we prepared Sc₂O₃ samples prepared by the ST method at the same sintering temperature 700 °C for 2 h [17].

Measurements and Characterization

Powder X-ray diffraction (XRD) datum was collected using Cu-Kα radiation (λ = 1.54056 Å) on an X-ray powder diffractometer (Rigaku D/Max IIA). Transmission electron microscopy (TEM) image was obtained by using a transmission electron microscope (JEM-2000EX) operating at an acceleration voltage of 200 kV. The UCL spectra were recorded with a spectrophotometer (Hitachi F-7000) and infrared spectra were performed by using a Triax 550 spectrometer (Jobin-Yvon) pumped with a power-controllable 980 nm diode laser at room temperature. Infrared spectra in transmission mode were measured on a Thermofisher Nicolet IS50 FT-IR spectrometer, using pressed KBr tablets. In fluorescence lifetime measurements, an optical parametric oscillator (OPO) was tuned to 980 nm as an excitation source, and the signals were detected by a Tektronix digital oscilloscope (TDS 3052).

Results and Discussion

The structures characterized by the XRD patterns are shown in Fig. 1a for samples by HT method with the nominal compositions of Sc₂O₃: 1%Er³⁺, y%Yb³⁺ (x = 0, 5, 10, 15). The pure phase Sc₂O₃ was synthesized in agreement with JCPDS card 84-1884. The host lattice exhibits the mineral bixbyite structure with the $Ia\overline{3}$ (T^h₂) symmetry [25]. In this structure, Sc³⁺ is sixfold with the effective ionic radius (0.745 Å). The Yb³⁺ ions owned the large ionic radius (0.868 Å) occupy Sc³⁺ sites to expand the lattice cell volume, making XRD peaks shift to smaller angles as Yb³⁺ concentration increases as shown in the magnified patterns of Fig. 1b. To further reveal the morphology and size distribution, the as-prepared Sc₂O₃ samples were characterized by TEM. Figure 2a shows the TEM image of HT-Sc₂O₃: 1%Er³⁺, 5%Yb³⁺. We obtained the sphered NPs with relatively uniform size and good monodispersity. Figure 2b depicts the histogram of the size distribution; these data were obtained from the TEM image of more than 300 NPs. The average diameter of NPs was determined to be about 19 nm.

Fig. 1

a XRD patterns for HT-Sc₂O₃: 1%Er³⁺, y%Yb³⁺ (x = 0, 5, 10, 15) NPs. b Magnified patterns in the diffraction angle ranged from 30° to 33°

Fig. 2

a TEM image and b histogram of size distribution of HT-Sc₂O₃: 1%Er³⁺, 5%Yb³⁺ NPs

Figure 3 shows the UCL spectra of Sc₂O₃: 1%Er³⁺, 10%Yb³⁺ (a) and Sc₂O₃: 1%Er³⁺, 5%Yb³⁺ (b) samples prepared by HT and ST methods under 980 nm excitation with an output power density of 3 mW mm⁻². The strong emission bands centered at ~ 550 and 660 nm are attributed to the 4f - 4f electronic transitions of Er³⁺: (²H_11/2, ⁴S_3/2) → ⁴I_15/2 and ⁴F_9/2 → ⁴I_15/2 transitions, respectively. The insets present the digital photographs of corresponding samples. It reveals that UCL has been dramatically enhanced for the HT sample, compared with the ST one. For HT-Sc₂O₃ samples, the calculated enhancement factor of red UCL is around 4, compared with corresponding ST-Sc₂O₃ samples. It is known that the size of samples has an influence on UCL intensity, which decreased with the decreasing of the size. However, for HT-Sc₂O₃ sample, it owns smaller size and more intensive UCL. It indicates the HT-Sc₂O₃ sample is an excellent material owned intense UCL with small size for the biological fields.

Fig. 3

UCL spectra of Sc₂O₃: 1%Er³⁺, 10%Yb³⁺ (a) and Sc₂O₃: 1%Er³⁺, 5%Yb³⁺ (b) samples prepared by HT and ST methods, respectively, pumped under 980 nm excitation. The insets present the digital photographs of corresponding samples

The FTIR spectra of HT-Sc₂O₃: 1%Er³⁺, 5%/10%Yb³⁺ and ST-Sc₂O₃: 1%Er³⁺, 5%Yb³⁺/10%Yb³⁺ samples are shown in Fig. 4. The broad band around 3429 cm^− 1 is attributed to the stretching vibration of –OH in the oleic acid (OA) and water [26, 27]. The 2925 and 2850 cm^− 1 absorption bands are assigned to the asymmetric and symmetric stretching vibrations of the methylene (CH₂) in the long alkyl chain of the OA molecules. The sharpness of the bands indicates that the hydrocarbon chains are well ordered. The anti-symmetric methyl stretch (CH₃) is seen as a shoulder on the peak at 2975 cm⁻¹. The bands at 1200–1750 cm⁻¹ can be assigned to the vibrations of C=O in the oleic acid molecule and CO₂ in the air [28]. The transformation to carbonate might have occurred on the surface of crystallites during the heat treatment. These results evidence the existence of capping ligands on the surfaces of samples. Figure 4 shows the absorption intensities of –OH vibration for ST-Sc₂O₃ samples are stronger. The intensities of surface groups for HT/ST-Sc₂O₃: 1%Er³⁺, 10%Yb³⁺ samples are both stronger than that in co-doped 5%Yb³⁺ samples. The abundant surface groups with available large vibrational quanta may efficiently enhance the MPR processes, inducing the decline of luminescence.

Fig. 4

FTIR spectra of HT-Sc₂O₃: 1%Er³⁺, 5%/10%Yb³⁺ (a) and ST-Sc₂O₃: 1%Er³⁺, 5%Yb³⁺/10%Yb³⁺ (b) samples

In order to exactly describe the population mechanism in Er³⁺/Yb³⁺ co-doped HT-Sc₂O₃ sample, the dependence of spectral distributions on the Er³⁺/Yb³⁺ concentrations has been studied in detail.

The UCL spectra of HT-Sc₂O₃: x%Er³⁺, 10%Yb³⁺ (x = 0, 0.5, 1, 2) under 980 nm excitation are presented in Fig. 5a. For the fixed Yb³⁺ concentration at 10%, the strongest UCL is observed for Er³⁺ concentration around 1%. When Er³⁺ concentration exceeds 1%, the intensity begins to diminish because of the cross relaxation (CR) of Er³⁺ ions [17]. The UCL spectra of HT-Sc₂O₃: 1%Er³⁺, y%Yb³⁺, (y = 0, 5, 10, 15) are presented in Fig. 5b. For the Er³⁺ singly doped Sc₂O₃, its UC emission is very faint, which has been magnified 100 times. The ET process of Yb³⁺ → Er³⁺ plays a dominant role for UCL enhancement. The strongest UCL is observed for Yb³⁺ concentration 5% when fixed the optimal Er³⁺ concentration 1%.

Fig. 5

UCL spectra of HT-Sc₂O₃: x%Er³⁺, 10%Yb³⁺ (x = 0, 0.5, 1, 2) (a) and HT-Sc₂O₃: 1%Er³⁺, y%Yb³⁺, (y = 0, 5, 10, 15) (b) under 980 nm excitation

The near infrared emission spectra in the range of 1000–1700 nm for the same variety samples are shown in Fig. 6. In the Er³⁺/Yb³⁺ co-doped samples, 980-nm photon excites Yb³⁺: ²F_7/2 → ²F_5/2 which exhibits fluorescence at 1000–1200 nm exciting Er³⁺ ions into ⁴I_11/2 level through a nonresonant phonon-assisted ET process [9]. The Er³⁺ ions in ⁴I_11/2 level decay nonradiatively to ⁴I_13/2 level, then radiatively to the ground state emitting the photon around 1550 nm [9]. In Fig. 6a, as Er³⁺concentration increases, the Yb³⁺ emission has a steady decline which evidences the efficient Yb³⁺ → Er³⁺ ET. The Er³⁺ emission gradually increases when Er³⁺ concentration increases from 0 to 1%, then declines slightly as a result of the self-absorption of Er³⁺ ions. In Fig. 6b, Er³⁺: ⁴I_13/2 emission gradually enhances when Yb³⁺ concentration increases from 0 to 5% but subsequently begins to decrease. As Yb³⁺ concentration increases, Yb³⁺ capacity of 980 nm photon absorption is enhanced. The Yb³⁺ emission intensity is shown to increase. Meantime, as the distance of Yb-Yb and Yb-Er pairs decreases, the enhanced energy migration among Yb³⁺ ions speeds up ET from Yb³⁺ to Er³⁺. It leads to the increased population of Er³⁺: ⁴I_13/2 level but the decreased one of Yb³⁺: ²F_5/2level. Due to the quenching of Er³⁺ by Yb³⁺ ions, the emission of Er³⁺: ⁴I_13/2 → ⁴I_15/2 reaches a maximum then drops down.

Fig. 6

Near-infrared emission spectra in the range of 1000–1700 nm for HT-Sc₂O₃: x%Er³⁺, 10%Yb³⁺ (x = 0, 0.5, 1, 2) (a) and HT-Sc₂O₃: 1%Er³⁺, y%Yb³⁺, (y = 0, 5, 10, 15) (b) under 980 nm excitation

The pumping power dependences of Er³⁺: (²H_11/2, ⁴S_3/2) → ⁴I_15/2 and Er³⁺: ⁴F_9/2 → ⁴I_15/2 intensities in HT-Sc₂O₃: 1%Er³⁺, 10%Yb³⁺ are measured under 980 nm excitation and plotted in a double logarithmic scales in Fig. 7. For the UCL processes, the UCL intensity (I_UCL) depends on the pumping laser power (P) as the equation: I_UCL ∝ Pn where n is the number of pumping photons absorbed per upconverted photon emitted [29]. The n value can be obtained from the slope of the linear plots between log (I) and log (P). For the two-step ET process, the n value is theoretically less than 2 due to the competition between linear decay and UC processes. Figure 7 shows the slope n values for red and green emissions are 2.5 and 2.1 in the low pump power density, respectively. It indicates, except for two-step process, that there are also the three-photon processes in HT-Sc₂O₃: 1%Er³⁺, 10%Yb³⁺ NPs [30, 31].

Fig. 7

Power dependence curves for Er³⁺: (²H_11/2, ⁴S_3/2) → ⁴I_15/2 and ⁴F_9/2 → ⁴I_15/2 transitions in HT-Sc₂O₃: 1%Er³⁺, 10%Yb³⁺ NPs

The upconversion mechanism is drawn in Fig. 8. The ET process is as follows:

• The ET①: Yb³⁺: ²F_5/2 + Er³⁺: ⁴I_15/2 → Yb³⁺: ²F_7/2 + Er³⁺: ⁴I_11/2

  • Er³⁺: ⁴I_11/2 → Er³⁺: ⁴I_13/2 (MPR)

• The ET②: Yb³⁺: ²F_5/2 + Er³⁺: ⁴I_13/2 → Yb³⁺: ²F_7/2 + Er³⁺: ⁴F_9/2

• The ET③: Yb³⁺: ²F_5/2 + Er³⁺: ⁴I_11/2 → Yb³⁺: ²F_7/2 + Er³⁺: ⁴F_7/2

  • Er³⁺: ⁴F_7/2 → Er³⁺: (²H_11/2, ⁴S_3/2) (MPR)

• The ET④: Yb³⁺: ²F_5/2 + Er³⁺: ⁴F_9/2 → Yb³⁺: ²F_7/2 + Er³⁺: ²H_9/2

  • Er³⁺: ²H_9/2 → Er³⁺: (²H_11/2, ⁴S_3/2) (MPR)

  • Er³⁺: (²H_11/2, ⁴S_3/2) → Er³⁺: ⁴F_9/2 (MPR)

• The ET⑤: Yb³⁺: ²F_5/2 + Er³⁺: (²H_11/2, ⁴S_3/2) → Yb³⁺: ²F_7/2 + Er³⁺: ²G_7/2

Fig. 8

The energy level diagrams and dominant upconversion mechanism in Sc₂O₃: Er³⁺, Yb³⁺ NPs under 980 nm pump

To verify and make a theoretical interpretation of the UCL results mentioned above, we utilize the simplified steady-state equations.

$$ \frac{dn_0}{dt}=0 $$

(1)

$$ \frac{dn_1}{dt}={n}_2{W}_{21}-{C}_2{N}_1{n}_1-\frac{n_1}{\tau_1} $$

(2)

$$ \frac{dn_2}{dt}={C}_1{N}_1{n}_0-{C}_3{N}_1{n}_2-{n}_2{W}_{21}-\frac{n_2}{\tau_2} $$

(3)

$$ \frac{dn_3}{dt}={C}_2{N}_1{n}_1-{C}_4{N}_1{n}_3-\frac{n_3}{\tau_3} $$

(4)

$$ \frac{dn_4}{dt}={C}_3{N}_1{n}_2-{C}_5{N}_1{n}_4-\frac{n_4}{\tau_4} $$

(5)

$$ \frac{dN_1}{dt}=\sigma {IN}_0-{C}_1{N}_1{n}_0-{C}_2{N}_1{n}_1-{C}_3{N}_1{n}_2-{C}_4{N}_1{n}_3-{C}_5{N}_1{n}_4-\frac{N_1}{\tau_{Yb}}=0 $$

(6)

Where σ is the absorption cross section of Yb³⁺ ions, I is the incident pumping power, N_i is the population density of the ith level of Yb³⁺, n_i is the population density of ith level of Er³⁺ involved in the upconversion process, τ_i is the lifetime of ith level of Er³⁺and τ_Yb is the lifetime of ²F_5/2 level of Yb³⁺, C_i represents the ET coefficient of Yb³⁺ → Er³⁺ for steps i = 1, 2, 3, 4, 5, and W₂₁ represents the nonradiative rate between 1 and 2 levels of the Er³⁺ ions.

Compared with two-step process, the UC efficiency of three-photon processes from NIR to visible is decreased [32]. Additionally, the high-photon process is prominent when pumping power is high enough. The excitations of Er³⁺: ⁴F_9/2 by ET to Er³⁺: ²H_9/2 can be neglected due to the weak pump in our experiment. By Eq. (4), the red emission intensity (I_Red) can be obtained by

$$ {I}_{Red}={\gamma}_3{n}_3={\gamma}_3{C}_2{\tau}_3{I}_{Yb}{I}_{n_1} $$

Due to the CR of the Er³⁺– Er³⁺ interaction is not considered, the lifetime, τ₃, is a constant. That is to say, ${\mathrm{I}}_{\mathrm{Red}}\propto {\mathrm{I}}_{\mathrm{Yb}}{\mathrm{I}}_{{\mathrm{n}}_1}$, where I_Yb and I_n1 represent the emission intensity of Yb³⁺: ²F_5/2 and Er³⁺: ⁴I_13/2, respectively. The γ₃ is radiative rate of red emission. The calculated I_Redvalues at various Er³⁺/Yb³⁺concentrations are presented in Fig. 9, scaled to the maximum. For comparison, the I_Red values obtained directly from the UCL emission spectra are also depicted. The calculated and experimental I_Red trends are consistent with each other and obtain the best value at the same Er³⁺/Yb³⁺ concentrations, demonstrating the validity of experimental data.

Fig. 9

Calculated and experimental red emission intensities (I_Red) values at various Er³⁺/Yb³⁺concentrations. The intensities are scaled to the maximum

The three-photon green and red UC processes occurred simultaneously result in the increase of the corresponding n values. Meanwhile, the n value of red UC process increases more effectively than that of green UC process. In Fig. 8, the green and red UCL can be populated by CR, as Er³⁺: ⁴G_11/2 + Er³⁺: ⁴I_15/2 → Er³⁺: (²H_11/2, ⁴S_3/2) + Er³⁺: ⁴I_13/2 and Er³⁺: ⁴G_11/2 + Yb³⁺: ²F_7/2 → Er³⁺: ⁴F_9/2 + Yb³⁺: ²F_5/2, respectively [31]. The three-photon green UCL is via a cross-relaxation process between two Er³⁺ ions; however, the cross-relaxation in the three-photon red UCL is between Yb³⁺ and Er³⁺ ions. Since the Yb³⁺ concentration is much higher than Er³⁺ in our experiment, the three-photon red UC process is more effective than the three-photon green UC process, resulting in a rapid increase of n value for red UCL. In addition, it should be noted that all the three-photon processes are few, so the n values deviate obviously from 3. At the high pump power density, two slopes gradually drop to 1 because UC process becomes dominant [33].

The decay curves of the Er³⁺: (²H_11/2, ⁴S_3/2) → ⁴I_15/2 and ⁴F_9/2 → ⁴I_15/2 transitions in HT-Sc₂O₃ and ST-Sc₂O₃ samples under the 980 nm excitation wavelength have been measured and shown in Fig. 10. The decay times for red and green emissions are calculated by integrating the area under the corresponding decay curves with the normalized initial intensity. Figure 10a, b shows the green and red emission lifetimes in HT-Sc₂O₃: 1%Er³⁺, 5%Yb³⁺ are longer than those in ST-Sc₂O₃: 1%Er³⁺, 5%Yb³⁺. The lifetime is proportional to population of level. The longer values indicate the stronger red and green UCL in HT-Sc₂O₃ sample. In our previous report, we found our samples own the shorter decay lifetime values than that in the literature. Actually, the decay times of Er³⁺: (²H_11/2, ⁴S_3/2) → ⁴I_15/2 and ⁴F_9/2 → ⁴I_15/2 emissions for HT/ST-Sc₂O₃: 1%Er³⁺, 5%Yb³⁺ samples are both close to each other. If Er³⁺: ⁴F_9/2 level is populated by the MPR process from Er³⁺: (²H_11/2, ⁴S_3/2) levels, the decay time of Er³⁺: ⁴F_9/2 level approaches to that of Er³⁺: ⁴S_3/2 level. However, this MPR process is inefficient for population of Er³⁺: ⁴F_9/2 level [17]. There is another non-MPR mechanism for populating the Er³⁺: ⁴F_9/2level from Er³⁺: ⁴S_3/2 level. The mechanism involves CR ET: Er³⁺: (²H_11/2, ⁴S_3/2) + Yb³⁺: ²F_7/2 → Er³⁺: ⁴I_13/2 + Yb³⁺: ²F_5/2; then, in the same Er³⁺–Yb³⁺ pair, an energy back transfer (CRB) Yb³⁺: ²F_5/2 + Er³⁺: ⁴I_13/2 → Yb³⁺: ²F_5/2 + Er³⁺: ⁴F_9/2occurs [1]. If the CRB process dominates the main way for the population of Er³⁺: ⁴F_9/2 level, the decay time of Er³⁺: ⁴F_9/2 level should be almost equal to the decay time of Er³⁺: ⁴S_3/2 level. The CRB process is fast and efficient at low excitation density.

Fig. 10

Decay curves of the Er³⁺: (²H_11/2, ⁴S_3/2) → ⁴I_15/2 and ⁴F_9/2 → ⁴I_15/2 transitions in HT-Sc₂O₃ and ST-Sc₂O₃ samples under the 980 nm excitation wavelength

Figure 11 shows the UCL spectra of three typical sesquioxides under 980 nm excitation. The Sc₂O₃: 1%Er³⁺, 5%Yb³⁺ sample exhibits the strongest UCL in the series of spectra. Furthermore, the emission line of Er³⁺: ⁴F_9/2 level at the lowest energy side in Sc₂O₃ shifts to the longer wavelength side by 8 nm relative to that in Y₂O₃. The nearest Sc-Sc distance is 3.27 Å in Sc₂O₃ shorter than the Y-Y distance (3.752 Å) in Y₂O₃ [3, 17]. The mean Sc–O bond length (2.121 Å) in Sc₂O₃ is shorter than the mean Y–O bond length (2.263 Å) in Y₂O₃. The Er³⁺/Yb³⁺ on Sc³⁺ site in Sc₂O₃ experiences a stronger crystal field than on Y³⁺ site in Y₂O₃. The red shift of spectrum can be attributed to the large Stark splitting of Er³⁺ ions in Sc₂O₃ host. The morphologies of Y₂O₃ and Lu₂O₃ samples were also characterized by TEM as shown in the inset of Fig. 11a, b, respectively, for comparison. The obtained spherical particles are both agglomerated to bulk. The better dispersion and uniformity of Sc₂O₃ NPs synthesized by HT method favor its application in biological assays and medical image.

Fig. 11

UCL spectra of Er³⁺/Yb³⁺ co-doped typical sesquioxide materials under 980 nm excitation

Conclusions

In summary, Sc₂O₃: Er³⁺, Yb³⁺ NPs about 19 nm were synthesized by a simple oleic acid-mediated HT process. The Sc₂O₃: Er³⁺, Yb³⁺ NPs by HT method shows the stronger UCL, of which the red UCL are enhanced by a factor of 4, in comparison with that in the same optimized concentration Sc₂O₃ samples by ST method. The UCL enhancement can be attributed to the reduced surface groups and longer lifetimes. The surface groups enhanced the MPR, inducing the decline of luminescence. Under the 980 nm excitation, the decay curves of Er³⁺: (²H_11/2, ⁴S_3/2) → ⁴I_15/2 and ⁴F_9/2 → ⁴I_15/2 emissions for HT-Sc₂O₃: 1%Er³⁺, 5%Yb³⁺ samples are close to each other, resulting from the non-MPR mechanism for populating the Er³⁺: ⁴F_9/2 level from Er³⁺: ⁴S_3/2 level. The mechanism involves CR ET: Er³⁺: (²H_11/2, ⁴S_3/2) + Yb³⁺: ²F_7/2 → Er³⁺: ⁴I_13/2 + Yb³⁺: ²F_5/2; then, in the same Er³⁺–Yb³⁺ pair, an energy back transfer (CRB) Yb³⁺: ²F_5/2 + Er³⁺: ⁴I_13/2 → Yb³⁺: ²F_5/2 + Er³⁺: ⁴F_9/2 occurs. Under the relatively low-power density, the slopes of the linear plots of log(I) vs log(P) for red and green emissions are 2.5 and 2.1, respectively, which are larger than 2 because of the existence of three-photon processes. Compared with the typical sesquioxides (Y₂O₃ and Lu₂O₃), the Sc₂O₃: 1%Er³⁺, 5%Yb³⁺ NPs exhibits the stronger UCL. Furthermore, in Sc₂O₃ the emission line of Er³⁺: ⁴F_9/2 level at the lowest energy side shifts to the longer wavelength side by 8 nm relative to that in Y₂O₃ owing to the large Stark splitting of Er³⁺ ions in Sc₂O₃ host. Results show the Sc₂O₃: Er³⁺, Yb³⁺ nanoparticles (NPs) is an excellent material for achieving intense UCL with small size in the biological fields.